*Review* **Structural Characterization of the Millennial Antibacterial (Fluoro)Quinolones—Shaping the Fifth Generation**

**Aura Rusu <sup>1</sup> , Ioana-Andreea Lungu <sup>2</sup> , Octavia-Laura Moldovan <sup>2</sup> , Corneliu Tanase 3,\* and Gabriel Hancu <sup>1</sup>**


**Abstract:** The evolution of the class of antibacterial quinolones includes the introduction in therapy of highly successful compounds. Although many representatives were withdrawn due to severe adverse reactions, a few representatives have proven their therapeutical value over time. The classification of antibacterial quinolones into generations is a valuable tool for physicians, pharmacists, and researchers. In addition, the transition from one generation to another has brought new representatives with improved properties. In the last two decades, several representatives of antibacterial quinolones received approval for therapy. This review sets out to chronologically outline the group of approved antibacterial quinolones since 2000. Special attention is given to eight representatives: besifloxacin, delafoxacin, finafloxacin, lascufloxacin, nadifloxacin and levonadifloxacin, nemonoxacin, and zabofloxacin. These compounds have been characterized regarding physicochemical properties, formulations, antibacterial activity spectrum and advantageous structural characteristics related to antibacterial efficiency. At present these new compounds (with the exception of nadifloxacin) are reported differently, most often in the fourth generation and less frequently in a new generation (the fifth). Although these new compounds' mechanism does not contain essential new elements, the question of shaping a new generation (the fifth) arises, based on higher potency and broad spectrum of activity, including resistant bacterial strains. The functional groups that ensured the biological activity, good pharmacokinetic properties and a safety profile were highlighted. In addition, these new representatives have a low risk of determining bacterial resistance. Several positive aspects are added to the fourth fluoroquinolones generation, characteristics that can be the basis of the fifth generation. Antibacterial quinolones class continues to acquire new compounds with antibacterial potential, among other effects. Numerous derivatives, hybrids or conjugates are currently in various stages of research.

**Keywords:** fluoroquinolones; quinolones; structure-activity relationship; DNA gyrase; topoisomerase IV; antibacterial activity

#### **1. Introduction**

The historical moment of the emergence of a new class of antibacterial compounds was in 1945 when George Lesher and his team discovered the antimicrobial potential of 7-chloro-quinoline. This molecule was a compound with bactericidal action isolated during the synthesis and purification of chloroquine (antimalarial agent). Nalidixic acid, the first antibacterial quinolone (QN) derivative introduced in therapy, was discovered based on this compound (characterized by a naphthyridine nucleus) and was introduced into therapy in 1963 [1–4].

**Citation:** Rusu, A.; Lungu, I.-A.; Moldovan, O.-L.; Tanase, C.; Hancu, G. Structural Characterization of the Millennial Antibacterial (Fluoro)Quinolones—Shaping the Fifth Generation. *Pharmaceutics* **2021**, *13*, 1289. https://doi.org/10.3390/ pharmaceutics13081289

Academic Editor: Tihomir Tomašiˇc

Received: 9 July 2021 Accepted: 14 August 2021 Published: 18 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The identification of a compound which is efficient against Gram-negative bacteria led to new derivatives as pipemidic acid, piromidic acid, oxolinic acid, cinoxacin and flumequine, the first generation of antibacterial quinolones (QNs) [5]. Flumequine was the first compound with a fluorine atom in the structure [6]. This optimization proved to be valuable for the next generation of antibacterial QNs. New quinoline derivatives were synthesized with superior pharmacokinetics and pharmacodynamic properties and a broader antibacterial spectrum [7,8]. Thus, the second generation of QNs was synthesized, obtained by introducing a fluorine atom in the sixth position of the quinolinic nucleus (Figure 1). These new QNs called generically "fluoroquinolones" (FQNs) presented an improved biological activity [9]. Numerous FQNs have been synthesized and studied. New compounds with an extended antibacterial spectrum, being active on both Gram-positive and Gram-negative bacteria (including *Pseudomonas aeruginosa*), have become valuable tools in therapy. The second generation comprises of both representatives for human use (norfloxacin, ciprofloxacin, ofloxacin), and for veterinary use (enrofloxacin) [3].

**Figure 1.** The general chemical structure of FQNs (1,4-quinolones) and numbering (X and Y = C or N).

More valuable representatives were included in the third generation as levofloxacin (the *L*-enantiomer of ofloxacin) and gatifloxacin, which presented increased activity against Gram-positive bacteria (*Streptococcus* sp.), increased tissue penetration and half-life. Due to severe side effects (hypoglycemia), gatifloxacin is used only topically as eye drops [3,10,11]. Fourth generation FQNs have, in addition, acquired activity against anaerobic bacteria (e.g., moxifloxacin). Also, levofloxacin and moxifloxacin were included in the therapeutic protocols used in second-line multidrug-resistant tuberculosis [12,13]. The optimization of the chemical structure also led to a long half-life in moxifloxacin (13 h) [3,11]. Data on the discovery of antibacterial QN class representatives and their approval in therapy by the U.S. Food and Drug Administration (FDA) and/or European Medicines Agency (EMA) are briefly presented in Table 1.

The aim of this review is to present the progress in FQNs class since 2000. The newest FQNs introduced in therapy are highlighted and critically analyzed. Special attention is given to eight selected representatives (besifloxacin, delafloxacin, finafloxacin, lascufloxacin, levonadifloxacin, nadifloxacin, nemonoxacin, and zabofloxacin), approached from the perspective of physicochemical properties, antibacterial activity spectrum and advantageous structural modifications which influence for antibacterial efficiency.


**Table 1.** The representatives of the class of antibacterial QNs and their approval in therapy.


<sup>1</sup> first year reported, <sup>2</sup> racemic, <sup>3</sup> diastereoisomers, <sup>4</sup> *R* or *S* isomer.

#### **2. Research Methodology**

The literature research was conducted mainly on Clarivate Analytics and ScienceDirect databases using relevant keywords: (a) topic "fluoroquinolones", "quinolones", "antibacterials"; (b) title: "besifloxacin", "delafloxacin", "finafloxacin", "lascufloxacin", "nadifloxacin", "levonadifloxacin", "nemonoxacin", "zabofloxacin", and other relevant representatives of the FQNs class.

The articles were selected if they included relevant data regarding the aspects referred to in our review: discovery of the compound and the entities involved, data on approval in therapy, pharmaceutical formulations, infections treated by the targeted representatives, antibacterial activity spectrum, physicochemical properties, structure-activity relationships, elements of the safety profile related to chemical structure optimizations, bacterial resistance, and new QN derivatives. The manuscript contains relevant references, including those published in the first part of 2021.

Biovia Draw 2019 was used for drawing chemical structures (https://discover.3ds. com/biovia-draw-academic, accessed on 6 July 2021) [60]. MarvinSketch was used for drawing, displaying and characterizing chemical structures MarvinSketch 20.20.0, ChemAxon (https://www.chemaxon.com, accessed on 11 June 2021) [61].

#### **3. Mechanism of Action**

The FQNs mechanism of action is well known and described in the literature [62–66]. It is known that FQNs act on two bacterial DNA enzymes: gyrase and topoisomerase IV (Figure 2) [67]. Thus, due to the covalent enzyme-DNA complex stabilization, DNA is cleaved. After this interaction, depending on the concentration, the death of the bacterial cell occurs in two ways: (1) at low concentration by blocking replication and transcription [62,68] and (2) at higher concentration (over the minimum inhibitory concentration) when the DNA topoisomerase is dissociated/removed [69], the DNA strands remain free, which leads to the chromosome fragmentation [70–72]. The advantage of new representatives is the action on both target enzymes and broadening the spectrum of activity against several types of pathogens [73,74]. In general, DNA gyrase from Gram-negative bacteria is more susceptible to inhibition than topoisomerase IV. On the other hand, topoisomerase IV from Gram-positive bacteria is more susceptible to inhibition than DNA gyrase [75].

**Figure 2.** The mechanism of action of antibacterial (fluoro)quinolones.

Some studies have shown a correlation between FQNs lethality and reactive oxygen species (ROS) formation [76–79]. On the other hand, some issues about how do FQNs induce ROS accumulation remain unclear [80]. For example, Rodríguez-Rosado et al. (2018) studied the mechanisms of FQN-induced mutagenesis and the role of N-acetylcysteine in FQNs therapy to inhibit FQN-induced mutagenesis [81].

## **4. Classification into Generations of FQNs Used in Therapy**

The most widely used classification of FQNs is the classification into generations based on of antibacterial activity and therapeutic use (Table 2).


**Table 2.** Classification into generations of the main FQNs for human use in therapy based on antibacterial spectrum and therapeutic indications (FDA and EMA approved).


#### **5. Compounds in Therapy Since 2000**

The class of FQNs has evolved significantly since 2000, acquiring valuable representatives for therapy, with a low risk of occurrence of antibacterial resistance (Figure 3). The fourth-generation antibacterial QNs are very active on the DNA gyrase and topoisomerase IV, enzymes involved in bacterial DNA replication, transcription, repair and recombination. Recently approved by FDA or EMA are besifloxacin (2009), finafloxacin (2014) and delafloxacin (2017). The action on the two target enzymes confers the advantage of being effective on bacteria resistant to FQNs from previous generations; the development of

bacterial resistance to the fourth generation representatives with multi-target properties is more difficult [42].

**Figure 3.** New FQNs chronology in therapy (since 2000) and essential structural characteristics.

The newest antibacterial 1,4-QNs used in therapy have diverse structural characteristics (Figure 4). According to the chemical structure of the base nucleus, these new compounds have a QN nucleus (besifloxacin, delafloxacin, finafloxacin, and nemonoxacin), a tricyclic ring including a QN nucleus (nadifloxacin), and naphthyridine nucleus (zabofloxacin). Regarding the presence of halogen atoms in the chemical structure, these compounds contain one fluorine atom (nadifloxacin and levonadifloxacin, finafloxacin, and zabofloxacin), one fluorine and one chlorine atom (besifloxacin), three fluorine atoms (lascufloxacin), three fluorine and one chlorine atoms (delafloxacin), but there is also an exception without any halogen atom (nemonoxacin).

**Figure 4.** Chemical structures of newer approved antibacterial QNs (\*final stage of approval).

Other FQNs from different generations that will be referred to for similarity to specific structural fragments, antibacterial activity or safety profile are described structurally in Figure S1—Supplementary Materials.

#### *5.1. Besifloxacin*

Besifloxacin is a chloro-FQN included in the fourth generation [101]. This new antibacterial molecule was developed for ophthalmic use by the SS Pharmaceutical SSP Co.Ltd. from Tokyo, Japan (former SS734). It has been approved by the FDA in 2009 and registered under the trade name Besivance (Bausch & Lomb Inc., Rochester, NY, USA) [46,98,101,102]. Besifloxacin is indicated for the treatment of bacterial conjunctivitis [98,103].

Besifloxacin's spectrum of activity includes various bacterial species (broadspectrum) [104–106] described in Table S1—Supplementary Materials. As for its mechanism of action, besifloxacin inhibits the two target enzymes, DNA gyrase and topoisomerase IV, essential in DNA replication [107]. Physicochemical properties of besifloxacin are comprised in Table 3. Exclusive topical administration is a peculiarity in the class of FQNs [101]. Most ophthalmic FQNs are also systemically (e.g., ciprofloxacin, ofloxacin, levofloxacin, moxifloxacin) adminstered. Gatifloxacin is administered for ophthalmic use only after the withdrawal from systemic use due to its side effects (hypo- and hyperglycemia) [43,108,109]. The approved pharmaceutical formulation of besifloxacin is an ophthalmic suspension 0.6%, which contains 6.63 mg of besifloxacin hydrochloride (equivalent to 6 mg of besifloxacin) [46,95–97,107]. At present, attempts are being made to develop several ophthalmic pharmaceutical formulas with besifloxacin, such as nanoemulsions [110], positively charged liposomes [111], and for the treatment of bacterial keratitis new loaded nanofibrous ocular inserts [112].

Substitution at the N1 position in the FQN structure is essential for its antimicrobial activity. The N1 position substituent has been shown to control bacterial activity (potency) and some pharmacokinetic properties, like the increased volume of distribution and bioavailability [3]. That is why this substituent is common with other valuable FQNs such as ciprofloxacin and moxifloxacin (Figure S1—Supplementary Materials) [6,113,114]. It was considered that the cyclopropyl moiety from the N1 position of the QN nucleus confers besifloxacin activity against aerobic bacteria (Figure 5) [115].

**Figure 5.** Chemical structure of besifloxacin.


**Table 3.** Physicochemical properties of besifloxacin.

<sup>1</sup> DMSO—Dimethyl sulfoxide.

Regarding the relationship between chemical structure and biological activity, substitution with a halogen (fluorine or chlorine) leads to decreased solubility, increased lipophilicity and increased penetration of the drug through cell membranes [120,121]. The electronic effects (inductive electron-attracting properties) are maximal for chlorine and very weak for fluorine [122]. The introduction of a fluorine atom at position C6 led to a spectacular increase in antimicrobial activity comparative to non-fluorinated QNs from the first generation. One fluorine atom in position C6 increased the degree of penetration into the bacterial cell and, at the same time, the activity against Gram-negative bacteria [6,113,114]. The fluorine atom appears to be essential in the mechanism of action.

A second substitution in the C8 position with a chlorine atom add an increased antimicrobial potency through the action on the target enzymes DNA gyrase and topoisomerase IV [46,123]. Also, C8 chlorine increases the antibacterial activity against FQN-resistant mutants of *Mycobacterium smegmatis* and *Staphylococcus aureus* [121].

Representatives of the second generation (ciprofloxacin, norfloxacin) with a piperazinyl group in the C7 position (Figure S1—Supplementary Materials) exhibit antibacterial activity against Gram-negative bacteria [6]. The C7 amino ring is a key substituent related to toxicity and solubility for analogues as clinafloxacin (with a 3-amino-1-pyrrolidinyl substituent) and sitafloxacin (with a (7*S*)-7-amino-5-azaspiro[2.4]heptan-5-yl) (Figure S1— Supplementary Materials). Unfortunately, clinafloxacin presented some side effects as phototoxicity and hypoglycemia. In addition, the solubility of clinafloxacin is poor and has inadequate stability in an aqueous solution [124]. Besifloxacin is administered only topically without systemic adverse reactions, being similar in terms of solubility. Assessment of besifloxacin toxicity conducted in silico presented a mutagenicity alert for two degradation products [125]. The replacement of the traditional piperazinyl group of the second generation with a hexahydro-1H-azepine cycle led to broadening the spectrum of activity on Gram-positive bacteria. The 3-aminohexahydro-1H-azepine ring contributes to specific action on the target enzyme DNA-gyrase, besifloxacin being superior to other FQNs in terms of antibacterial activity [126,127].

#### *5.2. Delafloxacin*

Delafloxacin is a recently approved FQN with an anionic chemical structure, from the fourth generation [38,128]. This new antibacterial molecule was developed for systemically use, both for oral and intravenous administration [39] by the Abbott Laboratories, Wakunaga Pharmaceutical (as ABT-492 compound or WQ-3034) and Melinta Therapeutics (former Rib-X Pharmaceuticals). It has been approved by the FDA in 2017 and registered under the trade name Baxdela for the treatment of acute bacterial skin and skin structure infections [39,129,130]. Physicochemical properties of delafloxacin are comprised in Table 4. The new product has the advantage of both oral and intravenous administration [131]. The parenteral form contains 433 mg delafloxacin meglumine (equivalent to 300 mg of delafloxacin) while the oral tablets contain 649 mg delafloxacin meglumine (equivalent to 300 mg of delafloxacin) [94]. Meglumine (1-deoxy-1-(methylamino)-D-glucitol) is a counterion used to increase the solubility of delafloxacin [132,133].

As a mechanism of action, delafloxacin inhibits the target enzymes DNA gyrase and topoisomerase IV, having a similar affinity for both [39,41]. An increased activity at acidic pH is an essential characteristic of this new chloro-FQN. Delafloxacin presents a broad spectrum of activity being active against both Gram-positive and Gram-negative bacteria, including methicillin-resistant *Staphylococcus aureus* (MRSA) and *Pseudomonas aeruginosa* (Table S1—Supplementary Materials) [130,134].

New formulations are being created to increase the effectiveness of delafloxacin. Optimized delafloxacin-loaded stearic acid (lipid) chitosan (polymer) hybrid nanoparticles proved to be superior comparative to delafloxacin standard suspension [135].


**Table 4.** Physicochemical properties of delafloxacin.

<sup>1</sup> DMSO—Dimethyl sulfoxide.

Delafloxacin differs from other FQNs by the substituent 3-hydroxyazetidinyl at the C7 position. Also, in the N1 position, delafloxacin has an unusual 6-amino-3,5 difluoropyridinyl moiety that substantially enlarges the molecule's molecular surface. This group is responsible for activity against Gram-positive bacteria [131,139]. This unique 3-hydroxyazetidinyl moiety on the C7 position confers acidic properties, and consequently, delafloxacin behaves as a weak acid (a non-zwitterion molecule with p*K*a 5.4) [131,139].

At acidic pH, delafloxacin is an uncharged molecule, which is favorable for its passage through biological membranes (Figure 6).

These properties give delafloxacin increased activity in an acidic pH environment with decreased minimal inhibitory concentrations (MIC). Intracellularly, at neutral pH delafloxacin will be ionized into the anionic form and thus remain inside the pathogen agent [140]. So, this drug is beneficial against abscesses produced in the infection with *Staphylococcus aureus* [131,140].

**Figure 6.** The macroprotonation scheme of delafloxacin and the step-wise protonation constants K<sup>1</sup> , K<sup>2</sup> and K<sup>3</sup> . The carboxylate, hydroxyl, and the pyridine ring's nitrogen atom (N1 position) are the most acidic, respectively, the most basic functions. All data were calculated with the MarvinSketch 20.20.0 version from ChemAxon [61].

In the C8 position, delafloxacin presents a chlorine substituent with an electronwithdrawing effect on the aromatic fragment of the QN nucleus (Figure 6), like besifloxacin (Figure 5). The chlorine substituent stabilizes delafloxacin molecule and could have a role in the reduction of the development of bacterial resistance. Thus, the whole polar molecule has increased activity [131,139]. C7 and C8 substitutions influence potency and spectrum of activity; both substitutions provide activity against anaerobic bacteria [139]. As a consequence, delafloxacin has proved activity against Gram-positive bacteria, especially against MRSA [139,141].

In the history of the development of FQNs, several trifluorinated molecules (e.g., fleroxacin, temafloxacin, trovafloxacin; Figure S1—Supplementary Materials) have been withdrawn due to severe side effects (Table 1). Fleroxacin (with N1-fluoroethyl, C6-fluor, C8-fluor) was the first promising trifluorinated representative but it was withdrawn due to severe phototoxic reactions [142]. Also, temafloxacin (with N1-difluorophenil) has been withdrawn due to severe hemolysis [19]. Finally, trovafloxacin (with N1-difluorophenil) has been withdrawn due to hepatotoxicity [142]. Unlike temafloxacin and trovafloxacin, delafloxacin contains a 6-amino-3,5-difluoropyridinyl substituent. This substituent appears to be more advantageous in reducing possible adverse reactions that have led to the withdrawal of the other trifluorinated compounds. However, the effects imprinted by the other substituents and the type of base nucleus must also be considered.

#### *5.3. Finafloxacin*

Finafloxacin (BAY35-3377) is a recent cyano-FQN included in the fourth generation. This new antibacterial molecule was developed by Bayer HealthCare Pharmaceuticals, Byk Gulden and MerLion Pharmaceuticals [42,50,99,143]. Relevant physicochemical properties are listed in Table 5. The FDA approved an otic suspension in 2014 and registered under the trade name Xtoro (developed by Novartis's division, Alcon, Geneva, Switzerland). Finafloxacin is indicated for the treatment of acute otitis externa [42,99].

At the same time, finafloxacin is in various stages of clinical trials to evaluate the efficacy of oral and intravenous formulations. These forms are intended for the treatment of uncomplicated and complicated urinary tract infections, pyelonephritis and *Helicobacter pylori* infections [99,144–147]. Finafloxacin has demonstrated broad-spectrum activity against a range of pathogens [148]. This cyano-FQN is active both in vitro and in vivo against *Pseudomonas aeruginosa* and *Staphylococcus aureus* [99].

As for the mechanism of action, similar to fourth-generation representatives, finafloxacin has a high affinity for the two target enzymes, DNA-gyrase and topoisomerase IV [99]. Antimicrobial activity of finafloxacin is enhanced in acidic conditions (pH 5.8) against multiple pathogens, including skin and urinary pathogens. Finafloxacin exhibits activity at neutral pH comparable to previous generations of FQNs. Also, a more prolonged post-antibacterial effect against multiple species was observed compared to other FQNs at acidic pH. The development of bacterial resistance to finafloxacin is less likely in acidic conditions [99,100,149,150].

**Table 5.** Physicochemical properties of finafloxacin.


<sup>1</sup> DMSO—Dimethyl sulfoxide.

The molecule optimizations include a pyrrolo-oxazinyl moiety at the C7 position and a cyano-substituent at the C8 position (Figure 7). Bearing a zwitterion chemical structure (carboxylate at C3 position and pyrrolo-oxazinyl at C7 position) finafloxacin presents two dissociation constants (Table 5) [143]. The voluminous C7 substituent confers to the molecule's unique characteristic of not being recognized by efflux transporters, the key to decreased bacterial resistance development [48]. The C7 pyrrolo-oxazinyl fragment emerged from the C7 azabicycle (pyrrolidine-piperidine) fragment of moxifloxacin and pradofloxacin (Figure 7), which confers the ability to remain longer in the bacterial cell (difficult to efflux molecules) [154,155].

**Figure 7.** Relevant structural elements to the antibacterial activity of moxifloxacin (C7—pyrrolo-piperidinyl, C8—methoxi), pradofloxacin (C7—pyrrolo-piperidinyl, C8—cyano) and finafloxacin (C7—pyrrolo-oxazinyl, C8—cyano).

A cyano-substituent at C8 is also present in the chemical structure of pradofloxacin, a veterinary-approved FQN classified in the third generation. This compound can be considered an analogue of moxifloxacin (fourth generation) due to the methoxy group at the C8 position which has been replaced by a cyano group. Pradofloxacin is more active against Gram-positive bacteria comparative to previous generations. Also, pradofloxacin exhibits good activity against anaerobic bacteria, similar to moxifloxacin, and an equal or lower activity against Gram-negative bacteria [36,156,157]. The C8 cyano group appears to play an essential role in activity against Gram-positive when comparing finafloxacin with pradofloxacin (Figure 7).

At the N1 position, finafloxacin has a cyclopropyl substituent similar to secondgeneration ciprofloxacin and fourth-generation besifloxacin.

Finafloxacin and delafloxacin in acidic conditions (pH 5.0–6.0 and respectively pH ≤ 5.5) are more active than other FQNs, but for different reasons. In the key C7 position of delafloxacin, the 3-hydroxyazetidine without a basic group is the substituent that confers acidic properties. In finafloxacin, the nitrogen atom from the oxazine fragment is responsible for the great activity in acidic conditions (can accept protons) [42,61]. In acidic conditions, finafloxacin is very active against *Staphylococcus aureus* due to an increased uptake in the bacteria [48].

#### *5.4. Lascufloxacin*

Lascufloxacin (KRP-AM1977) is a new FQN (Figure 8) developed in Japan by Kyorin Pharmaceutical Co., Ltd. [158]. Some physicochemical properties of lascufloxacin are comprised in Table 6.

**Figure 8.** Chemical structure of lascufloxacin.

**Table 6.** Physicochemical properties of lascufloxacin.


<sup>1</sup> DMSO—Dimethyl sulfoxide.

This new antibacterial agent was approved recently in Japan (2019) as a hydrochloride salt (oral formulation, Lasvic® 75 mg tablets) for the treatment of respiratory in-

fections (including community-acquired pneumonia (CAP)) and ear, nose and throat infections [163,164]. Lascufloxacin acts by binding to the target enzymes, DNA gyrase and topoisomerase IV (inhibiting DNA synthesis), similar to other antibacterial FQNs [165]. Also, lascufloxacin demonstrated a high binding capacity to phosphatidylserine (a component of human cell membranes; primary surfactant of alveolar epithelial fluid). Lascufloxacin is superior in tissue penetration (head and neck infections) compared with levofloxacin, garenoxacin, and moxifloxacin [158].

Lascufloxacin proved to be very active against Gram-positive bacteria, including resistant species (Table S1—Supplementary Materials). Also, lascufloxacin is very promising against FQN-resistant pathogens located in the respiratory tract [166–168]. For example, a potent activity of lascufloxacin was proved against first-step mutants of *Streptococcus pneumoniae*. Being a new FQN, lascufloxacin has a significant potential to fight against the installation of bacterial resistance in pneumococcal infections [169].

Another formulation for parenteral administration (KRPAM1977Y) was recently approved in Japan (November 2020) [56,163,164]. Lasvic® is the generic brand and contains lascufloxacin hydrochloride 150 mg [163]. A phase I clinical study of lascufloxacin was recently performed in Japan. The pharmacokinetic and safety profile was assessed in non-elderly healthy men comparative to elderly healthy men. The obtained results proved that lascufloxacin has a safe pharmacokinetic profile without dose adjustments for the two groups of men [170]. The average half-life of lascufloxacin is about 16.1 h after 100 mg (orally administered) [168]. This new FQN presented an extensive distribution into the lungs [171].

The substitution with a fluoroethyl of the N1 position is similar to the fleroxacin, the first trifluorinated antibacterial QN, whose use has been limited by the severe phototoxicity [142]. The fluoroethyl substituent was correlated with the photosensitising effect [172,173]. However, according to the data published so far lascufloxacin has a good safety profile [170].

In the C7 position, FQNs usually have nitrogen heterocycles (five or six atoms), aminopyrrolidines and piperazines. In lascufloxacin chemical structure, an unusual structural fragment is present in the C7 position, a main pyrrolidine heterocycle substituted with a (cyclopropylamino)methyl moiety. This position is essential for interaction with DNA gyrase or topoisomerase IV [114,174,175]. Also, an aminopyrrolidine improves Gram-positive activity, proven by the clinafloxacin representative [176]. Thus, clinafloxacin (with a C8 chlorine atom) was associated with severe side effects (phototoxicity hypoglycaemia) [18]. Sitafloxacin, a FQN approved in Japan (2008) and Thailand (2012) contains in the C7 position a pyrrolidinyl fragment included in a spiro substituent, [(7S)-7-amino-5-azaspiro[2.4]heptanyl] [57,177]. This FQN produces mild to moderate adverse reactions (mostly gastrointestinal disorders and laboratory abnormalities, phototoxicity potential) [178,179]. Another pyrrolidinyl fragment is found in the structure of zabofloxacin, also included in a spiro substituent, [(8*Z*)-8-methoxyimino-2,6-diazaspiro[3.4]octanyl] (chapter 5.7). Unlike sitafloxacin, zabofloxacin is considered a well-tolerated FQN with acceptable side effects [143].

The methoxy substituent in the C8 position improves activity and enhances antimicrobial potency, especially against anaerobic bacteria [158], similar to moxifloxacin and pradofloxacin [34,36,114].

#### *5.5. Nadifloxacin and Levonadifloxacin*

Nadifloxacin is the first FQN approved for dermatological use, being classified in the second generation. This new antibacterial molecule was developed by the Otsuka Pharmaceuticals from Japan (former OPC-7251) [27]. The physicochemical properties of nadifloxacin are comprised in Table 7. It has been approved in Japan in 1993 (Aqutim) and in several countries in the European Union (2000). Nadifloxacin was initially approved for the treatment of acne vulgaris, and then for other skin infections (1998) [26,89,180,181]. The approved topical formulation is a cream containing 1% nadifloxacin [89,180].

Nadifloxacin proved to be effective against Gram-positive (including MRSA and coagulase-negative staphylococci), aerobic Gram-negative, and anaerobic bacteria (Table S1—Supplementary Materials) [88]. Superior antibacterial activity of nadifloxacin has been reported comparative with ciprofloxacin, clindamycin and erythromycin against *Propionibacterium acnes, Staphylococcus epidermidis*, methicillin-susceptible *Staphylococcus aureus* (MSSA), and MRSA. Moreover, nadifloxacin did not have an additional effect on resistance [182].

Regarding the mechanism of action, nadifloxacin inhibits the enzyme DNA gyrase, involved in the synthesis and replication of bacterial DNA [180,183]. Also, nadifloxacin proved to have inhibitory effects upon activated T cells and keratinocytes, as a part of the mechanism involved in its effect against inflammatory acne [87].


**Table 7.** Physicochemical properties of nadifloxacin.

<sup>1</sup> DMF—Dimethylformamide, <sup>2</sup> DMSO—Dimethyl sulfoxide.

Nadifloxacin (Figure 9) is a tricyclic FQN very similar to ofloxacin (Figure S1— Supplementary Materials) [184,190]. The essential modification is the replacement in the C8 position of the methyl-piperazine moiety from the ofloxacin structure with a 4-hydroxypiperidine moiety. Nadifloxacin is considered a lipophilic compound compared to ofloxacin (logP = −0.39) [191,192]. In therapy, nadifloxacin is used as a racemic [26]. However, the two enantiomers have different biological activities. Thus, it is known that the levorotatory (*S*)-isomer is 64- to 256-times more potent than the (*R*)-isomer. Also, the levorotatory (*S*)-isomer is approximately twice as active as the racemate against Gram-positive and Gram-negative pathogens [180]. This stereoisomer is under study as an arginine salt for intravenous administration (WCK 771) and is known as levonadifloxacin [193]. In general, the introduction of hydroxyl groups into the structure of a compound will produce analogues with increased hydrophilicity and low solubility in lipids. The hydroxyl group into the chemical structure provides a new center for hydrogen bonding, which can influence the binding of the analogue to the active center of the target, the biological activity, and metabolism [120]. The introduction of a hydroxyl group on the piperidine heterocycle at position C8 confers a slight increase of the hydrophilic character of nadifloxacin and an increase of acidic properties. However, the molecule has a LogP 2.47, which denotes increased lipophilia and low aqueous solubility. This structural optimization is present in the structure of delafloxacin, but on an azetidine heterocycle on C7 position. Nevertheless, the LogP value of delafloxacin is lower (1.67—predicted value) than that of nadifloxacin.

**Figure 9.** Chemical structures of nadifloxacin (**1**) and levonadifloxacin (**2**).

Levonadifloxacin ((12*S*)-7-fluoro-8-(4-hydroxypiperidin-1-yl)-12-methyl-4-oxo-1 azatricyclo[7.3.1.05,13]trideca-2,5,7,9(13)-tetraene-3-carboxylic acid) is the active *S*(−) isomer of nadifloxacin recently approved in India (Figure 9) [194,195]. The *S*(−) isomer of nadifloxacin, has been shown to be more potent than the *R*(+) isomer and twice as active as the racemic form of nadifloxacin against Gram-positive and Gram-negative bacteria. It is a new broad-spectrum anti-MRSA agent belonging to the benzoquinolizine subclass of QN [180,195].

Levonadifloxacin (WCK 771) (*S*-(−)-9-fluoro-6,7-dihydro-8-(4-hydroxypiperidin-1-yl)- 5-methyl-1-oxo-1H,5H-benzo[i,j] quinolizine-2-carboxylic acid *L*-arginine salt tetrahydrate) is administered parenterally (intravenous) in the form of an *L*-arginine salt while its prodrug alalevonadifloxacin (WCK 2349) ((*S*)-(−)-9-fluoro-8-(4-*L*-alaninyl oxypiperidin-1-yl)-5 methyl-6,7-dihydro-1-oxo-1H,5H-benzo[i,j] quinolizine-2-carboxylic acid, methane sulfonic acid salt) in the form developed of an *L*-alanine ester mesylate salt can be administered orally. Both substances are being developed by Wockhardt Limited (India) [193,196].

Both levonadifloxacin and alalevonadifloxacin have successfully completed phase II and phase III trials, indicating that they are clinically appealing therapeutic alternatives for infections caused by multidrug-resistant Gram-positive pathogens. Due to simultaneous inhibition of DNA gyrase and topoisomerase IV, both representatives exhibit significant antibacterial activity against Gram-negative and Gram-positive bacteria, with an emphasis on MRSA [193,197,198]. Levonadifloxacin has the advantage of being potent against resistant pathogens with a very low frequency of mutation [199,200]. Both substances have been studied for the treatment of acute skin and skin structure bacterial infections, community-acquired bacterial pneumonia, and other infections in both non-clinical and clinical studies [193,197,199].

Because of its non-basic hydroxy piperidine side chain, levonadifloxacin remains un-ionized at acidic pH, allowing it to enter the bacterial cell more easily. As a result, levonadifloxacin's efficacy in acidic conditions increases significantly; this characteristic might be helpful for intracellular activity and antibacterial action [197]. Various in vitro and in vivo investigations have established levonadifloxacin's antibacterial spectrum against Gram-positive, Gram-negative, atypical, and anaerobic pathogens [201].

The excellent bioavailability of oral formulations can be helpful in the smooth switch from parenteral to oral therapy. Both medication forms have well-established pharmacokinetics and safety; in the phase I trial, there were no notable severe or unfavourable clinical or laboratory side effects, indicating that both formulations are well tolerated [193].

#### *5.6. Nemonoxacin*

Nemonoxacin is a new non-fluorinated QN chemotherapeutic (Figure 10). Nemonoxacin (TG-873870) was developed by TaiGen Biotechnology under the commercial name of Taigexyn® for the treatment of CAP, both orally and intravenously as well as the treatment of diabetic foot ulcer infections and skin and soft tissue infections [202]. Procter & Gamble initially developed nemonoxacin, and TaiGen Biotechnology was granted a worldwide license in October 2004. In March 2014, it gained its first global approval in Taiwan to

treat CAP in adults. TaiGen Biotechnology holds the nemonoxacin patent portfolio, which protects the drug's use, composition, and manufacturing techniques until 2029 [53,54].

**Figure 10.** Chemical structure of nemonoxacin.

A clinical study (phase II) regarding the safety and efficacy of nemonoxacin in diabetic foot infections was completed [203,204]. As a result, the FDA authorized oral administration of nemonoxacin to treat CAP and bacterial skin infections [166,205,206].

Taigexyn product contains nemonoxacin malate hemihydrate salt [207,208]. The understudy intravenously administered formula contains nemonoxacin malate sodium chloride [209]. Physicochemical properties of nemonoxacin are comprised in Table 8.


**Table 8.** Physico-chemical properties of nemonoxacin.

The QN ring's C8 methoxy substituent improves antibacterial efficacy against Grampositive bacteria and lowers the selection of resistant variants. The fluorine substituent absence may reduce the frequency of dangerous side effects [143]. The addition of a methoxy group at position C8 allows nemonoxacin to target both DNA gyrase and topoisomerase IV, resulting in a broader spectrum of activity and less mutant selection [213,214].

Nemonoxacin is similar to gatifloxacin (a fourth-generation FQN) (Figure S1—Supplementary Materials), except for the lack of C6 substitutions of fluorine and the 50 -methyl piperidinyl ring at C7 position of the QN ring [215,216]. Gatifloxacin proved to be more active against *Streptococcus pneumoniae* than second generation ciprofloxacin or the third generation levofloxacin [34,216]. This increased activity against *Streptococcus pneumoniae* is similar to moxifloxacin, another fourth-generation FQN with a methoxy group at C8 [34]. The C8 methoxy group of nemonoxacin probably potentiates the same level of inhibition of DNA

gyrase and topoisomerase IV in *Streptococcus pneumoniae* cells, and confers low mutant selectivity [216].

The introduction of a piperidine substituent to C7 has not been common in the past. Few representatives were obtained with a piperidine substituent in the C7 position. Among them is balofloxacin (Figure S1—Supplementary Materials), developed by Choongwae Pharma and approved only in Korea to treat urinary tract infections [32]. At the C7 position, balofloxacin exhibits a 3-(methylamino)piperidinyl moiety [217]. Balofloxacin did not have the expected success. This FQN from the third generation reported total adverse drug reaction rates of 5.4% compared to levofloxacin (1.3%). The side effects reported were gastrointestinal, CNS and skin-related [18,23,218]. Shankar et al. (2018) published the predicted toxicity of balofloxacin and its metabolites (*in silico* study); most of the metabolites are found to be immunotoxic [219]. Avarofloxacin (acorafloxacin, JNJ-Q2) is a new promising FQN in development with a piperidine substituent in the C7 (discussed in a later chapter) [143,220].

*In vitro* investigations have shown that nemonoxacin has broad-spectrum antibacterial activity, including activity against microorganisms resistant to other antibacterial drugs, including multidrug-resistant *Streptococcus pneumoniae* and MRSA [221,222]. It is used to treat Gram-positive and Gram-negative bacterial infections, including MSSA and MRSA, with once-daily oral and intravenous preparations. Nemonoxacin presented higher activity than levofloxacin and ciprofloxacin against a variety of Gram-positive bacteria, including resistant species. For Gram-negative bacteria such as *Escherichia coli, Hemophilus influenzae*, *Klebsiella pneumoniae*, and *Pseudomonas aeruginosa*, nemonoxacin activity was equivalent to levofloxacin and ciprofloxacin [223]. In vitro testing of nemonoxacin against 2440 clinical isolates revealed that it had better efficacy against most Gram-positive species than levofloxacin and moxifloxacin [224]. In the murine model of systemic, pulmonary, or ascending urinary tract infection, nemonoxacin outperforms the most commonly used FQNs [225]. Compared to other FQNs, nemoxacin has a low predisposition for generating resistant infections because bacteria develop resistance to nemonoxacin only when three distinct mutations happen in the QN resistance-determining region of the relevant gene [214].

Nemonoxacin has a favourable pharmacokinetic profile, being rapidly absorbed, having a high bioavailability, and a large distribution volume; it has a relatively long elimination half-life of more than 10 h and achieves maximum concentration (Cmax) 1–2 h after oral administration. Approximately 60–75% of the given dosage is excreted in an unaltered state; only a minor metabolite (5%) was identified due to metabolic processes [213,226]. Nemonoxacin is well tolerated, the gastrointestinal and neurological system-related are the most prevalent side effects of oral administration, with a frequency equivalent to that of levofloxacin therapy [227].

Nemonoxacin may play a significant role in the treatment of many infectious illnesses due to its equivalent or higher potency against Gram-positive bacteria and similar activity against Gram-negative pathogens when compared with other classic FQNs.

#### *5.7. Zabofloxacin*

Zabofloxacin is a FQN approved in 2015 only in South Korea [55,56]. The new compound (PB-101, DW224a, DW224aa) was developed by Dong Wha Pharm. Co. Ltd. (Seoul, Korea) [56]. There were two salts in development: DW224a as zabofloxacin hydrochloride, and DW224aa as zabofloxacin aspartate [228]. The physicochemical properties of zabofloxacin are comprised in Table 9. Zabofloxacin is marketed under the name Zabolante to treat acute bacterial exacerbation of chronic obstructive pulmonary disease by oral administration. Zabolante contains 512.98 mg zabofloxacin aspartate hydrate (equivalent to 366.69 mg of zabofloxacin) [56,229].

Zabofloxacin activity is mainly against Gram-negative and Gram-positive respiratory pathogens, especially against *Streptococcus pneumoniae*, and drug-resistant *Neisseria gonorrhoeae* (Table S1—Supplementary Materials) [230–232]. In Phase III clinical trial zabofloxacin

(367 mg once daily; 5 days) proved to be as efficient as moxifloxacin (400 mg once daily; 7 days) in treating chronic obstructive pulmonary disease exacerbations [92].

Zabofloxacin mechanism of action is similar to other FQNs with a broad spectrum against respiratory pathogens [233].


**Table 9.** Physicochemical properties of zabofloxacin.

Unlike previous compounds in the class of new antibacterial QNs, zabofloxacin is a fluoronaphthyridone (Figure 11) [56].

**Figure 11.** Chemical structure of zabofloxacin.

At the C7 position, zabofloxacin has an unusual heterocycle, a spiro substituent (2,6 diazaspiro[3.4]octan) substituted in C80 with an imino methoxy group. This new compound could be considered an analogue of gemifloxacin (from the fourth-generation) by optimizing the heterocycle from position C7 (Figure S1—Supplementary Materials) [34]. Although the antibacterial activity of gemifloxacin was superior to moxifloxacin, unfortunately, due to side effects (mainly rash), it was withdrawn [34,45]. Increasing the volume of the C7 heterocycle by maintaining the substituted methoxy-imino pyrrolidine ring resulted in a compound with acceptable side effects [228]. Various spiro compounds with antibacterial activity have been published [237–240].

#### **6. Is the Fifth Generation of Antibacterial FQNs Outlined?**

The therapeutic value of the newer FQNs discussed in this review is undeniable. This recent evolution in the class of FQNs is based on several essential elements of the chemical structure, which are further analyzed (Table 10).


**Table 10.** Essential moieties on the QN nucleus for the newer compounds.


<sup>1</sup> Similar to C7 on QN nucleus; <sup>2</sup> Tricyclic structure; <sup>3</sup> Similar to C8 on QN nucleus.

<sup>1</sup> Similar to C7 on QN nucleus; <sup>2</sup> Tricyclic structure; <sup>3</sup> Similar to C8 on QN nucleus.

Jones et al. (2016) consider that avarofloxacin (acorafloxaxin, JNJ-Q2) is a new FQN from the fifth generation (chapter 8). This new compound is highly active against drugresistant pathogens as MRSA, ciprofloxacin-resistant MRSA, and drug-resistant *Streptococcus pneumoniae* [248]. Although the mechanism of action of new FQNs(QNs) is based on the activity on the two target enzymes, DNA gyrase and topoisomerase IV, some particular aspects emerge from the structural and biological properties of the new compounds: the majority of the new representatives have a broad spectrum of activity, including activity against anaerobic bacteria (except nemonoxacin); the new representatives are active against many resistant bacteria (including resistant to FQNs); this is the main advantage of the newly approved compounds; some representatives are very active in the environment with acidic pH (delafloxacin, finafloxacin), this being an advantage over previous generations' representatives; some representatives were approved only for a specific type of administration (topic); these are very effective in the treatment of targeted infections (besifloxacin, The substituent at the N1 position increased potency, antibacterial activity and pharmacokinetic properties [241]. Cyclopropyl is known as the most potent optimization at the N1 position [113]. Thus, the substitution to N1 with cyclopropyl was preferred in four chemical structures of new FQNs (Table 10). An interesting issue is that in the past difluorophenyl in the N1 position was associated with several side effects (temafloxacin, trovafloxacin) [177]. This substitution is optimized in the chemical structure of delafloxacin with a 6-amino-3,5-difluoropyridinyl moiety. Basic groups are known to form salts in biological media. Substitution with basic groups will produce analogues with lower lipophilia and increased solubility in water. The more basic is the optimized molecule, the more likely it is to form salts and the less likely it is to be transported through a lipid membrane. The introduction of an amino group is likely to increase the binding of delafloxacin to target enzymes via hydrogen bonds [120]. The substituted pyridine residue proved to be more advantageous for the safety profile of delafloxacin comparative to older FQNs [134]. However, it should be noted that the incorporation of an aromatic amine (considered a toxophore) into the structure of a compound is avoided because aromatic amines are often highly toxic and carcinogenic [120].

finafloxacin); for these compounds, there are numerous ongoing clinical trials for oral or parenteral administration; lascufloxacin has superior tissue penetration due to its high binding capacity to phosphatidylserine. Given these aspects, we believe that there are premises to classify these new compounds in a new generation (the fifth). However, these new representatives still require supervision and further studies considering the fate of the many representatives withdrawn from previous generations due to the severe side-effects. **7. Antimicrobial Resistance to the Newer FQNs** Bacterial resistance to FQNs is a worldwide growing phenomenon; new resistant strains to FQNs have emerged in the last twenty years. The enhancement of bacterial resistance to FQNs will change patient management. This threatening phenomenon will produce changes in the therapeutic guidelines [249]. In this context, the newer FQNs aimed to reduce bacterial resistance in both humans The fluorine atom has an essential role in medicinal chemistry. Comparative to hydrogen, fluorine atom has small size, van der Waals radius of 1.47 Å versus van der Waals radius of 1.20 Å. In addition, the fluorine atom is highly electron-withdrawing (with impact on p*K*a), the C-F bond is more stable than the C-H bond, and the lipophilicity of the fluorinated molecule is higher than the non-fluorinated version. Also, substitution with a fluorine atom confers metabolic stability, influences the metabolic pathways and pharmacokinetic properties, increases the permeability of the molecule through cell membranes and the binding affinity to the target proteins [242,243]. Changes in potency produced by the introduction of a halogen-containing substituent or halogen group depend on the substitution position [120]. In the C6 position, the fluorine substituent increased the potency of FQNs versus non-fluorinated QNs. The fluorine atom increased the bacterial cell penetration and the affinity to the DNA-gyrase [113,241]. Over time, most synthesized compounds retain fluoride substitution at C6. All new compounds discussed in this review have a fluorine atom in position 6 (respectively 7 for nadifloxacin), except nemonoxacin. The other structural optimizations in the case of nemonoxacin compensated for the effect that the fluorine atom would have brought.

and animals. However, the increase in bacterial resistance to FQNs has led to researchers' efforts to understand resistance mechanisms and to identify new FQNs to combat the growing resistance. Mainly, the mechanisms of bacterial resistance to FQNs include: (1) The substituent from the C7 position increased potency, the spectrum of antibacterial activity, safety profile, and pharmacokinetic properties. This position on the QN nucleus was most often targeted for structural optimizations. Advantageous optimizations for

mutations in topoisomerase II; (2) decreased drug absorption by upregulation of efflux

and DNA topoisomerase IV. These mutations affect the interactions between FQNs and DNA enzymes [63]. Plasmid-mediated resistance encodes proteins that disrupt FQNs-enzyme interactions, increase FQNs efflux, or alter FQNs metabolism [250]. Chromosomemediated resistance affects cellular efflux pumps, decreasing cellular concentrations of

FQNs [251,252].

antibacterial activity were a five or six-membered nitrogen heterocycle, four-membered heterocycle, piperazinyl, fluorine or chlorine atoms, substituted hydrazine fragment or bicyclic substitution [241].

A C7 pyrrolidine substituent increases the activity against Gram-positive bacteria. An attempt to optimize the structure of FQNs with a C7 pyrrolidine substituent was clinafloxacin [244,245]. Although it had potential antibacterial clinafloxacin was associated with phototoxicity and hypoglycaemia [18].

Lascufloxacin contains an optimized pyrrolidine nucleus which confers great potential for treating respiratory infections (including CAP) and ear, nose and throat infections [158,246]. Regarding the new compounds, the pyrrolidine nucleus is found condensed with another heterocycle (morpholine) in finafloxacin's chemical structure. In zabofloxacin's chemical structure, the pyrrolidine nucleus is part of a spiro fragment.

A potential increase of antibacterial activity may appear with C7 and C8 cyclization. C8 substituents are essential for target affinity, because of the planar configuration of the molecule. Fluorine or chlorine, methyl or methoxy substituents proved to enhance antibacterial potency [247]. Out of these, the methoxy substituent is found in the structure of the representatives with potent anaerobic activity (e.g., moxifloxacin). Furthermore, the carbon atom at the C8 position can be replaced with nitrogen in naphthyridonic representatives with broad-spectrum activity (gemifloxacin, zabofloxacin) [244].

The third generation levofloxacin, and fourth generation moxifloxacin, are used against *Mycobacterium tuberculosis* [12,13]. The fourth generation exhibits broad-spectrum activity against Gram-positive and Gram-negative bacteria. Also, these representatives are active against anaerobes and atypical bacteria [243].

Jones et al. (2016) consider that avarofloxacin (acorafloxaxin, JNJ-Q2) is a new FQN from the fifth generation (chapter 8). This new compound is highly active against drugresistant pathogens as MRSA, ciprofloxacin-resistant MRSA, and drug-resistant *Streptococcus pneumoniae* [248].

Although the mechanism of action of new FQNs(QNs) is based on the activity on the two target enzymes, DNA gyrase and topoisomerase IV, some particular aspects emerge from the structural and biological properties of the new compounds:


Given these aspects, we believe that there are premises to classify these new compounds in a new generation (the fifth). However, these new representatives still require supervision and further studies considering the fate of the many representatives withdrawn from previous generations due to the severe side-effects.

#### **7. Antimicrobial Resistance to the Newer FQNs**

Bacterial resistance to FQNs is a worldwide growing phenomenon; new resistant strains to FQNs have emerged in the last twenty years. The enhancement of bacterial resistance to FQNs will change patient management. This threatening phenomenon will produce changes in the therapeutic guidelines [249].

In this context, the newer FQNs aimed to reduce bacterial resistance in both humans and animals. However, the increase in bacterial resistance to FQNs has led to researchers'

efforts to understand resistance mechanisms and to identify new FQNs to combat the growing resistance. Mainly, the mechanisms of bacterial resistance to FQNs include: (1) mutations in topoisomerase II; (2) decreased drug absorption by upregulation of efflux pumps; and (3) plasmid-mediated resistance [62,63].

Mutations cause the most significant form of antimicrobial resistance in DNA gyrase and DNA topoisomerase IV. These mutations affect the interactions between FQNs and DNA enzymes [63]. Plasmid-mediated resistance encodes proteins that disrupt FQNsenzyme interactions, increase FQNs efflux, or alter FQNs metabolism [250]. Chromosomemediated resistance affects cellular efflux pumps, decreasing cellular concentrations of FQNs [251,252].

It is known that older FQNs act on a single target enzyme [253]. On the other hand, it is currently considered that newer FQNs drugs, such as besifloxacin [105], delafloxacin or zabofloxacin [254] can act on both DNA topoisomerases [255,256]. Thus, antimicrobial activity increases and the spontaneous occurrence of FQNs resistance is reduced [257]. For example, some studies on *Staphylococus pneumoniae* have concluded that besifloxacin has a higher inhibitory activity against DNA gyrase and DNA topoisomerase IV than ciprofloxacin and moxifloxacin. In the case of DNA gyrase, the inhibitory concentration of besifloxacin against *Staphylococus pneumoniae* was up to eight times lower comparing with moxifloxacin and 15 times lower comparing with ciprofloxacin [126]. These results suggest that besifloxacin is less affected by target enzymes mutations than earlier FQNs [258]. The same conclusion was presented by Roychoudhury et al., following in vitro study with nemonoxacin on resistant *Streptococcus pneumoniae* [259].

It was shown that drug efflux pumps do not contribute significantly to antibiotic resistance for newer FQNs, such as besifloxacin [260]. Besifloxacin is administered only ophthalmically. This can be considered an advantage due to the less likely risk of the development of microbial resistance [101].

Other in vitro studies have also shown that MRSA is less likely to develop resistance to delafloxacin compared to older FQNs. Regarding nadifloxacin, Alba et al. [182] demonstrated no increase in resistance of *Propionibacterium acnes*, *Staphylococcus aureus* (MRSA and MSSA) and *Staphylococcus epidermidis*, showing much better antimicrobial activity compared to other antibiotics. The reduction in resistance to nadifloxacin appears probably because it is not influenced by overexpression of the NorA efflux pump on the bacterial cell membrane [88].

Predicting resistance potential is based on some essential aspects. Among them are determinants of bacterial resistance, dual activity on target enzymes, and effects on bacterial efflux systems. In addition, the newer FQNs seem to have the advantage to maintain concentrations higher than MIC of first-step resistant mutants. The detection of all gyrA mutations which confer resistance is helpful in rapid molecular diagnosis of FQN resistance [261]. Mismatch amplification mutation assay-polymerase chain reaction (MAMA-PCR) technique may serve as a tool to identify the multiple point mutations in the FQN resistance in Gram-negative bacteria [262].

Therefore, the double targeting and low resistance of bacteria are specific features of the new FQNs. Future studies are needed to complete the description of the resistance mechanism of new FQNs.

#### **8. Compounds in Development Based on Antibacterial QNs Structures**

There are numerous compounds in development that have been included in several recently published review articles [247,263]. The discovery of new potential drugs is in continuous progress. Below are briefly presented some relevant compounds under development.

#### *8.1. Avarofloxacin (Acorafloxacin)*

Avarofloxacin (acorafloxacin, JNJ-Q2) or (7-[3-[2-Amino-1(*E*)-fluoroethylidene]piperidin-1 yl]-1-cyclopropyl-6-fluoro-8-methoxy-4-oxo-1,4-dihydroquinoline-3-carboxylic acid) [143,220] is a new FQN with a zwitterionic aminoethylidenylpiperidine structure [233]. It is currently

undergoing clinical testing (phase III) to treat acute bacterial skin and skin-structure infections, CAP. It has shown improved antibacterial effectiveness against pathogens resistant to current FQNs [143].

It has antibacterial activity against a wide range of Gram-positive bacteria, including *Streptococcus pneumoniae*, MRSA, *Enterococcus* sp., *Escherichia coli*, *Klebsiella* spp., *Haemophilus influenzae* and *Pseudomonas aeruginosa* making it more potent than previously used FQNs [264].

Avarofloxacin can be administered orally and parenterally; the bioavailability is around 65% in parenteral oral administration. The fact that avarofloxacin is accessible in both parenteral and oral formulations sets it apart from several other MRSA treatments that are only available via injection [265].

In vitro investigations show that avarofloxacin has significant efficacy against pathogens including *Staphylococcus aureus* and *Streptococcus pneumoniae*, which cause acute bacterial skin and skin structure infections and community-acquired bacterial pneumonia; it was also demonstrated to have a more considerable resistance barrier than other drugs in the class, and it is still effective against drug-resistant organisms like MRSA, ciprofloxacin-resistant MRSA. Avarofloxacin was found to be as effective as linezolid for bacterial skin and skin structure infections and moxifloxacin for community-acquired bacterial pneumonia in two Phase II investigations [248]. Avarofloxacin has been granted Qualified Infectious Disease Product and Fast Track designations from the FDA [266].

#### *8.2. Other Derivatives of Antibacterial QNs*

Darehkordi et al. (2011) used N-substituted trifluoroacetimidoyl chlorides to synthesize piperazinyl QN derivatives. Out of the obtained compounds, two exhibited superior antibacterial activity against strains of *Escherichia coli*, *Klebsiella pneumonia* (compared to ciprofloxacin) and *Staphylococcus aureus* (compared to vancomycin) [267].

Sweelmeen et al. (2019) synthesized a novel derivative with antimicrobial potential (7-chloro-1-alkyl-6-fluoro-8-nitro-4-oxo-1,4-dihydroquinoline-3-carboxylic acid). This compound has been shown to be active against *Pseudomonas aeruginosa, Staphylococcus aureus, and Streptococcus agalactiae* [268]. In a review article, Zhang Bo (2014) highlighted different series of QN derivatives with antifungal potential in terms of structure-activity relationship: 2-quinolone, 4-quinolone, and FQN derivatives and FQN-metal complexes [263]. Lapointe et al. (2021) recently published the discovery and optimization of a novel series of compounds that inhibit the two bacterial target enzymes and stabilize the DNA cleavage complexes [269].

#### *8.3. Hybrids*

Numerous studies have aimed to obtain hybrid compounds that combine the properties of FQN with other types of active molecules [237,241,270,271]. In addition to broadening the spectrum of activity, the decrease in susceptibility to the installation of bacterial resistance is also pursued. Several hybrids were obtained with other antibiotics (e.g., oxazolidinones, tetracyclines, and aminoglycosides).

Gordeev et al. (2003) synthesized several compounds that incorporated pharmacophore structures of FQNs and oxazolidinones and demonstrated superior potency to linezolid against Gram-positive and Gram-negative bacteria, even for linezolid and ciprofloxacin-resistant strains of *Staphylococcus aureus* and *Enterococcus faecium*. The mechanism of action combined the inhibition of protein synthesis but also of DNA gyrase and topoisomerase IV [272]. Sriram et al. (2007) combined representatives from the tetracyclines class (tetracycline, oxytetracycline, and minocycline) with the secondary amino (piperazine) function of FQNs (norfloxacin, lomefloxacin, ciprofloxacin, and gatifloxacin). The results revealed anti-HIV and antitubercular activities, most significant for one of the compounds (minocycline-lomefloxacin derivative), making it a promising candidate in treating patients with HIV-1, co-infected with *Mycobacterium tuberculosis* [273].

Pokrovskaya et al. (2009) synthesized a series of hybrids with ciprofloxacin and neomycin. The antibacterial activity of most of the synthesized compounds was signif-

icantly higher on *Escherichia coli* and *Bacillus subtilis*, compared to that of the two free antibiotics. This case also showed that the combinations presented a dual mechanism of action, namely the inhibition of protein synthesis and target enzymes of FQNs [274]. Gorityala et al. (2016) studied an antibacterial hybrid consisting of moxifloxacin and tobramycin that acts against multidrug-resistant strains of *Pseudomonas aeruginosa,* by improving membrane permeability and reducing efflux [275]. Shavit et al. (2017) synthesized a series of hybrids composed of ciprofloxacin and kanamycin A, which showed superior action on Gram-negative bacteria. These hybrids delayed the emergence of resistance for strains of *Escherichia coli* and *Bacillus subtilis* compared to the 1:1 mixture of the two antibiotics [276].

In addition to the hybridization of antibacterial QNs with other antibiotics, several studies have included different types of drugs with biological potential in the design of hybrids. For example, Chugunova et al. (2016) synthesized a series of FQN hybrids with benzofuroxane derivatives. Some combinations showed superior antibacterial activity on *Bacillus cereus* 8035 strains compared to the free FQN [270]. Wang YN et al. (2018) synthesized a series of hybrids between QN derivatives and benzimidazole. One of the compounds showed unusual activity on the resistant strains of *Pseudomonas aeruginosa* and *Candida tropicalis* strains. It also caused a decrease in the resistance of *Pseudomonas aeruginosa*, compared to norfloxacin [277].

A series of 34 clinafloxacin-azole conjugates were synthesized and tested in vitro against *Mycobacterium tuberculosis* (H37Rv) and other Gram-negative and Gram-positive bacteria. A particular conjugate (TM2l) has been the most promising delimited in terms of a great activity against *Mycobacterium tuberculosis* (MIC = 0.29 µM), good safety predicted profile, and good drug-likeness values [124].

Yi-Lei Fan et al. (2018) review the numerous FQN derivatives as antituberculosis agents. Among them are FQN-isatin hybrids, FQN-azole hybrids, FQN-amide/azetidinone derivatives, FQN-quinoline/phenanthridine hybrids, FQN-hydrazone/hydrazide hybrids, dimeric FQN derivatives, FQN-oxime hybrids, FQN-sugar/coumarin/dihydroartemisinin/ tetracycline hybrids, and other FQN derivatives [241].

A whole decade has been reviewed from the perspective of hybrid compounds and dual-action molecules by Fedorowicz and S ˛aczewski (2018) [271].

#### **9. Concerning Side Effects**

Currently, FQNs are a valuable class of drugs used to treat infections with Grampositive and Gram-negative bacteria (Table S1—Supplementary Material). However, the new generations of FQNs have a broad spectrum of activity, including drug-resistant bacterial species (see recent authorized FQNs previously discussed). Unfortunately, this antibacterial class has been overused in therapy over time. It is known that FQNs could produce a series of severe side effects, which vary from one representative to another, mainly if they are not used judiciously [18,278–283]. These side effects occur at the gastrointestinal tract level (nausea and diarrhea), central nervous system (headache, dizziness, confusion, seizures, and insomnia), joints (Achilles tendon rupture), and muscles (neuromuscular blocking activity), cardiovascular system (QT prolongation and arrhythmias). Also, the FQNs could produce dysglycemia, hepatotoxicity, renal toxicity, phototoxicity, rush, anaphylactoid reactions, and anaphylaxis [11,18,282,284–287].

FDA has approved labeling changes of FQNs (black box warning) [288,289] and has issued a series of warnings about FQNs side effects [290], as tendinopathy and tendon rupture [291], aortic rupture or tears [292], and the negative impact on mental health and glucose homeostasis (dysglycemia) [293]. EMA has also issued similar warnings, suspensions, or restrictions of FQNs due to their potentially permanent side effects [294–298].

However, FQNs proved to be a beneficial antibacterial class, safe in the low doses and short course [281]; these drugs have potential side effects, especially in long or high doses, limiting their use. Therefore, FQNs of the new generations must be used responsibly, only in severe life-threatening infections with no alternative treatment options [281,288,299,300].

#### **10. Conclusions**

Antibacterial QNs had developed spectacularly over time, many compounds being approved and used successfully in therapy. Therefore, identifying novel antibacterial compounds has been a priority in recent years to produce effective treatments against bacteria that have gained resistance to classic FQNs. However, more information on efficacy against multidrug-resistant organisms is still needed, as these new medications are primarily aimed at these resistant strains.

Structure-activity relationship investigations were crucial in identifying substituents with a high affinity for binding to two target enzymes, the DNA gyrase and the topoisomerase IV enzymes. We have critically analyzed the structural changes in the new compounds compared to analogues from previous generations. Substitutes and combinations of substituents on the QN nucleus proved to confer to these new FQNs an acceptable safety profile by exceeding the possible side-effects identified in older compounds. These new representatives were highlighted by a broad spectrum of activity, including activity against anaerobic bacteria (except nemonoxacin). Many resistant bacteria (including resistant to FQNs) are susceptible to these compounds. Delafloxacin and finafloxacin have the advantage of being very active in an environment with acidic pH. Lascufloxacin has superior tissue penetration due to its high binding capacity to phosphatidylserine. Besifloxacin and finafloxacin were approved only for topic administration and are very effective in treating targeted infections. Thus, several positive aspects are added to the fourth generation FQNs, characteristics that can be the basis of a new generation (the fifth).

New molecules are in different phases of research, derivatives of FQNs (e.g., levonadifloxacin, avarofloxacin), and their conjugates or hybrids. This class of antimicrobials remains in the attention of researchers focused on developing new drugs efficient against resistant pathogens. However, the maximum therapeutic potential of this antimicrobials class has not been reached yet.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/pharmaceutics13081289/s1, Figure S1: Chemical structures of other FQNs from different generations, Table S1: Activity spectrum of major QNs approved for use in therapy after 2000.

**Author Contributions:** Conceptualization: A.R.; methodology, A.R., writing—original draft preparation, A.R., G.H., C.T., O.-L.M. and I.-A.L.; writing—review and editing, A.R., C.T., G.H. and I.-A.L.; visualization O.-L.M., I.-A.L. and A.R.; supervision, A.R.; All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Aura Rusu \* and Emanuela Lorena Buta**

Pharmaceutical and Therapeutical Chemistry Department, Faculty of Pharmacy, George Emil Palade University of Medicine, Pharmacy, Science and Technology of Targu Mures, 540142 Targu Mures, Romania; lorrenush@yahoo.com

**\*** Correspondence: aura.rusu@umfst.ro; Tel.: +40-766-600-898

**Abstract:** The tetracycline antibiotic class has acquired new valuable members due to the optimisation of the chemical structure. The first modern tetracycline introduced into therapy was tigecycline, followed by omadacycline, eravacycline, and sarecycline (the third generation). Structural and physicochemical key elements which led to the discovery of modern tetracyclines are approached. Thus, several chemical subgroups are distinguished, such as glycylcyclines, aminomethylcyclines, and fluorocyclines, which have excellent development potential. The antibacterial spectrum comprises several resistant bacteria, including those resistant to old tetracyclines. Sarecycline, a narrow-spectrum tetracycline, is notable for being very effective against *Cutinebacterium acnes*. The mechanism of antibacterial action from the perspective of the new compound is approached. Several severe bacterial infections are treated with tigecycline, omadacycline, and eravacycline (with parenteral or oral formulations). In addition, sarecycline is very useful in treating acne vulgaris. Tetracyclines also have other non-antibiotic properties that require in-depth studies, such as the anti-inflammatory effect effect of sarecycline. The main side effects of modern tetracyclines are described in accordance with published clinical studies. Undoubtedly, this class of antibiotics continues to arouse the interest of researchers. As a result, new derivatives are developed and studied primarily for the antibiotic effect and other biological effects.

**Keywords:** tetracyclines; structure-activity relationship; mechanism; antibacterial activity; resistance; fluorocycline; aminomethylcycline; glycylcycline

#### **1. Introduction**

Tetracyclines are an important class of broad-spectrum antibiotics that prevent bacterial growth by inhibiting protein biosynthesis. This large family includes compounds with bacteriostatic activity and a wide range of uses, from Gram-positive and Gram-negative bacterial infections to those caused by a protozoan parasite and intracellular organisms [1]. Sarecycline is unique, being the only narrow-spectrum antibiotic in the tetracycline-class family. The basic structure of tetracyclines consists of four linearly condensed benzene rings in a hydronaphtacene nucleus. The essential differences between the analogues of this class are given by the C5, C6, C7, and C9 substituents (Figure 1) [2].

#### *1.1. Brief History of Tetracycline Antibiotics*

The emergence of tetracycline development is due to the contribution of hundreds of dedicated researchers, scientists, and clinicians over more than 60 years [3]. Since their discovery (1948, aureomycin), tetracyclines have played an essential role in treating bacterial infections [4]. Stimulated by the extraordinary success of penicillins, several companies and academic institutions have focused on discovering new antibiotics produced by microorganisms, analysing numerous samples of soil sent from different parts of the world. It was observed that actinomycete bacteria produced a yellow colony, with a remarkable inhibitory effect against many pathogenic strains, including rickettsia and Gram-positive

**Citation:** Rusu, A.; Buta, E.L. The Development of Third-Generation Tetracycline Antibiotics and New Perspectives. *Pharmaceutics* **2021**, *13*, 2085. https://doi.org/10.3390/ pharmaceutics13122085

Academic Editor: Teresa Cerchiara

Received: 29 October 2021 Accepted: 3 December 2021 Published: 5 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

and Gram-negative bacteria. This actinomycete bacteria became famous for its broadspectrum antibiotic. The first tetracycline was extracted from *Streptomyces aureofaciens* and was named aureomycin (syn. chlortetracycline) [5–8]. Professor Benjamin Minge Duggar supervised the discovery of the first tetracycline. After the Food and Drug Administration (FDA) approval in 1948, aureomycin saved many lives and brought fame and profit to the Cyanamid (Lederle Laboratories Division) manufacturing company, being successfully marketed [3,7,9,10]. After Pfizer isolated *Streptomyces rimosus*, the aureomycin and terramycin (syn. oxytetracycline) were discovered [7]. This new compound was the second representative of this class of antibiotics, similar in chemical structure but with superior bioavailability and water solubility. The FDA approved Terramycin in 1950 [3,9,11]. for its broad-spectrum antibiotic. The first tetracycline was extracted from *Streptomyces aureofaciens* and was named aureomycin (syn. chlortetracycline) [5–8]. Professor Benjamin Minge Duggar supervised the discovery of the first tetracycline. After the Food and Drug Administration (FDA) approval in 1948, aureomycin saved many lives and brought fame and profit to the Cyanamid (Lederle Laboratories Division) manufacturing company, being successfully marketed [3,7,9,10]. After Pfizer isolated *Streptomyces rimosus*, the aureomycin and terramycin (syn. oxytetracycline) were discovered [7]. This new compound was the second representative of this class of antibiotics, similar in chemical structure but with superior bioavailability and water solubility. The FDA approved Terramycin in 1950 [3,9,11].

world. It was observed that actinomycete bacteria produced a yellow colony, with a remarkable inhibitory effect against many pathogenic strains, including rickettsia and Gram-positive and Gram-negative bacteria. This actinomycete bacteria became famous

*Pharmaceutics* **2021**, *13*, 2085 2 of 31

**Figure 1.** Tetracyclines—the general chemical structure and conventional numbering of the condensed rings and key positions. **Figure 1.** Tetracyclines—the general chemical structure and conventional numbering of the condensed rings and key positions.

Tetracycline was discovered in 1953, on the basis of the chemical structure of chlortetracycline, by catalitic hidrogenation (with palladium and hydrogen). The new antibiotic agent presented and improved the pharmacokinetic profile, which quickly became a favourite in therapy [12,13]. This remarkable success has proven for the first time in history that other biologically active and valuable antibiotics can be obtained by operating changes on the basic structure (molecule optimisations) [3,14]. After discovering chlortetracycline, oxytetracycline, and tetracycline (first generation of tetracyclines), the chemists of Pfizer and Lederle Laboratories began the development of new tetracyclines, with superior pharmacokinetic properties, wider antimicrobial spectrum, and lower toxicity [15]. Among the discovered representatives were methacycline (1966), doxycycline (1967), and minocycline (1972) [13,16]. Doxycycline is a semisynthetic analogue based on the chemical structure of the metacycline, approved in 1967 by the FDA. These tetracyclines are classi-Tetracycline was discovered in 1953, on the basis of the chemical structure of chlortetracycline, by catalitic hidrogenation (with palladium and hydrogen). The new antibiotic agent presented and improved the pharmacokinetic profile, which quickly became a favourite in therapy [12,13]. This remarkable success has proven for the first time in history that other biologically active and valuable antibiotics can be obtained by operating changes on the basic structure (molecule optimisations) [3,14]. After discovering chlortetracycline, oxytetracycline, and tetracycline (first generation of tetracyclines), the chemists of Pfizer and Lederle Laboratories began the development of new tetracyclines, with superior pharmacokinetic properties, wider antimicrobial spectrum, and lower toxicity [15]. Among the discovered representatives were methacycline (1966), doxycycline (1967), and minocycline (1972) [13,16]. Doxycycline is a semisynthetic analogue based on the chemical structure of the metacycline, approved in 1967 by the FDA. These tetracyclines are classified in the second generation (Table S1—Supplementary Materials) [3,14,17].

fied in the second generation (Table S1—Supplementary Materials) [3,14,17]. The further development of semisynthetic analogues of the second generation, and, more recently, of the third generation (Table 1), reveals the evolution of this class. The modern tetracyclines had acquired high potency and an increased efficacy, even against bacteria resistant to tetracyclines [18–20]. Therefore, biochemical mutants of *Streptomycetes*  The further development of semisynthetic analogues of the second generation, and, more recently, of the third generation (Table 1), reveals the evolution of this class. The modern tetracyclines had acquired high potency and an increased efficacy, even against bacteria resistant to tetracyclines [18–20]. Therefore, biochemical mutants of *Streptomycetes* strains have been created for a higher production yield and to discover novel tetracyclines [3,15].

strains have been created for a higher production yield and to discover novel tetracyclines

[3,15]. **Table 1.** Tetracyclines—classification into generations [14,21–24].


Therefore, compounds such as demeclocycline were discovered, being later converted in 1971 to a C7-amino-derivative known as minocycline. Some authors classified minocy-

cline as the last tetracycline in the second generation. Moreover, demeclocycline was a precursor of sancycline (obtained by reduction), a tetracycline with a simplified chemical structure and retained biological activity [1,25]. Thus, the main advantage of minocycline is a broader spectrum of activity compared to previously representatives. Furthermore, minocycline presented a better pharmacokinetic profile and was the most potent representative at the time, being the last introduced on the market in the next three decades [3,17,26]. was a precursor of sancycline (obtained by reduction), a tetracycline with a simplified chemical structure and retained biological activity [1,25]. Thus, the main advantage of minocycline is a broader spectrum of activity compared to previously representatives. Furthermore, minocycline presented a better pharmacokinetic profile and was the most potent representative at the time, being the last introduced on the market in the next three decades [3,17,26].

#### *1.2. The Discovery of Modern Tetracyclines 1.2. The Discovery of Modern Tetracyclines*

*Pharmaceutics* **2021**, *13*, 2085 3 of 31

**Table 1.** Tetracyclines—classification into generations [14,21–24].

**Generations Obtaining Method Representatives** 

First Biosynthesis Chlortetracycline, oxytetracycline, tetracy-

Second Semisynthesis Doxycycline, minocycline, lymecycline,

Third Semisynthesis Tigecycline, omadacycline, sarecycline

Therefore, compounds such as demeclocycline were discovered, being later converted in 1971 to a C7-amino-derivative known as minocycline. Some authors classified minocycline as the last tetracycline in the second generation. Moreover, demeclocycline

Total synthesis Eravacycline

cline, demeclocycline

meclocycline, methacycline, rolitetracycline

The growing occurrence of bacterial resistance to antibiotics has once again aroused the interest of scientists in the development of new tetracyclines. Thus, at the end of the 1980s, the programs were reopened for the synthesis of new compounds that could be classified into a new generation (the third one), re-evaluating the compounds already synthesised (Table S2—Supplementary Materials) [3]. The main interest was about the modification of C7 and C9 positions of the D ring in the sancycline structure (Figure 2, Table S3—Supplementary Materials). These steps have led to the discovery of a novel class of C9-aminotetracyclines, which bear a glycyl moiety known as *glycylcycline* [3,27–29]. The growing occurrence of bacterial resistance to antibiotics has once again aroused the interest of scientists in the development of new tetracyclines. Thus, at the end of the 1980s, the programs were reopened for the synthesis of new compounds that could be classified into a new generation (the third one), re-evaluating the compounds already synthesised (Table S2—Supplementary Materials) [3]. The main interest was about the modification of C7 and C9 positions of the D ring in the sancycline structure (Figure 2, Table S3—Supplementary Materials). These steps have led to the discovery of a novel class of C9-aminotetracyclines, which bear a glycyl moiety known as *glycylcycline* [3,27–29]*.* 

**Figure 2.** The chemical structures of sancycline; key positions highlighted C7 and C9 for structural design optimisation to obtain new derivatives. **Figure 2.** The chemical structures of sancycline; key positions highlighted C7 and C9 for structural design optimisation to obtain new derivatives.

Modern tetracyclines include derivatives with more or less similar chemical structures: a glycylglicine (tigecycline), an aminomethylcycline (omadacycline), a fluorocycline (eravacycline), and a 7-[(methoxy-(methyl)-amino)-methyl]methyl] derivative (sarecycline) (Figure 3). Modern tetracyclines include derivatives with more or less similar chemical structures: a glycylglicine (tigecycline), an aminomethylcycline (omadacycline), a fluorocycline (eravacycline), and a 7-[(methoxy-(methyl)-amino)-methyl]methyl] derivative (sarecycline) (Figure 3).

Tigecycline is a synthetic derivative of minocycline discovered in 1993. Tigecycline was the first tetracycline introduced in therapy after more than 30 years. Thus, tigecycline could be considered the prototype of a new subclass of tetracyclines [27,29]. This new tetracycline has the advantage of a superior potency over Gram-positive and Gram-negative multidrug-resistant bacteria (multiple drug resistance, MDR) [26,30]. Tigecycline was discovered by Wyeth Pharmaceuticals Inc. and approved by the FDA in 2005 [31] and later by the European Medicine Agency (EMA) in 2006, under the trade name Tygacil [32]. Tigecycline is a synthetic derivative of minocycline discovered in 1993. Tigecycline was the first tetracycline introduced in therapy after more than 30 years. Thus, tigecycline could be considered the prototype of a new subclass of tetracyclines [27,29]. This new tetracycline has the advantage of a superior potency over Gram-positive and Gram-negative multidrug-resistant bacteria (multiple drug resistance, MDR) [26,30]. Tigecycline was discovered by Wyeth Pharmaceuticals Inc. and approved by the FDA in 2005 [31] and later by the European Medicine Agency (EMA) in 2006, under the trade name Tygacil [32]. Tygacil received approval for complicated intra-abdominal and complicated skin and soft tissue infections [33,34]. Likewise, in 2008, the FDA approved the use of tigecycline to treat community-acquired pneumonia [35,36]. Once placed in the market, several other uses have been investigated: nosocomial pneumonia, diabetic foot infections, emergency therapy for MDR pathogens, nosocomial urinary tract infections, and *Clostridium difficile* infections [26]. A disadvantage of tigecycline is its exclusive parenteral use due to its poor bioavailability [37].

Omadacycline is one of the newest and most popular tetracyclines and the first in the aminomethylcycline subclass [3,38]. It has a broad spectrum of activity, proving in vitro effects against Gram-positive and Gram-negative bacteria, anaerobic bacteria, and atypical bacteria. In addition, the activity of this compound extends to methicillinresistant *Staphylococcus aureus* (MRSA); penicillin-resistant, MDR *Streptococcus pneumoniae*; and vancomycin-resistant enterococci [22]. Unlike tigecycline, omadacycline is available for both oral and parenteral administration. Both forms were approved in 2018 by the

FDA for the treatment of complicated intra-abdominal and complicated skin and soft tissue infections and community-acquired pneumonia. Currently, omadacycline is in phase II of clinical trials to treat urinary tract infections, such as acute pyelonephritis and cystitis [38]. The pharmaceutical product Nuzyra was approved in the United States of America (USA) [39]. community-acquired pneumonia [35,36]. Once placed in the market, several other uses have been investigated: nosocomial pneumonia, diabetic foot infections, emergency therapy for MDR pathogens, nosocomial urinary tract infections, and *Clostridium difficile* infections [26]. A disadvantage of tigecycline is its exclusive parenteral use due to its poor bioavailability [37].

Tygacil received approval for complicated intra-abdominal and complicated skin and soft tissue infections [33,34]. Likewise, in 2008, the FDA approved the use of tigecycline to treat

*Pharmaceutics* **2021**, *13*, 2085 4 of 31

Omadacycline is one of the newest and most popular tetracyclines and the first in the aminomethylcycline subclass [3,38]. It has a broad spectrum of activity, proving in vitro effects against Gram-positive and Gram-negative bacteria, anaerobic bacteria, and atypical bacteria. In addition, the activity of this compound extends to methicillin-resistant *Staphylococcus aureus* (MRSA); penicillin-resistant, MDR *Streptococcus pneumoniae*; and vancomycin-resistant enterococci [22]. Unlike tigecycline, omadacycline is available for Eravacycline is a synthetic fluorocycline, obtained by total synthesis, that contains a basic chemical structure of the tetracyclines class [40,41]. In addition, particular modifications on the D ring of the naphtacen nucleus were introduced. Those chemical optimisations give it a remarkable activity against Gram-positive and Gram-negative bacteria that developed specific resistance mechanisms to the tetracycline antibiotic class to treat complicated intra-abdominal infections in adults. It is available for parenteral administration in many countries in Europe, as well as in the USA [41].

both oral and parenteral administration. Both forms were approved in 2018 by the FDA for the treatment of complicated intra-abdominal and complicated skin and soft tissue infections and community-acquired pneumonia. Currently, omadacycline is in phase II of clinical trials to treat urinary tract infections, such as acute pyelonephritis and cystitis [38]. The pharmaceutical product Nuzyra was approved in the United States of America (USA) [39]. Eravacycline is a synthetic fluorocycline, obtained by total synthesis, that contains a basic chemical structure of the tetracyclines class [40,41]. In addition, particular modifications on the D ring of the naphtacen nucleus were introduced. Those chemical optimisations give it a remarkable activity against Gram-positive and Gram-negative bacteria that Sarecycline is an analogue of tetracycline specifically designed for the treatment of acne [42]. It is available as an oral formulation to treat inflammatory lesions of moderate to severe non-nodular acne vulgaris. The main advantage of this new tetracycline is a higher selective activity against *Cutinebacterium acnes* comparative to older tetracyclines (doxycycline and minocycline) used in acne therapy. Due to this selectivity, the probability of developing antibiotic resistance is lower than minocycline and doxycycline [43]. Sarecycline was developed by Paratek Pharmaceuticals and Allergan but acquired by Almirall S.A by purchasing the dermatological portfolio [44]. The FDA approved sarecycline in 2018 under the trade name Seysara [45,46].

#### developed specific resistance mechanisms to the tetracycline antibiotic class to treat com-**2. Research Methodology**

plicated intra-abdominal infections in adults. It is available for parenteral administration in many countries in Europe, as well as in the USA [41]. Sarecycline is an analogue of tetracycline specifically designed for the treatment of acne [42]. It is available as an oral formulation to treat inflammatory lesions of moderate to severe non-nodular acne vulgaris. The main advantage of this new tetracycline is a higher selective activity against *Cutinebacterium acnes* comparative to older tetracyclines (doxycycline and minocycline) used in acne therapy*.* Due to this selectivity, the probability of developing antibiotic resistance is lower than minocycline and doxycycline [43]. The relevant primary data were found on Clarivate Analytics and ScienceDirect databases using the following keywords: (i) topic: "tetracyclines", "antibacterials"; (ii) title: "tigecycline", "omadacycline", "eravacycline", "sarecycline", and other classic derivatives of the tetracycline class. In the second stage, the articles were selected if they comprised development of tetracyclines class, physicochemical properties, aspects related to structure–activity relationships, new tetracyclines design, mechanism of action, antibacterial spectrum, therapeutical value, safety profile, bacterial resistance, and new derivatives in development. The paper includes significant references, including the latest articles published in 2021.

All chemical structures were drowned with Biovia Draw 2019 (San Diego, CA, USA) [47].

#### **3. Overview of Modern Tetracyclines**

This paper's primary targeted the new tetracyclines classified as the third generation: tigecycline, omadacycline, eravacycline, and sarecycline.

#### *3.1. Considerations Regarding the Chemical Structure and Physicochemical Properties of the New Tetracyclines*

Recently introduced compounds in the tetracycline class contain the basic chemical structure specific to this class, four condensed rings (A, B, C, and D) into a naphtacencarboxamide system. Other common structural elements are a dimethyl-amino group at the C4 position, an amidic group at the C2 position, a keto–enol alternation system (C11, C12, and 12a positions), and asymmetric carbons at the junction of rings A-B (stereochemical configurations) (Figures 1 and 3) [21,48,49]. X-ray crystallography of tetracycline, doxycycline, and sancycline revealed that the C2 amide group is oriented to form an intramolecular hydrogen bond with oxygen atom from C3 position [50]. The above elements are considered the minimum pharmacophore (6-deoxy-dimethyltetracycline) required for antimicrobial activity and a start point for inserting other substituents [21,48,49].

Depending on the radicals grafted on the tetracyclic system, these new molecules introduced on the market after 2000 present different physicochemical and pharmacological characteristics and changes in the antimicrobial spectrum [13,21,38,41,44]. The optimisation of the basic structure consisted of C7 and C9 substitutions. Position C7 is subject to substitution with electron acceptor or donor groups [40]. Thus, tigecycline and omadacycline have a dimethyl-amino group in this position. Eravacycline has a fluorine atom at the C7 position, being an electron-withdrawing substituent [40,51]. In the same place, sarecycline has a more voluminous radical, methoxy-methyl-amino-methyl [42]. The radicals in the C9 position are distinct for each of the new representatives. Tigecycline and eravacycline are synthetic analogues that contain a glycyl-amide substituent [40,52].

Tigecycline was synthesised by adding a *tert*-butyl-glycyl-amide substituent, while in eravacycline this group was replaced with a pyrrolidinyl-acetamide group [32,40]. However, tigecycline is an analogue of minocycline formed by adding the *tert*-butyl-glycylamide substituent at the C9 position (Figure 4). It is the first glycylcycline tetracycline discovered [29,31]. Currently, tigecycline is manufactured as a lyophilised powder form because it undergoes a degradation process. Tigecycline is commercialised under the trade name Tygacil (pharmaceutical form for intravenous infusion). The recommended doses regimen is 100 mg initial dose, followed by 50 mg every 12 h [31]. *Pharmaceutics* **2021**, *13*, 2085 6 of 31 [38,53,54]. The optimisation at the C9 atom was based on a methyl(2,2-dimethylpropylamino) fragment that replaces the glycylamide group present in the case of its homologues (tigecycline and eravacycline) (Figure 5) [52]. Omadacycline is formulated as tosylate salt for intravenous or oral administration under the trade name Nuzyra [38].

**Figure 4.** The chemical structures of the tigecycline and conventional numbering. **Figure 4.** The chemical structures of the tigecycline and conventional numbering.

**Figure 5.** The chemical structures of the omadacycline and conventional numbering.

**Figure 6.** The chemical structures of the eravacycline and conventional numbering.

[46].

Sarecycline has no substitute in the C9 position [56]. Sarecycline is chemically distinguishable from other tetracycline-class antibiotics by the 7 [[methoxy(methyl)amino]methyl] group attached at the C7 position of the ring D. This stable modification represents the longest and the largest C7 moiety among all of the tetracyclines (Figure 7). Sarecycline inhibits bacterial ribosomes through interactions with the mRNA as a consequence of C7 optimisation. This new tetracycline blocks accommodation into the A site of the first aminoacyl transfer RNA and appears to be a more potent initiation inhibitor comparative to previous analogues [57]. Sarecycline is manufactured as hydrochloride salt [42]. Pharmaceutical formula Seysara tablets for oral use contains sarecycline (60 mg, 100 mg, 150 mg)

Omadacycline (sin. amadacycline) is the first aminomethylcycline of this new subclass, for which the glycyl-amide group was changed to an alkyl-amino-methyl group [38,53,54]. The optimisation at the C9 atom was based on a methyl(2,2-dimethylpropylamino) fragment that replaces the glycylamide group present in the case of its homologues (tigecycline and eravacycline) (Figure 5) [52]. Omadacycline is formulated as tosylate salt for intravenous or oral administration under the trade name Nuzyra [38].

tion of a fluorine atom in the C7 position and the substitution of the *tert*-butyl-aminoacetamide group in the C9 position with a pyrrolidin-acetamido group (Figure 6) [41]. Pharmaceutical formulation under the trade name Xerava contains eravacycline, powder for concentrate for solution for infusion (50 mg and 100 mg) for intravenous use [55].

**Figure 5.** The chemical structures of the omadacycline and conventional numbering. **Figure 5.** The chemical structures of the omadacycline and conventional numbering. **Figure 5.** The chemical structures of the omadacycline and conventional numbering.

The primary structure of tetracyclines is maintained in the chemical structure of eravacycline, which is an analogue of tigecycline, with two changes on the D ring: the addition of a fluorine atom in the C7 position and the substitution of the *tert*-butyl-aminoacetamide group in the C9 position with a pyrrolidin-acetamido group (Figure 6) [41]. Pharmaceutical formulation under the trade name Xerava contains eravacycline, powder for concentrate for solution for infusion (50 mg and 100 mg) for intravenous use [55]. The primary structure of tetracyclines is maintained in the chemical structure of eravacycline, which is an analogue of tigecycline, with two changes on the D ring: the addition of a fluorine atom in the C7 position and the substitution of the *tert*-butyl-aminoacetamide group in the C9 position with a pyrrolidin-acetamido group (Figure 6) [41]. Pharmaceutical formulation under the trade name Xerava contains eravacycline, powder for concentrate for solution for infusion (50 mg and 100 mg) for intravenous use [55]. The primary structure of tetracyclines is maintained in the chemical structure of eravacycline, which is an analogue of tigecycline, with two changes on the D ring: the addition of a fluorine atom in the C7 position and the substitution of the *tert*-butyl-aminoacetamide group in the C9 position with a pyrrolidin-acetamido group (Figure 6) [41]. Pharmaceutical formulation under the trade name Xerava contains eravacycline, powder for concentrate for solution for infusion (50 mg and 100 mg) for intravenous use [55].

[38,53,54]. The optimisation at the C9 atom was based on a methyl(2,2-dimethylpropylamino) fragment that replaces the glycylamide group present in the case of its homologues (tigecycline and eravacycline) (Figure 5) [52]. Omadacycline is formulated as tosylate salt

[38,53,54]. The optimisation at the C9 atom was based on a methyl(2,2-dimethylpropylamino) fragment that replaces the glycylamide group present in the case of its homologues (tigecycline and eravacycline) (Figure 5) [52]. Omadacycline is formulated as tosylate salt

for intravenous or oral administration under the trade name Nuzyra [38].

for intravenous or oral administration under the trade name Nuzyra [38].

*Pharmaceutics* **2021**, *13*, 2085 6 of 31

**Figure 4.** The chemical structures of the tigecycline and conventional numbering.

**Figure 4.** The chemical structures of the tigecycline and conventional numbering.

**Figure 6.** The chemical structures of the eravacycline and conventional numbering. **Figure 6.** The chemical structures of the eravacycline and conventional numbering.

**Figure 6.** The chemical structures of the eravacycline and conventional numbering. Sarecycline has no substitute in the C9 position [56]. Sarecycline is chemically distinguishable from other tetracycline-class antibiotics by the 7 [[methoxy(methyl)amino]methyl] group attached at the C7 position of the ring D. This stable modification represents the longest and the largest C7 moiety among all of the tetracyclines (Figure 7). Sarecycline inhibits bacterial ribosomes through interactions with the mRNA as a consequence of C7 optimisation. This new tetracycline blocks accommodation into the A site of the first aminoacyl transfer RNA and appears to be a more potent initiation inhibitor comparative to previous analogues [57]. Sarecycline is manufactured as hydrochloride salt [42]. Pharmaceutical formula Seysara tablets for oral use contains sarecycline (60 mg, 100 mg, 150 mg) Sarecycline has no substitute in the C9 position [56]. Sarecycline is chemically distinguishable from other tetracycline-class antibiotics by the 7 [[methoxy(methyl)amino]methyl] group attached at the C7 position of the ring D. This stable modification represents the longest and the largest C7 moiety among all of the tetracyclines (Figure 7). Sarecycline inhibits bacterial ribosomes through interactions with the mRNA as a consequence of C7 optimisation. This new tetracycline blocks accommodation into the A site of the first aminoacyl transfer RNA and appears to be a more potent initiation inhibitor comparative to previous analogues [57]. Sarecycline is manufactured as hydrochloride salt [42]. Pharmaceutical formula Seysara tablets for oral use contains sarecycline (60 mg, 100 mg, 150 mg) [46]. Sarecycline has no substitute in the C9 position [56]. Sarecycline is chemically distinguishable from other tetracycline-class antibiotics by the 7 [[methoxy(methyl)amino]methyl] group attached at the C7 position of the ring D. This stable modification represents the longest and the largest C7 moiety among all of the tetracyclines (Figure 7). Sarecycline inhibits bacterial ribosomes through interactions with the mRNA as a consequence of C7 optimisation. This new tetracycline blocks accommodation into the A site of the first aminoacyl transfer RNA and appears to be a more potent initiation inhibitor comparative to previous analogues [57]. Sarecycline is manufactured as hydrochloride salt [42]. Pharmaceutical formula Seysara tablets for oral use contains sarecycline (60 mg, 100 mg, 150 mg) [46]. *Pharmaceutics* **2021**, *13*, 2085 7 of 31

**Figure 7.** The chemical structures of the sarecycline and conventional numbering. **Figure 7.** The chemical structures of the sarecycline and conventional numbering.

Essential physicochemical properties of the third generation tetracyclines are shown Essential physicochemical properties of the third generation tetracyclines are shown in Table S4 (Supplementary Materials).

in Table S4 (Supplementary Materials). Tetracyclines are optically active substances. The X-ray diffraction analysis (XRD) established the stereochemistry of the basic structure of these compounds. Depending on Tetracyclines are optically active substances. The X-ray diffraction analysis (XRD) established the stereochemistry of the basic structure of these compounds. Depending on

substitution, several chiral atoms are C4, C4*a*, C5, C5*a*, C6, and C12*a* (Figure 8a). Some

In acidic conditions, tetracyclines epimerase reversibly at the C4 position (Figure 8b). The resulted isomers are known as "epitetracyclines", founded in equal amounts after establishing the equilibrium. The formation of 4-epitetracyclines is notable because they

Tetracyclines are amphoteric compounds due to the characteristic structural elements (hydroxyls and dimethylamino substituents and the conjugated keto-enolic system). In reaction with an acid or a base, tetracyclines form salts. In pharmaceutical formulations, tetracyclines are most commonly used in the form of hydrochloric salts (e.g., eravacycline, sarecycline). Depending on the solvent, the tetracyclines' structure changes from an ionised to a non-ionised state (protonation–deprotonation equilibria). At the neutral pH, tetracyclines mainly adopt the zwitterion form. It is known that acid salts of tetracyclines exhibit a minimum of three acidity constants in aqueous solutions [14,50,59,60]. The main protonation sites of tetracyclines are the tricarbonyl system (C1-C2-C3), phenolic diketone-system (C10-C11-C12), and dimethylamino group (C4) (Figure 9). [58,61–63]. Depending on the substituents on the basic chemical structure, the protonation state of the compound also changes [64]. Tetracyclines are multiprotic compounds. Put simply, tetracyclines can be considered to behave similar to triprotic acids [63]. Other authors have suggested that tetracyclines have four ionisation equilibria and four correspondent pKa values (at pH values of 3.2, 7.6, 9.6, and 12) and five protonation states [63].

**Figure 8.** The chiral atoms on the chemical structure of tetracyclines and conventional numbering (**a**); epimerisation of

are less active than non-epimerised isomers [58,59].

nucleus (C10 to C12; C1 to C3) [58].

tetracyclines (**b**); \*—chiral centers.

substitution, several chiral atoms are C4, C4*a*, C5, C5*a*, C6, and C12*a* (Figure 8a). Some derivatives, such as oxytetracycline and doxycycline, have six chiral carbon atoms, due to the C5α-hydroxyl substituent. Moreover, a conjugated system is known in the naphtacene nucleus (C10 to C12; C1 to C3) [58]. substitution, several chiral atoms are C4, C4*a*, C5, C5*a*, C6, and C12*a* (Figure 8a). Some derivatives, such as oxytetracycline and doxycycline, have six chiral carbon atoms, due to the C5α-hydroxyl substituent. Moreover, a conjugated system is known in the naphtacene nucleus (C10 to C12; C1 to C3) [58].

Essential physicochemical properties of the third generation tetracyclines are shown

Tetracyclines are optically active substances. The X-ray diffraction analysis (XRD) established the stereochemistry of the basic structure of these compounds. Depending on

**Figure 7.** The chemical structures of the sarecycline and conventional numbering.

in Table S4 (Supplementary Materials).

*Pharmaceutics* **2021**, *13*, 2085 7 of 31

**Figure 8.** The chiral atoms on the chemical structure of tetracyclines and conventional numbering (**a**); epimerisation of tetracyclines (**b**); \*—chiral centers. **Figure 8.** The chiral atoms on the chemical structure of tetracyclines and conventional numbering (**a**); epimerisation of tetracyclines (**b**); \*—chiral centers.

In acidic conditions, tetracyclines epimerase reversibly at the C4 position (Figure 8b). The resulted isomers are known as "epitetracyclines", founded in equal amounts after establishing the equilibrium. The formation of 4-epitetracyclines is notable because they are less active than non-epimerised isomers [58,59]. In acidic conditions, tetracyclines epimerase reversibly at the C4 position (Figure 8b). The resulted isomers are known as "epitetracyclines", founded in equal amounts after establishing the equilibrium. The formation of 4-epitetracyclines is notable because they are less active than non-epimerised isomers [58,59].

Tetracyclines are amphoteric compounds due to the characteristic structural elements (hydroxyls and dimethylamino substituents and the conjugated keto-enolic system). In reaction with an acid or a base, tetracyclines form salts. In pharmaceutical formulations, tetracyclines are most commonly used in the form of hydrochloric salts (e.g., eravacycline, sarecycline). Depending on the solvent, the tetracyclines' structure changes from an ionised to a non-ionised state (protonation–deprotonation equilibria). At the neutral pH, tetracyclines mainly adopt the zwitterion form. It is known that acid salts of tet-Tetracyclines are amphoteric compounds due to the characteristic structural elements (hydroxyls and dimethylamino substituents and the conjugated keto-enolic system). In reaction with an acid or a base, tetracyclines form salts. In pharmaceutical formulations, tetracyclines are most commonly used in the form of hydrochloric salts (e.g., eravacycline, sarecycline). Depending on the solvent, the tetracyclines' structure changes from an ionised to a non-ionised state (protonation–deprotonation equilibria). At the neutral pH, tetracyclines mainly adopt the zwitterion form. It is known that acid salts of tetracyclines exhibit a minimum of three acidity constants in aqueous solutions [14,50,59,60].

racyclines exhibit a minimum of three acidity constants in aqueous solutions [14,50,59,60]. The main protonation sites of tetracyclines are the tricarbonyl system (C1-C2-C3), phenolic diketone-system (C10-C11-C12), and dimethylamino group (C4) (Figure 9). [58,61–63]. Depending on the substituents on the basic chemical structure, the protonation state of the compound also changes [64]. Tetracyclines are multiprotic compounds. Put simply, tetracyclines can be considered to behave similar to triprotic acids [63]. Other authors have suggested that tetracyclines have four ionisation equilibria and four correspondent pKa values (at pH values of 3.2, 7.6, 9.6, and 12) and five protonation states [63]. The main protonation sites of tetracyclines are the tricarbonyl system (C1-C2-C3), phenolic diketone-system (C10-C11-C12), and dimethylamino group (C4) (Figure 9) [58,61–63]. Depending on the substituents on the basic chemical structure, the protonation state of the compound also changes [64]. Tetracyclines are multiprotic compounds. Put simply, tetracyclines can be considered to behave similar to triprotic acids [63]. Other authors have suggested that tetracyclines have four ionisation equilibria and four correspondent pKa values (at pH values of 3.2, 7.6, 9.6, and 12) and five protonation states [63]. *Pharmaceutics* **2021**, *13*, 2085 8 of 31

[61–65].

tion percentage.

pH 6.60 (32.92%; highest) pH 7.40 (17.87%)

> **O H O**

**N H3C C H3**

pH 4.00 (98.86%; highest) pH 7.40 (2.11%)

**O N C H3 C H3**

**O**

**<sup>O</sup>– O H**

**1 2 4** 

**3 5 6** 

**NH2**

**O**

**O**

**Figure 9.** The tetracycline (TC) structural sites and the correspondent acidic dissociation constants **Figure 9.** The tetracycline (TC) structural sites and the correspondent acidic dissociation constants [61–65].

and C9 [65]. Using MarvinSketck (ChemAxon, Budapest, Hungary) for sarecycline, researchers found 17 possible microspecies depending on the pH value. Table 2 comprises the predicted microspecies and the highest value of ionisation (%) at a specific pH; additionally, the degree of ionisation at the physiologic pH (7.4) was highlighted. The highest ionisation percentages of the microspecies were identified as follows: no. 3 at pH 0–1 (98.15%), no. 6 at pH 4 (98.66%), no. 14 at pH 11 (96.30%), no. 17 at pH 14 (95.55%), and no. 7 at pH 8.6 (49.38%). At the physiologic pH (7.4), no. 2 was 21.59%, the highest ionisa-

> pH 7.60 (12.59%; highest) pH 7.40 (12.19%)

> pH 0.00 (98.16%; highest) pH 7.40 (0.00%)

**Table 2.** Sarecycline microspecies and the degree of ionisation as a function of pH (calculated) [66].

pH 7.60 (22.31%; highest) pH 7.40 (21.59%)

> **O H O**

**N H3C C H3** **NH2**

**O**

**O**

pH 7.60 (3.84%; highest) pH 7.40 (3.72%)

**No. Microspecies No. Microspecies No. Microspecies** 

**O N C H3 C H3**

**O**

**O O H –**

Thus, tigecycline poses five main ionisation groups, specifically, five values of p*K*<sup>a</sup> (at pH values of 2.8, 4.4, 7.4, 8.9, and 9.5), an important role having the substitutes from C7 and C9 [65]. Using MarvinSketck (ChemAxon, Budapest, Hungary) for sarecycline, researchers found 17 possible microspecies depending on the pH value. Table 2 comprises the predicted microspecies and the highest value of ionisation (%) at a specific pH; additionally, the degree of ionisation at the physiologic pH (7.4) was highlighted. The highest ionisation percentages of the microspecies were identified as follows: no. 3 at pH 0–1 (98.15%), no. 6 at pH 4 (98.66%), no. 14 at pH 11 (96.30%), no. 17 at pH 14 (95.55%), and no. 7 at pH 8.6 (49.38%). At the physiologic pH (7.4), no. 2 was 21.59%, the highest ionisation percentage. Thus, tigecycline poses five main ionisation groups, specifically, five values of p*K*a (at pH values of 2.8, 4.4, 7.4, 8.9, and 9.5), an important role having the substitutes from C7 and C9 [65]. Using MarvinSketck (ChemAxon, Budapest, Hungary) for sarecycline, researchers found 17 possible microspecies depending on the pH value. Table 2 comprises the predicted microspecies and the highest value of ionisation (%) at a specific pH; additionally, the degree of ionisation at the physiologic pH (7.4) was highlighted. The highest ionisation percentages of the microspecies were identified as follows: no. 3 at pH 0–1 (98.15%), no. 6 at pH 4 (98.66%), no. 14 at pH 11 (96.30%), no. 17 at pH 14 (95.55%), and no. 7 at pH 8.6 (49.38%). At the physiologic pH (7.4), no. 2 was 21.59%, the highest ionisation percentage. Thus, tigecycline poses five main ionisation groups, specifically, five values of p*K*a (at pH values of 2.8, 4.4, 7.4, 8.9, and 9.5), an important role having the substitutes from C7 and C9 [65]. Using MarvinSketck (ChemAxon, Budapest, Hungary) for sarecycline, researchers found 17 possible microspecies depending on the pH value. Table 2 comprises the predicted microspecies and the highest value of ionisation (%) at a specific pH; additionally, the degree of ionisation at the physiologic pH (7.4) was highlighted. The highest ionisation percentages of the microspecies were identified as follows: no. 3 at pH 0–1 (98.15%), no. 6 at pH 4 (98.66%), no. 14 at pH 11 (96.30%), no. 17 at pH 14 (95.55%), and no. 7 at pH 8.6 (49.38%). At the physiologic pH (7.4), no. 2 was 21.59%, the highest ionisation percentage. Thus, tigecycline poses five main ionisation groups, specifically, five values of p*K*a (at pH values of 2.8, 4.4, 7.4, 8.9, and 9.5), an important role having the substitutes from C7 and C9 [65]. Using MarvinSketck (ChemAxon, Budapest, Hungary) for sarecycline, researchers found 17 possible microspecies depending on the pH value. Table 2 comprises the predicted microspecies and the highest value of ionisation (%) at a specific pH; additionally, the degree of ionisation at the physiologic pH (7.4) was highlighted. The highest ionisation percentages of the microspecies were identified as follows: no. 3 at pH 0–1 (98.15%), no. 6 at pH 4 (98.66%), no. 14 at pH 11 (96.30%), no. 17 at pH 14 (95.55%), and no. 7 at pH 8.6 (49.38%). At the physiologic pH (7.4), no. 2 was 21.59%, the highest ionisation percentage.

**Figure 9.** The tetracycline (TC) structural sites and the correspondent acidic dissociation constants

**Figure 9.** The tetracycline (TC) structural sites and the correspondent acidic dissociation constants

**Figure 9.** The tetracycline (TC) structural sites and the correspondent acidic dissociation constants

**Table 2.** Sarecycline microspecies and the degree of ionisation as a function of pH (calculated) [66]. **Table 2.** Sarecycline microspecies and the degree of ionisation as a function of pH (calculated) [66]. **Table 2.** Sarecycline microspecies and the degree of ionisation as a function of pH (calculated) [66]. **Table 2.** Sarecycline microspecies and the degree of ionisation as a function of pH (calculated) [66].

*Pharmaceutics* **2021**, *13*, 2085 8 of 32

*Pharmaceutics* **2021**, *13*, 2085 8 of 32

*Pharmaceutics* **2021**, *13*, 2085 8 of 32

[63].

[63].

[63].

[61–65].

pH 7.40 (4.50%)

pH 7.40 (4.50%)

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

pH 7.40 (4.50%)

pH 7.40 (4.50%)

pH 7.40 (4.50%)

**O H O**

pH 7.40 (4.50%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– O H**

**O**

**<sup>O</sup>– O H**

**O**

**<sup>O</sup>– O H**

**O**

**<sup>O</sup>– O H**

**O**

**<sup>O</sup>– O H**

**O**

**<sup>O</sup>– O H**

**O**

**<sup>O</sup>– O H**

**O**

**O**

**<sup>O</sup>– O H**

**<sup>O</sup>– O H**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**16 17** 

**16 17** 

**16 17** 

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**16 17** 

**16 17** 

**16 17** 

**16 17** 

**16 17** 

**16 17** 

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 7.60 (7.85%; highest) pH 7.40 (7.60%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

pH 8.60 (0.69%; highest) pH 7.40 (0.17%)

[61–65].

[61–65].

[58,61–63]. Depending on the substituents on the basic chemical structure, the protonation state of the compound also changes [64]. Tetracyclines are multiprotic compounds. Put simply, tetracyclines can be considered to behave similar to triprotic acids [63]. Other authors have suggested that tetracyclines have four ionisation equilibria and four correspondent pKa values (at pH values of 3.2, 7.6, 9.6, and 12) and five protonation states

[58,61–63]. Depending on the substituents on the basic chemical structure, the protonation state of the compound also changes [64]. Tetracyclines are multiprotic compounds. Put simply, tetracyclines can be considered to behave similar to triprotic acids [63]. Other authors have suggested that tetracyclines have four ionisation equilibria and four correspondent pKa values (at pH values of 3.2, 7.6, 9.6, and 12) and five protonation states

[58,61–63]. Depending on the substituents on the basic chemical structure, the protonation state of the compound also changes [64]. Tetracyclines are multiprotic compounds. Put simply, tetracyclines can be considered to behave similar to triprotic acids [63]. Other authors have suggested that tetracyclines have four ionisation equilibria and four correspondent pKa values (at pH values of 3.2, 7.6, 9.6, and 12) and five protonation states

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**<sup>O</sup>– <sup>O</sup>–**

**O**

**O**

**O– O–**

**O**

**O– O–**

**O**

**O– O–**

**O**

**O– O–**

**O**

**O– O–**

**O**

**O– O–**

**O**

**O– O–**

**O**

**O**

**O– O–**

**O– O–**

pH 7.40 (0.26%)

pH 11.00 (96.30%; highest)

**<sup>O</sup>– <sup>O</sup>**

**<sup>O</sup>– <sup>O</sup>**

**<sup>O</sup>– <sup>O</sup>**

pH 7.40 (0.26%)

**<sup>O</sup>– <sup>O</sup>**

**<sup>O</sup>– <sup>O</sup>**

**<sup>O</sup>– <sup>O</sup>**

**<sup>O</sup>– <sup>O</sup>**

**<sup>O</sup>– <sup>O</sup>**

**<sup>O</sup>– <sup>O</sup>**

pH 7.40 (0.26%)

pH 11.00 (96.30%; highest)

pH 11.00 (96.30%; highest)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

pH 14.00 (95.55%; highest) pH 7.40 (0.00%)

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

molecules of tigecycline (MH+ and MH22+) are predominantly formed [67].

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**C–**

**C–**

**C–**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**OC–**

**O**

**C–**

**C–**

**C–**

**C–**

**C–**

**C–**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**C–**

**C–**

**13 14** 15

**13 14** 15

**13 14** 15

**13 14** 15

**13 14** 15

**13 14** 15

**13 14** 15

**13 14** 15

**13 14** 15

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**C–**

**C–**

**C–**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**C–**

**C–**

**C–**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**C–**

**C–**

**C–**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3** pH 7.40 (1.52%)

pH 7.40 (1.52%)

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

pH 7.40 (1.52%)

pH 7.40 (1.52%)

pH 7.40 (1.52%)

**O H O**

pH 7.40 (1.52%)

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3** pH 7.40 (0.44%)

pH 7.40 (0.44%)

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

pH 7.40 (0.44%)

pH 7.40 (0.44%)

pH 7.40 (0.44%)

**O H O**

pH 7.40 (0.44%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

pH 7.60 (2.82%; highest) pH 7.40 (2.73%)

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O**

**<sup>O</sup> O H –**

**O**

**<sup>O</sup> O H –**

**O**

**O**

**<sup>O</sup> O H –**

**O**

**<sup>O</sup> O H –**

**O**

**<sup>O</sup> O H –**

**<sup>O</sup> O H –**

**O**

**<sup>O</sup> O H –**

**O**

**O**

**<sup>O</sup> O H –**

**<sup>O</sup> O H –**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3** **NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**C–**

**C–**

**C–**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3** **N C H3**

**N C H3**

**O C H3**

**O C H3**

**N C H3**

**O C H3**

**O C H3**

**O C H3**

**N C H3**

**N C H3**

pH 4.00 (98.86%; highest) pH 7.40 (2.11%)

pH 4.00 (98.86%; highest) pH 7.40 (2.11%)

pH 4.00 (98.86%; highest) pH 7.40 (2.11%)

pH 4.00 (98.86%; highest) pH 7.40 (2.11%)

pH 4.00 (98.86%; highest) pH 7.40 (2.11%)

> **O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**O O H**

**O O H**

**O O H**

**O O H**

**O O H**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O**

**O**

**O– O H**

**O– O H**

**O**

**O**

**O– O H**

**O– O H**

**O**

**O– O H**

**O**

**O**

**<sup>O</sup>– O H**

**<sup>O</sup>– O H**

**O**

**O**

**<sup>O</sup>– O H**

**<sup>O</sup>– O H**

**O**

**<sup>O</sup>– O H**

pH 8.60 (49.38%; highest) pH 7.40 (11.90%)

pH 8.60 (49.38%; highest) pH 7.40 (11.90%)

**O H O**

**O H O**

**O H O**

pH 8.60 (49.38%; highest) pH 7.40 (11.90%)

pH 8.60 (49.38%; highest) pH 7.40 (11.90%)

pH 8.60 (49.38%; highest) pH 7.40 (11.90%)

**O H O**

**O– O**

**O H O**

**O– O**

**O– O**

**O– O**

**O– O**

Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated molecules of tigecycline (MH+ and MH22+) are predominantly formed [67]. Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated molecules of tigecycline (MH+ and MH22+) are predominantly formed [67]. Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated molecules of tigecycline (MH+ and MH22+) are predominantly formed [67]. Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated molecules of tigecycline (MH+ and MH22+) are predominantly formed [67]. Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated molecules of tigecycline (MH+ and MH22+) are predominantly formed [67]. Protonation equilibria and formed microspecies play an essential role in the bioavailability of tetracyclines [63]. The protonated state of tetracyclines is also essential in the analysis. For example, in electrospray mass spectrometry, both protonated molecules of tigecycline (MH+ and MH<sup>2</sup> 2+) are predominantly formed [67].

Electron-rich functional groups depending on pH can be protonated or deprotonated. Consequently, the tetracyclines have excellent chelating properties with several bivalent or trivalent metal cations [60]. Thus, tetracyclines form stable complexes with metal ions due to the characteristic substituents (Table S3—Supplementary Materials) [60].

First-generation tetracyclines form insoluble complexes with metal ions (Ca2+, Mg2+ , Fe3+, and Al3+) and consequently present reduced absorption [49]. Doxycycline and minocycline (second generation) are known for their excellent ability to chelate Fe3+ . Both doxycycline and minocycline absorption are impaired by ferrous sulphate; bismuth; and other antacids containing aluminium, calcium, and magnesium salts, such as coadministration of pharmaceuticals with Fe3+ and antacids (rich in Ca2+, Mg2+) [49,68].

Minocycline (second generation) and tigecycline (third generation) are more chelated by Ca2+ than tetracycline due to the C7 dimethylamino group. This moiety increased the electron density at the Ca2+ coordination site for the two studied tetracyclines. In addition, it was observed that tigecycline formed a different higher-order complex comparative to minocycline through the C9 N-t-butylglycylamido substituent in Ca2+ coordination [65]. Complexes with magnesium ions inhibit bacterial growth by impairing protein synthesis; these tetracycline complexes with magnesium act by binding to the 30S ribosomal subunits [69]. In plasma, tetracyclines are mainly chelated with Ca2+ and Mg2+ ions. A known mechanism of bacterial resistance to tetracyclines involves metal complexation. A possible strategy to combat bacterial resistance is to use in therapy the metal complexes of tetracycline (e.g., Pt2+ or Pd2+ complexes) [69,70].

#### *3.2. Structure-Activity Relationships*

As a result of studies of the relationship between chemical structure and biological activity, several aspects related to the class of tetracyclines are already known [3,13,48,49,71]. Next, the structural elements with an impact on the biological properties of the new tetracyclines are targeted.

**O– O O–**

**O– O O–**

**O– O O–**

**O– O O–**

**O– O O–**

**O**

**O C H3**

**O C H3**

**O C H3**

**O C H3**

*Pharmaceutics* **2021**, *13*, 2085 9 of 32

*Pharmaceutics* **2021**, *13*, 2085 9 of 32

*Pharmaceutics* **2021**, *13*, 2085 9 of 32

*Pharmaceutics* **2021**, *13*, 2085 9 of 32

*Pharmaceutics* **2021**, *13*, 2085 9 of 32

pH 7.60 (3.84%; highest) pH 7.40 (3.72%)

pH 7.60 (3.84%; highest) pH 7.40 (3.72%)

pH 7.60 (3.84%; highest) pH 7.40 (3.72%)

pH 7.60 (3.84%; highest) pH 7.40 (3.72%)

pH 7.60 (3.84%; highest) pH 7.40 (3.72%)

**O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3** **NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O**

**O**

**O O H –**

**O O H –**

**O**

**O**

**O O H –**

**O O H –**

**O**

**O O H –**

pH 6.80 (24.07%; highest) pH 7.40 (13.10%)

pH 6.80 (24.07%; highest) pH 7.40 (13.10%)

pH 6.80 (24.07%; highest) pH 7.40 (13.10%)

pH 6.80 (24.07%; highest) pH 7.40 (13.10%)

pH 6.80 (24.07%; highest) pH 7.40 (13.10%)

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**O H**

**O H**

**O H**

**O H**

**O H**

**N**

**N**

**N**

**N**

**N**

**O O**

**O O**

**O O**

**O O**

**O O**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**O**

**O**

**O**

**O**

**O**

pH 0.00 (98.16%; highest) pH 7.40 (0.00%)

pH 0.00 (98.16%; highest) pH 7.40 (0.00%)

pH 0.00 (98.16%; highest) pH 7.40 (0.00%)

pH 0.00 (98.16%; highest) pH 7.40 (0.00%)

pH 0.00 (98.16%; highest) pH 7.40 (0.00%)

> **O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**N H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH+ H3C C H3**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**O**

**O**

**C–**

**C–**

**C–**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**C–**

**C–**

pH 8.80 (17.75%; highest) pH 7.40 (4.50%)

pH 8.80 (17.75%; highest) pH 7.40 (4.50%)

pH 8.80 (17.75%; highest) pH 7.40 (4.50%)

pH 8.80 (17.75%; highest) pH 7.40 (4.50%)

pH 8.80 (17.75%; highest) pH 7.40 (4.50%)

> **O H O**

**O H O**

**O H O**

**O H O**

**O H O**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O N C H3 C H3**

**O**

**O**

**O O H –**

**O O H –**

**O**

**O**

**O O H –**

**O O H –**

**O**

**O O H –**

**O**

**O**

**<sup>O</sup> O H –**

**<sup>O</sup> O H –**

**O**

**O**

**<sup>O</sup> O H –**

**<sup>O</sup> O H –**

**O**

**<sup>O</sup> O H –**

**3 5 6** 

**3 5 6** 

**3 5 6** 

**3 5 6** 

**3 5 6** 

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**NH2**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**C–**

**C–**

**C–**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**O**

**C–**

**C–**

**7 8 9** 

**7 8 9** 

**7 8 9** 

**7 8 9** 

**7 8 9** 

Tigecycline was discovered as a result of chemical structure–activity studies [72]. Due to structural modifications made to C9 position (Figure 10), tigecycline has an affinity for the ribosomal target five times higher than tetracycline or minocycline. Therefore, this change is responsible for broadening the antibacterial spectrum and combating ribosomal protection, one of the two mechanisms of bacterial resistance specific to tetracycline [29,73,74]. Moreover, this radical is a bulky steric hindrance that prevents the expulsion of the substance out of the bacterial cell by effluent tet proteins, thus reducing the susceptibility of developing antibiotic resistance [28,75]. of the new tetracyclines are targeted. Tigecycline was discovered as a result of chemical structure–activity studies [72]. Due to structural modifications made to C9 position (Figure 10), tigecycline has an affinity for the ribosomal target five times higher than tetracycline or minocycline. Therefore, this change is responsible for broadening the antibacterial spectrum and combating ribosomal protection, one of the two mechanisms of bacterial resistance specific to tetracycline [29,73,74]. Moreover, this radical is a bulky steric hindrance that prevents the expulsion of the substance out of the bacterial cell by effluent tet proteins, thus reducing the susceptibility of developing antibiotic resistance [28,75].

Minocycline (second generation) and tigecycline (third generation) are more chelated by Ca2+ than tetracycline due to the C7 dimethylamino group. This moiety increased the electron density at the Ca2+ coordination site for the two studied tetracyclines. In addition, it was observed that tigecycline formed a different higher-order complex comparative to minocycline through the C9 N-t-butylglycylamido substituent in Ca2+ coordination [65]. Complexes with magnesium ions inhibit bacterial growth by impairing protein synthesis; these tetracycline complexes with magnesium act by binding to the 30S ribosomal subunits [69]. In plasma, tetracyclines are mainly chelated with Ca2+ and Mg2+ ions. A known mechanism of bacterial resistance to tetracyclines involves metal complexation. A possible strategy to combat bacterial resistance is to use in therapy the metal complexes of tetracy-

As a result of studies of the relationship between chemical structure and biological activity, several aspects related to the class of tetracyclines are already known [3,13,48,49,71]. Next, the structural elements with an impact on the biological properties

*Pharmaceutics* **2021**, *13*, 2085 10 of 31

cline (e.g., Pt2+ or Pd2+ complexes) [69,70].

*3.2. Structure-Activity Relationships* 

**Figure 10.** Optimisation of C7 and C9 positions in the development of new tetracyclines. **Figure 10.** Optimisation of C7 and C9 positions in the development of new tetracyclines.

Researchers observed some structural features of glycylcyclines that are essential in maintaining biological activity [13]. A critical element is the basic nitrogen atom from the glycyl unit; derivatives containing low volume alkyl-amino or cyclic amine groups have shown optimal results. Attempts to replace the radical from the C9 position with other amino acids such as alanine, phenylalanine, and leucine failed because the resulting compounds were much less effective [72]. Similarly, substitutions with smaller groups than Researchers observed some structural features of glycylcyclines that are essential in maintaining biological activity [13]. A critical element is the basic nitrogen atom from the glycyl unit; derivatives containing low volume alkyl-amino or cyclic amine groups have shown optimal results. Attempts to replace the radical from the C9 position with other amino acids such as alanine, phenylalanine, and leucine failed because the resulting compounds were much less effective [72]. Similarly, substitutions with smaller groups than the *tert*-butyl-amino group led to compounds with low potency, while the attempt to substitute the amine with n-propyl, n-butyl, and n-hexyl has not brought improvements [76]. The antimicrobial activity and the pharmacokinetics of tigecycline are considerably influenced by the ability to form complexes with metal ions (calcium, magnesium, and iron). The target of the cations is the keto–enol system (C11 and C12 positions), the enol in position C1, and the carboxamide in position C2. Therefore, water-insoluble chelates are formed with a low absorption [13].

Chemical modulations performed to obtain omadacycline on C9 led to an increase in antimicrobial potency by overcoming the resistance to the efflux mechanisms and overcoming ribosomal protection [22,77]. Furthermore, the aminomethyl moiety from C9 position provides improved pharmacokinetic parameters, such as dose reduction (high doses cause side effects such as nausea and vomiting, often encountered in C9 glycylcyclines), and, in particular, oral bioavailability [78]. Due to these changes, omadacycline has a pharmacokinetic profile (absorption, distribution, metabolism, and excretion, ADME) that distinguishes it from the glycylcycline subgroup [77].

A study was conducted on aminomethycyclines with in vitro potency (with a minimum inhibitory concentration (MIC) ≤ 0.06–2.0 µg/mL) against Gram-negative bacteria

with different mechanisms of resistance by ribosomal protection (Tet(M)): *Staphylococcus aureus, Enterococcus faecalis,* and *Streptococcus pneumoniae,* and on the efflux mechanisms Tet(K) in *Staphylococcus aureus*, Tet(L) in *Enterococcus faecalis*) [51]. Compounds with lipophiliaenhancing or benzyl substitutions in the aminomethyl side group showed the highest potency against ribosomal alteration and efflux of resistant strains. However, high-polarity analogues or electrically charged groups, as well as acyl derivatives, showed a significant decrease in antibacterial activity. Although alkyl substituents (e.g., *tert*-butyl group) showed moderate potency, they were chosen for further optimisation and screening. It was found that analogues with the alkyl group, which extend with at least three carbon atoms to the aminomethyl group; those with branched alkyl chains; and piperidine analogues have superior activity. The ramification in the alkyl chain from position 1 has a detrimental impact due to steric hindrance. The introduction of two methyl groups in position 2 showed a significant improvement in antibacterial activity. Finally, residues containing more than five carbon atoms had reduced activity in the presence of plasma, indicating a high percentage binding to plasma proteins [38,51].

Therefore, following classical studies on the chemical structure–biological activity relationships, omadacycline, a compound containing a neopentyl moiety in the aminomethyl group, has been identified as the most valuable aminomethylcycline in this series, becoming a new subclass of tetracyclines [77].

It has been observed on fluorocyclines that as the substituents attached to the carbon atom at the C9 positions are more polar or more basic, the microbiological activity of the compound will increase, especially on Gram-negative bacteria. A study conducted in 2012 examined the behaviour of analogues of 7-fluoro-9-amino-acetamido-6-demethyl-6-deoxytetracyclines on various Gram-negative and Gram-positive bacteria, but also on gene isolates resistant to tetracycline class [79]. In general, less voluminous secondary or tertiary amine analogues from C9 position were found to have a lower MIC compared to analogues with substituents such as aromatic amines or alkylamines with lower basicity. Compared to tertiary alkylamine, dimethyl, azetidine, and piperidine analogues, eravacycline, a compound bearing a pyrrolidine nucleus, showed a 8 to 16 times higher potency against *K. pneumoniae* (tet(A)) and 4 to 8 times higher potency against *E. coli* (tet(A)). In addition, eravacycline is 4 to 64 times more potent than piperidine and azetidine omologues against bacterial isolates that have been tested (except for *S. pneumoniae* expressing or not expressing tet(M) protein, were it showing an equivalent response). The addition of polar substituents, fluorine atoms, or pyrrolidine bicycles produced no improvements, but neither did negative influence on the activity against pneumococcal bacteria when compared to unsubstitued pyrrolidine analogues [40,80]. The pyrrolidine substituent at the C9 the fluoro substituent at C7, the main optimisations in eravacycline, positively influenced the potency and the antibacterial spectrum [19].

Unlike other tetracyclines, the chemical structure of sarecycline includes a unique modification at the C7 position (the longest and most voluminous of the whole class), a 7-[(methoxy-(methyl)-amino)-methyl]methyl group (Figure 10). As a result of this chemical optimisation, the activity of this compound is enhanced, binding to the codon of the A site, interfering with the movement of messenger RNA (mRNA) along the channel, or disrupting the codon A-anticodon interaction [42].

The substituted tetracycline system at positions C7 and C9 is the basis of compounds with increased antibacterial activity, while any modification made at the C1-C4, C10-C12, C-11a, and C-12a will have a negative consequence on their action. Other important aspects regarding the relationship between chemical structure and biological activity are presented in Figure 11 [3,38,40,51].

In addition, tetracyclines contain a 4S(α)-dimethyl-amino group in the C4 position, an absolute necessity for optimal antibacterial activity. On the other hand, the epimerisation of the 4R(β) isomer will lead to a decrease in antibacterial activity, especially against Gramnegative bacteria. The epimerisation process from position C4 takes place during harsh

chemical reactions, in vivo metabolism phenomenon, but also under changes in the pH values [25]. Gram-negative bacteria. The epimerisation process from position C4 takes place during harsh chemical reactions, in vivo metabolism phenomenon, but also under changes in the pH values [25].

the C9 the fluoro substituent at C7, the main optimisations in eravacycline, positively in-

Unlike other tetracyclines, the chemical structure of sarecycline includes a unique modification at the C7 position (the longest and most voluminous of the whole class), a 7- [(methoxy-(methyl)-amino)-methyl]methyl group (Figure 10). As a result of this chemical optimisation, the activity of this compound is enhanced, binding to the codon of the A site, interfering with the movement of messenger RNA (mRNA) along the channel, or

The substituted tetracycline system at positions C7 and C9 is the basis of compounds with increased antibacterial activity, while any modification made at the C1-C4, C10-C12, C-11a, and C-12a will have a negative consequence on their action. Other important aspects regarding the relationship between chemical structure and biological activity are

In addition, tetracyclines contain a 4S(α)-dimethyl-amino group in the C4 position, an absolute necessity for optimal antibacterial activity. On the other hand, the epimerisation of the 4R(β) isomer will lead to a decrease in antibacterial activity, especially against

*Pharmaceutics* **2021**, *13*, 2085 12 of 31

fluenced the potency and the antibacterial spectrum [19].

disrupting the codon A-anticodon interaction [42].

presented in Figure 11 [3,38,40,51]**.** 

**Figure 11.** The essential relationship between chemical structure and biological activity of modern tetracyclines. **Figure 11.** The essential relationship between chemical structure and biological activity of modern tetracyclines.

The C4 β-epimers have noticeably different properties from those of compounds with a normal configuration. The most significant difference is observed in antibacterial activity manifested in vitro. β-Epimers have been found to be responsible for approximately 5% of normal tetracycline activity. It has been observed that the epimerisation phenomenon takes place in different solvent systems, at variations of pH between 2 and 6 [14]. Tetracyclines are prone to epimer formation, particularly under weak acidic conditions. The epimers have distinct toxicological and antibacterial properties, and therefore selective biosynthesis is a major challenge. This is due to the fact that the epimers are isobars with the parent compound, having very similar physico-chemical properties. Epimerisation can occur in vivo, even in the bladder [59]. The C4 β-epimers have noticeably different properties from those of compounds with a normal configuration. The most significant difference is observed in antibacterial activity manifested in vitro. β-Epimers have been found to be responsible for approximately 5% of normal tetracycline activity. It has been observed that the epimerisation phenomenon takes place in different solvent systems, at variations of pH between 2 and 6 [14]. Tetracyclines are prone to epimer formation, particularly under weak acidic conditions. The epimers have distinct toxicological and antibacterial properties, and therefore selective biosynthesis is a major challenge. This is due to the fact that the epimers are isobars with the parent compound, having very similar physico-chemical properties. Epimerisation can occur in vivo, even in the bladder [59].

#### *3.3. Mechanism of Action*

Tetracyclines inhibit protein synthesis by inhibiting the association of aminoacyl-tRNA with bacterial ribosome [49,75]. Tetracyclines bind with high affinity to a specific locus (16S) on the 30S ribosomal unit during translation. In this way, the penetration of aminoacyl transporter RNA (tRNA) into the acceptor site (A) on the bacterial ribosome is blocked, the consequence being the cessation in the incorporation of amino acids residues in the process of elongation of the polypeptide chain. Thus, the protein synthesis is stopped (Figure 12) [18,81,82]. Commonly, at therapeutic concentrations, tetracyclines are consider bacteriostatic antibiotics [18], but late studies have described their bactericidal effects in vitro, especially in the case of tigecycline (studies on mice) [83].

*3.3. Mechanism of Action* 

Tetracyclines inhibit protein synthesis by inhibiting the association of aminoacyltRNA with bacterial ribosome [49,75]. Tetracyclines bind with high affinity to a specific locus (16S) on the 30S ribosomal unit during translation. In this way, the penetration of aminoacyl transporter RNA (tRNA) into the acceptor site (A) on the bacterial ribosome is blocked, the consequence being the cessation in the incorporation of amino acids residues in the process of elongation of the polypeptide chain. Thus, the protein synthesis is

consider bacteriostatic antibiotics [18], but late studies have described their bactericidal

effects in vitro, especially in the case of tigecycline (studies on mice) [83].

**Figure 12.** Scheme of the tetracyclines' mechanism of action, where AA—aminoacids, TCs—tetracyclines, tRNA—transfer ribonucleic acid, mRNA—messenger ribonucleic acid, 30S and 50S—ribosomal subunits (created with BioRender.com (accessed on 30 September 2021) [84]. **Figure 12.** Scheme of the tetracyclines' mechanism of action, where AA—aminoacids, TCs—tetracyclines, tRNA—transfer ribonucleic acid, mRNA—messenger ribonucleic acid, 30S and 50S—ribosomal subunits (created with BioRender.com (accessed on 30 September 2021) [84].

It is well known that tetracyclines cross the membranes of Gram-negative bacteria through a cationic complex with Mg2+, using the OmpF and OmpC porins in the outer membrane [38,85]. Later on, the Donnan potential generated along the outer membrane causes the accumulation of the complex in the periplasmic space, where the dissociation from the Mg2+ ion of tetracycline takes place and there is a release of an electrically uncharged molecule that is lipophilic enough to diffuse through the inner membrane into the cytoplasm [86]. The uptake of tetracyclines in the cytoplasm is partially energy-dependent, involving, in addition to passive diffusion, the proton-motive force and the hydrolysis of phosphate bonds [18]. For Gram-positive bacteria, it has been reported that these agents reach the cytoplasm by passive diffusion and/or active transport. In the cytoplasm, tetracyclines chelate Mg2+ ions again and, in this form, attack the ribosomal target [75]. Hence, bivalent ions are a vital element in the transport and efficiency of these compounds [85]. It is well known that tetracyclines cross the membranes of Gram-negative bacteria through a cationic complex with Mg2+, using the OmpF and OmpC porins in the outer membrane [38,85]. Later on, the Donnan potential generated along the outer membrane causes the accumulation of the complex in the periplasmic space, where the dissociation from the Mg2+ ion of tetracycline takes place and there is a release of an electrically uncharged molecule that is lipophilic enough to diffuse through the inner membrane into the cytoplasm [86]. The uptake of tetracyclines in the cytoplasm is partially energydependent, involving, in addition to passive diffusion, the proton-motive force and the hydrolysis of phosphate bonds [18]. For Gram-positive bacteria, it has been reported that these agents reach the cytoplasm by passive diffusion and/or active transport. In the cytoplasm, tetracyclines chelate Mg2+ ions again and, in this form, attack the ribosomal target [75]. Hence, bivalent ions are a vital element in the transport and efficiency of these compounds [85].

Tigecycline is not affected by most common antibiotic resistance mechanisms because it binds to the 30S subunit (five times stronger than tetracyclines) [28], even in the presence of ribosomal protection, being excepted from membrane efflux [87]. This is due to the voluminous substituent in the C9 position of the naphtacenic nucleus, representing a steric hindrance [88,89]. Tigecycline's activity was evaluated by a study on *Escherichia coli* derivatives containing plasmids expressing different specific efflux genes (tet[B], tet[C], and tet[K]). An unchanged MIC value confirmed tigecycline's protection against these efflux genes [90]. Moreover, glycylcyclines also manifest resistance to less common Tigecycline is not affected by most common antibiotic resistance mechanisms because it binds to the 30S subunit (five times stronger than tetracyclines) [28], even in the presence of ribosomal protection, being excepted from membrane efflux [87]. This is due to the voluminous substituent in the C9 position of the naphtacenic nucleus, representing a steric hindrance [88,89]. Tigecycline's activity was evaluated by a study on *Escherichia coli* derivatives containing plasmids expressing different specific efflux genes (tet[B], tet[C], and tet[K]). An unchanged MIC value confirmed tigecycline's protection against these efflux genes [90]. Moreover, glycylcyclines also manifest resistance to less common mechanisms, such as altered target site conformation, enzymatic degradation, and mutations in DNA gyrase [87].

> Similar to tigecycline, omadacycline possesses excellent activity against bacterial isolates carrying a wide variety of resistance mechanisms, including both tet[K] and tet[O] genes simultaneously [22,91]. Due to the reversible binding of tetracyclines to ribosomes, they act as bacteriostatic agents. Instead, in vitro omadacycline has demonstrated bactericidal activity against *Haemophilus influenzae*, *Streptococcus penumoniae*, and *Moraxella catarrhalis* [78]. Omadacycline has no significant effect on the synthesis of RNA, DNA, and peptidoglycan. Like tigecycline, omadacycline binds to the 30S subunit of the bacterial ribosome with enhanced binding based on other molecular interactions [77]. Having a D-ring modified with a pyrrolidinacetamide side chain, eravacycline was designed

to maintain its activity against resistant bacteria (e.g., carbapenem-resistant, MDR, and extended-spectrum cephalosporin-resistant Enterobacteriaceae and extended-spectrum, β-lactamase-producing Enterobacteriaceae) [41,92]. Eravacycline has 10 times the affinity for the ribosomal target in vitro and inhibits translation at four times lower concentrations than tetracycline [93]. A study conducted by *Batool* et al. in 2020 concluded that sarecycline differs slightly from other tetracycline derivatives in terms of mechanism of action, emphasising its unique role in this large family, a role that clinicians should take into account when evaluating its therapeutic potential. By analysing the crystal structure of sarecycline related to the bacterial initiation complex, it has demonstrated that in addition to binding to the same site of the small ribosomal subunit, sarecycline, due to the C7 moiety, expands and establishes uncommon interactions with mRNA. This contact leads to the stabilisation of the substance on the ribosome and an increased inhibitory effect. Thus, sarecycline overcome the mechanisms of bacterial resistance to tetracyclines [42,57].

#### Other Biological Effects

In addition to the approved therapeutical uses as antibiotics, tetracyclines have other biological effects, which have been investigated and exploited. These non-antibiotic properties comprise anti-inflammatory effects; anti-apoptotic activity; immunomodulatory properties; inhibitory effects on proteolysis, angiogenesis, and tumour metastasis; and a neuroprotector effect [17].

The anti-inflammatory activity of this class is mediated by a large number of mechanisms such as inhibition of neutrophil activation and migration; T lymphocyte activation and proliferation; inhibition of phospholipase, angiogenesis, nitric oxide synthesis, and granuloma formation; suppression of inflammatory cytokine release (TNFα, IL-1β, IL-6, IL-8); and decrease of reactive oxygen species [94,95]. However, the best-known mechanism of anti-inflammatory action is the inhibition of matrix metalloproteinases (MMPs). This mechanism occurs both directly and indirectly by inhibiting the MMPs synthesis [96]. Therefore, tetracyclines, such as doxycycline, minocycline, and sarecycline, are frequently prescribed for acne vulgaris when topical treatment is unsuccessful [94]. In a study using the carrageenan-induced rat paw oedema inflammation model, sarecycline presented an anti-inflammatory effect comparable to doxycycline and minocycline at all the tested doses [42]. Moreover, sarecycline has been proven to be a valuable alternative for treating papulopustular rosacea in terms of efficiency, safety, and tolerability [97].

The non-antibiotic properties of tetracyclines and their analogues were studied in both dermatological and non-dermatological diseases, as presented in Table 3 [98].


**Table 3.** Other therapeutical uses of tetracyclines related to their non-antibiotic properties.

Research has shown that minocycline has several non-antibiotic biological effects that are beneficial in experimental models of various inflammatory diseases. These include dermatitis, periodontitis, atherosclerosis, and autoimmune diseases such as rheumatoid arthritis and inflammatory bowel disease. Due to its high lipophilicity, minocycline readily penetrates the blood–brain barrier and achieves high concentrations in the brain; hence, it

has been effective in neuroprotection. However, because of its lipophilicity, vestibular side effects such as dizziness and vertigo have been associated with minocycline therapy. This outcome has been confirmed by experimental studies of ischemia, traumatic brain injury, and neuropathic pain, as well as several neurodegenerative diseases (e.g., Parkinson's disease, Huntington's disease, Alzheimer's disease, amyotrophic lateral sclerosis, spinal cord injury, and multiple sclerosis). In addition, other pre-clinical studies have shown the ability of minocycline to inhibit malignant cell growth and activation, HIV replication, and prevent bone resorption [17].

#### *3.4. Spectrum of Antibacterial Activity*

Detailed antibacterial spectrum of the new generation of tetracyclines is shown in Table S5 (Supplementary Materials).

Tigecycline presents activity against MDR pathogens, such as MRSA; *Staphylococcus epidermidis*; vancomycin-resistant Enterococcus; *Acinetobacter* spp.; *Stenotrophomonas maltophilia*; penicillin-resistant *Streptococcus pneumoniae*; and Enterobacteriaceae resistant to aminoglycosides, carbapenems, fluoroquinolones, and β-lactamase producers [118,119]. Tigecycline is very active against *Neisseria gonorhoae* and *Eikenella corrodens*, as well as on rapidly growing species of mycobacteria (*M. chelonae*, *M. abscessus*, *M. fortuitum)* [120]. Although against most of tigecycline's action is bacteriostatic, there are pathogens on which it acts bactericidally, such as *Legionella pneumophila* and *Streptococcus pneumoniae*. However, tigecycline is not effective against *Pseudomonas aeruginosa*, *Proteus mirabilis*, *Providencia* spp., or *Morganella morganii* [121].

Omadacycline is very potent against atypical bacteria, as well as Gram-positive and Gram-negative aerobic pathogens [22,122,123], and further against anaerobic bacteria that cause infections from dog or cat bites (except *Eikenella corrodens*); however, omadacycline is not effective on the species of *Proteus, Providencia*, *Morganella*, and *Pseudomonas* [124]. Generally, omadacycline acts as bacteriostatic agent, but against *Escherichia coli, Streptococcus pneumoniae*, and *Haemophilus influenzae* acts bactericidal [125]. This new tetracycline is more active than tigecycline and acts similar to eravacycline against Gram-positive pathogens [40].

Eravacycline is a broad-spectrum tetracycline that has showed a great activity against aerobic and anaerobic Gram-negative and Gram-positive bacteria, except *P. aeruginosa* and *Burkholderia cenocepacia*. Eravacycline also shows good activity against MDR bacteria, including Enterobacteriaceae and *A. baumannii*, expressing extended spectrum βlactamases, carbapenem resistance, and mechanisms conferring resistance to other antibiotic classes [18]. Eravacycline is more effective than omadacycline against Gram-negative and broad-spectrum beta-lactamase-producing bacteria. Eravacycline is two to four times more active than tigecycline on clinically relevant Gram-positive species [40].

The spectrum of sarecycline is narrow, with little activity against aerobic and anaerobic Gram-negative bacteria and microflora commonly found in the gastrointestinal tract. The activity of this compound specifically targets *Cutinebacterium acnes*, but also some clinically relevant Gram-positive bacteria, including MRSA [126]. It is noteworthy that the prolonged and intermittent use of broad-spectrum antibiotics such as doxycycline and minocycline in acne vulgaris has been associated with the development of antimicrobial resistance and permanent perturbation of the gut and cutaneous microbiome. Although no causal relationship has been definitively established, the use of doxycycline in patients with acne was found to be associated with a 2.25-fold greater risk of developing Crohn's disease [127,128].

#### *3.5. Bacterial Resistance to New Tetracyclines*

The widespread use of these antibacterial agents has unavoidably led to the development of bacterial resistance through plasmid-encoded tetracycline resistance genes (tet), conjugated transposons and integrons, which allow tet genes to be transmitted from one

species/generation to another species/generation through conjugation [13,72,129]. The four main types of tetracycline resistance are outlined in Table 4 [18,27,38].

**Table 4.** Mechanisms of resistance and resistance determinants of tetracyclines.


The expression of these genes leads to the production of proteins that contribute to the two primary mechanisms of resistance: ribosomal protection by dissociating tetracyclines from their target (e.g., tet[M], tet[O]) and the efflux of the substance out of the cell by active transport (e.g., tet[A], tet[B]) [28,118]. The efflux pumps are located in the cytoplasmic membrane and act through the antiport of a proton in exchange with a tetracycline–magnesium monocationic complex. Thus, the intracellular concentration of tetracyclines decreases [13]. The most common pumps are part of the Major Facilitator Superfamily of carriers [75]. In contrast to efflux pumps, for which the mechanism is elucidated, the ribosomal protection mechanism is not fully known. However, studies suggest that genes involved (e.g., tet[M], tet[O], and tet[S]) alter the conformation of ribosomes and displace the drug from the active site [38,85,130]. These genes are protein molecules GTPases (guanosine triphosphatases) that have structures and sequences similar to elongation factors (EF-G and EF-Tu) [131].

The other two less common mechanisms include two distinct genes that modify tetracyclines, leading to their degradation and mutations of the ribosomal 16S subunit, decreasing the affinity of the compounds for the ribosomal target [27]. The first of the mechanisms, chemical inactivation, is caused by a FAD-monooxygenase encoded by the tet[X] and tet [37] genes. These hydroxylate the C11 position, altering its structure and coordination with the magnesium ion, and therefore the affinity for the ribosome [38,132]. Moreover, the hydroxylated version degrades even in the absence of enzymes [133]. Because monooxygenase uses NADPH and O<sup>2</sup> in its activity, this resistance occurs only in aerobic organisms [27]. The second mechanism of resistance, the site-binding mutation, is found predominantly in bacteria with a small number of copies of rRNA [18]. The first case was described by Ross et al. (1998) on a bacterial strain of *Propionibacterium acnes*. In addition, there are innate resistance mechanisms, with some bacteria being more immune to the tetracycline class due to differences in membrane permeability. For example, Gramnegative bacteria are more resistant due to the outer wall containing lipopolysaccharide molecules [27].

Although tigecycline is not affected by the main resistance mechanisms, since its introduction in the clinic (2005), the number of pathogens that have developed resistance is continuously increasing. In Gram-negative bacteria, most cases were caused by overexpression of resistance-nodulation-cell division (RND) pumps. For example, MexXY-OprM for *Pseudomonas aeruginosa* strains and AdeIJK/AdeABC for *Acinetobacter baumannii* [134–137]. Something similar occurs for some Gram-positive bacteria by overexpressing the Multidrug and Toxic Compound Extrusion (MATE) (case of MepA pump for Staphylococcus aureus) [138]. Other mechanisms also decrease the susceptibility of tigecycline to other microorganisms: mutations in ribosomal protein genes, mutations in the 16S subunit of rRNA [139], inactivation by FAD-dependent monooxygenase [140], and tetX gene-carrying plasmids [141]. A recent study on *Acinetobacter baumannii* found a resistance plasmid

containing the tet(X5) gene, with similarity regarding structure and function of other tet(X) variants, probably using the same transfer elements for spreading [142].

The antimicrobial activity of omadacycline is not disturbed by the two significant resistance mechanisms in therapy [143]. This aspect is due to the functional groups in the C7 (dimethylamino) and C9 (aminomethyl) positions, which prevent the expulsion of the antibiotic by the efflux pumps and the ribosomal protection, respectively [22]. Omadacycline has demonstrated in vitro activity against bacterial strains containing efflux and ribosomal protection genes (*Staphylococcus aureus* expressing the tet(K) and tet(M) genes, *Enterococcus faecalis* expressing the tet(L) and tet(M) genes) [125]. However, the activity of omadacycline is decreased by mutations in the ribosomal RNA of some microorganisms. According to a study conducted by Heidrich et al. (2016), the MIC for tetracycline, tigecycline, and omadacycline on species containing mutations (G1055C, G996U) of the 16S ribosomal subunit increased four to eightfold, suggesting that they are susceptible to these changes, regardless to their affinity for the target [27,144]. The omadacycline's action and the action of glycylcyclines, in general, are also affected by chemical inactivation [27].

Eravacycline avoids tet(A) efflux pumps, maintains activity against staphylococcicontaining tet(K) genes, and successfully binds to bacterial ribosomes modified by tet(M) proteins [98,145]. However, eravacycline remains vulnerable to overexpression of MDR efflux pumps belonging to Gram-negative bacteria, change in ribosomal target (16S or 10S), and enzymatic degradation sometimes encountered in Bacterioides spp. The resistance to eravacycline has also been observed in mutant species of Enterococcus, mutations encoded by the rpsJ gene [41,146,147].

In the case of sarecycline, the probability of inducing bacterial resistance is low due to the narrow spectrum of activity and the unique structural modifications at the C7 position. The rate of spontaneous mutations varies from 10−<sup>9</sup> to 10−<sup>11</sup> in *Cutinebacterium acnes*, and 10−<sup>9</sup> and 10−<sup>8</sup> in *Staphylococcus aureus* and *Staphylococcus epidermidis*, respectively (at increases in MIC values between four and eight times) [126,148].

#### **4. Therapeutic Use of the New Tetracyclines**

In the past, tetracyclines have been widely used for various genitourinary, gastrointestinal, respiratory tract, and dermatological diseases. However, the tremendous onset of bacterial resistance, as well as the emergence of new antibacterial agents, has diminished the area of infections for which tetracyclines are considered the first therapeutic option [1,18,27].

The activity of tigecycline has been evaluated in several in vivo clinical trials in human subjects, following which the three FDA-approved indications were formulated. In the treatment of hospitalised patients with complicated skin and soft tissue infections, tigecycline has not only been shown to be effective but has also shown a favourable pharmacokinetic profile [149]. In another phase 2 open-label clinical trial, which included patients with complicated intra-abdominal infections (gangrenous perforated appendicitis, cholecystitis, diverticulitis, peritonitis), tigecycline was a safe and effective treatment [120]. Last but not least, tigecycline is indicated and approved in the treatment of communityacquired pneumonia. For all research, tigecycline met all non-inferiority criteria [31,150].

Omadacycline, unlike tigecycline or eravacycline, brings an advantage through oral formulation, facilitating patient compliance and hospitalisation costs [38]. Omadacycline is recommended in treating complicated skin and soft tissue infections and communityacquired pneumonia, but there are ongoing studies for its use in urinary tract infections [39,151–153].

Eravacycline, due to its broad antibacterial spectrum, in vitro activity, and superior tolerability in comparison with tigecycline, is an appropriate solution for treating complicated intra-abdominal infections in adults, especially when the pathogen possesses resistance mechanisms to other tetracyclines or classes of antibiotics [41]. Infection, for which the efficacy of eravacycline has been studied by comparison with the beta-lactam antibiotics

meropenem or ertapenem, include: appendicitis, cholecystitis, diverticulitis, gastric or duodenal perforation, intra-abdominal abscess, intestinal perforation, and peritonitis [145].

The FDA-approved sarecycline is used for treating acne vulgaris in patients aged nine years or above, demonstrating efficacy against moderate-to-severe, inflammatory, or non-inflammatory (comedones) forms. In addition, due to its targeted action on *Cutinebacterium acnes* and low blood–brain penetration, sarecycline has a good safety profile (minimal side effects), low potential to induce bacterial resistance, and also potentially low impact on the gut microbiota when compared to the broad-spectrum doxycycline and minocycline [154,155].

*Pharmacokinetic properties.* Tigecycline is highly bound to plasma proteins and has a large volume of distribution (above plasmatic volume), which indicates its concentration in tissues. Moreover, tigecycline is rapidly distributed in tissues; the highest concentrations were observed in the bone marrow, thyroid gland, salivary glands, spleen, and kidneys. Tigecycline metabolises independent of cytochrome P450 enzymes, but not extensively. Consequently, tigecycline does not interfere with the metabolism of other substances mediated by the six cytochrome P450 isoforms (1A2, 2C8, 2C9, 2C19, 2D6, and 3A4) [31]. The pharmacokinetics of tigecycline is linear—this may be influenced by the coadministration of P-glycoprotein inhibitors or inducers; tigecycline acts as a substrate of these [156]. Similarly, omadacycline presented a low probability of interactions through transport mechanisms [133]. The omadacycline rate of absorption decreases if a high-fat meal is consumed two hours earlier. Thus, omadacycline must be taken after a fasting period of at least 4 h, followed by 2 h without ingestion of drinks and food (apart from water), as well as 4 h without administration of antacids, multivitamins, and dairy products [157]. The liver metabolises eravacycline, but none of the metabolites are pharmacologically active. Therefore, caution is required in CYP3A4 inducers to increase the extent of eravacycline metabolism to a clinically relevant rate [41]. Sarecycline inhibits P-glycoprotein in vitro. Consequently, decreasing dose and toxicity examination is required when it is coadministered with substrate substances [56]. Generally, the pharmacokinetics of modern tetracyclines are not remarkably influenced by age, sex, or renal function (including renal failure and haemodialysis) [29,39,158]. The pharmacokinetic parameters of the four modern tetracyclines are shown in Table S6 (Supplementary Materials).

#### **5. Side Effects of the Third-Generation Tetracyclines**

As an antibiotic class, tetracyclines are generally well tolerated. However, there is a diversity of side effects and contraindications, with these compounds affecting several systems of the human body. For example, tetracyclines often cause gastrointestinal disorders such as abdominal discomfort, nausea, vomiting, and epigastric pain. Moreover, typical side effects are photosensitivity, manifested by erythema and skin blisters, discolouration of the teeth, and inhibition of bone growth in children. Rarely, tetracyclines may cause increased intracranial pressure (pseudotumor cerebri), renal toxicity, hepatotoxicity, and *Clostridium difficile* infections [159].

Similar to tetracyclines in the first generations, the most common side effects of modern tetracyclines are those of the gastrointestinal tract [75,160–162]. Other side effects of the tetracycline new generation are shown in Table 5. Nausea and vomiting may occur in the first two days of treatment and are usually mild to moderate. In the case of tigecycline, these effects are correlated with the dose administered, the highest tolerated doses being 100 mg for healthy subjects without fasting, 200 mg postprandial [156]. Diarrhoea has been reported, and is associated, in the vast majority of cases, with *Clostridium difficile* superinfection, ranging from mild forms to severe or fatal colitis. Other less common side effects are constipation, anorexia, dyspepsia, dry mouth, acute pancreatitis, and pancreatic necrosis [1,31,39,46,145]. Regarding the pharmacotoxicology of sarecycline, due to its narrower spectrum, it does not affect the intestinal flora as much, and therefore adverse effects such as diarrhoea and fungal infections have been observed less clinically [126].


**Table 5.** Other side effects of modern tetracyclines [31,33,38,39,46,145,152,154,160,163,164].

Tetracyclines affect teeth and bones by forming stable complexes with calcium ions, accumulating in deposits at these levels. Thus, the teeth may acquire a yellow to brown colour, sometimes even permanent, due to the formation of chelates of tetracycline-calcium orthophosphate, which darkens after exposure to the sun [126]. This phenomenon has effects from an aesthetic point of view but can be aggravated, leading to demineralisation and hypoplasia of tooth enamel with decreased resistance to caries attack [165]. Tooth staining is more common in long-term treatment with tetracycline derivatives but has also been observed with repeated short-term administrations. Children who receive tetracyclines in the first part of life and children whose mothers have used them since the second trimester of pregnancy tend to have tetracyclines deposited at the level of baby teeth. There was also a decrease in the rate of fibula growth and ossification processes for the foetus exposed in utero due to accumulation in the tissues [39,46,145].

Another subgroup of side effects is skin damage. Tetracyclines may cause allergic-type side effects with pruritus, transient rash, or itchy skin and hyperhidrosis [31,145]. These reactions are due to the increased sensitivity of the skin to light during systemic tetracycline therapy. Therefore, patients undergoing treatment should avoid excessive exposure to natural or artificial sunlight (ultraviolet radiation) [46]. Hypersensitivity reactions (Stevens– Johnson syndrome, anaphylaxis), sepsis, and death have been reported with low frequency when using tigecycline [166].

Hepatobiliary disorders due to tetracyclines are not very common in this class but are reflected in increased plasma concentrations of aspartate aminotransferase (AST), alanine aminotransferase (ALT), bilirubin, and hepatic transaminases (TGP, TGO). Other laboratory parameters that may change are increased amylase, lipase, gamma-glutamyltransferase, urea nitrogen (class effect), creatinine phosphokinase, alkaline phosphatase, and decreased creatinine clearance. These reactions occur relatively infrequently, with a frequency <2% for tigecycline and omadacycline, and <1% for eravacycline (following clinical trials) [31,39,145]. Cases of cholestatic jaundice and mild pancreatitis induced by tigecycline have been reported [73,167]. At the vascular level, forms intended for the intravenous route produce reactions at the site of administration. With the exception of sarecycline (orally administration), cases of extravasation of the infusion solution, hypoesthesia, pain, erythema, swelling, inflammation, irritation, phlebitis, and thrombophlebitis have been reported [29,39,145,168]. Other side effects that may occur with omadacycline are cardiovascular, with grouped clinical trials showing a frequency of over 2% of hypertension and <2% of tachycardia and atrial fibrillation. No adverse cardiovascular reactions are known for tigecycline. A study in healthy subjects showed that this compound has no significant effect on the QT interval [169].

At the haematological and lymphatic levels, tigecycline and omadacycline may lead to anaemia but have opposite effects on platelets, with tigecycline decreasing their number (thrombocytopenia) [170], compared to omadacycline, which may cause thrombocytosis [157]. Other common side effects of tigecycline are prolongation of partially activated thromboplastin time (aPTT) and prothrombin time. However, an increase in the international normalised ratio (INR) is less common [31].

The primary concern with tigecycline is the increased mortality due to its use, compared with other anti-infective agents [73]. The results obtained in phases 3 and 4 of 13 clinical trials alerted the FDA, which issued a black box warning about the increased risk of mortality of patients treated with this drug [171]. Consequently, several meta-analyses of all controlled and randomised clinical trials were performed, concluding that tigecycline is not indicated in severe infections and should only be reserved for use in situations where alternative treatments are not appropriate [170,172–174]. Because these studies also found higher rates of clinical failure, superinfections, and septic shock compared to the comparator group, several hypotheses were postulated in an attempt to find a cause. These could be low efficacy, low plasma concentrations (could explain persistent bacteremia), and low alveolar concentrations (partly explains the low efficacy in patients with pneumonia associated with mechanical ventilation) [175,176]. In addition, on the basis of animal studies (rats), eravacycline and omadacycline have been shown to have undesirable effects on fertility, affecting sperm production, maturation, morphology, and motility [39,145].

During pregnancy, tetracyclines are not recommended, due to fetotoxicity and teratogenicity (tigecycline, according to the recommendations given by the FDA is classified as risk category D). Tetracyclines are used only when the benefit to the mother outweighs the potential risk for the foetus [171]. The results of animal studies indicate that tetracyclines cross the placenta, reach therapeutic concentrations in the foetal circulation, and may have toxic effects on foetal growth (often related to delayed skeletal development) [43,177]. Cases of embryotoxicity in animal models treated at the beginning of pregnancy were also highlighted. In addition, tetracyclines are excreted in human milk. Although the rate of absorption for infants is unknown, it is recommended that they not be used in breastfeeding either to avoid the risk of tooth discolouration and damage to osteogenesis [31,39,46,145].

The absorption of modern oral tetracyclines (omadacycline, sarecycline), similar to older members of this class, may be affected by the concomitant use of multivitamins; antacids (containing aluminium, calcium); or those in the composition of which magnesium, iron, and/or zinc are found. In these situations, non-absorbable chelating complexes are formed. Tetracyclines may also increase the anticoagulant effect of warfarin [13]. Several reports note decreased coagulation efficiency and bleeding in some patients after starting tetracycline therapy [38]. In a pharmacokinetic-pharmacodynamic study, tigecycline decreased the clearance of warfarin, and therefore careful monitoring of anticoagulant levels is indicated by more frequent international normalised ratio (INR) and prothrombin time (PT) [31,52,178]. The use of tetracyclines may decrease the effectiveness of oral contraceptives, but this interaction is quite controversial due to limited information [38,179]. There is no clinically significant effect of sarecycline on the efficacy of oral contraceptives containing ethinyl estradiol and norethindrone acetate [46]. Other interactions recorded in the literature are increased serum digoxin concentration, interference with penicillin activity, and a synergistic effect with oral retinoids on increased intracranial pressure [38,46].

#### **6. New Compounds under Development**

Sriram et al. (2007) synthesised tetracycline derivatives with anti-HIV, antimycobacterial, and HIV-1 integrase inhibitory properties. This was achieved by the reaction between certain tetracyclines (minocycline, tetracycline, and oxytetracycline), formaldehyde, and the secondary amine function (piperazine) of some fluoroquinolones (norfloxacin, lomefloxacin, ciprofloxacin, gatifloxacin), with the help of microwave radiation. Compound no. 10 (a tetracycline hybrid with lomefloxacin) represented in Figure 13 demonstrated the most potent effect on HIV-1 replication. These studies show that the combination of tetra-

cyclines with fluoroquinolones has resulted in both anti-HIV and anti-tuberculosis activity (*Mycobacterium tuberculosis*) and has a promising prospect in treating AIDS [180,181]. *Pharmaceutics* **2021**, *13*, 2085 22 of 31

**Figure 13.** Promising derivatives of tetracyclines. **Figure 13.** Promising derivatives of tetracyclines.

Currently, Tetraphase Pharmaceuticals holds two compounds, phase I of clinical trials, TP-271 and TP-6076, whose structures are represented in Figure 13 [183–186]. Thanks to a research program opened in the mid-1990s, synthetic routes with increased scalability Using total synthesis, Sun et al. (2015) projected a series of tetracycline analogues with six fused rings called hexacyclines. Their structure consists of the classical skeleton of tetracyclines, having attached a bicyclic ring EF at the level of ring D (Figure 13) [170].

and efficacy of obtaining tetracycline analogues have been discovered. To date, more than 3000 analogues have been synthesised using these methods, including the two previously mentioned [187]. TP-271 is a new, clinically developing fluorocycline with promising activity against bacteria that cause respiratory infections, community-acquired pneumonia, anthrax, bubonic plague, and tularemia [187]. Following both in vivo studies and in vivo evaluations, TP-271 has shown an increased potential against susceptible and multidrugresistant pathogens associated with moderate to severe community-acquired pneumonia. These include key bacteria in respiratory infections, *Streptococcus pneumoniae* (MIC90 = 0.03 The relationships between chemical structure and antibacterial activity were tested, with substitutions in positions C7, N8, C9, and C10, evaluating the efficacy of various analogues on a wide range of Gram-positive and Gram-negative bacteria, including tetracycline-resistant or multidrug-resistant strains. Of all the compounds studied, the best results were recorded for C7-fluorohexacycline and C7-trifluoromethoxyhexacycline, with a broad antibacterial activity in vitro and good activity in vivo on *Pseudomonas aeruginosa*. The promising data extracted from this study support the optimisation of this type of skeleton to discover and obtain in the future some new tetracyclines that are clinically valuable [182].

μg/mL), methicillin-sensitive *Staphylococcus aureus* (MIC90 = 0.25 μg/mL), methicillin-resistant *Staphylococcus aureus* (MIC90 = 0.12 μg/mL), *Streptococcus pyogenes* (MIC90 = 0.03 μg/mL), *Moraxella catarrhalis* (MIC90 ≤ 0.016 μg/mL), and *Haemophilus influenzae* (MIC90 = 0.12 μg/mL) [185]. TP-271 has also shown strong activity against important pathogens: *Yersinia pestis*, *Bacillus anthracis*, *Francisella tularensis, Burkholderia mallei*, and *Burkholderia pseudomallai* [188]*.* Furthermore, TP-271 has been shown to be effective in animal studies of immunocompetent pneumonia and neutropenia with *Streptococcus pneumoniae*, MRSA, and *Haemophilus influenzae*. Regarding the mechanism of action, this compound binds to the 30S ribosomal subunit and maintains its activity, even in the presence of the ribosomal protective protein Tet (M). Therefore, due to the positive results obtained on animal models and the broad spectrum, TP-271 is a promising candidate for treating moderate to severe community-acquired pneumonia [185]. Tetraphase Pharmaceuticals investigated analogues of C4, C7, and C8 trisubstituted tetracyclines. Thereby, some of these tetracyclines have demonstrated increased in vitro potency against clinically significant pathogens, including *Acinetobacter baumannii* (MIC90 = 0.063 μg/mL) and carbapenem-resistant Enterobacteriaceae (MIC90 = 0.5 μg/mL). The C4 positioned diethyl-amine analogue, TP-6076 (Figure 13), currently in phase I of clinical trials, showed the highest activity of all the three-substituted compounds analysed. The phase I clinical trial is focused on pharmacokinetics and safety studies to assess the bron-Currently, Tetraphase Pharmaceuticals holds two compounds, phase I of clinical trials, TP-271 and TP-6076, whose structures are represented in Figure 13 [183–186]. Thanks to a research program opened in the mid-1990s, synthetic routes with increased scalability and efficacy of obtaining tetracycline analogues have been discovered. To date, more than 3000 analogues have been synthesised using these methods, including the two previously mentioned [187]. TP-271 is a new, clinically developing fluorocycline with promising activity against bacteria that cause respiratory infections, community-acquired pneumonia, anthrax, bubonic plague, and tularemia [187]. Following both in vivo studies and in vivo evaluations, TP-271 has shown an increased potential against susceptible and multidrug-resistant pathogens associated with moderate to severe community-acquired pneumonia. These include key bacteria in respiratory infections, *Streptococcus pneumoniae* (MIC<sup>90</sup> = 0.03 µg/mL), methicillin-sensitive *Staphylococcus aureus* (MIC<sup>90</sup> = 0.25 µg/mL), methicillin-resistant *Staphylococcus aureus* (MIC<sup>90</sup> = 0.12 µg/mL), *Streptococcus pyogenes* (MIC<sup>90</sup> = 0.03 µg/mL), *Moraxella catarrhalis* (MIC<sup>90</sup> ≤ 0.016 µg/mL), and *Haemophilus influenzae* (MIC<sup>90</sup> = 0.12 µg/mL) [185]. TP-271 has also shown strong activity against important pathogens: *Yersinia pestis*, *Bacillus anthracis*, *Francisella tularensis, Burkholderia mallei*, and *Burkholderia pseudomallai* [188]. Furthermore, TP-271 has been shown to be effective in animal studies of immunocompetent pneumonia and neutropenia with *Streptococcus pneumoniae*, MRSA, and *Haemophilus influenzae*. Regarding the mechanism of action,

chopulmonary disposition of intravenous TP-6076 in healthy subjects, in order to assess the potential utility in *Acinetobacter baumannii* pneumonia. According to Tetraphase, TP-6076 is in development for the treatment of serious and life-threatening bacterial infec-

this compound binds to the 30S ribosomal subunit and maintains its activity, even in the presence of the ribosomal protective protein Tet (M). Therefore, due to the positive results obtained on animal models and the broad spectrum, TP-271 is a promising candidate for treating moderate to severe community-acquired pneumonia [185].

Tetraphase Pharmaceuticals investigated analogues of C4, C7, and C8 trisubstituted tetracyclines. Thereby, some of these tetracyclines have demonstrated increased in vitro potency against clinically significant pathogens, including *Acinetobacter baumannii* (MIC<sup>90</sup> = 0.063 µg/mL) and carbapenem-resistant Enterobacteriaceae (MIC<sup>90</sup> = 0.5 µg/mL). The C4-positioned diethyl-amine analogue, TP-6076 (Figure 13), currently in phase I of clinical trials, showed the highest activity of all the three-substituted compounds analysed. The phase I clinical trial is focused on pharmacokinetics and safety studies to assess the bronchopulmonary disposition of intravenous TP-6076 in healthy subjects, in order to assess the potential utility in *Acinetobacter baumannii* pneumonia. According to Tetraphase, TP-6076 is in development for the treatment of serious and life-threatening bacterial infections. The potency of this compound on Gram-negative, multidrug-resistant bacteria is 2 to 64 times higher than that of tigecycline. TP-6076 also retains a high efficacy against isolates with intrinsic resistance mechanisms that generally affect the tetracycline class (e.g., carbapanemase-producing Enterobacteriaceae and carbapanemase-producing *Acinetobacter baumannii*). Antibacterial activity was not affected by the type of carbapenem resistance determinant or international clone. The increased in vitro potency also resulted in high in vivo efficiency in models of mice infected with resistant Gram-negative multi-drug isolates [185,189,190].

#### **7. Conclusions**

The evolution of the tetracycline class is remarkable, through its development of semisynthetic analogues of the second generation, and, more recently, of the third generation. The new tetracyclines had acquired high potency and increased efficacy, even against resistant bacteria to tetracyclines. On the basis of the classical method of production, biosynthesis, and semi-synthesis, we are able to obtain the new compounds by total chemical synthesis, such as eravacycline.

The main focus in optimising the chemical structure was on modifying the C7 and C9 positions of the D ring in the simplest tetracycline with biological activity (sancycline). Thus, a new tetracycline class was discovered, based on C9-aminotetracyclines, which bear a glycyl moiety known as glycylcyclines. First in class was the tigecycline representative. Recently approved tetracyclines include beside tigecycline, omadacycline (an aminomethylcycline), eravacycline (a fluorocycline), and sarecycline (a 7-[(methoxy- (methyl)-amino)-methyl]methyl] derivative).

Tigecycline has the advantage of a superior potency over Gram-positive and Gramnegative MDR bacteria; its pharmaceutical formulation is only parenteral. Omadacycline has a broad spectrum of activity, including MRSA, penicillin-resistant, MDR *Streptococcus pneumoniae* and vancomycin-resistant enterococci. This new tetracycline drug is more advantageous in therapy than tigecycline because it can be administered orally and parenterally. Eravacycline is a synthetic fluorocycline with a great activity against Grampositive and Gram-negative bacteria that developed specific resistance mechanisms to tetracyclines. Similar to tigecycline, eravacycline is administered exclusively parenterally. The main advantage of sarecycline is the narrow-spectrum activity and the higher selective activity against *Cutinebacterium acnes*. Sarecycline is available as an oral formulation to treat inflammatory lesions of moderate-to-severe non-nodular acne vulgaris.

Although tetracyclines currently act bacteriostatically, omadacycline has demonstrated bactericidal activity in vitro against some bacterial agents. It was proven that glycylcyclines manifest resistance to less common mechanisms, such as altered target site conformation, enzymatic degradation, and mutations in DNA gyrase. Therefore, eravacycline was designed to maintain its activity against resistant bacteria. Sarecycline expands and establishes uncommon interactions with mRNA.

Currently, some studies confirm other biological effects in tetracyclines class that require in-depth future studies. Therefore, through its newly acquired members, this class of antibiotics arouses the interest of researchers in the field. Consequently, new derivatives have been already developed, and many are in development. These are studied primarily studied for the antibiotic effect and also for other biological effects.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/pharmaceutics13122085/s1, Table S1. The representatives of tetracyclines class from first and second generations and their approval in therapy (FDA—USA Food and Drug Administration, EMA—European Medicine Agency, MHRA—UK Medicines and Healthcare Products Regulatory Agency). Table S2. Modern tetracyclines of the third generation introduced in therapy. Table S3. Essential structural features of tetracycline antibiotics. Table S4. Physicochemical properties of thirdgeneration tetracyclines. Table S5. The antibacterial spectrum of the newly approved tetracyclines. Table S6. Pharmacokinetics parameters of modern tetracyclines.

**Author Contributions:** Conceptualisation, methodology, writing—review and editing, supervision, A.R.; writing—original draft preparation, visualisation, E.L.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **Ruthenium Complexes in the Fight against Pathogenic Microorganisms. An Extensive Review**

**Alexandra-Cristina Munteanu \* and Valentina Uivarosi \***

Department of General and Inorganic Chemistry, Faculty of Pharmacy, "Carol Davila" University of Medicine and Pharmacy, 020956 Bucharest, Romania

**\*** Correspondence: alexandra.ticea@umfcd.ro (A.-C.M.); valentina.uivarosi@umfcd.ro (V.U.)

**Abstract:** The widespread use of antibiotics has resulted in the emergence of drug-resistant populations of microorganisms. Clearly, one can see the need to develop new, more effective, antimicrobial agents that go beyond the explored 'chemical space'. In this regard, their unique modes of action (e.g., reactive oxygen species (ROS) generation, redox activation, ligand exchange, depletion of substrates involved in vital cellular processes) render metal complexes as promising drug candidates. Several Ru (II/III) complexes have been included in, or are currently undergoing, clinical trials as anticancer agents. Based on the in-depth knowledge of their chemical properties and biological behavior, the interest in developing new ruthenium compounds as antibiotic, antifungal, antiparasitic, or antiviral drugs has risen. This review will discuss the advantages and disadvantages of Ru (II/III) frameworks as antimicrobial agents. Some aspects regarding the relationship between their chemical structure and mechanism of action, cellular localization, and/or metabolism of the ruthenium complexes in bacterial and eukaryotic cells are discussed as well. Regarding the antiviral activity, in light of current events related to the Covid-19 pandemic, the Ru (II/III) compounds used against SARS-CoV-2 (e.g., BOLD-100) are also reviewed herein.

**Keywords:** ruthenium; antimicrobial; antibacterial; antiviral; antiparasitic; COVID-19

#### **1. Introduction**

The alarming pace at which microorganisms are evading antibiotics constitutes a challenge for modern medicine [1]. The phenomenon of multidrug resistance has generated a sense of urgency around the development of new classes of antibiotics. Yet most of the drugs under clinical development for the treatment of bacterial infections are organic derivatives of currently used antibiotics, which suggests that these molecules are susceptible to in place mechanisms of bacterial resistance [2].

Although the pipeline for new antibiotics is running dry, the coordination chemistry field is still largely underexplored for antibacterial drug development, with limited clinical use for bismuth and silver-based antimicrobials. Bismuth compounds, for instance, are used for the treatment of *H. pylori* infections and diarrhea and in wound dressings [3], while silver compounds are used for wound healing applications and management of topical infections [4]. The focus of current research is directed towards the development of metal-based nanoparticles (NPs), with special interest being given to AgNPs following their introduction to the U.S. market in 2016 [5].

It is rather unfortunate that less attention is being given to metal complexes. It should be noted that metal-based compounds offer a vast structural diversity of three-dimensional (3D) scaffolds due to the variety of metal ions, ligands, and possible geometries [2,6,7]. While most organic fragments have linear (1D) or planar (2D) shapes, more complex 3D fragments are desirable for the molecular recognition by biomolecules and optimal interaction with intracellular targets [6]. Furthermore, increasing the 3D chemical topology of molecules has been correlated with a broader activity spectrum [7,8]. Therefore, metal

**Citation:** Munteanu, A.-C.; Uivarosi, V. Ruthenium Complexes in the Fight against Pathogenic Microorganisms. An Extensive Review. *Pharmaceutics* **2021**, *13*, 874. https://doi.org/ 10.3390/pharmaceutics13060874

Academic Editor: Carlos Alonso-Moreno

Received: 1 May 2021 Accepted: 9 June 2021 Published: 13 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

complexes are ideal candidates for future drug discovery pursuits meant to access the underexplored 3D chemical space [6]. In addition, metal complexes possess unique mechanisms of action that are not readily available to organic compounds: ROS generation, redox activation, ligand exchange, and depletion of substrates involved in vital cellular processes [2,9,10]. When compared with solely organic molecules, metal-based compounds were found to display a significantly higher hit-rate against critical antibiotic-resistant pathogens (0.87% vs. 9.9%). Moreover, the percentages of toxic to healthy eukaryotic cells and/or hemolytic compounds in the two groups were found to be nearly identical. Therefore, a generally higher degree of toxicity cannot explain the remarkably high antimicrobial activity of the metal-based set of compounds compared with the organic molecules [2].

The potential of metal complexes has been acknowledged over the last two decades through several platinum-, ruthenium-, copper-, iron-, and gallium-based drugs, which have reached different stages in clinical trials for the treatment of cancer, neurodegenerative diseases, and malaria [11,12]. Several ruthenium (Ru) complexes have been evaluated in clinical trials for the treatment of cancer, namely NAMI-A [13,14], KP1019 [15,16] and its water-soluble sodium salt IT-139 (formerly KP1339) [17], and, more recently, TLD-1433 [18]. Previous knowledge of their chemical properties and biological behavior, gained from the research directed towards the development of novel anticancer compounds, has led to increased focus on tailoring ruthenium complexes as antimicrobial agents [1]. Moreover, a recent study screening 906 metal-containing compounds for antimicrobial activity identified ruthenium as the most frequent element found in active compounds that are nontoxic to eukaryotic cells, followed by silver, palladium, and iridium [2]. Therefore, ruthenium-based compounds hold promise for potential antimicrobial applications, which will be extensively reviewed in this paper.

In order to clarify the use of the terms 'antibacterial', 'antibiotic', and 'antimicrobial' in this manuscript, definitions are given below. The term antibacterial refers to substances, materials, or assemblies that kill or inhibit the growth of bacteria. WHO defines an antibiotic as a substance with a direct action on bacteria that is used for the treatment or prevention of infections or infectious diseases [19]. Although we recognize the distinction between these two terms, in order to avoid repetition, we have occasionally used the terms 'antibiotic' and 'antibacterial' interchangeably. Antimicrobials, on the other hand, will be used generically for compounds or materials that act against microorganisms (bacteria, fungi, viruses, protozoa, parasites, etc.). Consequently, antimicrobials will include antibacterials, antifungals, antivirals, antiprotozoals, and antiparasitics.

#### **2. General Remarks on Bacterial Cell Structure. Gram-Positive vs. Gram-Negative Strains**

The bacterial cell structure comes as a result of the extreme conditions they must survive in, which are inhospitable for eukaryotes. For instance, the rigid cell wall that covers the cell membrane is vital for protection from physical, chemical, and mechanical stressors. Based on the Gram staining procedure, bacteria are classified into two groups: Gram-positive and Gram-negative bacteria [1].

Gram-positive strains retain the Crystal Violet stain due to the presence of a thick layer of peptidoglycan in their cell walls, which is densely embedded with negatively charged glycopolymers called wall teichoic acids (Figure 1). The fairly porous cell wall structure generally allows for passage for exogenous molecules into the bacterial cells [20].

Gram-negative bacteria, however, have more complex cell wall structures (Figure 1). Due to the absence of inlaid teichoic acid molecules, their layer of peptidoglycan is thin, yet bound to an outer membrane coated with lipopolysaccharides (LPSs). LPSs are amphiphiles, consisting of a hydrophobic lipidic domain (lipid A) covalently bound to a polysaccharide, which comprises the O antigen and the inner and outer cores; these negatively charged (due to the presence of the phosphate and acid groups) macromolecules are stabilized by divalent cations such as calcium and magnesium. LPSs greatly decrease bacterial permeability to antibiotics and play a crucial role in the development of resistance mechanisms for many pathogenic Gram-negative bacteria [1,20].

Additionally, on the cell surface of some bacteria (e.g., *Streptococcus pneumoniae*) a slime layer or a capsule can offer additional protection against desiccation or phagocytosis by host cells. Flagella, fimbriae, and pili are external filamentous appendages that serve as organelles of locomotion or assist with bacterial attachment and adhesion to a surface or genetic exchange [1,21].

**Figure 1.** Comparison between Gram-negative and Gram-positive bacteria cell walls. Adapted from [22] with permission. Copyright © 2020 Huan, Kong, Mou and Yi.

At physiological pH, the high content of zwitterionic phosphatidylcholine confers an overall neutral charge to the eukaryotic cell membranes. In contrast, bacterial outer cell walls and membranes are usually negatively charged due to the presence of negatively charged components (phospholipids, teichoic acids, and lipopolysaccharides) [1,23]. Hence, in order to increase selectivity, new antibacterial drugs (including ruthenium complexes) are generally designed so as to possess a cationic component.

#### **3. Mechanisms of Action of Current Drugs**

Antibiotics are classified into four major groups (Figure 2), based on their intracellular target and mechanism of action: (1) inhibition of bacterial cell wall synthesis (penicillin and its derivatives, cephalosporins, carbapenems, and glycopeptides—these drugs are more active against Gram-positive bacteria); (2) disruption of bacterial membranes (polymyxins these are active against Gram-negative bacteria and considered a last-line therapy against Gram-negative 'superbugs'); (3) inhibition of nucleic acid synthesis (quinolones, rifampicin, and sulphonamidesare—these are broad-spectrum synthetic antibiotics); and (4) inhibition of protein synthesis (tetracycline, aminoglycosides, chloramphenicol, and macrolides these inhibit protein synthesis by targeting the RNA-rich surfaces of ribosomes) [1].

**Figure 2.** Mechanisms of action of currently used antibiotics (Image by Kendrick Johnson, licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license).

Several new classes of antibiotics have been discovered over the last two decades. Gepotidacin, for instance, belongs to a new chemical class of antibiotics called triazaacenaphthylene. It is a topoiosomerase inhibitor, which is currently being investigated in a phase III clinical study in patients with uncomplicated urinary tract infection and urogenital gonorrhoea [24]. Other current strategies include the use of phages (viruses that kill specific bacterial strains) [25], various types of engineered nanoparticles [25], and cationic materials, including cationic polypeptides, polymers, copolymers, and dendrimers [26]. Furthermore, several natural products, e.g., teixobactin, have been identified as lead compounds in the fight against antimicrobial resistance [27].

#### **4. Mechanisms of Resistance to Antibiotics**

Bacterial resistance to antibiotics can result from intrinsic or acquired antibioticresistant mechanisms. *P. aeruginosa* and other Gram-negative pathogens are intrinsically more resistant to antibiotics due to the reduced permeability of their outer membranes. These bacterial strains have porins of unusually low permeability. In addition, the outer membranes of mycobacteria have a high lipid content that allows for hydrophobic drugs such as fluoroquinolones to enter the cell but limits the access of hydrophilic drugs.

Acquired bacterial resistance is caused by alterations in microorganisms that result in drug inactivation or a decrease in therapeutic efficacy. Improper prescribing and overuse of antibiotics are factors that have contributed to the growing issue of microbial resistance. Consequently, infections have become increasingly difficult or even impossible to treat [28].

Bacterial resistance can emerge as a result of various biochemical mechanisms, including decreased drug uptake, modification of a specific bacterial target, enzymatic inactivation of the drug, and modifications to the bacterial efflux systems [1,28]. For instance, a common resistance mechanism is the alteration of the bacterial membrane permeability, resulting in limited uptake of an antibiotic. Modification of the drug's target can involve mutations in DNA gyrase and topoisomerase IV or alterations in the structure and/or number of penicillin-binding proteins [5]. Drug inactivation occurs via mutations in genes coding for key enzymes, such as β-lactamases, acetyltransferases, adenylyltransferases, and aminoglycoside-30 -phosphotransferase. These mutations can occur either inside the bacterial chromosomal DNA or as a result of foreign genetic material acquisition. Acquisition of genetic material that confers resistance is possible through horizontal gene transfer, which is mediated either by plasmids or bacteriophages [28].

Another common mechanism of resistance used by many pathogens involves the association of multiple bacterial cells in matrices called biofilms. The bacterial cells within the biofilm have a slow metabolism rate and slow cell division. Therefore, antimicrobials targeting growing and dividing bacterial cells are rendered ineffective. Moreover, the thick biofilm extracellular matrix consists of bacterial polysaccharides, proteins, and DNA, which hinder access of the antimicrobial agent to the bacteria. It is also likely that the proximity of the bacterial cells facilitates horizontal gene transfer. Therefore, the antimicrobial resistance genes can be shared between the cells forming the biofilm [28–30].

Nosocomial infections or hospital-acquired infections are a growing threat worldwide and are often caused by multidrug-resistant bacteria. Interestingly, a small group of microorganisms, known as ESKAPE pathogens, are responsible for most antibiotic-resistant infections. These pathogens include: *Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa*, and *Enterobacter* spp., which possess innate resistance or can acquire resistance against multiple antibiotics [31].

#### **5. Antibacterial and Antifungal Activities of Ruthenium Complexes**

Based upon their chemical stability, Ru complexes can be classified as either stable, relatively inert compounds, and prodrugs. A metal complex is inert when the ligand framework remains unaltered in biological media. The ruthenium ion in these compounds acts merely as a central scaffold that carries the bioactive ligands to their target. Consequently, the properties of the coordinated ligands are essential to the antibacterial activity [32]. The presence of the ruthenium ion, however, provides the molecule with a positive charge, which aids in targeting the negatively charged cell wall structures of bacteria. The antibacterial activity of these complexes depends on their lipophilicity and charge, which in turn shape their ability to interact with specific targets (e.g., DNA, RNA, proteins, bacterial membranes).

Prodrugs are labile complexes that release the ligand/s when exposed to solvents and/or media and generate species that can bind to various biological targets or photoactivated drugs. The latter become active upon light irradiation and act as photosensitizers. Since this behavior is somewhat unconventional for the general understanding of the term 'prodrug' in the traditional medicinal chemistry sense, 'prodrug-like molecules' seems more appropriate to describe this type of metal complex. In the case of labile complexes, active species are released as a result of either partial or total ligand exchange in biological media. These active species are either ruthenium species resulting from ligand exchange with media components or the released ligands. In the latter case, the ruthenium compounds are called 'carrier' complexes; one such example is the Ru(II) chelate–chloroquine complex, [RuCl2(CQ)]2, where CQ = chloroquine (see 6. Antiparasitic activity of ruthenium complexes). In the following sections, ruthenium complexes will be classified based on their structure. Details and comments with regard to their mechanisms of action will be provided wherever such information is available.

#### *5.1. Mononuclear Ruthenium (II) Complexes*

Mononuclear polypyridylruthenium (II) complexes with antimicrobial activities were first reported in the 1950s and 1960s by Dwyer et al. [33,34]. With the general interest shifting towards discovering new analogues of existing classes of antibiotics, their impressive seminal work was unfortunately not further pursued. However, the advancement into clinical trials of NAMI-A, KP1019, and TLD1433 for the treatment of cancer and the urge to develop new classes of antibiotics have led, over the last two decades, to an increased focus on research and development of ruthenium-based antimicrobials [35].

Dwyer et al. made the first steps towards the development of kinetically inert Ru(II) complexes and the study of their in vitro and in vivo antimicrobial activities. The addition of methyl groups to the phenanthroline ligands enhanced lipophilicity and increased the activity of [Ru(Me4phen)3] 2+ (Figure 3) against Gram-positive bacteria, as compared with [Ru(phen)3] 2+ (Figure 3) [36]. More recent studies [37,38], however, have shown that these

complexes are much less active against various antibiotic-resistant ESKAPE pathogens. Additionally, their activity in vivo has been proven to be unsatisfactory, as they caused severe neurotoxic effects when injected into mice [39].

**Figure 3.** Examples of inert structural mononuclear polypyridylruthenium (II) complexes.

Following up on this remarkable work, various heteroleptic mononuclear polypyridyl Ru (II) complexes were tested for antibacterial activity. Their activities (MIC values) against various bacterial strains, as well as toxicity towards healthy eukaryotic cells and modes of action, where available, are listed in Table 1.

**Table 1.** Activities of selected ruthenium complexes against bacteria, toxicity to healthy mammalian cells, and mode of action.





#### 5.1.1. Mononuclear Polypyridyl Ru (II) Complexes

R-825 (Figure 3) was shown to interfere with the iron acquisition systems in *S. pneumoniae*, which led to a dramatic decrease in intracellular iron, correlated with a bactericidal effect. In addition, R-825 was essentially non-toxic to human A549 non-small-cell lung cancer cells in vitro [41]. Iron is an essential nutrient for the development and survival of bacteria, as well as a key factor in host infection. In order to scavenge iron from their surroundings, bacteria make use of highly effective iron acquisition systems. In *S. pneumoniae*, the ABC transporters PiaABC, PiuABC, and PitABC play a major role in the acquisition of

heme, ferrichrome, and ferric irons, respectively [72]. The deletion of the *piuA* gene in a mutant strain of *S. pneumoniae* resulted in a significant decrease in ruthenium uptake, leading to an increased resistance of the mutant to R-825 treatment. These results suggest that the mechanism of uptake for R-825 appears to involve active transport via the PiuABC iron uptake pathway [41]. Note that this mechanism of uptake is different than those used by the currently approved antibiotics. Generally, due to the chemical similarity between iron and ruthenium, the ability of novel antibiotics to interfere with iron acquisition systems in bacteria (including ABC transporters) is considered to be a viable strategy for the discovery of new antibacterial drugs.

A variety of mononuclear heteroleptic polypyridyl ruthenium (II) chelates bearing bpy, phen, dmp (4,40 -dimethyl-2,20 -bipyridine), or hdpa (2,2'-dipyridylamine) and other mono/bidentate ligands were active in various degrees against Gram-positive and Gramnegative bacteria and fungi [73–81]. Although their mechanisms of action have not been determined, all complexes were shown to interact with DNA duplexes and several exerted photoactivated cleavage of plasmid DNA in vitro [75,77,79–81] with singlet oxygen (1O2) probably playing a significant role in the cleavage mechanism.

Mononuclear Ru(II) Heteroleptic Complexes Bearing 2,2'-Bipyridine (bpy) Ligands

Numerous octahedral heteroleptic Ru(II) complexes containing 2,2'-bipyridine (bpy), with the general formula [Ru(bpy)2L]Y<sup>n</sup> (where L = a mono/bidentate ligand, note that when L is monodentate, the first coordination sphere of Ru(II) is saturated with chloride ions; Y = counterion) have been synthesized and tested against bacteria. Generally, these complexes showed moderate to high activity on Gram-positive bacteria, but were inactive against Gram-negative strains. X-03 (Figure 4), for instance, was active against several Gram-positive bacteria, *S. pneumoniae*, *Listeria monocytogenes*, and *S. aureus*, but showed no toxicity at the tested concentrations against Gram-negative microorganisms. X-03 appears to interfere with iron acquisition systems in *S. pneumoniae* cells, in a similar manner to R-825. Proteomic data revealed that X-03 caused the downregulation of several proteins involved in oxidative stress response and fatty acid biosynthesis, suggesting a mechanism of action based on increased susceptibility to oxidative stress and membrane damage. Additionally, X-03 displayed low toxicity even at a concentration 8 times higher than the MIC value to the A549 alveolar and HBE bronchial epithelial cell lines, indicating selective toxicity against bacteria [42].

Complexes with photolabile ligands, in which L is unidentately coordinated, L = 4- (4-chlorobenzoyl)pyridine (clbzpy), Y = PF6−, *<sup>n</sup>* = 1 ([Ru(bpy)2Cl(clbzpy)]<sup>+</sup> , Figure 4), was moderately active against *S. aureus* and *S. epidermidis*. Additionally, the complex was shown to suffer blue light photolysis (453 nm) in aqueous solution and the resulting photoproduct, *cis*-[Ru(bpy)2(H2O)Cl]<sup>+</sup> , displayed high binding affinity towards DNA in vitro. The antibacterial activity, however, was not influenced by blue light irradiation, which indicates that the antibacterial activity is not due to DNA damage, but might be the result of bacterial membrane disruption [43]. Blue LED irradiation, however, has been shown to enhance the activity of [Ru(bpy)2(methionine)]2+, albeit not drastically, against *S. aureus* and *S. epidermidis* [44]. Methionine release and subsequent exchange with water molecules via photolysis at 453 and 505 nm in aqueous solution lead to *cis*- [Ru(bpy)2(H2O)2] 2+, which can bind covalently to double-stranded DNA [44,82] and promote photocleavage [44].

**Figure 4.** Chemical structures of heteroleptic Ru(II) complexes bearing 2,2'-bipyridine (bpy) ligands. BTPIP = (2-(4-(benzo[b]thiophen-2-yl)phenyl)-1*H*-imidazo [4,5-*f*][1,10]phenanthroline); ETPIP = 2-(4-(thiophen-2-ylethynyl)phenyl)-1*H*-imidazo[4,5-*f*][1,10]phenanthroline); CAPIP = *(E)*-2-(2-(furan-2-yl)vinyl)-1*H*-imidazo[4,5-*f*][1,10]phenanthroline; dmp = 4,4'-dimethyl-2,2'-bipyridine; bpy = 2,2' bipyridine; phen = 1,10-phenanthroline.

[Ru(bpy)2L]Y<sup>n</sup> complexes, where L = BTPIP, ETPIP, CAPIP, Y = ClO4−, *n* = 2, [Ru(dmb)<sup>2</sup> (ETPIP)]2+, and [Ru(phen)2(ETPIP)]2+ (see Figure 4 for the chemical structures and the IUPAC names of the ligands) displayed good activities against drug-susceptible *S. aureus*. [Ru(bpy)2(BTPIP)]2+ was the most active compound of the series (MIC = 0.016 mg/mL) and was shown to inhibit biofilm formation and, thus, prevent bacteria from developing drug resistance. [Ru(bpy)2(BTPIP)]2+ [46] and [Ru(phen)2(ETPIP)]2+ [45] increased the susceptibility of *S. aureus* to certain aminoglycosidic antibiotics (kanamycin and gentamicin). [Ru(phen)2(ETPIP)]2+ was found to suppress the gene regulatory activity of the catabolite control protein A (CcpA) in *S. aureus*, which can explain the synergistic effects observed for this complex and kanamycin [45]. Studies conducted on a murine skin infection model for Ru(bpy)2(BTPIP)]2+ showed that Ru(bpy)2(BTPIP)]2+ ointments were effective as topical products against skin infection [46]. These complexes, however, have proven to be cytotoxic to A549 cancer cell lines, with IC<sup>50</sup> values lower than those required for the antibacterial activity [83–86], which might indicate poor selectivity towards bacteria. To the extent of our knowledge, no cytotoxic tests on normal cell lines have been performed.

The corresponding ruthenium(II) bipyridine complex in which L = curcumin and Y = PF6<sup>−</sup> (Figure 5) was tested against various ESKAPE pathogens. It displayed bactericidal activity against methicillin and vancomycin-resistant *S. aureus* strains (MIC = 1 µg/mL) and high selectivity towards bacteria as compared with eukaryotic Vero cells (SI > 80). Moreover, the complex strongly inhibited biofilm formation in *S. aureus* cells and displayed in vivo antibacterial activity against *S. aureus* comparable to that of vancomycin in a murine

neutropenic thigh infection model. However, [Ru(bpy)2curcumin]<sup>+</sup> was not toxic to the Gram-negative *E. coli*, *K. pneumoniae*, *A. baumanii*, and *P. aeruginosa* cells. In comparison, the corresponding Ru(II) complex, [Ru(phen)2curcumin]<sup>+</sup> , bearing 1,10-phenanthroline (Figure 5), was also active against the Gram-negative *A. baumanii* with a MIC value comparable to that of levofloxacin, in addition to its activity on the Gram-positive *S. aureus* bacteria and lack of toxicity against eukaryotic cells [47].

**Figure 5.** [Ru(N-N)2curcumin]<sup>+</sup> , where N-N is either 2,2'-bypiridine (bpy) or 1,10-phenanthroline (phen).

Mononuclear Ru(II) Heteroleptic Complexes Bearing 1,10-phenanthroline (phen)

Mononuclear Ru(II) complexes bearing phenanthroline ligands have also been investigated as potential antibacterial agents. Amongst these complexes, mono-bb<sup>n</sup> ([Ru(phen)2bbn] 2+) (Figure 6), where bb<sup>n</sup> is bis[4(4'-methyl-2,2'-bipyridyl)]-1,*n*-alkane and *n* stands for the number of methylene groups in the alkane chain of bb<sup>n</sup> (*n* = 7 or 10), have been extensively investigated. Although mono-bb<sup>10</sup> has a larger alkane chain and therefore is more lipophilic, it was less active than mono-bb<sup>7</sup> against drug-susceptible *S. aureus* [38,87,88]. The bactericidal activity of mono-bb<sup>7</sup> was linked to the extent of cellular accumulation, since its activity on Gram-negative strains is low and the uptake in *Staphylococcus* strains is much higher than in *E. coli* or *P. aeruginosa* [37,38]. Mono-bb<sup>7</sup> caused membrane depolarization in *S. aureus* cells and increased membrane permeability, which might suggest the membrane damage as part of its mode of action [88]. Morphological changes indicative of membrane damage have also been reported for a similar complex, [Ru(phen)2(BPIP)]2+, where BPIP = 2-(40 -biphenyl)imidazo[4,5-*f*][1,10]phenanthroline (Figure 6), in Gram-positive (*Micrococcus tetragenus* and *S. aureus*) bacteria [76]. Mono-bb<sup>7</sup> displayed selective activity against bacterial over healthy mammalian cells [38,89].

**Figure 6.** Chemical structures of heteroleptic Ru(II) complexes bearing 1,10-phenanthroline (phen) ligands.

A complex in which the bb<sup>12</sup> ligand is tetradentately bound to Ru (II), *cis*-α-[Ru(phen)bb12] 2+ (Figure 7a, see for comparison the other isomers of the compound, depicted in Figure 7b,c), was found to be more active against the Gram-negative *P. aeruginosa* than the more lipophilic mono-bb7. The activity was found to be positively correlated with the uptake of the complex into the cells. Nonetheless, *cis*-α-[Ru(phen)bb12] 2+ was still considerably more active against Gram-positive bacteria as compared with *P. aeruginosa*, the compound being more active against MRSA than ampicillin and gentamicin. Interestingly, *cis*-α- [Ru(phen)(bb12)]2+ was found to be two to four times more active than its geometric isomer, *cis*-β-[Ru(phen)(bb12)]2+, against the Gram-negative strains (*E. coli* and *P. aeruginosa*), while no difference in activity was found for the Gram-positive bacteria (*S. aureus* and MRSA). It is unclear why the *cis*-α isomer is more active, since no significant difference in cellular accumulation was observed for the two isomers. Moreover, both geometric isomers were shown to bind tightly and with similar potency to duplex DNA in vitro, but no correlation between the binding constants and activity was found [48]. It should be noted that DNA/RNA binding is a possible mechanism of action for these complexes, since several reports indicate that various inert Ru(II) polypyridyl complexes bearing phenanthroline ligands target DNA and RNA in bacterial and eukaryotic cells [76,90,91]. Notably, the similar complex *cis*-α-[Ru(Me4phen)(bb7)]2+ displayed similar activity towards Gram-positive and Gram-negative bacteria as *cis*-α-[Ru(phen)(bb12)]2+ and remarkably high DNA binding affinity (~10<sup>7</sup> ) [92].

**Figure 7.** The ligand bb<sup>n</sup> and the possible isomeric forms of the mononuclear complex [Ru(phen)(bbn)]2+ with bb<sup>n</sup> as a tetradentate ligand: (**a**) *cis*-α isomer, (**b**) *cis*-β isomer, and (**c**) a form in which the central polymethylene chain spans the *trans*. Reproduced from [48] with permission. Copyright © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

Mononuclear Ru (II) Heteroleptic Complexes Bearing Pyridophenazine Ligands

[Ru(phen)2(dppz)]2+ (Figure 8), where dppz = dipyrido[3,2-*a*:2',3'-c]phenazine and phen = 1,10-phenanthroline, displayed good bactericidal activity against *M. smegmatis* (MIC = 2 µg/mL). Its mechanism of action was suggested to be linked to ROS generation and DNA intercalation [93]. A similar complex, [Ru(2,9-Me2phen)2(dppz)]2+, was active against MRSA and *B. subtilis*, and displayed time–kill curves that were similar to those of currently used antibiotics, but displayed no activity against *E. coli*. The activity appeared to be correlated with the ability to intercalate into DNA double strands in vitro. In vivo antibacterial activity has been assessed using the nematode *Caenorhabditis elegans* infection model and [Ru(2,9-Me2phen)2(dppz)]2+ proved to be non-toxic to the nematodes [40].

[Ru(bb7)(dppz)]2+ (Figure 8) (bb<sup>7</sup> = bis[4(4'-methyl-2,2'-bipyridyl)]-1,7-alkane) was 2–8 fold more active than its parent compound [Ru(phen)2(dppz)]2+ against both Grampositive (*S. aureus*, MRSA) and Gram-negative bacteria (*E. coli*, *P. aeruginosa*). Although the two complexes have comparable lipophilicity, [Ru(bb7)(dppz)]2+ accumulated in *P. aeruginosa* to the same degree as in MRSA and was shown to permeabilize a model membrane system to a higher degree than [Ru(phen)2(dppz)]2+. Therefore, its higher cellular uptake might be responsible for the increase in activity. However, Ru(bb7)(dppz)]2+ was also ~3-fold more toxic to healthy eukaryotic cells than [Ru(phen)2(dppz)]2+, while still being more active against bacterial cells [49].

Complexes bearing tetrapyridophenazine (tpphz) are more lipophilic relative to their dppz analogues and generally more active. For instance, the luminescent, mononuclear ruthenium(II) complex bearing the tpphz ligand, [Ru(Me4phen)2(tpphz)]2+ (Figure 8), displayed a comparable activity to that of ampicillin and oxacillin in drug-sensitive strains and the activity was retained in resistant strains. The complex was taken up by both Gram-positive (*E. faecalis*, *S. aureus*) and Gram-negative (*E. coli*, *A. baumannii*, *P. aeruginosa*) bacteria in a glucose-independent manner and was shown to target chromosomal DNA in both Gram-positive and Gram-negative strains. Moreover, model toxicity screens showed that the compound is non-toxic to *Galleria mellonella* larvae at concentrations that are 3– 25 times higher than the MIC values [50]. This complex represents the starting point for the kinetically inert dinuclear polypyridylruthenium(II) complex [Ru2(Me4phen)2(tpphz)]4+ (see below), which displayed higher antibacterial activity (Table 1), except against *S. aureus*. Unlike the dinuclear derivative, [Ru(Me4phen)2(tpphz)]2+ does not cause membrane damage.

**Figure 8.** Chemical structures of heteroleptic Ru (II) complexes bearing pyridophenazine ligands.

#### 5.1.2. Mononuclear Ru (II)–arene Complexes

Due to the promising anticancer activities of some representatives, the potential antibacterial properties of piano-stool Ru(II)-η <sup>6</sup>–arene complexes, with the general structure shown in Figure 9, have also been considered for antimicrobial applications [94–103]. While some of them displayed modest activity [76,79,80], complexes of the general formulae [Ru(η 6 -*p*-cymene)X2(PTA)] (RAPTA-C complexes), where X = Cl, Br, I, NCS (labile) and PTA = 1, 3, 5-triaza-7-phosphaadamantane, were active in different degrees against bacteria (*E. coli*, *B. subtilis*, *P. aeruginosa*) and fungi (*Candida albicans*, *Cladosporium resinae,* and *Trichrophyton mentagrophytes*). The PTA ligand was suggested to play a role in facilitating the uptake of the complex into bacterial cells, while the antimicrobial activity was suggested to be mediated by the interaction of the Ru(II) ion with intracellular proteins. Although the complexes were found to cause DNA damage in vitro, their affinity towards DNA was not correlated with their antibacterial activities. Interestingly, extracts from *E. coli* cells treated with a PTA derivative show specific protein–ruthenium interactions, suggesting that the intracellular proteins are most likely targets of these complexes [94].

**Figure 9.** Representative 'piano stool' RuII -η <sup>6</sup>–arene complex, where X, Y, and/or Z is a labile ligand.

Relying on potential interference with the iron-acquisition systems and in order to increase internalization of the complexes in bacteria, a Trojan Horse strategy was applied for three Ru (II)–arene complexes and one RAPTA-like complex bearing derivatives of deferoxamine B (DFO) (Figure 10) [104]. DFO is a commercially available siderophore, namely an iron chelator that is secreted by microorganisms to bind extracellular Fe (III) and aid in its transport across bacterial membranes inside the cells [105]. These compounds displayed only modest activity against three ESKAPE pathogens (*S. aureus*, *K. pneumoniae*, *A. baumannii*) and one fungal strain (*C. albicans*) when Fe (III) ions were present in the medium. Absence of iron in the media led to an increase in activity, particularly for *K. pneumoniae*. All Ru (II) complexes of this series, however, showed little to no activity against *P. aeruginosa*, *E. coli*, and *C. neoformans*, presumably because these bacterial and fungal strains are more susceptible to internalizing DFO. Antiproliferative studies on normal cells (HEK-293) showed that these complexes were essentially non-toxic towards normal eukaryotic cells in the presence of iron [104].

**Figure 10.** General structure of deferoxamine B (DFO)-derived Trojan Horse antibacterial drugs and some DFO-derived Ruthenium(II)–Arene Complexes [104].

Various Ru(II)–arene complexes with thiosemicarbazone ligands were more active against Gram-positive bacteria than Gram-negative bacteria and/or fungi, but were still less active than the antibiotics used as controls (ampicillin, streptomycin, or ciprofloxacin) [95,98,100,106]. As was seen for other ruthenium complexes, they were shown to bind DNA

and human serum albumin with significant affinity in vitro, suggesting that DNA and/or proteins are potential targets of these complexes in bacterial cells. Several complexes were shown to exert low cytotoxicity towards healthy cell lines [95].

Ru(II)-η 6 -*p*-cymene complexes bearing pyrazole derivatives containing *N,S* donor atoms exerted moderate antibacterial activity against Gram-positive strains, including *S. aureus*, *S. epidermidis*, and *E. faecalis*, while displaying very weak to no activity against Gram-negative bacteria (*P. vulgaris*, *P. aeruginosa*). Notably, the complexes were non-toxic against the healthy human fibroblast HFF-1 cells [107]. Other Ru(II)–arene complexes with various *N,N*- or *N,O*- bidendate ligands displayed moderate activity against various Grampositive bacterial strains and, notably, were found to be more active against *P. aeruginosa* than various clinically used antibiotics used as controls [96,99].

While it is well known that Ru(II)–arene complexes have been widely investigated as potential anticancer agents, their clinical use as antibacterial drugs may be limited by their cytotoxic effects (and generally the poor selectivity for cancerous over healthy cells). Some of these complexes, however, exhibited dual antibacterial and anticancer activities [104]. This constitutes a desirable trait as current anticancer therapy weakens the immune system and often leaves patients susceptible to opportunistic infections. Conversely, patients suffering from a chronic infection are more prone to develop cancer due to certain defects in the immune response [108].

#### 5.1.3. Other Mononuclear Ru Complexes

Various other Ru(II/III) complexes have been reported to possess antibacterial activity. However, microbiological studies for these complexes mainly involved disc diffusion assays or MIC testing, without any further research with regard to their modes of action [109–120]. These complexes were generally more active against Gram-positive strains, with little to no activity against Gram-negative or drug-resistant bacteria. However, a Ru(III) complex, [Ru(L)Cl2]Cl, where L is a N,N,N,N- tetradentate macrocyclic ligand derived from 2,6 diaminopyridine and 3-ethyl-2,4-pentanedione, was moderately active against the Gramnegative bacteria *Xanthomonas campestris* and *P. aeruginosa* and displayed higher activity than the corresponding Pd(II), Pt(II), and Ir(III) complexes [114]. Three ruthenium halfsandwich complexes containing phenyl hydrazone Schiff base ligands also displayed good activity against the Gram-negative *P. aeruginosa*, comparable to that of the positive control, gentamicin, and generally higher than the corresponding Ir(III) and Rh(III) complexes [111].

There are few examples of Ru(II) complexes that display antimycobacterial activity. However, 'SCAR' compounds, consisting of a series of Ru(II) complexes containing phosphine/picolinate/diimine ligands (Figure 11), had low MIC values against multidrugresistant strains of *M. tuberculosis* [51,121,122]. Moreover, the SCAR complexes exerted synergistic interactions with first-line antibiotics, with the best overall synergistic activity observed with isoniazid [122]. Although these complexes displayed some selectivity towards bacterial over healthy eukaryotic cells, an increase in the toxic effects against bacteria was correlated with higher toxicity against eukaryotic cells. *Cis*-[RuCl2(dppb)(bpy)] (SCAR6), where dppb = 1,4-bis(diphenylphosphino)butane and bpy = 2,2'-bipyridine, the least active compound of the series, was found to be the least stable in aqueous solutions [121]. Upon dissolution in water, the chlorido ligands are released, and the resulting species was shown to bind covalently to DNA and induce DNA damage in a similar manner to cisplatin [51,121]. Moreover, the metabolic products of SCAR6 were responsible for the mutagenic effects of the compound observed in *Salmonella typhimurium*. In contrast, SCAR4 and SCAR5 did not display any mutagenic effect [51].

A biphosphinic ruthenium complex, *cis*-[Ru(dppb)(bqdi)Cl2] 2+ (Figure 11, RuNN), where dppb = 1,4-bis(diphenylphosphino)butane and bqdi = o-benzoquinonediimine, displayed bacteriostatic and bactericidal activity against Gram-positive bacteria (*S. aureus*, including MRSA, and *S. epidermidis*). Time–kill kinetics studies indicated that RuNN displayed bactericidal activity in the first 1–5 h [52]. Note that this is a much shorter time than that reported for vancomycin or telavancin (24 h) [123]. Additionally, the combination

treatment of RuNN and ampicillin (but not tetracycline) resulted in a dramatic increase in activity, highlighting the synergistic effect of the two drugs against *Staphylococcus* spp. For the drug-resistant *S. epidermidis* ATCC 35,984 strain, the MIC value for the RuNN + ampicillin treatment was 1/16 of that of ampicillin alone. Furthermore, RuNN inhibited the formation of *S. aureus* biofilms and reduced the total biomass of mature biofilms by ~50%. The complex displayed no hemolytic activity on erythrocytes [52].

**Figure 11.** Chemical structures of selected SCAR complexes and RuNN.

Several ruthenium complexes with antibiotics have been reported. The activity of trimethoprim was, unfortunately, significantly decreased upon complexation with Ru(III) [124]. Complexes of the half-sandwich Ru(II)–arene complex [Ru(η 6 -*p*-cymene)] with a ciprofloxacin derivative, CipA, exhibited higher activity against *E. coli* and *S. aureus* than CipA. These complexes are labile in aqueous solutions and, therefore, their activity is probably the result of additive or synergistic effects of the [Ru(η 6 -*p*-cymene)] complex and CipA [125]. Ru(II) complexes with clotrimazole were active against mycobacteria, but were also found to be significantly toxic to mammalian cells [126]. Three Ru(III) complexes of ofloxacin, namely [Ru(OFL)2(Cl)2]Cl [Ru(OFL)(AA)(H2O)2]Cl2, where OFL = ofloxacin and AA is either glycine or alanine, were active against Gram-negative bacteria (*E. coli* and *K. pneumoniae*), but showed little to no activity on Gram-positive bacteria (*S. epidermidis*, *S. aureus*) [127]. This is unsurprising, given that fluoroquinolones are particularly effective against Gram-negative microorganisms [128].

Homo- and hetero-leptic ruthenium(II) complexes with "click" pyridyl-1,2,3-triazole ligands with various aliphatic and aromatic substituents (generally denoted as Ru-pytri and Ru-tripy, Figure 12) have been reported to possess good antibacterial activity. Generally, the most active complexes displayed high activity against Gram-positive strains, including MRSA (MIC = 1−8 µg/mL), but were less effective against Gram-negative bacteria (MIC = 8−128 µg/mL) [53,54]. The Ru-tripy series was generally more effective against Gram-negative bacteria than the Ru-pytri compounds [54]. Notably, the water-soluble chloride salts of the most active Ru-pytri complexes ([Ru(hexpytri)3] 2+ and Ru(octpytri)3] 2+ , Figure 12) displayed higher activity than the gentamicin control against two strains of MRSA (MR 4393 and MR 4549). Moreover, the Ru-pytri complexes exhibited only modest cytotoxic effects at concentrations higher than the MIC values on Vero (African green monkey kidney epithelial) and human dermal keratinocyte cell lines [53]. For the Ru-tripy series, the activity appears to be closely linked to the length of the alkyl chain, with hexyl or heptyl substituents on the "click" ligands resulting in the highest activity of the corresponding homo- and hetero- leptic Ru(II) complexes. The MIC values for the most active complex of the Ru-tripy series, [Ru(hexyltripy)(heptyltripy)]Cl2, were 2 µg/mL and 8 µg/mL, respectively, against *S. aureus* and *E. coli*. Despite being generally more active than the Ru-pytri series, the Ru-tripy complexes demonstrated little to no selectivity for prokaryotic vs. eukaryotic cells (IC<sup>50</sup> = 2–25 µM on eukaryotic cells lines—cancer and skin). With regard to their mechanism of action, transmission electron microscopy (TEM) experiments and propidium iodide assays identified cell wall/cytoplasmic membrane disruption as the main mechanism for the Ru-pytri complexes [53], while [Ru(hexyltripy)(heptyltripy)]Cl<sup>2</sup> appears to cause abnormal cellular division [54].

Chitosan Schiff base derivatives conjugated to Ru(III) ions give polymers enhanced water solubility and antibacterial activity against Gram-positive (*B. subtilis* and *S. aureus)* and Gram-negative (*E. coli*, *K. pneumoniae,* and *P. aeruginosa*) bacteria [79].

**Figure 12.** Chemical structures of ruthenium(II) complexes with "click" pyridyl-1,2,3-triazole ligands with various aliphatic and aromatic substituents (generally denoted as Ru-pytri [53] and Ru-tripy [54]). Adapted with permission from [53], Copyright © 2016, American Chemical Society and [54], © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

#### *5.2. Polynuclear Ruthenium (II) Complexes*

5.2.1. Kinetically Inert Dinuclear Polypyridylruthenium (II) Complexes

The ruthenium polynuclear complexes, commonly known as Rubbn, are the most investigated ruthenium-based compounds with regard to their antimicrobial activities. Rubb<sup>n</sup> are kinetically inert dinuclear polypyridylruthenium (II) complexes with the general formula [(Ru(phen)2)2(µ-bbn)]4+ (Figure 13), where bb<sup>n</sup> = bis[4(4'-methyl-2-2'-bipyridyl)]-1, *n*-alkane. In the dinuclear Rubb<sup>n</sup> complexes, two mononuclear mono-bb<sup>n</sup> fragments (described above) are bridged by a flexible methylene linker, bbn, where *n* represents the number of methylene groups in the alkyl chain. Rubb<sup>n</sup> are moderately active against Gram-negative bacteria (*E. coli*, *P. aeruginosa*) and exhibit excellent activity against Gram-positive strains (including MRSA—MIC Rubb12/16 = 1 mg/L, while MIC gentamicin = 16 mg/L) [37,38]. The antibacterial activity appears to be closely linked to cellular uptake, which was, in turn, shown to be directly proportional to the length of the alkyl chain and therefore the lipophilicity of the compounds [38]. Of note, a follow-up study comparing the mononuclear [Ru(Me4phen)3] 2+ (Figure 3) with the dinuclear Rubb<sup>n</sup> complexes reported significant differences in the cellular uptake and mode of action. While Rubb<sup>n</sup> are taken up by *S. aureus* cells via a passive transport mechanism, the cellular uptake of [Ru(Me4phen)3] 2+ appears to be protein-mediated (active transport) [88]. In eukaryotic cells, however, Rubb<sup>n</sup> complexes are transported via either an active or a passive mechanism depending on the cell type and have been shown to localize to the mitochondria or the RNA-rich nucleolus [56,91,129].

**Figure 13.** Chemical structures of the inert dinuclear Rubb<sup>n</sup> ([Ru<sup>2</sup> (phen)<sup>2</sup> (tpphz)]4+, [Ru<sup>2</sup> (5- Mephen)<sup>2</sup> (tpphz)]4+, [Ru<sup>2</sup> (2,9-Me2phen)<sup>2</sup> (tpphz)]4+, and [Ru<sup>2</sup> (Me4phen)<sup>2</sup> (tpphz)]4+) and mononuclear ([Ru(phen)<sup>2</sup> (tpphz)]2+) complexes.

The large positive charge (+4) and the hydrophobic alkyl chain are key structural features that contribute to the activity of the Rubb<sup>n</sup> complexes, allowing these compounds to pierce the bacterial cell walls and exert antibacterial activity. Based on the knowledge gained so far, two modes of action have been reported for dinuclear Rubb<sup>n</sup> complexes: membrane damage and/or interaction with nucleic acids, specifically ribosomal RNA.

Rubb<sup>n</sup> complexes were found to depolarize and permeabilize the membranes of *S. aureus* cells, while no membrane permeabilization was observed for [Ru(Me4phen)3] 2+ , although it did cause depolarization [88]. Additionally, Rubb<sup>12</sup> was shown to embed via a pore-formation mechanism into negatively charged phospholipid multilamellar vesicles, an artificial model generally used to study drug–membrane interactions in vitro [130]. Interestingly, the corresponding Ir(III) complex, Irbb<sup>12</sup> (with a formal charge of +6), was not taken up by cells and was inactive [60]. Molecular dynamics (MD) simulations showed that the bulky, positively charged Rubb<sup>12</sup> spanned the bacterial membrane model at the negatively charged glycerol backbone and the bb<sup>12</sup> linker threaded the hydrophobic core. It is yet to be determined whether the interaction with bacterial membranes results in a change of state (fluidity, charge) of the membrane and if it plays a part in the activity of Rubb12. It should be noted that the complex only interacted at the surface level with a neutrally charged eukaryotic membrane model, which could explain its lower toxicity towards healthy cells vs. bacteria (see below) [130]. This does not exclude the possibility of a protein-mediated transport of Rubb<sup>12</sup> inside eukaryotic cells.

The bactericidal mechanism of these complexes [38] was originally presumed to be linked to their ability to bind DNA [131,132]. Indeed, the dinuclear polypyridyl complex [(phen)2Ru-(µ-tpphz)-Ru(phen)2] 4+ [133] and Rubb<sup>7</sup> [132] were found to localize to *S. aureus* chromosomal DNA. However, despite binding with reasonably high affinity to double-stranded DNA in vitro, Rubb<sup>n</sup> complexes prefer non-duplex structures such as bulges and hairpins[132,134,135]. Live cell microscopy experiments on *E. coli* cells showed that Rubb<sup>16</sup> was found to localize at polysomes, with negligible binding to chromosomal DNA. Polysomes are formed when multiple ribosomes associate along the coding region of mRNA and therefore play an essential role in protein synthesis. The cationic charge of Rubb<sup>16</sup> is thought to promote its interaction with the highly negatively charged polysomes. Furthermore, Rubb<sup>16</sup> was found to induce condensation of the polysomes, an effect which is thought to hinder protein production and therefore inhibit bacterial growth [90]. Rubb<sup>n</sup> also displayed high affinity towards the serum transport proteins albumin and transferrin in vitro, which suggests that these complexes could potentially target intracellular proteins [88].

As was shown for Rubb16, targeting ribosomal RNA (rRNA) in bacteria can be advantageous for the development of selective antibacterial agents, since there are significant differences between prokaryotic and eukaryotic rRNA [136]. Moreover, in vitro experiments and MD simulations have shown that Rubb<sup>12</sup> only interacts at a surface level with a neutral membrane bilayer mimic of a eukaryotic membrane [130]. Indeed, these inert Ru(II) complexes generally display selectivity for bacteria over normal eukaryotic cells. Although toxic to cancer cells, Rubb12/16 were much less active (up to 100-fold) against healthy cell lines [89,90,129]. In spite of the fact that Rubb<sup>16</sup> is slightly more active against bacteria than Rubb12, the higher in vitro toxicity of Rubb<sup>16</sup> to both healthy eukaryotic cells and red blood cells makes Rubb<sup>12</sup> a more promising drug candidate [37].

Rubb<sup>12</sup> injected intramuscularly was not toxic to mice at concentrations up to 64 mg/kg. Moreover, pharmacokinetic experiments have shown that 30 min post-administration, serum concentrations of Rubb<sup>12</sup> are higher than the MIC values for Gram-positive bacteria and were maintained for more than 3 h [55]. Encapsulation of Rubb<sup>12</sup> in cucurbit[10]uril (Rubb12⊂Q[10]) resulted in a two-fold decrease in toxicity (free Rubb12—1 mg/kg, Rubb<sup>12</sup> ⊂Q[10]—2 mg/kg) when administered intravenously to mice. Interestingly, while free Rubb<sup>12</sup> accumulated predominantly in the liver, Rubb12⊂Q[10] was found to be distributed in comparable amounts in both the liver and kidneys. A substantial reduction (∼2-fold) in the ruthenium concentrations (quantified using Inductively Coupled Plasma Mass Spectrum, ICP-MS) found in the liver was reflected by an increase (∼4-fold) in the kidneys. The significant increase in kidney accumulation is the result of the renal excretion of Rubb12⊂Q[10]. The encapsulation in cucurbit[10]uril resulted in higher cellular accumulation, lower toxicity, and faster clearance of Rubb<sup>12</sup> [137].

As opposed to Rubbn, which bear flexible linkers, systems bridged by a rigid, extended aromatic ligand possess a property that is rather unusual for this class of complexes, that is a generally higher activity against pathogenic Gram-negative as compared with Grampositive bacteria. The more rigid structure of these complexes is thought to play an essential role in their activity against Gram-negative strains, as well as the presence of potentially ionizable nitrogen sites and the more complex 3D structure when compared with typical drug architectures [57,138]. Thus, a range of luminescent dinuclear Ru(II) complexes bearing tetrapyridophenazine (tpphz) (Figure 13) were found to be more active against Gram-negative (both a wild-type and a multidrug-resistant strain of *E. coli*) than Grampositive (a vancomycin resistant strain of *E. faecalis*) bacteria. [(Ru2(5-Mephen)2)2(tpphz)]4+ was the least active compound of the series, most likely due to its low water solubility. For the other three complexes, a direct, positive relationship was observed between lipophilicity and activity. The lead compound of the series, [Ru2(Me4phen)2(tpphz)]4+, was also nontoxic to healthy eukaryotic cells (Table 1). Of note, all complexes showed appreciable activity against the ESKAPE pathogens and [Ru2(Me4phen)2(tpphz)]4+ even displayed higher activity than ampicillin against the wild-type strain of *E. coli* and against *E. faecalis*. Selectivity towards the Gram-negative strains has also been observed for the mononuclear parent compound, [Ru(phen)2(tpphz)]2+, even though it was found to be significantly less active than its dinuclear derivatives against all bacterial strains [57].

[Ru2(Me4phen)2(tpphz)]4+ was shown to be actively taken up into Gram-negative bacterial cells and to disrupt the bacterial membrane structure before internalization [57], results which were further substantiated by transcriptomic analysis. Thus, the complex caused a significant downregulation of genes involved in membrane transport and the tricarboxylic acid cycle and upregulation of the *spy* gene [58]. The *spy* gene, encoding a periplasmic chaperone, is involved in zinc homeostasis and in maintaining the homeostasis of protein folding under cellular stress [139]. Thus, overexpression of the *spy* gene in the [Ru2(Me4phen)2(tpphz)]4+-stressed cells indicates protein damage in the outer membrane. Moreover, multi-drug resistant *E. coli* cells developed resistance to [Ru2(Me4phen)2(tpphz)]4+ much slower, and only in low levels, in comparison with various clinically available antibiotics. Encouragingly, [Ru2(Me4phen)2(tpphz)]4+ was active at

low micromolar concentrations against other Gram-negative ESKAPE pathogens, including *P. aeruginosa* and *A. baumannii* [58].

A similar mode of action involving membrane and DNA damage was reported in the less susceptible, Gram-positive *S. aureus* cells. However, [Ru2(Me4phen)2(tpphz)]4+ was found to accumulate to a lower extent in Gram-positive when compared with Gramnegative bacteria, which may account for the lower efficacy of these complexes against the former. This was shown to be related to a resistance mechanism developed by Grampositive bacteria against cationic species, which involves upregulation of the *mprF* gene. Overexpression of this gene leads to the accumulation of positively charged phospholipids on the outer leaflet of the cytoplasmic membrane, which repel cationic molecules, such as metal complexes. Consequently, it was found that [Ru2(Me4phen)2(tpphz)]4+ was more active against a *mprF*-deficient *S. aureus* strain and in mutant *S. aureus* strains missing, or with altered, wall teichoic acids [59].

This class of compounds, particularly [Ru2(Me4phen)2(tpphz)]4+, shows remarkable promise for the treatment of infections caused by Gram-negative pathogens. In addition, the lead compound displays good kinetic solubility, which suggests good bioavailability and possible oral administration [58]. Clearly, animal experiments are needed to further assess the efficacy of this class of compounds as novel antibacterial agents in vivo.

#### 5.2.2. Chlorido Dinuclear Polypyridylruthenium (II) Complexes

A range of symmetrical dinuclear polypyridylruthenium(II) complexes with the general formula [(Ru(terpy)Cl)2(µ-bbn)]2+ (where terpy = 2,2':6',2"-terpyridine) have been reported [55,60]. These labile complexes are commonly denoted as Cl-Rubbn-Cl (Figure 14). These complexes have a positive charge of +2; however, upon dissolution in water followed by the substitution of the chloride ions with solvent molecules, their charge increases to +4 [60]. The Cl-Rubb7/12/16-Cl complexes exert bactericidal activity against Gram-positive strains (*S. aureus* and MRSA), *E. coli*, and *P. aeruginosa*, with Cl-Rubb12-Cl being the lead compound of the series. Cl-Rubb7/12-Cl are more active than their dinuclear inert analogues; however, the Cl-Rubb16-Cl complex was significantly less active than Rubb<sup>16</sup> [60]. It is uncertain why this variation occurs, but a possible reason is speculated to be that the enhanced cellular uptake of the Cl-Rubb7/12-Cl complexes can compensate for a reduction in activity. Since Rubb<sup>16</sup> readily accumulates into cells, the addition of chlorido groups only results in a lower activity.

Asymmetrical chloride-containinig dinuclear polypyridylruthenium(II) complexes, Rubb7/12/16-Cl (Figure 14), have also been reported. The Rubbn-Cl complexes contain two ruthenium centers bridged by a flexible methylene linker. One ruthenium center bears a labile chlorido ligand, while the second is kinetically inert. The MIC values calculated for these complexes are comparable with those reported for the previously described Cl-Rubb7/12/16-Cl series. Furthermore, with the exception of Rubb16-Cl, Rubbn-Cl complexes exert superior antibacterial activities as compared with their inert dinuclear analogues, with Rubb12-Cl being the most active against both Gram-positive and Gram-negative strains. Rubb12-Cl was found to be 30- to 80-fold more toxic to the bacteria than to eukaryotic cell lines—two healthy cell lines (baby hamster BHK and embryonic HEK-293 kidney) and one cancerous cell line (liver carcinoma HepG2). Interestingly, large differences were found in the cytotoxic activity of Rubb7-Cl as compared with Rubb12/16-Cl. It was significantly more active towards the Gram-negative *E. coli* than against the Gram-positive *S. aureus* and MRSA, significantly more toxic to HepG2 (IC<sup>50</sup> = 3.7 µM), and far less toxic to BHK (IC<sup>50</sup> = 238 µM) cells than Rubb12/16-Cl. Cellular localization studies in HepG2 cells suggest that all complexes of this series were shown to accumulate preferentially in the rRNA-rich nucleolus. In addition, the large differences in the toxicity profile of the Rubbn-Cl complexes might be related to the fact that Rubb7-Cl binds to chromosomal DNA to a greater extent than Rubb12/16-Cl [56].

**Figure 14.** Chemical structures of labile dinuclear Cl-Rubbn-Cl and Rubbn-Cl complexes, where *n* = 7, 12, 16.

#### 5.2.3. Tri-/Tetra-Nuclear Polypyridylruthenium(II) Complexes

Generally, the more lipohilic tri- and tetra- nuclear complexes, Rubb7/10/12/16-tri and Rubb7/10/12/16-tetra (Figure 15), displayed higher activities against Gram-positive and Gram-negative strains, as well as a range of bacterial clinical isolates, than the dinuclear Rubb<sup>n</sup> complexes [37,61]. The linear tetranuclear [Rubbn-tetra]8+ complexes were more active, with MIC values < 1 µM against Gram-positive bacteria, than their non-linear trinuclear [Rubbn-tri]6+ analogues. Time–kill curve experiments showed that Rubb12-tri and Rubb12-tetra exert bactericidal activity and kill bacteria within 3–8 h [55].

As opposed to the inert dinuclear Rubb<sup>n</sup> complexes, there is no apparent relationship between the antibacterial activities of the Rubbn-tri and Rubbn-tetra complexes and lipophilicity or cellular uptake. The more active tetranuclear [Rubbn-tetra]8+ complexes are less lipophilic than their trinuclear [Rubbn-tri]6+ analogues, despite the additional non-polar bb<sup>n</sup> ligand. This is unsurprising, considering the difference in the overall charge of the tri- and tetra- nuclear complexes. Moreover, even though Rubbn-tri and Rubbn-tetra complexes were more active against Gram-positive bacteria, they were shown to accumulate to a greater extent in Gram-negative bacteria [61]. Within eukaryotic Hep-G2 cells, Rubb12-tri and Rubb12-tetra have been shown to accumulate preferentially in the RNA-rich nucleolus, as was previously described for the dinuclear Rubb<sup>n</sup> complexes [91].

The mechanism of action of the Rubbn-tri and Rubbn-tetra complexes is yet to be determined, but it is thought to be related to their abilities to bind to nucleic acids and/or proteins [1,61]. The Rubbn-tri and Rubbn-tetra complexes have been shown to interact with single-stranded oligonucletides and proteins in vitro, with significantly higher affinities than their dinuclear analogues [87,88,91]. The mechanism underlying their interactions with the DNA backbone may differ for the linear tetranuclear and the three dimensional non-linear trinuclear species [87].

Two inert polypyridylruthenium(II) tetranuclear complexes, containing the bridging ligand bis[4(4'-methyl-2,2'-bipyridyl)]-1,7-heptane, with linear (Rubb7-tetra or Rubb7-TL) and non-linear (Rubb7-TNL) (Figure 15) structures, displayed good antibacterial activity against both Gram-positive (*S. aureus*, MRSA) and Gram-negative (*E. coli*, *P. aeruginosa*) bacteria. The non-linear (branched) species displayed slightly higher activity than the corresponding linear analogue and accumulated in the nucleolus and cytoplasm but not in the mitochondria [62].

Rubbn-tri and Rubbn-tetra complexes were found to be more toxic than Rubb<sup>n</sup> to carcinoma and healthy mammalian cell lines in vitro, with IC<sup>50</sup> values lower than or comparable to that of cisplatin [55,89,91]. However, the tri- and tetra- nuclear complexes were still ~50-fold more toxic to Gram-positive bacteria and 25 times more toxic to the susceptible Gram-negative strains than to eukaryotic cells [55]. Rubb7-TNL was slightly less toxic to healthy eukaryotic BHK cells than its linear analogue (Table 1), yet still more toxic than cisplatin [62]. In comparison, the dinuclear ∆∆-Rubb<sup>12</sup> complex was ~100-fold more toxic to Gram-positive bacteria. The cytotoxic effects of the tri- and tetra- nuclear ruthenium complexes towards eukaryotic cells suggest that merely increasing the lipophilicity and charge is likely to result in decreased selectivity. Therefore, the general goal now is the development of new ruthenium complexes that are highly selective towards bacteria.

**Figure 15.** Chemical structures of inert tri- and tetra- nuclear ruthenium complexes, where *n* = 7, 10, 12 or 16.

#### 5.2.4. Other Polynuclear Complexes

The dinuclear [Ru2L3] 4+ ruthenium(II) triply stranded helicate, bearing bidentate "click" pyridyl-1,2,3-triazole ligands, displayed modest antimicrobial activity (MIC > 256 µg/mL) [140] as compared with similar mononuclear Ru(II) complexes bearing "click" ligands [53,54]. However, in contrast to the similarly structured [Fe2L3] 4+ helicates, the more kinetically inert [Ru2L3] 4+ system proved stable over a period of at least 3 days in DMSO solutions [140].

The binuclear ruthenium (III) complexes [RuX3L]<sup>2</sup> (X = Cl, X = Br), [RuX3L1.5]<sup>2</sup> (X = Br), and [RuX3L2]<sup>2</sup> (X = Br), where L stands for 2-substituted benzimidazole derivatives, were moderately active against Gram-negative bacteria (*E. coli* and *S. typhi*) as tested by the agar diffusion method. The activity on the Gram-positive bacteria *S. aureus* and *Bacillus aureus* was, however, low when compared with the standard antibiotics ampicillin and fluconazole [141].

#### *5.3. Hetero-bi/tri-Metallic Complexes*

The organometallic complex containing ruthenocene (Compound 1, Figure 16a) was moderately active against MRSA, admittedly less so than the ferrocene derivative (Compound 2, Figure 16a). The organometallic complex containing ferrocene (2) (Figure 16a) was found to generate ROS, in contrast to 1, as indicated by oxidative stress assays. Consequently, the difference in activity was suggested to result from their differing abilities to generate ROS [142].

**Figure 16.** Chemical structures of hetero- (**a**) trimetallic complexes bearing ruthenocene or ferrocene moieties and (**b**) bimetallic complex bearing a ferrocenyl–salicylaldimine moiety.

Incorporation of ferrocene as well as ruthenium in a half-sandwich heterobimetallic complex bearing a ferrocenyl–salicylaldimine moiety (Figure 16b) showed promising activity against *Mycobacterium tuberculosis*. Due to the observed glycerol-dependent antimycobacterial activity, a possible mechanism of action involves disruption of glycerol metabolism and accumulation of toxic intermediate metabolites. The complex was found to possess relatively low cytotoxicity in vitro against normal microbial flora, which also suggests selectivity [143].

A Ru(II)–Pt(II) bimetallic complex, [RuCl(tpy)(dpp)PtCl2](PF6), where dpp = 2,3- bis(2 pyridyl)pyrazine and tpy = 2,2':6',2"-terpyridine, was reported to inhibit the growth of *E. coli* cells (albeit at a relatively high concentration of 400 µM). In contrast, its monometallic Ru(II) precursor, [RuCl(tpy)(dpp)](PF6), was inactive against *E. coli*. The improved activity of the Ru(II)/Pt(II) heteronuclear complex was attributed to the *cis*-PtCl<sup>2</sup> moiety, although the heterobimetallic complex was still less effective than cisplatin alone [144]. Although [RuCl(tpy)(dpp)PtCl2](PF6) was reported in a follow-up study to induce DNA photocleavage, the effect of light irradiation on its antibacterial activity was not assessed [145].

#### *5.4. Ruthenium-Based Carbon-Monoxide-Releasing Molecules (CORMs)*

With a unique mode of action involving ligand exchange, carbon-monoxide-releasing molecules (CORMs) represent an emerging class of biologically active organometallic derivatives. Although their mechanism of action is fairly complex and not yet fully understood (Figure 17), CORMs release carbon monoxide (CO) to bind to intracellular targets, which is partially responsible for their activity. The chemistry and antimicrobial activity of ruthenium-based CORMs have already been reviewed in several excellent works [32,146–149]. The reader can find in the following pages a summary of what has already been reviewed, as well as references to more recent research.

**Figure 17.** Modes of action and intracellular targets of CORMs. The bacterial membrane includes the inner membrane (IM), the outer membrane (OM), and periplasm (P), which are represented at the top. *1.* CORMs enter bacteria by unknown pathways and mechanisms; CO enters the cells by diffusion. *2.* After they enter the cell, CORMs release CO, forming inactivated CORM (iCORM). *3.* CO, CORM, and iCORM are detected by transcription factors (TFs), causing transcriptional changes. *4.* TFs are activated by ROS that may be generated directly by CORMs or can be generated as a result of the interaction of CORMs with the respiratory chain. *5.* A simplified aerobic respiratory chain of bacteria is represented, consisting of a flavin-containing NADH dehydrogenase, a ubiquinone (Q) pool, and a terminal heme-containing quinol oxidase. *6.* CO binds to the heme-containing quinol oxidase active site, competing with oxygen and impeding respiration. *7.* Impairment of ATP generation by ATP synthase. *8.* CO or CORM may directly or indirectly interact with IM transporters. *9.* Diverse cellular responses to CO and CORM. Question marks represent unknown targets, effects, or mechanisms: transport into (or out of) cells; intracellular mechanisms of CO release from CORMs; interaction with TFs and modification of gene expression by CORMs; effects of CORMs on membrane transporters. Figure reproduced from [146].

CORM-2, a highly lipophilic dinuclear ruthenium(II) complex with the formula [Ru(CO)3Cl2]<sup>2</sup> (Figure 18), and the water-soluble mononuclear CORM-3, [Ru(CO)3Cl(glycinate)] (Figure 18), have been intensely investigated over the last two decades. Various reports have shown that CORM-2 and CORM-3 exhibit broad-spectrum antibacterial activity against several strains and clinical isolates of Gram-positive (*S. aureus, Lactobacillus lactis*) and Gram-negative (*E. coli*, *H. pylori*, *Campylobacter jejuni*, *P. aeruginosa*, *Salmonella enterica* serovar Typhimurium) bacteria [63,64,66,67,150–153]. Notably, CORM-3 displays bactericidal activity against antibiotic-resistant *P. aeruginosa* [66], *H. pylori* [64], and *E. coli* [67] strains.

**Figure 18.** Chemical structures of CORM-2 and CORM-3.

Studies in various buffers indicated that significant ligand exchange is likely to occur in biological media. Both the Cl− and glycinate ligands of CORM-3 are labile and undergo partial or full displacement by either water molecules or other counter-ions (e.g., phosphate) existing in the buffer or growth medium constituents [154].

#### 5.4.1. Mechanisms of Action

#### The Role of CO

CO is an inorganic compound that can bind hemoglobin with highly toxic effects. CO is produced endogenously as a result of heme breakdown by heme oxygenase. It is generally known to interact with metalloproteins due to its affinity towards transition metals (for instance, the ferrous ions in hemoglobin). Despite its notorious toxic effects, CO acts as a signaling molecule with important therapeutical properties, which include anti-inflammatory, anti-apoptotic and anti-proliferative effects [146]. The possibility of limiting its intrinsic severe toxicity and enhancing the biological activity has been explored with pro-drugs acting as CO-releasing molecules (CORMs), including transition metal (Mn, Ru, Fe, Mo) carbonyl complexes.

Ru-carbonyl CORMs were initially thought to act merely as vectors designed to deliver the toxic CO gas inside bacterial cells and, hence, the respiratory chain was presumed to be the main target of these molecules. The antibacterial activity of Ru-based CORMs was attributed to their ability to release CO in certain microenvironments of the cell, effecting an increase in the ratio of CO relative to O2, which eventually impedes the oxygen metabolism [35,142,150]. Indeed, there is substantial evidence in the literature that CORM-2 and CORM-3 impair aerobic respiration in *E. coli* [152,155,156], *P. aeruginosa* [66,157], *H. pylori* [64], *C. jejuni* [150], and *S. enterica* [152]. However, administration of BSA-Ru(II)(CO)2, an adduct formed between BSA and the hydrolytic decomposition products of CORM-3 in vitro, was demonstrated to release CO in a controlled manner in tumor-bearing mice, but did not produce any significant effect on bacterial growth in *E. coli* cells [158]. Additionally, in physiological conditions CORM-3 was found to release low amounts of CO inside bacterial cells (for 100 µM CORM-3, the concentration of CO detected in cells was < 0.1 µM) [159]. Thus, the toxicity of CO alone appears to be insufficient to explain the antibacterial activity of these compounds.

#### ROS Generation

ROS-induced oxidative stress has also been assessed as a possible mechanism of action responsible for the antimicrobial activity of CORMs. This assumption was based on the positive correlation observed in *E. coli* between the bactericidal activity and the ROS levels generated upon treatment with CORMs [35,160]. In vitro studies performed in aqueous solutions indicated that CORM-2 and CORM-3 are able to generate OH• [160,161] and O<sup>2</sup> •− [151,161] radicals. However, the amount of superoxide ions was measured to be only ~1% of the total CORM-3 concentration, which does not account for the bactericidal activity of the compound [151]. In airway smooth muscle cells, CORM-2 stimulated ROS production through inhibition of cytochromes on both NAD(P)H oxidase and the respiratory chain [162,163]. Furthermore, *E. coli* mutant strains in which genes encoding catalases and superoxide dismutases (SODs) have been deleted are more susceptible to CORM-2 treatment due to an increase in intracellular ROS content; this effect is alleviated upon supplementation of the culture medium with antioxidants (reduced glutathione or cysteine) [160]. For CORM-3, however, addition of catalase or SOD did not have any significant impact on its respiratory effects in *E. coli*, implying that peroxide or superoxide are not involved in the activity of CORM-3 in these cells [152,155]. In *C. jejuni*, however, CORM-3 was shown to inhibit respiration and generate hydrogen peroxide, although no effect on cell growth was observed even at concentrations as high as 500 µM [150]. Addition of various sulfur-containing antioxidants, namely cysteine, N-acetyl cysteine (NAC), or glutathione (GSH), abolished the respiratory and growth inhibitory effects of ruthenium–carbonyl CORMs in *E. coli* and *P. aeruginosa* [66,151,157,160,164]. However, this

effect is presumed to be independent of the antioxidant activity of CORMs, based on two reports showing that NAC strongly inhibits the uptake of CORMs in *E. coli* cells [155] and a NAC–CORM-2 complex displays no activity against bacterial cells [165]. It is more likely that the Ru(II) species derived from CORMs in biological environments form adducts with exogenous compounds bearing thiol groups, which cannot be readily internalized into bacteria and are therefore less potent antibacterial agents. Non-thiol antioxidants do not alleviate the inhibitory effects of CORMs on respiration [155]. Moreover, CORM-3 was shown to impair the tricarboxylic acid (TCA) cycle, also known as the Krebs cycle, in *E. coli* cells treated under anaerobic conditions, suggesting that its activity extends beyond ROS generation [67]. Hence, it is unlikely that ROS-induced oxidative stress represents the main mechanism behind the CORMs' bactericidal activity, although ROS generation probably plays some part in inhibiting the growth and respiration of CORM-2 on *E. coli* cells.

#### Membrane Damage

The bactericidal activity of CORM-3 has also been linked to membrane damage in *E. coli* cells, as penetration of propidium iodide [156] and N-phenyl-1-napthylamine [166], fluorescent dyes that cannot pierce healthy membranes, is allowed after CORM-3 treatments. Clearly, loss of membrane integrity can occur in the aftermath of cell death; therefore, it is not necessarily part of the antibacterial mechanism.

#### The Role of the Ru(II) ion Interactions with Proteins and DNA

In ruthenium-based CORMs, the Ru ion was assumed to have more of a structural role. This paradigm was based on the assumption that ruthenium–carbonyl CORMs were stable enough to reach the intracellular environment, where reducing agents (e.g., sulfites) would trigger CO release [35]. However, more recent research suggests that CORM-2 and CORM-3 undergo ligand exchange and interact with serum proteins in vivo to form protein–Ru(CO)<sup>2</sup> adducts. CO release occurs following decomposition of these adducts [158,167–170]. Additionally, no CO release was detected in vitro upon addition of CORM-2 and CORM-3 in phosphate buffers and cell culture media in the absence of sulfur-containing reducing agents [159].

Therefore, CO release cannot be solely responsible for the cytotoxic effects of CORMs, which is further inferred by the fact that CORM-3 is toxic even for heme-deficient cells [166]. Moreover, Ru-carbonyl CORMs are considerably more active than other non-rutheniumbased CORMs [63,102,157] and inhibit aerobic respiration and bacterial growth more potently than CO gas alone [66,156]. Taking into account all of the above-mentioned arguments, it stands to reason that the ruthenium ion plays an essential role in the antimicrobial activity of these metal complexes.

The Ru(II) ion in CORM-3 was found to bind tightly to thiols. Addition of various compounds containing thiol groups in growth media protected both bacterial and mammalian cells against CORM-3. The binding affinities of CORM-3 for the compounds tested vary in the order cysteine ≈ GSH >> histidine > methionine. Moreover, a direct positive correlation was found between the protective effects of these compounds and the dissociation constants of the complexes formed between CORM-3 and the respective thiol compounds. Other amino acids (alanine and aspartate) did not exert significant protective effects. Southam et al. suggest a mechanism in which CORM-3 undergoes ligand displacement reactions in buffers or media to generate complex species in which the Ru(II) centers are readily available to bind to intracellular components such as glutathione. Another mode of action for CORMs is therefore presumed to involve Ru(II) binding to intracellular targets, impairment of glutathione-dependent systems, and disruption of redox homeostasis [159].

Indeed, CORMs have been shown to interact with various intracellular or membranebound proteins. CORM-3 has been shown to interact in vitro with the serum proteins myoglobin, hemoglobin, transferrin, and albumin, forming protein–Ru(II)(CO)<sup>2</sup> adducts [167,168]. As described above, CORM-3 possesses two labile ancillary ligands (Cl− and glycinate),

which can be readily released in aqueous media, allowing further interaction with serum proteins to occur [168]. With BSA, CORM-3 forms in vitro a [BSA-(Ru(II)(CO)2)16] complex, in which the Ru(II)(CO)<sup>2</sup> adducts bind to histidine residues exposed on the surface of the protein. As stated above, the CO-releasing protein–Ru(II)(CO)<sup>2</sup> complex did not have any significant effect on bacterial growth in *E. coli* cells [158]. The reason is unknown. In addition, CORM-2 has also been shown to inhibit urease activity in *H. pylori* [64] and lactate dehydrogenase in primary rat cardiocytes [171]. *H. pylori* urease is essential to the survival of the bacterium in the acidic gastric milieu [172]; therefore, its inhibition can represent a viable strategy against *H. pylori* infections. The histidine-rich active site involved in coordination of Ni(II) ions is presumed to be the target of CORM-2. It is uncertain whether urease inhibition occurs via direct binding of the Ru(II) ion to the active site accompanied by Ni(II) displacement, or CO binding to the Ni(II) ion in the active site.

Soft and borderline transition metals have been shown to bind to Fe–S clusters, which are important cofactors of various enzymes including several pertaining to the Krebs (or TCA) cycle [173]. CO is also reported to bind to iron–sulfur clusters in a redox-dependent manner [174]. Therefore, Fe–S enzymes have been studied as potential targets for CORMs. Indeed, treatment of *E. coli* cells with CORM-2 resulted in an increase in intracellular iron, suggesting degradation of the Fe–S clusters. This assumption was further supported by the significant inhibition of two Fe–S proteins, aconitase B and glutamate synthase, following exposure of *E. coli* extracts to CORM-2. Although the presence of intracellular Fe–S clusters was shown to correlate with the antimicrobial activity of CORM-2, it was not clearly determined whether the Ru(II) ion of CORM-2 binds directly to Fe–S clusters, or if the degradation of the clusters occurs indirectly as a result of other processes [160]. However, a cell extract from *E. coli* overexpressing aconitase B displayed a 50% decrease in the activity of the enzyme after incubation with CORM-3, relative to untreated cells, suggesting that the protein–CORM-3 complex occurs at a post-translational level. Additionally, recent metabolomics studies in *E. coli* cells revealed that CORM-3 inhibits the activity of several Fe–S proteins, namely the glutamate synthase GOGAT and enzymes of the TCA cycle (aconitase B, isocitrate dehydrogenase, and fumarase). In response to the severe imbalance in the energy and redox homeostasis caused by the Ru-carbonyl complex, activation of the glycolysis pathway was detected in the CORM-3-stressed cells. Notably, other non-COreleasing Ru(II) species, used as controls, were non-toxic to *E. coli* cells and had no effect on the Fe–S enzymes at the concentration used in this study (120 µM—a growth inhibitory but nonlethal concentration of CORM-3) [67].

Although numerous cytotoxic ruthenium complexes developed as anticancer agents have been shown to interact with DNA, no in vitro or in vivo studies clearly demonstrate whether Ru-based CORMs bind directly to DNA or not. However, CORM-2 has been shown to induce DNA damage and increase the expression of a double-strand break repair gene, *recA*, in *E. coli* [65,160]. DNA damage can be the result of CORM-2-induced generation of intracellular ROS, although this has not been clearly established [65].

#### Effects on Gene Expression

Transcriptome studies on *E. coli* revealed that CORM treatments under either aerobic or anaerobic conditions trigger complex transcriptional responses of gene expression [151,156,166,175,176] that exceed those induced by CO alone [177]. CORM-2 and CORM-3 downregulate genes involved in aerobic respiration, energy metabolism, and biosynthesis pathways and upregulate those involved in the SOS response and DNA damage and repair mechanisms. A recent gene profiling study analyzed the effects induced by CORM-2 exposure on a multidrug-resistant extended-spectrum beta-lactamase (ESBL)-producing uropathogenic *E. coli* clinical isolate [65]. Numerous genes encoding the NADH dehydrogenase complex were repressed by CORM-2 [65], as was previously shown for CORM-3 in the *E. coli* K12 strain [156]. Transcriptomics analysis of *E. coli* cells treated with CORM-3 indicated altered expression of the cytochrome genes *cyoABCDE* and

*cydAB* [151,156]. However, CORM-2 had no effect on the expression of cytochrome genes, which could be attributed to the differences in the growth media [65].

Exposure to CORM-2 and CORM-3 increased the expression of genes coding for proteins with roles in stress response and adaptation, e.g., *ibBA*, *ibpA,* and *spy* [65,151,156, 166,176]. The *spy* gene appears to be one of the main non-heme targets for CORMs. Several genes coding for multidrug efflux pump proteins were also upregulated by CORM-2 [65] and CORM-3 [166]. Upregulation of multidrug efflux pump systems has been shown to lead to the development of resistant phenotypes over time [178]. However, the growth inhibitory activity of CORM-2 was not diminished by repeated exposure (20 times), neither in the multidrug-resistant ESBL-producing *E. coli* strain, nor in two other antibiotic-susceptible *E. coli* strains [65].

Significant upregulation has also been found for genes involved in metal homeostasis, such as iron or zinc [151,156,166], and genes involved in the uptake and/or metabolism of sulfur compounds (sulphate-thiosulphate, methionine, cysteine, glutathione) and the sulfur starvation response [67,151,166,176]. In agreement with the already-discussed inhibitory effects of CORMs on Fe–S enzymes [67,160], genes involved in Fe–S cluster biosynthesis and repair are also upregulated by CORM-2 and CORM-3 [151,166,176]. Transcriptomic data, therefore, correlate well with the in vitro observation that sulfur species represent intracellular targets of Ru(II)-based CORM [151].

#### 5.4.2. Ruthenium-Based CORM Polymers

Encapsulation of drugs into polymers is a modern therapeutic strategy that makes use of building blocks with 3D structures that enable controlled ligand exchange [179]. Conjugation to lipophilic polymers reduces the access of water molecules to CORMs, causing the solvent-assisted ligand exchange reactions to occur at a slower, sustainable pace. A Ru-based CORM was conjugated to the side chain of polymeric fibers bearing different thiol moieties, yielding the three water-soluble CO-releasing macromolecules CORM-polymers 1–3 (Figure 19). The resulting polymers have been shown to exhibit bactericidal activity against *P. aeruginosa* and to prevent biofilm formation more efficiently than CORM-2, most likely due to their high CO-loading capacity, controlled release of CO, and prolonged half-lives. Notably, the antimicrobial activity was not directly proportional to the half-lives of the complexes, since CORM-polymer 2 was the most active compound of the series, while CORM-polymer 1 had the longest half-life [180].

**Figure 19.** Chemical structures of ruthenium-based CORM-polymers 1–3.

#### 5.4.3. Cellular Uptake

The mechanisms of uptake of CORM-2 and CORM-3 complexes are unknown. It is also unclear which are the Ru-CORM-derived species that pierce the bacterial membranes and it is likely that different species use different mechanisms of uptake. However, ruthenium species (quantified using ICP-MS) have been found to accumulate to high levels in *E. coli* cells treated with either CORM-2 or CORM-3 [155,156]. CORM-3 was found to be rapidly taken up by *E. coli* cells at an initial rate of 85 µM/min over the first 2.5 minutes after treatment, with intracellular Ru levels reaching a plateau after ∼40 min [151]. CORM-3 accumulated to higher levels in *S. enterica* serovar Typhimurium than in *E. coli*, and at a faster initial uptake rate (>120 µM/min). This may explain, at least partially, why *Salmonella* strains are more susceptible to CORM-3 than *E. coli* [153]. Notably, simultaneous addition of NAC and CORM-2 or CORM-3 reduced the ruthenium accumulation inside bacterial cells, which is probably why exogenous thiols, such as NAC, are able to interfere with the antibacterial activity of both CORM-2 and CORM-3 [155]. These findings also suggest that the bactericidal effects of these compounds are dependent on the ruthenium uptake by the bacterial cells [151]. It has been suggested that Ru–based CORMs could be transported actively, or diffuse, inside the cells via an unidentified route against the concentration gradient and, due to the reactions that occur inside the cells, uptake can continue passively [153].

#### 5.4.4. Toxicity and Pharmacokinetics

The more lipophilic, DMSO-soluble, CORM-2 is generally more toxic to mammalian cells than the water-soluble CORM-3. It has been suggested that the use of DMSO is at least partially accountable for the increased cytotoxicity [102,171,181]. Toxic concentrations reported for eukaryotic cells are considerably higher than the MIC values [66,102,153,164,169– 171,181]. For instance, the bactericidal effects against *P. aeruginosa* occurred at concentrations of CORM-3 that are 50-fold lower than the toxic concentrations for macrophages [66]. However, survival of the mammalian cells was drastically increased by the presence of exogenous thiols in the growth media. For instance, treatment with 25 µM CORM-3 in phosphate-buffered saline (PBS) decreased survival relative to untreated human colon carcinoma RKO cells in PBS by 92%, compared with only 23% in RPMI-1640 growth medium, whilst in DMEM the survival rate was enhanced relative to untreated controls [159].

In vivo studies revealed that a two-week CORM-3 treatment (with increasing doses from 7.5 to 22.5 mg/kg) caused no mortality or any apparent side effects to healthy mice [66]. In contrast, consecutive administrations of 15–37 mg CORM-3/kg in rats caused severe liver and kidney damage after 21 days of treatment. Biodistribution studies in CORM-3-treated mice concluded that the ruthenium species derived from CORM-3 mostly accumulated in the blood for the first hour after the intravenous administration and then were slowly distributed to the kidneys, liver, lungs, and heart. Notably, only trace amounts of ruthenium were found in the brain, suggesting that the complex and its derived species did not cross the blood–brain barrier. Both ruthenium and elevated levels of protein were found in the urine of the CORM-3-treated mice, indicating kidney damage. Moreover, the RuII center was oxidized to RuIII in vivo by enzymes such as cytochrome P450 [170].

#### 5.4.5. In Vivo Studies Regarding the Antibacterial Activity of CORMs

Several studies have reported the in vivo antibacterial activity of Ru–carbonyl CORMs [66,157,182]. In a murine model of polymicrobial sepsis, treatment with 10 µM CORM-2/kg 12 and 2 h before the inoculation of bacteria resulted in a significant decrease in bacterial counts relative to the vehicle-treated mice. CORM-2 improved the survival rates of heme oxygenase (HO)-1 null mice, mutants that are more susceptible to polymicrobial infection, even when administered intraperitoneally after the onset of sepsis [182]. Administration of CORM-2 (12.8 mg/kg) was also shown to significantly increase the survival rates of BALB/c mice infected with *P. aeruginosa* [157].

Injections of CORM-3 (7.5–22.5 mg/kg) in the murine model of *P. aeruginosa* infection reduced bacterial counts in the spleen and increased the survival rates of the infected mice (from 20% in the vehicle-treated group to 100% in the CORM-3-treated mice). Moreover, treatment with CORM-3 reduced bacterial counts in the spleen of immunosuppressed mice to a similar degree to in immunocompetent mice, suggesting a direct, rather than host-mediated, antibacterial effect of CORM-3 [66].

The modes of action in vivo of these compounds are still unknown and require further assessment. The promising results of these studies, however, pave the way for a more extensive preclinical evaluation of the antibacterial efficacy of Ru-based CORMs.

#### *5.5. Ruthenium Complexes in Antimicrobial Photodynamic Therapy*

Photodynamic therapy (PDT) is a therapeutic strategy that makes use of a combination of photosensitive molecules, light, and molecular oxygen. PDT has been investigated against a range of medical conditions, including atherosclerosis, psoriasis, and malignant cancers [183,184]. Antimicrobial photodynamic therapy (aPDT) has been used against a variety of microbial pathogens (bacteria, fungi, and viruses). It relies on the ability of a compound, a photosensitizer, to generate singlet oxygen (1O2) and other ROS upon light irradiation, causing bacteria inactivation [185].

Ru(II) complexes, particularly Ru(II)–polypyridyl complexes, have been intensively investigated for PDT applications against malignant cancers due to their optical properties, such as the long luminescence lifetimes of the triplet metal-to-ligand charge transfer (MLCT) excited state [184,186]. The remarkable potential of Ru complexes as PDT agents has been confirmed by TLD-1433 [18], which is currently undergoing phase II clinical studies as a photosensitizer for PDT against bladder cancer.

Taking into account the remarkable success of Ru(II)-based PDT agents in the treatment of cancer, several Ru(II) complexes have been considered as potential photosensitizers for aPDT. For instance, [Ru(dmob)3] 2+ (Figure 20), where dmob = 4,4'-dimethoxy-2,2' bipyridine, was more active than the corresponding complexes bearing bpy and phen ligands against *S. aureus*, *P. aeruginosa,* and *C. albicans* strains. The enhanced activity was attributed to the increased lipophilicity of the complex due to the presence of methoxy groups in its structure, which can translate to enhanced uptake by the bacterial cells [68].

**Figure 20.** Chemical structures of ruthenium complexes developed for Antimicrobial Photodynamic Therapy.

[Ru(bpy)2(dppn)]2+ (Figure 20), where bpy = 2,2'-bipyridine and dppn = 4,5,9,16 tetraazadibenzo[*a,c*]-napthacene, was shown to cause potent photoinactivation of *E. coli* cells, while dark incubation with the compound had no effect on the viability of the

microorganism. Treatment with [Ru(bpy)2(dppn)]2+ led to a 70% CFU decrease at 0.1 µM and complete inactivation at 0.5 µM following light activation [187].

The ruthenium(II) complex *cis*-[Ru(bpy)2(INH)2] 2+ (Figure 20), where INH = isoniazid, has been shown to undergo stepwise photoactivation in aqueous media after irradiation with 465 nm blue light. The resulting products of this process are two equivalents of the antituberculosis drug isoniazid and *cis*-[Ru(bpy)2(H2O)2] 2+ . *cis*-[Ru(bpy)2(INH)2] 2+ was inactive in the dark; however, upon photoactivation, it was 5.5-fold more efficient against *Mycobacterium smegmatis* in comparison with isoniazid. Notably, *cis*-[Ru(bpy)2(INH)2] 2+ displayed high selectivity towards mycobacteria over healthy MRC-5 human lung cells in vitro [69].

A heterobimetallic complex [Ru(Ph2phen)2(dpp)PtCl2] 2+ (Figure 20), where Ph2phen = 4,7-diphenyl-1,10-phenanthroline and dpp = 2,3-bis(2-pyridyl)pyrazine, has been reported to induce photocytotoxic effects in *E. coli* cells in the presence of oxygen and visible light. The dose required for complete cell growth inhibition under visible light irradiation was 5 µM, as opposed to 20 µM in the dark [70]. In comparison, cisplatin induced complete cell growth inhibition at 5 µM in the dark, but a similar complex, [RuCl(tpy)(dpp)PtCl2] + (see above), had the same effect at 200 µM [145]. Inside the cells, photoactivated [Ru(Ph2phen)2(dpp)PtCl2] 2+ was shown to bind to chromosomal DNA [70]. Further experiments are needed to assess the nature of the DNA binding, as well as what species are responsible for the activity.

Incorporation in or conjugation with biocompatible polymers has been used as an efficient strategy to increase the ability of ruthenium complexes to penetrate bacterial cell walls and therefore their antimicrobial activity. A Ru(II)–polypyridyl complex, [Ru(bpy)2-dppz-7-hydroxymethyl][PF6]<sup>2</sup> (RuOH), where bpy = 2,2'-bipyridine and dppz = dipyrido[3,2-*a*:2;2',3'-*c*]phenazine, was found to be inactive against Gram-positive and Gram-negative bacteria. This lack of activity was thought to stem from its low uptake by bacterial cells. In order to solve this issue, RuOH was conjugated to the end-chain of a hydrophobic polylactide (PLA) polymer to form ruthenium–polylactide (RuPLA) nanoconjugates (Figure 20). Although RuPLA nanoconjugates displayed superior photophysical properties, including luminescence and enhanced <sup>1</sup>O<sup>2</sup> generation, they were only moderately active against Gram-positive (*S. aureus*, *S. epidermidis*) bacteria, with MIC values of 25 µM. The RuPLA nanoconjugates remained non-toxic to the Gram-negative (*E. coli* and *P. aeruginosa*) bacterial strains and displayed phototoxicity against human cervical carcinoma cells (IC<sup>50</sup> = 4.4 µM) [185].

In a recent study, the antibacterial activity of the purely inorganic polymer [Ru(CO)2Cl2]<sup>n</sup> (Figure 20), with repeating dicarbonyldichlororuthenium (II) monomers, was studied against *E. coli* and *S. aureus*. Significant inhibitory effects were observed on both strains at concentrations as low as 6.6 ng/mL after irradiation with 365 nm UV light. Interestingly, the polymer displayed stronger photobactericidal activity against the Gram-negative *E. coli* (MIC ~33 ng/mL) than the Gram-positive *S. aureus* (MIC ~166 ng/mL) bacteria. In addition, [Ru(CO)2Cl2]<sup>n</sup> remained non-toxic to human dermal fibroblasts and red blood cells at concentrations much higher than the MIC values. Although the complex was considerably toxic to both bacterial strains in the dark, the antibacterial activity of [Ru(CO)2Cl2]<sup>n</sup> was significantly increased upon photoirradiation, which can be attributed to the enhanced generation of ROS under UV light. SEM analysis revealed that its mode of action might involve disruption of bacterial membranes. Moreover, [Ru(CO)2Cl2]<sup>n</sup> was able to cause morphological changes to biofilm structures and to disassemble the biofilm matrix [71]. It should be noted that although the structure of [Ru(CO)2Cl2]<sup>n</sup> is similar to that of CORMs, there is no information available in the literature on the ability of the polymer to release CO or undergo ligand exchange in aqueous media.

The antibacterial photosensitizing activity towards a panel of bacterial strains has been assessed for seventeen homo- or heteroleptic polypyridyl Ru(II) complexes with the following formulae: [Ru(Phen)3](PF6)2, [Ru(Phen)2(Phen-X)](PF6)2, [Ru(Phen)(Phen-X)2](PF6)2, [Ru(Phen-X)3](PF6)2, [Ru(Phen-X)2Cl2], or [Ru(Phen)2Cl2] (Figure 21), varying

due to the number and the nature of the substituents. With regard to the most active complexes, **2**, **5**, and **6** stood out, **5** was highly efficient against MRSA N315 even without light irradiation, and **2** demonstrated activity against four *S. aureus* strains, one *E. coli* strain, and three *P. aeruginosa* strains. However, **2** and **5** were more toxic towards eukaryotic cells upon light irradiation, with **6** being non-toxic. The counterion (PF<sup>6</sup> − vs. Cl−) did not appear to have a significant effect on the antibacterial activity. In contrast, a dicationic charge was vital to the activity, taking into account that the two neutral Ru(II) complexes, **16** and **17**, were inactive. Surprisingly, the best photosensitizers for <sup>1</sup>O<sup>2</sup> production (**8**, **9**, **10**, **15**) did not correspond to the most efficient aPDT agents (**2**, **5**, **6**). The ability of the complexes to interact efficiently with bacteria seems to be crucial for aPDT activity, considering the short half-life of ROS generated upon light irradiation. Thus, solely increasing <sup>1</sup>O<sup>2</sup> production is not sufficient to yield more efficient aPDT agents. Parameters impacting the interactions with bacteria, such as lipophilicity and ability to form aggregates, should also be considered in the development of optimized future compounds for aPDT [184].

**Figure 21.** Chemical structures of the homo- or heteroleptic polypyridyl Ru(II) complexes **(1)**–**(17)** with the general formulae [Ru(Phen)<sup>3</sup> ](PF<sup>6</sup> )2 , [Ru(Phen)<sup>2</sup> (Phen-X)](PF<sup>6</sup> )2 , [Ru(Phen)(Phen-X)<sup>2</sup> ](PF<sup>6</sup> )2 , [Ru(Phen-X)<sup>3</sup> ](PF<sup>6</sup> )2 , [Ru(Phen-X)2Cl<sup>2</sup> ], or [Ru(Phen)2Cl<sup>2</sup> ]. The core structures of the complexes **(1)**–**(17)** correspond to either (**a**) or (**b**), as denoted in the top right corner of the figure. The fluorene unit was bonded to the 1,10-phenanthroline moiety ligand either directly (Fluorenyl, bottom right corner) or via a triple bond (T-Fluorenyl).

#### **6. Antiparasitic Activity of Ruthenium Complexes**

Parasitic infections, including malaria (*Plasmodium* sp.), Chagas' disease (*Trypanosoma cruzi*), African trypanosomiasis (*Trypanosoma brucei*), and leishmaniasis (*Leishmania* sp.), mainly affect the tropical and subtropical regions of Africa and Asia and only a narrow spectrum of effective drugs is available for treatment. In this context, several ruthenium complexes have been reported as efficient antiparasitic agents active against malaria, leishmaniasis, trypanosomiasis, and Chagas' disease [188]. Generally, the enhanced antiplasmodial activity of these complexes when compared to the free ligands is thought to be related to their increased lipophilicity, which translates to increased uptake into the parasite's cells and/or increased ability to evade the parasite's drug efflux pumps.

#### *6.1. Antiplasmodial Activity*

Malaria is a highly infectious parasitic disease, with over 40% of the world's population living in an endemic region. Malaria parasites belong to the genus *Plasmodium*, the most virulent strain being *P. falciparum* [189,190]. Conventional treatment strategies use either quinoline-based drugs, such as chloroquine (CQ)) and its derivatives, or fixed-dose combination therapies containing a derivative of the Chinese natural product artemisinin. The increasing widespread resistance to these compounds requires urgent attention to the development of new therapeutic strategies [190,191].

An organometallic complex, [RuCl2(CQ)]<sup>2</sup> (Figure 22a), where CQ = chloroquine, displayed 2–5-fold increased activity against *P. falciparum* compared with chloroquine diphosphate in vitro [192]. Moreover, the complex was significantly more active when compared with its organic derivative in mice infected with *Plasmodium berghei*, with no apparent signs of acute toxicity up to 30 days after treatment [193]. [RuCl2(CQ)]<sup>2</sup> was shown to bind to hematin and inhibit aggregation of β-hematin (synthetic hemozoin—a target of the malaria parasite) in vitro, albeit to a slightly lower extent than chloroquine diphosphate. However, the heme aggregation inhibitory activity of the complex is significantly higher than that displayed by chloroquine, suggesting that the main target of the complex is the heme aggregation process. [RuCl2(CQ)]<sup>2</sup> was shown to be significantly more lipophilic than chloroquine diphosphate, suggesting that the addition of Ru(II) induced drastic changes in the pharmacokinetic profile of the organometallic compound. One chlorido ligand from each of the two Ru(II) centers is displaced by water molecules upon addition to aqueous solutions. The resulting species, [RuCl(OH2)3(CQ)]2[Cl]2, is considered to be the active species in vitro and in vivo. The enhanced activity of the complex against CQ-resistant strains of *P. falciparum* was suggested to relate to its lipophilicity. This can be explained by the fact that the parasite efflux pump, usually involved in the resistance mechanism to chloroquine, has a lower ability to bind to highly lipophilic drugs [192].

**Figure 22.** Chemical structures of ruthenium complexes with antiplasmodial activity. (**a**) [RuCl<sup>2</sup> (CQ)]<sup>2</sup> , (**b**) cyclometallated Ru(II) complexes of 2-phenylbenzimidazoles, and (**c**) PTA-derived ruthenium(II) quinoline complexes.

CQ has also been used as a chelating ligand in a series of organoruthenium complexes with the general formula [RuCQ(η 6 -C10H14)(N–H)]2+, where η 6 -C10H<sup>14</sup> is *α*-phellandrene and N–H is either 2'-bipyridine (BCQ), 5,5'-dimethyl-2,2'-bipyridine (MCQ), 1,10-phenanthroline (FCQ), or 4,7-diphenyl-1,10-phenanthroline (FFCQ). As was previously shown for [RuCl2(CQ)]2, the Ru–CQ bonds are stable, and CQ is not released upon aquation. The organoruthenium complexes displayed intraerythrocytic activity against CQ-sensitive and -resistant strains of *P. falciparum*. Unlike CQ, the complexes exerted moderate activity against the liver stage and potent activity against the sexual stage of the parasite, suggesting that they operate via a different mechanism than that of CQ. It has been shown that [RuCQ(η 6 -C10H14)(N–H)]2+ induces oxidative stress in the parasite, which might be linked to their mode of action. In addition, the organoruthenium complexes displayed low mammalian cytotoxicity and inhibited parasitemia in mice infected with *P. berghei* [194].

A range of Ru(II)–arene complexes were developed in the knowledge that increasing the lipophilic properties of a drug is likely to increase passive diffusion through membranes and hence the antiplasmodial activity. For instance, a series of half-sandwich Ru(II) complexes with aryl-functionalized organosilane thiosemicarbazone ligands were more active against *P. falciparum* at low micromolar concentrations (2.29–6.66 µM) and less cytotoxic to the Chinese Hamster Ovarian (CHO) cell line in comparison with the corresponding free ligands. It should be stated that the activity of the complexes was still much lower than that of both CQ and artesunate, which were used as controls. However, the complexes also displayed much lower resistance index values relative to the control drugs, which suggests that the parasites are less likely to develop cross-resistance to the metal complexes [195].

An enhancement of the antiplasmodial activity has also been observed for cyclometallated Ru(II) complexes of 2-phenylbenzimidazoles (Figure 22b), when compared with the free ligands. These complexes were found to be active against CQ-sensitive and multidrug-resistant *P. falciparum* strains, with IC<sup>50</sup> values in the low to submicromolar range (0.12–3.02 µM). The nature of the substituent on the η 6 -*p*-cymene moiety does not seem to influence the activity to a great extent. Although CQ was still more active than the cyclometallated complexes against both strains, the latter displayed lower resistance index values relative to CQ. In addition, the metal complexes displayed relatively low cytotoxicity against the mammalian CHO cells. Notably, the Ru(II) complexes were found to be more active than the Ir(III) analogues on the resistant strain [196], which was also reported for other Ru(II)–arene complexes [197]. PTA-derived ruthenium(II) quinoline complexes (Figure 22c) were, however, generally less effective against CQ-sensitive and resistant strains of *P. falciparum* than their Ir(III) correspondents, but were also much less toxic to the CHO cells. In addition, these RAPTA-like complexes inhibited β-hematin formation, suggesting that their mechanism of action is similar to that of CQ [198].

Di- and tri- nuclear Ru(II)-η 6 -*p*-cymene complexes (Figure 23a), in which the ruthenium centers are bridged by pyridyl aromatic ether ligands, were evaluated against CQsensitive and -resistant *P. falciparum* strains. While the dinuclear derivative displayed only moderate activity, the trinuclear complex proved to be highly active in both strains, displaying activities in the nanomolar range (IC<sup>50</sup> = 240 nM and 670 nM for the CQ-sensitive and -resistant *P. falciparum* strains, respectively). The trinuclear complex was also able to inhibit more efficiently β-hematin formation in vitro, in comparison with the dinuclear derivative, which suggests that hemozoin might be a target of the complexes in vivo. Notably, the trinuclear Ru(II) complex was only slightly more toxic than the corresponding tripyridyl ether ligand, indicating that it was the incorporation of a triazine moiety that had a more significant impact on activity [197]. This was confirmed by the fact that trinuclear Ru(II) η 6 -*p*-cymene complexes, in which the ruthenium centers are bridged by pyridyl aromatic ester ligands lacking the triazine moiety (Figure 23b), are much less efficient antiparasitic agents [199].

**Figure 23.** Di- and tri- nuclear Ru(II)-η 6 -*p*-cymene complexes in which the ruthenium centers are bridged by (**a**) pyridyl aromatic ether ligands and (**b**) pyridyl aromatic ester ligands.

Using 'old' drugs to assist in the search for new agents that are more efficient for either common or rare diseases is the scope of a relatively new therapeutic strategy called drug repositioning/repurposing. For instance, a series of ferrocenyl and ruthenocenyl derivatives incorporating tamoxifen-based compounds were tested against CQ-resistant *P. falciparum* blood forms. Tamoxifen (Figure 24) is an anticancer agent used in current treatment plans to prevent and treat breast cancer. The results within this series indicated that the ruthenocenyl-containing complexes (Figure 24) were more active than their ferrocenyl analogues, but still only displayed moderate activity (IC<sup>50</sup> = 4.7–16.5 mM) against *P. falciparum*. The ruthenocenyl complexes were considered nontoxic to HepG2 cells [200].

**Figure 24.** Chemical structures of tamoxifen and the ruthenocenyl complexes incorporating tamoxifen-based ligands.

#### *6.2. Antitrypanosomal Activity*

Chagas' disease (American trypanosomiasis) affects millions of people worldwide, mainly in Central and South America, where the disease is endemic. It is a life-threatening

disease caused by the parasite *Trypanosoma cruzi*. No vaccines are currently available, and treatment options are limited to only two drugs, benznidazole and nirfurtimox [201]. Sleeping sickness (African trypanosomiasis) predominantly affects people living in sub-Saharan Africa and is transmitted by the bite of the tsetse fly. The disease is caused by the insect-borne *T. brucei* parasite [202].

Two Ru(II)–NO donor compounds, namely *trans*-[Ru(NO)(NH3)4(isn)](BF4)<sup>3</sup> (Figure 25) where isn = isonicotinamide and *trans*-[Ru(NO)(NH3)4(imN)](BF4)<sup>3</sup> (Figure 25) where imN = imidazole, displayed significant activity against *T. cruzi* both in vitro and in vivo. NO release upon reduction of the ruthenium nitrosyls in culture cells and animal models is thought to play an essential role in the antiproliferative and trypanocidal activities. Notably, *trans*-[Ru(NO)(NH3)4(imN)](BF4)<sup>3</sup> allowed for survival of up to 80% of infected mice at a much lower dose (100 nmol/kg/day) than that required for benznidazole (385 µmol/kg/day) [203,204]. Ru(II) complexes with the formulae *cis*-[Ru(NO)(bpy)2(imN)] (PF6)<sup>3</sup> and *cis*-[Ru(NO)(bpy)2SO3]PF<sup>6</sup> displayed inhibitory effects on the *T. cruzi* glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (IC<sup>50</sup> = 89 and 153 µM, respectively), which is a potential molecular target. These compounds exhibited in vitro and in vivo trypanocidal activities at doses up to 1000-fold lower than the clinical dose for benznidazole [205]. Furthermore, in a series of nitro/nitrosyl-Ru(II) complexes, *cis*, *trans*- [RuCl(NO)(dppb)(5,5'-mebipy)](PF6)2, where 5,50 -mebipy = 5,50 -dimethyl-2,20 -bipyridine and dppb = 1,4-*bis*(diphenylphosphino)butane, was the most active compound. The complex displayed an IC<sup>50</sup> of 2.1 µM against trypomastigotes (the form of the parasite during the acute stage of the disease) and an IC<sup>50</sup> of 1.3 µM against amastigotes (the form of the parasite during the chronic stage of the disease), while it was less toxic to macrophages. Moreover, the complex exerted synergistic activity with benznidazole in vitro against trypomastigotes and in vivo in infected mice [206].

**Figure 25.** Chemical structures of the ruthenium NO-donor complexes *trans*-[Ru(NO)(NH<sup>3</sup> )4 (isn)]3+ and *trans*-[Ru(NO)(NH<sup>3</sup> )4 (imN)]3+ .

A series of symmetric trinuclear ruthenium complexes bearing azanaphthalene ligands with the general formula [Ru3O(CH3COO)6(L)3]PF<sup>6</sup> (Figure 26), where L = (1) quinazoline (qui), (2) 5-nitroisoquinoline (5-nitroiq), (3) 5-bromoisoquinoline (5-briq), (4) isoquinoline (iq), (5) 5-aminoisoquinoline (5- amiq), and (6) 5,6,7,8-tetrahydroisoquinoline (thiq), were developed. All complexes presented in vitro trypanocidal activity, complex 6 being the lead compound of the series, with IC<sup>50</sup> values of 1.39 µM against trypomastigotes and 1.06 µM against amastigotes. Complex 6 was up to 10 times more effective than benznidazole, while being essentially non-toxic to healthy mammalian cells (SI trypomastigote: 160, SI amastigote: 209) [207].

**Figure 26.** Chemical structures of symmetric trinuclear ruthenium complexes bearing azanaphthalene ligands with the general formula [Ru3O(CH3COO)<sup>6</sup> (L)<sup>3</sup> ]PF<sup>6</sup> .

A range of Ru (II)–cyclopentadienyl thiosemicarbazone complexes displayed sub- or micromolar IC<sup>50</sup> values against *T. cruzi* and *T. brucei*. Notably, [RuCp(PPh3)L] (Figure 27), where HL is the *N*-methyl derivative of 5-nitrofuryl containing thiosemicarbazone and Cp is cyclopentadienyl, exhibited high (IC<sup>50</sup> *T. cruzi* = 0.41 µM; IC<sup>50</sup> *T. brucei* = 3.5 µM) and selective activity (SI *T. cruzi* > 49 and SI *T. brucei* SI > 6). These complexes had the ability to interact with DNA in vitro, but no correlation with the biological activity was observed [208]. A Ru(II)–cyclopentadienyl clotrimazole complex, [RuCp(PPh3)2(CTZ)](CF3SO3) (Figure 27), where Cp = cyclopentadienyl and CTZ = clotrimazole, was more cytotoxic on *T. cruzi* than nifurtimox. With regard to its mechanism of action, the complex was shown to impair the sterol biosynthetic pathway in *T. cruzi* [209]. In another series of Ru(II)–cyclopentadienyl clotrimazole complexes, [RuII(*p*-cymene)(bpy)(CTZ)][BF4]<sup>2</sup> (Figure 27) was found to be the most active compound, increasing the activity of CTZ by a factor of 58 against *T. cruzi* (IC<sup>50</sup> = 0.1 µM)*,* with no appreciable toxicity to human osteoblasts [210].

**Figure 27.** Chemical structures of Ru(II)–arene complexes with antitrypanosomal activity.

#### *6.3. Antileishmaniasis Activity*

Leishmaniasis is a disease caused by protozoan parasites of the genus *Leishmania* and is characterized by high morbidity. It is estimated that more than 1 billion people live in endemic areas, with more than 1 million new cases of leishmaniasis occurring annually. Current treatment for leishmaniasis relies on the use of pentavalent antimonials and other drugs, such as pentamidine isethionate, amphotericin B, and miltefosine. However, antileishmanial treatment cannot provide a sterile cure, and the parasite can cause a relapse when the human body is immunosuppressed [211,212].

An improved antiplasmodial activity in comparison with that of the free ligand has been reported for Ru(II)–lapachol complexes. [RuCl2(Lap)(dppb)] was active against *L. amazonensis* promastigotes and infected macrophages, with submicromolar IC<sup>50</sup> values comparable with that of the reference drug, amphotericin B. In addition, the complex was not toxic to macrophages at concentrations much higher than the IC<sup>50</sup> values [212].

[RuII(*p*-cymene)(bpy)(CTZ)][BF4]<sup>2</sup> (Figure 27) was found to be active against promastigotes of *L. major* at nanomolar concentrations (IC<sup>50</sup> = 15 nM) and displayed no appreciable toxicity against human osteoblasts (SI > 500). Moreover, in *L. major*-infected mice macrophages, the complex caused a significant inhibition of the proliferation of intracellular amastigotes (IC<sup>70</sup> = 29 nM) [210].

#### **7. Antiviral Activity of Ruthenium Complexes**

#### *7.1. Anti-HIV Activity*

The mixed-valent tetranuclear ruthenium–oxo oxalato cluster Na7[Ru4(µ3-O)4(C2O4)6] exerted promising anti-HIV-1 activity with over 98% inhibition of viral replication toward the R5-tropic HIV-1 strain at a 5 µM concentration and similar inhibitory activity toward X4-tropic viral replication. Moreover, the ruthenium–oxo oxalato cluster displayed selective anti-viral activity, with over 90% survival of the host cells registered at concentrations up to 50 µM. Notably, Na7[Ru4(µ3-O)4(C2O4)6] was 10-fold more effective against HIV-1 reverse transcriptase (IC<sup>50</sup> = 1.9 nM) than the commonly used HIV-1 RT inhibitor 3'-azido-3'-deoxythmidine-5'-triphosphate (IC<sup>50</sup> = 68 nM) [213].

Another ruthenium cluster, [H4Ru4(η 6 -*p*-benzene)4] 2+, displayed selective activity against Polio virus, without inhibiting the growth of healthy human cells. It has been suggested that the complex might only be cytotoxic in Polio-infected cells, as the virus alters cell membrane permeability, facilitating passage for the cluster [94].

[Ru(bpy)2eilatin]2+ (Figure 28), where eilatin = dibenzotetraazaperylene, inhibited HIV-1 replication in CD4+ HeLa cells and in human peripheral blood monocytes (IC<sup>50</sup> values ~1 µM). Eilatin is a fused, heptacyclic aromatic alkaloid that was isolated from the sea squirts belonging to *Eudistoma* sp., reported to possess cytotoxic and antiproliferative activities. [Ru(bpy)2eilatin]2+ is a kinetically inert complex and in vitro studies suggest that its mechanism of action relies upon inhibition of key protein–RNA interactions. The planar structure of the bidentate ligand, eilatin, was found to be essential for the activity of the complex [214].

**Figure 28.** Chemical structure of [Ru(bpy)2eilatin]2+ .

#### *7.2. Anti-SARS-Cov-2 Activity*

In spite of the extensive vaccination campaigns that are currently ongoing, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is spreading at an alarming pace across the world. The large number of mutations rendered several new variants less susceptible to treatment options, and possibly to vaccines. Thus, there is still an urgent need for the development of new drugs with a broader spectrum of activity [215,216].

BOLD-100 (sodium trans-[tetrachlorobis(1H-indazole)ruthenate(III)], KP1339, Figure 29), developed as an anticancer agent, was shown to selectively inhibit stress-induced upregulation of 78-kDa glucose-regulated protein (GRP78) [217–219], which is a master chaperone protein serving critical functions in the endoplasmic reticulum of normal cells [220,221]. The interaction of SARS-CoV-2 spike protein with the GRP78 protein located on the cell membrane can mediate viral entry. Therefore, disruption of this interaction may be used to develop novel therapeutic strategies against coronavirus [222]. Indeed, BOLD-100 was reported to reduce viral loads in various COVID-19 variants, including the more virulent B.1.1.7, originally identified in the United Kingdom. Unlike vaccines, which are more effective against certain viral variants, BOLD-100, with a broad antiviral mechanism of action, appears to remain active on all mutant strains [223]. In vivo studies are currently in progress.

**Figure 29.** Chemical structure of BOLD-100 (sodium trans-[tetrachlorobis(1H-indazole)ruthenate(III)], KP1339).

BOLD-100 has a tolerable safety profile (minimal neurological or hematological effects), as was shown in a recently completed phase I clinical study involving 41 patients with advanced cancer. Moreover, it is currently undergoing clinical trials in combination with FOLFOX chemotherapy (which includes folinic acid, 5-fluorouracil, and oxaliplatin) for gastrointestinal solid tumors [224]. Therefore, BOLD-100 has already been successfully developed as a clinical-stage product, which suggests its potential for rapid further clinical development against COVID-19.

Additionally, [Ru(bpy)3] 2+ is used in the Elecsys® Anti-SARS-CoV-2, a chemiluminescence immunoassay intended for qualitative detection of antibodies to SARS-CoV-2 in human serum and plasma, which has been approved worldwide. In this assay, the SARS-CoV-2-specific recombinant antigen is labeled with the ruthenium complex [225,226]. Other metal complexes identified as potential anti-SARS-Cov-2 agents include auranofin [227,228] and Re(I) tricarbonyl complexes [229].

#### **8. Conclusions**

Ruthenium-based antimicrobial agents have a fairly complex mode of action involving multiple mechanisms acting in synergy. The knowledge gained so far in this area suggests that the activity of ruthenium compounds against microbial cells is based upon their ability to induce oxidative stress, interact with the genetic material, proteins, or other intracellular targets, and/or damage the cell membranes. The complex interplay between these modes of action is likely responsible for the activity of some ruthenium-based compounds against drug-resistant strains.

Generally, ruthenium complexes exert excellent activity against Gram-positive bacteria (e.g., *S. aureus* and MRSA) and, with some exceptions (see, for instance, the dinuclear polypyridylruthenium(II) complexes and ruthenium-based CORMs), display lower activity towards Gram-negative strains (e.g., *E. coli* and *P. aeruginosa*). With regard to their activity against Gram-negative bacteria, one can notice a trend towards higher efficacy against *E. coli* when compared with *P. aeruginosa* and *K. pneumoniae*. For most classes of compounds, activity towards both Gram-negative and Gram-positive strains has been correlated to the uptake of the complex into the cells.

Additionally, this work highlights recent advances in ruthenium-based compounds that are active against neglected tropical diseases caused by parasites, such as malaria, Chagas' disease, and leishmaniasis. Notably, several complexes possess excellent activity, at submicromolar concentrations, results that raise awareness about the potential use of ruthenium compounds as effective antiparasitic agents. Moreover, the antiviral activity of ruthenium complexes, particularly the anti-HIV and anti-SARS-Cov-2 activities, has been reviewed herein. It is worth noting that BOLD-100 (formerly denoted KP1339) displays a broad antiviral mechanism of action and appears to remain active on all mutant strains of the SARS-Cov-2 virus.

In general terms, ruthenium complexes have been shown to display low levels of toxicity towards healthy eukaryotic cells in vitro and in vivo. This finding underlines the potential of these compounds for future clinical development, since selective toxicity against microbial over host cells in vitro and in vivo is imperative for a potential drug to advance in clinical trials. More in vivo studies are clearly needed in order to provide proof beyond a reasonable doubt that ruthenium complexes are strong candidates in the field of antimicrobial drug discovery.

In conclusion, this work aimed to highlight the potential of ruthenium-based compounds as novel antimicrobial agents due to the diverse range of complex 3D structures and modes of action they provide. Given that the pipeline of new antibiotics is running dry, the ruthenium species with high activity and selectivity presented herein may represent the starting point for a much-needed new class of antimicrobial agents. Therefore, we hope that this review will succeed in raising awareness about the potential of ruthenium complexes for antimicrobial applications and spur further research into their development.

**Author Contributions:** A.-C.M. performed a literature search, conceptualized the manuscript, and contributed to the writing of, corrections to, and the final shape of the manuscript. V.U. performed a literature search and contributed to the writing of and corrections to the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by The Executive Unit for the Financing of Higher Education, Research, Development, and Innovation (UEFISCDI), Project No. 383PED/2020 and Project No. PD 219/2020.

**Acknowledgments:** We thank Joseph Cowell for proofreading the manuscript and improving the use of English throughout the paper.

**Conflicts of Interest:** The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **References**

