**Contents**



## **About the Editors**

#### **C ´elia F. Rodrigues**

Celia F. Rodrigues (PharmD, PhD) is a *Candida* spp. expert, with extensive know-how in susceptibility assays, biofilm development, antimicrobial drugs, molecular techniques, alternative and novel treatments, and biomaterials. Presently, she is an invited assistant professor at CESPU, where she teaches future pharmacists, and she is a PhD member of TOXRUN (Toxicology Research Unit - CESPU). She has also worked at LEPABE, Faculty of Engineering, the University of Porto on a project related to microorganisms, FISH, and microfluidics, where she maintains a close collaboration. Celia is a reviewer for more than 50 international journals and has co-supervised/mentored MSc, PhD Students, and Post-Doctoral researchers, organized research conferences/seminars, and has been a jury of grants and congress. Finally, Celia has won several grants and awards from portuguese and international entities. (https://www.researchgate.net/profile/Celia Rodrigues2; Ciencia ID: 5F12-D3E1-E028).

#### **Jesus A. Romo**

Dr. Jesus Romo is originally from Mexico but moved to Texas at an early age. He received his PhD in Cell and Molecular Biology (Microbiology and Immunology Track) from The University of Texas at San Antonio, where his research focused on developing anti-virulence therapies for the treatment of fungal infections caused by opportunistic pathogenic fungi from the genus Candida. More specifically, he focused on inhibiting biofilm formation and preventing the development of drug resistance. Currently, he is an NIH IRACDA postdoctoral scholar at Tufts University School of Medicine in the department of Molecular Biology and Microbiology studying the role of fungal colonizers of the gastrointestinal tract during infection by the bacterial pathogen *Clostridioides difficile*. Jesus is pursuing a career in academia because of his passion for research and love for teaching and mentoring the next generation of underrepresented scientists.

### *Editorial* **Fungal Biofilms 2020**

**Célia F. Rodrigues 1,\* and Jesus A. Romo 2,\***


Fungal infections are an important and increasing global threat, carrying not only high morbidity and mortality rates, but also extraordinary healthcare costs. Without an effective response, it is predicted that 10 million people will die per year because of multidrug-resistant pathogens. A high percentage of the mortalities caused by fungi are known to have a biofilm etiology [1–4]. In fact, biofilms are the predominant mode of fungal growth. They have several ecologic benefits, for example higher nutrient availability, metabolic cooperation, protection from the environmental stresses, and acquisition of new and advantageous features. Besides, single-species and mixed-species biofilms are particularly problematic to eradicate, being, thus, the foundation of chronic infections, particularly if medical devices are existent [5].

A total of ten papers were published in this Special Issue including three reviews and six original articles. These cover a wide range of topics with original research on polymicrobial biofilms of fungi and bacteria, small molecule screening, characterization of the impact of current antifungals on biofilms of non-*albicans* species, characterization of non-*albicans* species biofilm matrix, and biofilms of *Aspergillus fumigatus.* Additionally, review articles cover the antifungal effect of *quorum sensing* molecules on *Candida* biofilms, sexual biofilms of *Candida albicans*, and a compilation of plant derived compounds and their activities against biofilms formed by *Candida* species.

The reports describe original research in the area of antimicrobials and include work involving individual and combinatorial efficacy of compounds with specific activity against fungi, bacteria, or both within polymicrobial biofilms [6], a screen of a small molecule library alone or in combination with current antifungals in search of compounds with antibiofilm and pre-formed biofilm activities [7], characterization of the effect of echinocandins against planktonic and biofilm lifestyles of clinical isolates from the *Candida haemulonii* complex [8], and the use of a membranotropic peptide to disrupt polymicrobial biofilms of *Candida albicans* and *Klebsiella pneumoniae* [9]. Additional reports phenotypically characterized colonies from *Candida parapsilosis* clinical isolates as a way to predict their biofilm formation capabilities [10], conducted analyses of the biofilm matrix composition from the *Candida haemulonii* species complex [11], and investigated the virulence and biofilm capabilities of an *Aspergillus fumigatus* environmental isolate with interest in the role of this isolate in the textile industry [12].

The reviews in this Special Issue covered recent developments in the area of *Candida albicans* sexual biofilms specifically focusing on how they are formed, their physical characteristics, and their role in *Candida* biology [13], the properties of the quorum sensing molecules farnesol and tyrosol secreted by *Candida* and their effect as anti-biofilm agents [14], and an extensive compilation of plant derived compounds with activities against biofilms of distinct *Candida* species [15]. Overall, this Special Issue is a great resource highlighting novel work on fungal biofilms.

**Citation:** Rodrigues, C.F.; Romo, J.A. Fungal Biofilms 2020. *J. Fungi* **2021**, *7*, 603. https://doi.org/10.3390/ jof7080603

Received: 22 July 2021 Accepted: 23 July 2021 Published: 26 July 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/).

**Acknowledgments:** C.F.R. would like to acknowledge the UID/EQU/00511/2020 Project—Laboratory of Process Engineering, Environment, Biotechnology and Energy (LEPABE)—financed by national funds through FCT/MCTES (PIDDAC).

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

#### **References**


## *Review* **Plant Preparations and Compounds with Activities against Biofilms Formed by** *Candida* **spp.**

**Tomasz M. Karpi ´nski 1,\* , Marcin Ozarowski ˙ <sup>2</sup> , Agnieszka Seremak-Mrozikiewicz 3,4,5, Hubert Wolski 3,6 and Artur Adamczak <sup>7</sup>**


**Abstract:** Fungi from the genus *Candida* are very important human and animal pathogens. Many strains can produce biofilms, which inhibit the activity of antifungal drugs and increase the tolerance or resistance to them as well. Clinically, this process leads to persistent infections and increased mortality. Today, many *Candida* species are resistant to drugs, including *C. auris*, which is a multiresistant pathogen. Natural compounds may potentially be used to combat multiresistant and biofilm-forming strains. The aim of this review was to present plant-derived preparations and compounds that inhibit *Candida* biofilm formation by at least 50%. A total of 29 essential oils and 16 plant extracts demonstrate activity against *Candida* biofilms, with the following families predominating: Lamiaceae, Myrtaceae, Asteraceae, Fabaceae, and Apiacae. *Lavandula dentata* (0.045–0.07 mg/L), *Satureja macrosiphon* (0.06–8 mg/L), and *Ziziphora tenuior* (2.5 mg/L) have the best antifungal activity. High efficacy has also been observed with *Artemisia judaica*, *Lawsonia inermis*, and *Thymus vulgaris*. Moreover, 69 plant compounds demonstrate activity against *Candida* biofilms. Activity in concentrations below 16 mg/L was observed with phenolic compounds (thymol, pterostilbene, and eugenol), sesquiterpene derivatives (warburganal, polygodial, and ivalin), chalconoid (lichochalcone A), steroidal saponin (dioscin), flavonoid (baicalein), alkaloids (waltheriones), macrocyclic bisbibenzyl (riccardin D), and cannabinoid (cannabidiol). The above compounds act on biofilm formation and/or mature biofilms. In summary, plant preparations and compounds exhibit anti-biofilm activity against *Candida*. Given this, they may be a promising alternative to antifungal drugs.

**Keywords:** *Candida*; biofilm; treatment; antifungals; natural compounds; essential oil; extract; minimal inhibitory concentration (MIC)

#### **1. Introduction**

The genus *Candida* contains about 150 species; however, most are environmental organisms. The most medically important is *Candida albicans*, which accounts for about 80% of infections. *C. albicans* causes more than 400,000 cases of bloodstream life-threatening infections annually, with a mortality rate of about 42% [1]. *Candida* non-*albicans* species that

**Citation:** Karpi ´nski, T.M.; Ozarowski, M.; ˙ Seremak-Mrozikiewicz, A.; Wolski, H.; Adamczak, A. Plant Preparations and Compounds with Activities against Biofilms Formed by *Candida* spp. *J. Fungi* **2021**, *7*, 360. https:// doi.org/10.3390/jof7050360

Academic Editors: Célia F. Rodrigues and Jesus A. Romo

Received: 20 March 2021 Accepted: 1 May 2021 Published: 5 May 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/).

are mainly responsible for infections are *C. glabrata*, *C. parapsilosis*, *C. tropicalis*, *C. krusei*, and *C. dubliniensis* [2]. Less frequently identified are *C. guilliermondii*, *C. lusitaniae*, *C. rugosa*, *C. orthopsilosis*, *C. metapsilosis*, *C. famata*, *C. inconspicua*, and *C. kefyr* [3].

*C. albicans* is a member of the commensal microflora. It colonizes the oral mucosal surface of 30–50% of healthy people. The rate of carriage increases with age and in persons with dental prostheses up to 60% [4–6]. Opportunistic infection caused by *Candida* species is termed candidiasis. At least one episode of vulvovaginal candidiasis (or thrush) concerns 50 to 75% of women of childbearing age [7]. Candidiasis can also affect the oral cavity, penis, skin, nails, cornea, and other parts of the body. In immunocompromised persons, untreated candidiasis poses the risk of systemic infection and fungemia [5,8]. *Candida* can be an important etiological factor in the infection of chronic wounds that are difficult to treat; this is mainly related to the production of biofilm [9].

Treatment of candidiasis depends on the infection site and the patient's condition. According to guidelines, vulvovaginal candidiasis should be treated with oral or topical fluconazole; however, regarding *C. glabrata* infection, topical boric acid, nystatin, or flucytosine is suggested. In oropharyngeal candidiasis, the treatment options include clotrimazole, miconazole, or nystatin, and in severe disease, fluconazole or voriconazole. In candidemia and invasive candidiasis, the drugs of choice are echinocandins (caspofungin, micafungin, anidulafungin), fluconazole, or voriconazole; in resistant strains, amphoteticin B is used. In selected cases of candidemia caused by *C. krusei*, voriconazole is recommended [10–12]. More details can be found in the Guidelines of the Infectious Diseases Society of America [12] and the European Society of Clinical Microbiology and Infectious Diseases [11]. Increasingly, *Candida* species are becoming resistant to drugs. Marak and Dhanashree [13] tested the resistance of 90 *Candida* strains isolated from different clinical samples, such as pus, urine, blood, and body fluid. Their study revealed that about 41% of *C. albicans* strains are resistant to fluconazole and voriconazole. Simultaneously, about 41% of *C. tropicalis* strains are resistant to voriconazole and about 36% of strains to fluconazole. In strains of *C. krusei*, about 23% are resistant to fluconazole and about 18% to voriconazole. Rudramurthy et al. [14] studied resistance in *C. auris*, which is considered a multiresistant pathogen. Among 74 strains obtained from patients with candidemia, over 90% of strains were resistant to fluconazole and about 73% to voriconazole. Virulence factors of *Candida* species include the secretion of hydrolases, the transition of yeast to hyphae, phenotypic switching, and biofilm formation [15,16]. All microorganisms in biofilm form are more resistant to antimicrobial and host factors, which leads to difficulties in eradication [17]. It has also been shown that resistance to drugs increases significantly in the case of *Candida* biofilm occurrence. Biofilm prevents the spread of antifungals; moreover, fluconazole is bound by the biofilm matrix [18]. The formation of a *Candida* biofilm during infection increases mortality, length of hospital stay, and cost of antifungal therapy [19].

Due to the above, new antifungal drugs are sought that could effectively combat not only planktonic fungi but also fungal biofilms. The natural compounds offer promise, with many acting on *Candida* species or biofilms in vitro [20].

The aim of this review was to present plant-derived natural compounds that have an effect against biofilms formed by *Candida* species.

#### **2. Materials and Methods**

In this review, publications available in PubMed and Scopus databases and through the Google search engine were taken into account. The following keywords and their combinations were used: "antifungal," "Candida," "anti-biofilm," "biofilm," "plant," "compound," "extract," and "essential oil." The principal inclusion criterion was the inhibition of biofilm formation by at least 50%. We focused on biofilm inhibition assays, in which the time of culture allowed for *Candida* biofilm maturation was at least 24 hours. Articles from the year 2000 to the present were taken into account. All articles published in predatory journals were rejected.

#### **3. Results and Discussion**

#### *3.1. Plant Preparations That Display Activity against Candida Biofilms*

The present review includes 60 articles in which *Candida* biofilm formation was inhibited by at least 50%. It has been shown that preparations from 34 plants demonstrate activity against *Candida* biofilms. Among them were 29 essential oils and 16 extracts. The plants from the following families dominated: Lamiaceae (6 species in 5 genera), Myrtaceae (5 species in 4 genera), Asteraceae (4 species in 4 genera), Fabaceae (4 species in 3 genera), and Apiacae (4 species in 2 genera).

Plants from the Lamiaceae family had the best antifungal activity, including *Lavandula dentata* (0.045–0.07 mg/L) [21], *Satureja macrosiphon* (0.06–8 mg/L) [22], and *Ziziphora tenuior* (2.5 mg/L) [23]. *Artemisia judaica* (2.5 mg/L) from the Asteraceae family [24], *Lawsonia inermis* (2.5–12.5 mg/L) from the Lythraceae family [25], and *Thymus vulgaris* (12.5 mg/L) from the Lamiaceae family [26] likewise exhibited good antifungal activity (Table 1). All preparations were essential oils, with the exception of *Lawsonia inermis*, which was an extract. Most of the plant preparations presented in Table 1 acted on biofilm formation and/or mature biofilms.

**Table 1.** Antifungal (MICs) and anti-biofilm (inhibition >50%) activity of plant preparations (essential oils or extracts).



#### **Table 1.** *Cont.*


#### **Table 1.** *Cont.*

Legend: MIC—minimal inhibitory concentration; XTT—reduction assay of 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-[carbonyl(phenylamino)]- 2H-tetrazolium hydroxide; MTT—reduction assay of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide [54,55].

> Antibiofilm activity may vary between plants in the same family. For example, in the Lamiaceae family, essential oil from *Lavandula dentata* acted against *C. albicans* biofilm at concentrations of 0.045–0.07 μL/mL [21], while essential oil from *Satureja hortensis* acted against the same biofilm at concentrations of 400–4800 mg/L [51]. There may also be large differences within the same species, due to various reasons. This may be influenced by, for example, different research methodologies, the use of different strains of fungi, and different chemical compositions depending on the plant variety, country, and season of harvest. A notable example of such a difference is observed with *Mentha* × *piperita*. In studies by Benzaid et al. [44], essential oil of *M. piperita* acted against *Candida* biofilm at a concentration of 10 μL/mL. However, the work of Agarwal et al. [38] showed that the same essential oil was active at 800 μL/mL.

> Changes in the content of active substances were described by Gonçalves et al. [56]. They showed that in essential oil from *Mentha cervina* collected in August, the amount of

isomenthone was 8.7% and pulegone was 75.1%. However, in essential oil collected in February, the ratio of the two compounds reversed and amounted to 77.0% for isomenthone and 12.9% for pulegone. The method of obtaining the compounds likewise had an influence on their content in the final essential oil. In a study by Cavar et al. [ ´ 57], the composition of essential oils of *Calamintha glandulosa* differed depending on the extraction method. The level of menthone was 3.3% using aqueous reflux extraction, 4.7% using hydrodistillation, and 8.3% using steam distillation, while the concentration of shisofuran was only 0.1% using hydrodistillation and steam distillation, while aqueous reflux yielded 9.7%.

#### *3.2. Plant Compounds That Display Activity against Candida Biofilm*

It has been shown that 69 compounds obtained from plants demonstrate activity against *Candida* biofilms (Table 2). Among these, the most common are monotherpenes (20), followed by sesquiterpene lactones (7) and sesquiterpenes (6). Another big group is also phenolic compounds, including phenols (6), phenolic acids (5), phenolic aldehydes (2), polyphenols (2), and phenolic alcohol (1).

In terms of activity, large differences were found, depending on the authors cited. Eugenol and thymol serve as good examples. Both compounds exhibited excellent activity in some studies (from 12.5 mg/L for eugenol [58] and 1.56 mg/L for thymol [26]), and in other studies, the activity was very poor (up to 80,000 for both [59]). These differences may be related, for example, to a different purity of the compound, a different fungal suspension density, or even to the use of other *Candida* strains with different sensitivities to chemical substances. A number of other factors, such as the type of culture medium, pH of the medium, incubation time, and temperature may likewise influence the antimicrobial activity [20].

According to the European Committee on Antimicrobial Susceptibility Testing (EU-CAST), the antifungal clinical breakpoints are between 0.001 mg/L and 16 mg/L [60]. Using EUCAST guidelines in this review, the most active compounds that inhibit (>50%) *Candida* biofilm formation are lichochalcone A (from 0.2 mg/L) [61], thymol (from 3.12 mg/L) [26], dioscin (from 3.9 mg/L) [31], baicalein (from 4 mg/L) [62], warburganal (4.5 mg/L) [52], pterostilbene, waltheriones and riccardin D (both from 8 mg/L) [63–65], polygodial (10.8 mg/L) [52], cannabidiol and eugenol (both from 12.5 mg/L) [58,66], and ivalin (15.4 mg/L) [67]. It is interesting that monotherpenes, which represent the highest percentage of substances listed in Table 2, are not the most active compounds. The two larger groups with the best activity are phenolic compounds (thymol, pterostilbene, and eugenol), and sesquiterpene derivatives (warburganal, polygodial, and ivalin). Single compounds with the highest observed activity belong to chalconoids (lichochalcone A), steroidal saponins (dioscin), flavonoids (baicalein), alkaloids (waltheriones), macrocyclic bisbibenzyls (riccardin D), and cannabinoids (cannabidiol). Most of the compounds presented in Table 2 acted on biofilm formation and/or mature biofilm.




#### **Table 2.** *Cont.*


**Table 2.** *Cont.*


**Table 2.** *Cont.*

Legend: MIC—minimal inhibitory concentration; XTT—reduction assay of 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5- [carbonyl(phenylamino)]-2H-tetrazolium hydroxide; MTT—reduction assay of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide [54,55].

#### **4. Conclusions**

Plant preparations (essential oils and extracts) and pure compounds exhibit antibiofilm activity against *Candida* species. Some of them are characterized by high activity in concentrations below 16 mg/L. Given this activity at relatively low concentrations, some may prove to be promising alternatives to antifungal drugs, especially in the cases of resistant or multiresistant strains of *Candida*. Moreover, the simple chemical structures involved and relative ease of extraction from natural sources warrant further research into the development of new, promising, and much-needed plant-based antifungals.

**Author Contributions:** Conceptualization, T.M.K. and M.O.; methodology, T.M.K.; analysis of results, T.M.K. and M.O.; writing—original draft preparation, T.M.K., M.O., A.S.-M., H.W., and A.A.; writing—review and editing, T.M.K. and M.O.; supervision, T.M.K.; funding acquisition, T.M.K. and H.W. 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.

**Acknowledgments:** We are very grateful to Mark Stasiewicz for English language corrections.

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

#### **References**


## *Article* **Compounds with Distinct Targets Present Diverse Antimicrobial and Antibiofilm Efficacy against** *Candida albicans* **and** *Streptococcus mutans,* **and Combinations of Compounds Potentiate Their Effect**

**Carmélia Isabel Vitorino Lobo , Ana Carolina Urbano de Araújo Lopes and Marlise Inêz Klein \***

Department of Dental Materials and Prosthodontics, School of Dentistry, São Paulo State University (Unesp), Araraquara. Rua Humaitá, 1680, Araraquara, São Paulo 14801-903, Brazil; carmelialobo@gmail.com (C.I.V.L.); carolina.urbano@unesp.br (A.C.U.d.A.L.)

**\*** Correspondence: marlise.klein@unesp.br; Tel.: +55-16-3301-6410

**Abstract:** *Candida albicans* and *Streptococcus mutans* interact synergistically in biofilms associated with a severe form of dental caries. Their synergism is driven by dietary sucrose. Thus, it is necessary to devise strategies to hinder the development of those biofilms and prevent cavities. Six compounds [*tt*-farnesol (sesquiterpene alcohol that decreases the bacterium acidogenicity and aciduricity and a quorum sensing fungal molecule), myricetin (flavonoid that interferes with *S. mutans* exopolysaccharides production), two 2'-hydroxychalcones and 4'-hydroxychalcone (intermediate metabolites for flavonoids), compound 1771 (inhibitor of lipoteichoic synthase in Gram-positive bacteria)] with targets in both fungus and bacterium and their products were investigated for their antimicrobial and antibiofilm activities against single-species cultures. The compounds and concentrations effective on single-species biofilms were tested alone and combined with or without fluoride to control initial and pre-formed dual-species biofilms. All the selected treatments eliminated both species on initial biofilms. In contrast, some combinations eliminated the bacterium and others the fungus in pre-formed biofilms. The combinations 4'-hydroxychalcone+*tt*-farnesol+myricetin, 4'-hydroxychalcone+*tt*-farnesol+fluoride, and all compounds together with fluoride were effective against both species in pre-formed biofilms. Therefore, combinations of compounds with distinct targets can prevent *C. albicans* and *S. mutans* dual-species biofilm build-up in vitro.

**Keywords:** biofilm; *Candida albicans*; *Streptococcus mutans*; extracellular matrix; antimicrobial agents; antibiofilm agents

#### **1. Introduction**

Several human diseases are caused by dysbiotic biofilms, including tooth decay, periodontal diseases, and oral candidiasis [1]. *Candida albicans* is an opportunistic species that, when associated with *Streptococcus mutans,* contributes to forming a complex and organized biofilm that is more tolerant to environmental stresses, including antimicrobial [2]. The interaction between these two species is synergistic in the presence of dietary sucrose and leads to severe dental caries lesions [3]. Therefore, it is necessary to devise strategies to hinder the development of those biofilms.

Within the complex oral microbiota, *S. mutans* is a producer of the extracellular matrix and modulates cariogenic biofilm formation when sucrose from the host's diet is available [4]. This bacterium is acidogenic and aciduric but not the most numerous species in the mouth, and there are other acidogenic and aciduric microorganisms [5,6]. However, its exoenzymes glucosyltransferases (Gtfs) and fructosyltransferase (Ftf) use sucrose as a substrate for the synthesis of exopolysaccharides (α-glucans and fructans), important components of biofilm construction [4]. Gtfs also adsorb on the surface of other oral microorganisms, converting them into glucan producers [4]. *C. albicans* is one of the microorganisms to which Gtfs

**Citation:** Lobo, C.I.V.; Lopes, A.C.U.d.A.; Klein, M.I. Compounds with Distinct Targets Present Diverse Antimicrobial and Antibiofilm Efficacy against *Candida albicans* and *Streptococcus mutans,* and Combinations of Compounds Potentiate Their Effect. *J. Fungi* **2021**, *7*, 340. https://doi.org/10.3390/ jof7050340

Academic Editors: Célia F. Rodrigues and Jesus A. Romo

Received: 31 March 2021 Accepted: 25 April 2021 Published: 28 April 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/).

binds and form high amounts of exopolysaccharides [7]. This fungus is also acidogenic and aciduric [8], and oral biofilms could serve as reservoirs for it.

The extracellular matrix of *C. albicans* biofilms contains extracellular DNA, β-glucans, mannans, proteins, and lipids [9–12]. This matrix has been associated with resistance against antifungals [13,14]. Moreover, the biogenesis of this matrix is coordinated extracellularly, reflecting cooperative actions in the biofilm community [14]. Therefore, the production of exopolysaccharides synthesized on surfaces in situ allows adhesion and microbial accumulation [4] and contributes to the construction of the 3D matrix that surrounds and supports cells, forming an environment with acidic niches and limited diffusion [6,15]. Thus, strategies that control the matrix formation could prevent pathogenic biofilms development [13] and perhaps control the amount of *C. albicans* in oral biofilms that could flourish when conditions are favorable.

The therapeutic modalities for controlling dental biofilm formation are still limited. Chlorhexidine is a broad-spectrum antimicrobial agent that suppresses microorganisms levels in saliva but is not effective against mature biofilms, and its daily and continuous use is not recommended [16]. Fluoride is the mainstay of caries prevention, but its protection against the disease is incomplete [17]. Therefore, an attractive and superior strategy would be to use or include bioactive agents targeting virulence factors and the mechanisms for biofilm development.

Several studies have prioritized finding new antibiofilm agents, including natural substances [18–20]. Among the promising agents, *tt*-farnesol (a membrane-targeting sesquiterpenoid) and myricetin (a flavonoid) hinder the development of cariogenic biofilm formed by *S. mutans*. Myricetin inhibits the *gtfs* gene expression and Gtfs activity, thereby hindering the exopolysaccharides synthesis in situ [18,19]. *tt*-farnesol targets the cellular membrane, affecting the tolerance of *S. mutans* to acid stress. Both agents have a moderate biocidal effect [18,19], and their combination with fluoride results in fewer and less severe carious lesions [19]. In addition, *tt*-farnesol is a derivative of the sterol biosynthesis pathway in eukaryotic cells and a *quorum-sensing* molecule of *C. albicans* [21], which keeps the fungus in yeast form. However, it appears not to affect *S. mutans* in concentrations produced when both microorganisms are co-cultivated in biofilms, possible due to the thickness and amount of biomass of these biofilms [2,3]. Additionally, at concentrations above the threshold (i.e., the physiological concentration of the species), *tt*-farnesol can induce apoptosis in planktonic cultures of *C. albicans* cells [22]. Therefore, the antibiofilm effect of *tt*-farnesol and myricetin against *C. albicans* and *S. mutans* biofilms still needs to be investigated [23].

Hydroxychalcones are precursor metabolic intermediates for flavonoids and isoflavonoids. These agents inhibit the streptococcal Gtfs activity [24], impair *S. mutans* survival in planktonic culture [25]; thus, possibly impairing the structure of biofilms. In addition, flavonoids interfere with *C. albicans* cell wall formation, cause disruption of the plasma membrane and mitochondrial dysfunction, affect cell division, protein synthesis, and the efflux-mediated pumping system [26]. Also, synthetic hydroxychalcones were shown to have anti-*Candida* activity [25,27]. Nevertheless, the efficacy of hydroxychalcones on dual-species *S. mutans* and *C. albicans* biofilms is unexplored.

Also, the interference in the metabolism of lipoteichoic acids (LTA) would affect the development of biofilms by Gram-positive bacteria. The compound 1771 targets LtaS, an LTA synthase enzyme in *S. aureus* [28] and *Enterococcus faecium* [29]. This compound also hinders *S. mutans* biofilm formation [30]. However, the effect of compound 1771 on *C. albicans* is unknown.

Thus, considering the virulence and difficulty of controlling mature biofilms, this study evaluated the antimicrobial and antibiofilm activities of six compounds (*tt*-farnesol, myricetin, two 2- -hydroxychalcones, 4- -hydroxychalcone, and compound 1771) against *C. albicans* and *S. mutans* single- and dual-species settings.

#### **2. Materials and Methods**

#### *2.1. Experimental Design*

Antimicrobial activity was evaluated using planktonic cultures of *C. albicans and S. mutans* in microdilution assay to determine the minimum inhibitory concentration (MIC) and minimum fungicidal and bactericidal concentrations (MFC and MBC). Next, singlespecies biofilms were used to investigate the antibiofilm activities of compounds during initial biofilm formation (24 h) and against pre-formed biofilms (48 h). Finally, promising compounds (and their corresponding concentrations) on both species were combined to analyze the antibiofilm activity on dual-species biofilms formed by *C. albicans* and *S. mutans* on initial biofilm formation (24 h) and pre-formed biofilms (48 h). At that time, fluoride was also added, and groups with and without fluoride were evaluated. All tests were performed using a 96-polystyrene microplate to determine viable microbial population (colony forming units or CFU). At least three independent experiments were performed in triplicate for the antimicrobial and antibiofilm tests (*n* = 9). The data were statistically analyzed according to the factorial design of this study, considering each microplate well as a statistical block. The hypothesis was that elimination or reduction of at least 50% of microbial cells (of both species) from biofilm using the proposed agents and their combinations substantially affect the development of dual-species biofilms.

#### *2.2. Test Agents*

An in vitro study with *C. albicans* and *S. mutans* was conducted to investigate the antimicrobial and antibiofilm activity of six compounds: *tt*-farnesol or (E,E)-3,7,11- Trimethyl-2,6,10-dodecatrien-1-ol, trans,trans-3,7,11-Trimethyl-2,6,10-dodecatrien-1-ol (Sigma-Aldrich Co., St Louis, MO, USA; Cat.#46193; 96% purity), myricetin or 3,5,7 trihydroxy-2-(3,4,5-trihydroxyphenyl)-4H-chromen-4-one (AK Scientific, Inc., Union City, CA, USA; Cat.#J10595; 95% purity), three hydroxychalcones [2- -hydroxichalcone or 1-(2-hydroxyphenyl)-3-phenylprop-2-en-1-one (Angene, Hong Kong, China; Cat.#AGN-PC-015IM; 95% purity), 2- -hydroxichalcone or (2E)-1-(2-hydroxyphenyl)-3-phenylprop-2-en-1-one (AK Scientific, Inc.; Cat.#R815; 98% purity), 4- -hydroxichalcone or (2E)-1-(4 hydroxyphenyl)-3-phenylprop-2-en-1-one (AK Scientific, Inc.; Cat.#C135; 98% purity)], and compound 1771 [(5-phenyl-1,3,4-oxadiazol-2-yl)carbamoyl]methyl 2-{naphtho[2,1 b]furan-1-yl}acetate)] (UkrOrgSynthesis Ltd., Kiev, Ukraine; Cat.#PB25353228; purity not available). The compounds were diluted with 84.15% ethanol (EtOH; Sigma-Aldrich; Cat.#E7023) and 15% dimethyl sulfoxide (DMSO; Sigma-Aldrich; Cat.#D8418) to have stock solutions at 15 mg/mL, except for compound 1771 that was used at 2 mg/mL. The concentration for compound 1771 was lower than the other agents because of solubility issues. Also, sodium fluoride was prepared at 5000 ppm (Sigma-Aldrich; Cat.# 71519). These stock solutions were diluted to distinct concentrations for assays. For antimicrobial assays, the agents with stock concentration at 15 mg/mL were tested using concentrations of 1250, 1000, 500, 250, 125, 62.5, 31.25, 15.625 μg/mL. For compound 1771 (stock concentration at 2 mg/mL), the concentrations used were 250, 125, 62.5, 31.25, 15.625, 7.813, 3.906, 1.953, 0.977, 0.488, 0.244 μg/mL.

#### *2.3. Microbial Strain and Growth Conditions*

*C. albicans* SC5314 and *S. mutans* UA159 (serotype c; ATCC 700610) strains preserved in a freezer −80 ◦C were thaw, platted on a blood agar plate (5% sheep's blood; Laborclin, Pinhais, PR, Brazil), and incubated (48 h, 37 ◦C, 5% CO2; Steri-Cult™ Thermo Scientific, Waltham, MA, USA). Next, five colonies of each microorganism were inoculated into a liquid culture medium (2.5% tryptone with 1.5% yeast extract or TY, pH 7.0; Becton, Dickinson and Company (BD), Sparks, MD, USA) containing 1% of glucose (*w*/*v*) (TY+1% glucose) and incubated (16 h, 37 ◦C, 5% CO2). After that, the starter cultures were diluted 1:20 in the same culture medium and grown until mid-log growth phase: *S. mutans* OD562 nm 0.500 (±0.100) and *C. albicans* OD562nm 0.482 (±0.058) (ELISA plate reader, Biochrom Ez, Cambourne, UK). These cultures were diluted in TY+1% glucose for each microorganism

inoculum with 2 × <sup>10</sup><sup>6</sup> colony-forming units per milliliter (CFU/mL) for antimicrobial evaluation. However, the mid-log growth phase cultures were diluted in TY+1% sucrose (*w*/*v*) to yield 2 × <sup>10</sup><sup>6</sup> CFU/mL for single-species antibiofilm assays.

#### *2.4. Antimicrobial Activity*

The antimicrobial activity was evaluated by determining the minimum inhibitory concentration (MIC) and minimum fungicidal and bactericidal concentration (MFC and MBC, respectively) using broth microdilution following the Clinical and Laboratory Standards Institute [31–33], with some modifications [34]. The compounds were evaluated according their stock concentration ranging from 0 to 1250 μg/mL (when stock concentration 15 mg/mL) and 0 to 250 μg/mL (for compound 1771 with stock concentration at 2 mg/mL). Of note, 0 μg/mL was the vehicle-control. For most of the newest compounds, MIC has been described with some caveats, as when the visual inspection and the optical density are compromised by precipitation, for example [35]. Here, most agents complexed when in contact with the culture medium, making visual inspection subjective and interfering with optical density readings, and most of them did not present a clear dose-dependent effect. Therefore, the abbreviation IC50 was defined as the inhibitory concentration of the agent that inhibited the growth of the microorganism by 50% [34], considering microbial growth as the CFU/mL counts. Thus, MIC abbreviation was not employed to state the outcomes.

All selected compounds [*tt*-farnesol, myricetin, 2- -hydroxychalcone (AGN), 2- hydroxychalcone (R815), 4- -hydroxychalcones (C135), and compound 1771 (1771)] were tested against *S. mutans* planktonic culture. However, only the effective compounds against the bacterium (*tt*-farnesol, myricetin, C135, and compound 1771) were used for *C. albicans* because when both species are together, they form robust biofilms mediated primarily by the exopolysaccharides produced by the bacterium in a rich-sucrose environment [3].

The inoculum culture (100 μL of bacterium or fungus) was transferred to 96-well plates containing TY+1% glucose and distinct concentrations of agents were arranged from the highest to the lowest for a final volume of 200 <sup>μ</sup>L (yielding 1 × 106 CFU/mL for each species). As controls for each experiment, wells containing only culture medium, only inoculum (growth control without treatment), and inoculum plus vehicle (final concentration of 7% EtOH and 1.25% DMSO) were added to rule out any effect of the vehicle on microbial cells. The assembled plates were incubated (24 h, 37 ◦C, 5% CO2), followed visual inspection (turbidity: microbial growth, clear: no growth), and OD562 nm readings (ELISA plate reader). Next, the plates were incubated to homogenize the cultures (5 min, 75 rpm, 37 ◦C) (Quimis, São Paulo, Brazil). After that, an aliquot from each well was removed for serial dilution in saline solution (0.89% NaCl; Synth, Diadema, SP, Brazil), plating (undiluted and 10−<sup>1</sup> to 10<sup>−</sup>5), and incubation (48 h, 37 ◦C, 5% CO2) to determine CFU/mL quantification and inhibitory concentration (IC50). The MBC and MFC were measured by CFU/mL count and defined as the lowest compound concentration that inhibited microbial growth (or absence of colony growth on agar). However, for some compounds that may target the extracellular matrix production in biofilms, the concentrations that inhibited at least 50% of microbial growth (i.e., 50% of CFU/mL reduction vs. vehicle) were considered promising antimicrobial activity [34,35].

#### *2.5. Antibiofilm Activity*

This analysis was conducted after determining antimicrobial activity and was performed using single- and dual-species settings and different exposure to compounds: activity against initial biofilm formation (24 h) and pre-formed biofilms (48 h). For initial biofilm formation, the agents were introduced at the time 0 h, and biofilms were evaluated at 24 h of development. For pre-formed biofilms, biofilms were grown for 24 h and then treated, being evaluated at 48 h of growth. Thus, it was evaluated the inhibition of biofilm formation for biofilms at 24 h and the eradication of biofilm growth for 48 h-old biofilms. The measurement was considered effective when the CFU/mL count was reduced by 50% (vs. vehicle) for 24 and 48 h-old biofilms [34,36].

The strains were grown and prepared using the methodology described above. However, the culture medium and inoculum of the experiments were prepared using TY+1% sucrose. The selected compounds and their concentrations were based on antimicrobial data: C135 (from 1250 to 62.5 μg/mL), myricetin (from 1250 to 125 μg/mL), *tt*-farnesol (from 1250 to 31.25 μg/mL), and compound 1771 (from 250 to 1.953 μg/mL). Previous studies evaluated the antibiofilm activity of myricetin, compound 1771, and *tt*-farnesol for *S. mutans* [18,19,30], but here distinct concentrations were tested. Tests with compound 1771 for *C. albicans* and C135 for both species will be presented for the first time.

Initially, single-species biofilm to prevent initial biofilm formation (24 h) were analyzed. On a 96-well plate, 100 <sup>μ</sup>L of final inoculum with 2 × <sup>10</sup><sup>6</sup> CFU/mL (for both species) were added to each well, containing test concentrations and culture medium (TY+1% sucrose), totalizing 200 <sup>μ</sup>L (1 × <sup>10</sup><sup>6</sup> CFU/mL). Controls of experiments were also added (wells containing only culture medium, wells containing only the inoculum of the experiment, and wells containing the inoculum plus vehicle or 0 μg/mL). The plate was incubated (24 h, 37 ◦C, 5% CO2). After incubation, visual inspection was performed, followed by orbital incubation (5 min, 75 rpm, 37 ◦C). The remaining biofilms on the wells were rinsed three times with 200 μL of 0.89% NaCl solution to remove any loose material. Next, these biofilms were scraped with pipet tips five times (5X) with 200 μL of 0.89% NaCl, totalizing 1 mL of biofilm suspension (from each well). This biofilm suspension was placed in a microtube, subjected to serial dilutions (10−<sup>1</sup> to 10−5), which were plated, as were the undiluted suspensions. The plates were incubated (48 h, 37 ◦C, 5% CO2), and the CFU/mL was calculated. Next, the microbial growth inhibition of each concentration was compared to vehicle control.

Subsequently, prevention of build-up pre-formed biofilm (48 h) was evaluated [36]. Here, 96-well plates were assembled using 100 μL final inoculum of *C. albicans* or *S. mutans* (2 × <sup>10</sup><sup>6</sup> CFU/mL) and 100 <sup>μ</sup>L of TY+1% sucrose (to reach 1 × <sup>10</sup><sup>6</sup> CFU/mL). The microplate was incubated (24 h, 37 ◦C, 5% CO2) without any treatment or vehicle control. After incubation and biofilm formation, visual inspection was performed, followed by culture medium removal and washing of remaining biofilms (three times with 200 μL 0.89% NaCl solution). Next, fresh culture medium TY+1% sucrose and test concentrations of agents or the vehicle were added. The microplate was incubated again (24 h, 37 ◦C, 5% CO2). After incubation (when biofilms were 48 h-old), the same processing protocol applied for 24 h-old biofilms was conducted until obtaining 1 mL of biofilm suspension. The biofilm suspensions were sonicated (30 s at 7 w, Sonicator QSonica, Q125, Newtown, CT, USA), subjected to serial dilutions (10−<sup>1</sup> to 10−5), and plated. The undiluted suspensions were also plated. The CFU/mL counts were evaluated and compared to vehicle control.

Finally, the antibiofilm activity for dual-species biofilms of *C. albicans* and *S. mutans* was also performed at 24 and 48 h. These analyzes were performed with the same methodology used for single-species biofilms (24 and 48 h). However, the inoculum of the dual-species culture was prepared with 2 × 106 CFU/mL of *S. mutans* and <sup>2</sup> × <sup>10</sup><sup>4</sup> CFU/mL of *C. albicans* [3] to reach 1 × <sup>10</sup><sup>6</sup> CFU/mL of *S. mutans* and 1 × 104 CFU/mL of *C. albicans* after adding culture medium or treatments.

The concentrations of agents with better results against each species in the singlespecies biofilm setting were selected: 125 μg/mL (C135 and *tt*-farnesol), 500 μg/mL (myricetin), and 3.906 μg/mL (1771). Then, compounds with selected concentrations were combined with each other and with or without sodium fluoride (250 ppm) or F totalizing 16 groups. The elected combinations included groups without fluoride (C135, C135+*tt*-farnesol, C135+1771, C135+myricetin, C135+*tt*-farnesol+1771, C135+*tt*-farnesol+myricetin, C135+*tt*-farnesol+myricetin+1771, C135+1771+myricetin, *tt*-farnesol+myricetin, 1771+myricetin, *tt*-farnesol+1771+myricetin) and groups with fluoride (250 ppm) (C135+fluoride, C135+*tt*-farnesol+fluoride, C135+1771+fluoride, C135+myricetin+fluoride, C135+*tt*-farnesol+1771+myricetin+fluoride). The concentration of fluoride was selected based on the commercially available fluoride-based mouth rinse [19,37].

#### *2.6. Statistical Analyses*

The statistical analyses for CFU/mL values were performed using descriptive and inferential statistics. Normality was evaluated with the Shapiro-Wilk test employing a significance level of 5%. The data were non-parametric; thus, the data were evaluated with Kruskal–Wallis test, followed by Dunn's post-test, considering α = 0.05 (Prism 9 software, GraphPad Software, Inc. 2021). The microbial growth inhibition of each agent and concentration was compared to vehicle control. In addition, the CFU/mL data were transformed to log10 or log to verify the log reduction.

#### **3. Results**

#### *3.1. Antimicrobial Activity*

#### 3.1.1. *S. mutans*

Three compounds (AGN, C135, R815) complexed with the culture medium, making the visual inspection analysis inaccurate; turbidity was also present for myricetin (but in minor proportion than for the three aforementioned compounds), and compound 1771 (at concentrations equal of more than 31.25 μg/mL). The observation of culture medium turbidity occurred immediately after adding the compound into the culture media, without microbial inoculation and incubation (controls per each concentration tested). The compound that did not complex with culture medium was *tt*-farnesol, and the absence of visual growth was observed at 31.25 μg/mL, which was also its IC50.

Regarding MBC, as per CLSI definition, the absence of colony growth on an agar plate was found for *tt*-farnesol and compound 1771, as 62.5 μg/mL and 250 μg/mL, respectively. Thus, a potential antimicrobial activity was achieved for the compounds when the compound at a specific concentration yielded a 50% reduction of CFU/mL counts compared to the vehicle control (IC50), as follows.

For C135, the lowest concentration that yielded 50% reduction was 62.5 μg/mL, but the same reduction was observed for 125, 250, and 500 μg/mL (Figure 1). Thus, the antibacterial activity of C135 was not dose-dependent.

There was an absence of expressive effect on growth inhibition for all concentrations of AGN and R815 (vs. vehicle). However, some concentrations of R815 showed statistical differences and some inhibition of growth at 500 and 250 μg/mL (2 and 1 log reduction, respectively), 125 and 62.5 μg/mL (0.5 log reduction). These reductions were not dosedependent and are not within the cutoff established for an effective compound (Figure 1). Thus, AGN and R815 were not used in the downstream assays.

Three concentrations of myricetin presented statistical differences compared to the vehicle (250, 500, and 1000 μg/mL) but were not dose-dependent. However, a better effect was obtained for 500 μg/mL, which was considered the IC50.

For compound 1771, the IC50 was 7.813 μg/mL; higher concentrations also demonstrated significative statistical differences (at least 4 log reduction vs. vehicle) and, as mentioned above, 250 μg/mL rendered absence of CFU/mL quantification on agar (MBC).

**Figure 1.** Antimicrobial activity of *S. mutans* using six compounds: C135, R815, AGN, *tt*-farnesol, myricetin and 1771. Data represents median and interquartile ranges (*n* = 9). Statistical differences are represented with \*\* (*p* < 0.05), \*\*\* (*p* < 0.001), and \*\*\*\* (*p* < 0.0001) (Kruskal Wallis, followed by Dunn's test). The tabulated results are in Table S1.

#### 3.1.2. *C. albicans*

The antimicrobial activity of *C. albicans* was analyzed using four agents (C135, *tt*farnesol, myricetin, and compound 1771) that were selected based on the effective antimicrobial activity founded for *S. mutans*. Among them, only *tt*-farnesol did not complex with the culture medium; the absence of visual growth was observed at 125 μg/mL, which was also its IC50.

C135 and *tt*-farnesol presented a dose-dependent effect on viable fungal growth. C135 was the most effective in inhibiting fungal viability (Figure 2). All its concentrations above 31.25 μg/mL hindered colony growth on agar plates; thus, its IC50 and MFC were determined as 62.5 μg/mL (absence of colony growth on agar plates). The MFC for *tt*-farnesol was 1000 μg/mL (Figure 2). Both myricetin and compound 1771 did not demonstrate significant antimicrobial activities as per the cutoff of 50% colony growth reduction (IC50), although statistical differences were observed, as depicted in Figure 2.

**Figure 2.** Antimicrobial activity of *C. albicans* with compounds: C135, *tt*-farnesol, myricetin and 1771. Data represents median and interquartile ranges (*n* = 9). Statistical differences are represented with \* (C135 *p* = 0.048; 1771 *p* = 0.013), \*\* (*p* < 0.05) and \*\*\*\* (*p* < 0.0001) (Kruskal Wallis, followed by Dunn's test). The tabulated results are in Table S2.

#### *3.2. Antibiofilm Activity*

3.2.1. Single-Species *S. mutans* Biofilm

On 24 h biofilm (initial biofilm formation)*,* all concentrations of tested compounds (C135, myricetin, *tt*-farnesol, and compound 1771) demonstrated antibiofilm activity, specially myricetin and *tt*-farnesol concentrations that eliminated bacterial growth (Figure 3). C135 eliminated the bacterium at 62.5, 125, and 250 μg/mL, but not at higher concentrations. Also, all concentrations of compound 1771 reduced about 5 logs of bacterium growth (vs. vehicle).

However, on *S. mutans* pre-formed biofilms (48 h) a greater inhibitory effect was achieved with *tt*-farnesol; where concentrations of 62.5 μg/mL and higher eliminated the bacterium. For C135 the best concentration was 125 μg/mL with 5 logs reduction (vs. vehicle). In contrast, a lower inhibitory activity was observed for myricetin and compound 1771 (although they presented statistical differences, the reduction was about 1 log vs. vehicle) (Figure 3).

**Figure 3.** Antibiofilm activity of *S. mutans* with compounds: C135, *tt*-farnesol, myricetin, and 1771. On the first line are presented data of 24 h biofilm (initial biofilm formation); and right below are the data of pre-formed biofilms (48 h). The data represents median and interquartile ranges (*n* = 9). Statistical differences are represented with \* (*p* = 0.026), \*\* (*p* < 0.05), \*\*\* (*p* < 0.001), and \*\*\*\* (*p* < 0.0001) (Kruskal Wallis, followed by Dunn's test). The tabulated results are in Table S3.

3.2.2. Single-Species *C. albicans* Biofilm

The best antibiofilm activity for *C. albicans* 24 h biofilm was observed for C135 and *tt*-farnesol; both eliminated the fungus, except at 31.25 μg/mL of *tt*-farnesol that reduced 4 logs (vs. vehicle). Compound 1771 reduced 3 logs from 3.906 μg/mL to higher concentrations. However, the lowest antibiofilm effect was observed for myricetin with about 1 log reduction (vs. vehicle) in all tested concentrations, although statistical differences were observed (Figure 4).

The antibiofilm activity against *C. albicans* pre-formed biofilm (48 h) was achieved effectively only by C135. C135 eliminated the fungus at 125, 500, and 1250 μg/mL; it also decreases CFU/mL counts by 4 logs (62.5 μg/mL) and 5 logs (250 and 1000 μg/mL) (vs. vehicle). A weak activity was observed using *tt*-farnesol, where concentrations above 125 μg/mL presented about 2 log reduction (vs. vehicle). However, no effect was observed using myricetin and compound 1771 (except 1771 at 250 μg/mL with 2 logs reduction vs. vehicle) (Figure 4).

**Figure 4.** Antibiofilm activity of *C. albicans* with compounds: C135, *tt*-farnesol, myricetin, and 1771. On the first line are presented data of 24 h biofilm (initial biofilm formation); and right below are the data of pre-formed biofilms (48 h). The data represents median and interquartile ranges (*n* = 9). Statistical differences are represented with \*\* (*p* = 0.002), \*\*\* (*p* ≤ 0.0003), and \*\*\*\* (*p* ≤ 0.0001) (Kruskal Wallis, followed by Dunn's test). The tabulated results are in Table S4.

3.2.3. Dual-Species *C. albicans* and *S. mutans* Biofilms.

Based on the previously presented data, compounds and their most effective concentrations were selected for combinations of compounds tested on dual-species *C. albicans* and *S. mutans* biofilms (see data summarized in Table 1). The selected concentrations were: 125 μg/mL for C135 and *tt*-farnesol, 500 μg/mL for myricetin, and 3.906 μg/mL for compound 1771. In addition, C135 was used alone or combined with the other agents with and without sodium fluoride because it was the most effective agent against the fungus growth.


**Table 1.** Summary of antimicrobial and antibiofilm activity on single-species cultures for selection of compounds concentrations (μg/mL) to test against dual-species biofilms.

\* represent selected concentrations that did not reduce 50% of CFU/mL but had a significative statistical reduction vs. vehicle between all tested concentrations. IC50: the inhibitory concentration of the agent that inhibited the growth of the microorganism by 50%, considering microbial growth as the CFU/mL counts. MBC: minimum bactericidal concentration. MFC: minimum fungicidal concentration. 24 h: 24 h-old biofilms. 48 h: 48 h-old biofilms.

> For 24 h-old biofilms (initial biofilm formation), all 16 formulations were effective, impeding the growth of both species (bacterium and fungus) (Figure 5). However, the microbial growth inhibition of pre-formed (48 h-old biofilms) was different between treatments and species (Figure 5). Among the 16 formulations tested, three inhibited the growth of both species completely: C135+*tt*-farnesol+myricetin, C135+*tt*-farnesol+fluoride, and C135+*tt*-farnesol+1771+myricetin+fluoride (all compounds combined with fluoride).

Furthermore, considering the total inhibition of bacterial growth in the dual-species setting, four formulations were effective (C135+*tt*-farnesol+1771+myricetin, C135+fluoride, C135+1771+fluoride, and C135+myricetin+fluoride). Considering the total inhibition of fungal growth in the dual-species setting, four treatments were effective (C135, C135+*tt*farnesol, C135+1771, and C135+myricetin). Also, some formulations reduced at least 50% CFU/mL (vs. vehicle) of *S. mutans* (C135+*tt*-farnesol, C135+*tt*-farnesol+1771), or *C. albicans* (C135+*tt*-farnesol+1771+myricetin and C135+myricetin+fluoride).

**Figure 5.** Antibiofilm activity of dual-species *S. mutans* and *C. albicans* biofilms with combined compounds (with and without sodium fluoride): C135 (C), *tt*-farnesol (Far), myricetin (Myr), 1771, and sodium fluoride (F). The top graphs presented data of 24 h biofilm (initial biofilm formation). The bottom graphs depict data of pre-formed biofilms (48 h). The data represents median and interquartile ranges (*n* = 9). Statistical differences are represented with \* (*p* = 0.04), \*\* (*p* ≤ 0.002), \*\*\* (*p* = 0.0007), and \*\*\*\* (*p* ≤ 0.0001) (Kruskal Wallis, followed by Dunn's test). The tabulated results are in Table S5.

#### **4. Discussion**

Several strategies can be used to control biofilms to prevent oral diseases. The classical strategies for oral biofilm control include brushing/flossing (mechanical removal of biofilms) and restricting dietary sugar intake (mainly frequency) to prevent biofilm build-up. Both diet restriction and brushing/flossing require behavioral compliance, which can be challenging. In addition, fluoride is used to avoid teeth demineralization and promote remineralization as part of oral hygiene products (toothpaste and mouthwashes) and/or supplied in tap drinking water. However, they may not be appropriate to all individuals, such as those without adequate dexterity (e.g., young children, elderly, people with disabilities, people in ICUs), which may require supervision for brushing/flossing

and aids to weaken the overall biofilm structure to facilitate its mechanical removal, or even substances to enhance biofilm control.

Single targets for biofilm prevention and antimicrobial control can be difficult and limit the treatment options. Thus, combining agents with distinct targets can be an effective approach to access different sites in biofilms, which present complex biological traits and protected niches [1,19]. Therefore, this study evaluated the antimicrobial and antibiofilm activities of six compounds [*tt*-farnesol, myricetin, two 2- -hydroxychalcones (AGN and R815), 4- -hydroxychalcone (C135), and compound 1771] with different targets against *C. albicans* and *S. mutans* single- and dual-species settings.

Of note, an antimicrobial substance/molecule may not be an antibiofilm agent, and a compound with antibiofilm activity may not be an antimicrobial per se (e.g., molecules that affect extracellular enzymes responsible for the extracellular matrix construction). This scenario is depicted by the findings in both antimicrobial and antibiofilm assays performed as some agents were effective against the microorganisms in planktonic cultures and were not in the biofilms' settings. Also, the effect of compounds was mostly non-dose-dependent for both *C. albicans* and *S. mutans*.

The antimicrobial outcome for *S. mutans* showed the lowest IC50 value for compound 1771 (7.813 μg/mL), followed by *tt*-farnesol (31.25 μg/mL), C135 (62.5 μg/mL), while the highest value was observed for myricetin (500 μg/mL). However, IC50 values for *C. albicans* were C135 (62.5 μg/mL) and *tt*-farnesol (125 μg/mL). Furthermore, *tt*-farnesol eliminated CFU/mL count of both species reaching a MBC (62.5 μg/mL) and a MFC (1000 μg/mL). The compound 1771 presented MBC (250 μg/mL) and C135 reached MFC (62.5 μg/mL). The findings for C135 for both species and 1771 for *C. albicans* are presented for the first time here. Thus, C135 presented a promising antimicrobial effect for both species, and compound 1771 did not inhibit *C. albicans* growth.

A previous study with *S. mutans* using different concentrations of compound 1771 did not eliminate the bacterium [30], but the total elimination of CFU/mL count was observed here using the greatest concentration. In the same study [30], the antimicrobial activity of myricetin for *S. mutans* was at a lower concentration than the one found here. The antimicrobial effect of *tt*-farnesol on both species corroborates previous findings [18,22]. Among the three chalcones tested, the antimicrobial activity was better for C135, a 4- hydroxychalcone. The other two 2- -hydroxychalcones (AGN and R815) did not present antimicrobial effect for *S. mutans*. These results can be explained by the differences between the chemical structure of the selected hydroxychalcones, suggesting that the presence of hydroxyl groups on the ring of the 4- -hydroxychalcone scaffold is crucial for the growth inhibition [24,38,39].

The presence of an extracellular matrix is essential for the existence of biofilm and the complete expression of virulence by microbial pathogens and pathobionts, hindering the action of antimicrobial agents and preventing their access to microbial cells [40]. Sucrose can modulate microbial synergism and ecology of the oral microbiota because its hexoses (glucose and fructose) are used for exopolysaccharides and organic acid production that influences the structure and composition of dental biofilms [1,6]. Cariogenic biofilms promote interactions and mechanisms that control dysbiosis [1,6] as observed on dualspecies biofilms of *S. mutans* and *C. albicans* in vitro [2,3]. Therefore, it is important to understand the mechanisms of possible antibiofilm compounds.

Myricetin and some hydroxychalcones inhibit *S. mutans* F-ATPase activity (acid tolerance mechanism) [19], glycolysis (organic acid production or acidogenicity) [19], and synthesis of extracellular matrix glucans (by interfering with *gtfs* gene expression and Gtfs activity) [19,24]. The deficit in glucan production can compromise the integrity and 3D structure biofilms [4], facilitating their disruption. These findings can explain the antibiofilm effect of myricetin on *S. mutans* initial biofilms (24 h) and the greatest potential of C135 (4- -hydroxychalcone) on initial (24 h) and mature *S. mutans* biofilm (48 h). The weaker inhibition of myricetin on pre-formed biofilms (48 h—biofilm eradication) can be promoted by the presence of pre-formed microcolonies and their 3D extracellular matrix.

*tt*-farnesol eliminated the bacterium in 24- and 48 h-old single-species biofilms (at 62.5 μg/mL and higher concentrations). The antimicrobial and antibiofilm effect of this compound can be related to the targets on *S. mutans* cytoplasmatic membrane, altering its proton permeability, decreasing its tolerance to acid stress [18]. Compound 1771 also had a promising antibiofilm effect for the bacterium, especially on initial biofilm (24 h-old single-species biofilms). This compound inhibits the LTA synthesis [28], interfering with the cell wall composition, making the cytoplasmatic membrane an easy target for environmental stresses. Also, LTA from the cell wall are released in the matrix and interact with exopolysaccharides during the development and maturation of biofilms [41]; hence, interfering with LTA metabolism can impair cell wall and extracellular matrix composition. Thus, *tt*-farnesol and compound 1771 hindered *S. mutans* biofilms by promoting antimicrobial and antibiofilm activities.

Hydroxychalcones inhibit the cell wall formation of *C. albicans* cells [42,43], while *tt*farnesol keeps the fungus in its yeast form [21]. However, it is unknown whether myricetin inhibits *C. albicans* biofilm formation or its extracellular matrix development or whether compound 1771 could target this fungus or its metabolism. The effect on the extracellular matrix components and construction could make the fungal cells more susceptible to antimicrobial agents. Here, single-species *C. albicans* biofilms were not greatly affected by myricetin. However, the 4- -hydroxychalcone C135 presented a promising effect in the initial (24 h) and mature (48 h) fungal biofilms. Also, *tt*-farnesol and 1771 were effective on initial fungal biofilm, but only *tt*-farnesol eliminated the fungus (at 62.5 μg/mL and higher concentrations). In addition, *tt*-farnesol had a weak effect, while 1771 had practically no effect on fungal growth on pre-formed biofilms. Thus, eradication of *C. albicans* biofilm (48 h) was achieved only with C135.

Previous studies with *C. albicans* biofilms demonstrate that chalcones inhibited enzymes involved in resistance pathways [42]. Thus, the effect of C135 on *C. albicans* biofilms can be related to targets on resistance pathways; nevertheless, this hypothesis must be better explained. In addition, as observed on the data above from antimicrobial activity of *tt*-farnesol, this compound can inhibit fungal growth, induce apoptosis in *C. albicans* cells, and inhibit the fungal hyphae [21,22]. As described for initial biofilms (24 h), the treatment was applied at 0 h, so it may be that the effect of *tt*-farnesol (preventing filamentous morphology) supports the inhibition of fungal growth in 24 h while hampered its effect on pre-formed biofilm (48 h). Therefore, combine compounds could improve the effect in mature biofilms. This hypothesis is confirmed by the potentiated effect on *C. albicans* cells in dual-species biofilm (48 h) when C135 and *tt*-farnesol were combined, suggesting inhibition of the resistance mechanisms when both compounds are present.

Altogether, the data from antimicrobial and single-species biofilms assays enabled the selection of specific concentrations of the four compounds (C135, myricetin, *tt*-farnesol, and compound 1771) that were combined, with or without sodium fluoride, to assess the antibiofilm activity of formulations against dual-species *S. mutans* and *C. albicans* biofilm. All formulations without fluoride [(C135, C135+*tt*-farnesol, C135+1771, C135+myricetin, C135+*tt*-farnesol+1771, C135+*tt*-farnesol+myricetin, C135+*tt*-farnesol+1771+myricetin, C135+ 1771+myricetin, *tt*-farnesol+1771, 1771+myricetin, *tt*-farnesol+1771+myricetin)] and with fluoride [(C135+fluoride, C135+*tt*-farnesol+fluoride, C135+1771+fluoride, C135+J10595+fluoride, C135+*tt*-farnesol+1771+myricetin+fluoride completely inhibited the initial biofilm formation (24 h). These findings showed that there might be a potential synergism between the compounds and a greater effect when they are combined and applied since the beginning of biofilm formation and during the 24 h of dual-species biofilm development. Part of the effect can be because of the antimicrobial effect per se, as microbial cells were exposed to formulations in their free form before adhesion to the surface. Also, the effect on extracellular matrix formation can not be ruled out.

In contrast, dual-species biofilm eradication (48 h) in which both species did not grow occurred for three formulations: C135+*tt*-farnesol+myricetin, C135+*tt*-farnesol+fluoride, and C135+*tt*-farnesol+1771+myricetin+fluoride. Four formulations only eradicated the

bacterial growth (C135+*tt*-farnesol+1771+myricetin, C135+fluoride, C135+1771+fluoride, and C135+myricetin+fluoride), while other four formulations eradicated fungal growth (C135, C135+*tt*-farnesol, C135+1771, and C135+myricetin). Of note, the formulations with fluoride exhibited a greater antibiofilm activity (mainly for the bacterium), reinforcing the importance of its inclusion in strategies to prevent and control cariogenic biofilms [18]; fewer and less severe carious lesions were observed using combined treatments and fluoride on a rodent model of dental caries [19]. Fluoride can interfere with microbial metabolism, especially on glycolytic enzymes and proton gradient dissipation on the cell cytoplasmatic membrane (when the extracellular pH is higher than the intracellular pH) [44]. This effect can hamper cell growth. Nevertheless, C135 alone or combined with other agents (even those without pronounced effect on singles-species 48 h-old biofilms) prevented fungal and bacterial growth in dual-species biofilms, making it a promising agent.

The present findings provided insights about: (i) compounds as inhibitors of biofilm formation of single-species biofilm (24 h); (ii) compounds that can eradicate pre-formed biofilm (48 h); and (iii) formulations with combined compounds for biofilm inhibition (24 h) and eradication (48 h) of both species in dual-species biofilms. C135 is a novel compound with possible distinct targets alone or in combination with other agents. The formulations that combined agents with distinct targets prevented *C. albicans* and *S. mutans* dual-species biofilm build-up in vitro. The formulation C135+*tt*-farnesol with or without fluoride may represent a potential alternative approach that deserves further investigation, including cytotoxicity to host [30,45]. Therefore, the outcomes of this study could be applied to future studies using the compound alone or combined as an adjuvant strategy to control oral biofilms using shorter exposure times, as mouthwashes.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/jof7050340/s1. Table S1: Antimicrobial activity of *S. mutans* using six compounds: C135, R815, AGN, *tt*-farnesol, myricetin, and 1771. Data shown in Figure 1 as CFU/mL. Table S2: Antimicrobial activity of *C. albicans* with compounds: C135, *tt*-farnesol, myricetin, and 1771. Data shown in Figure 2 as CFU/mL. Table S3: Antibiofilm activity of *S. mutans* with compounds: C135, *tt*-farnesol, myricetin, and 1771 of 24 h biofilm (initial biofilm formation) and pre-formed biofilms (48 h). Data shown in Figure 3 as CFU/mL. Table S4: Antibiofilm activity of *C. albicans* with compounds: C135, *tt*-farnesol, myricetin, and 1771 of 24 h biofilm (initial biofilm formation) and pre-formed biofilms (48 h). Data shown in Figure 4 as CFU/mL. Table S5. Antibiofilm activity of dual-species *S. mutans* and *C. albicans* biofilms with combined compounds (with and without sodium fluoride): C135 (C), *tt*-farnesol (Far), myricetin (Myr), 1771, and sodium fluoride (F). Data shown in Figure 5 as CFU/mL.

**Author Contributions:** Conceived and designed the experiments: C.I.V.L. and M.I.K. Carried out the experiments: C.I.V.L. Analyzed the data C.I.V.L. and M.I.K. Reagents/materials/analysis tools: M.I.K. Wrote the paper: C.I.V.L., A.C.U.d.A.L., and M.I.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by a research grant from the National Council for Scientific and Technological Development (CNPq #409668/2018-4 to M.I.K.) and scholarships (the Coordination of Superior Level Staff Improvement—CAPES #001 and CNPq #141316/2020-9 to C.I.V.L.; and CAPES #001 and the São Paulo Research Foundation—FAPESP #2020/02946-2 to A.C.U.A.L.). The funding body had no role in the design of the study or collection, analysis, and interpretation of data and in writing the manuscript.

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available as supplementary material. Additional data are available on request from the corresponding author.

**Acknowledgments:** The authors are thankful to CNPq, CAPES, and FAPESP for grant and scholarships funding. The present research will be part of the Ph.D. thesis by C.I.V.L.

**Conflicts of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

#### **References**


## *Article Candida parapsilosis* **Colony Morphotype Forecasts Biofilm Formation of Clinical Isolates**

**Emilia Gómez-Molero 1,2, Iker De-la-Pinta 3, Jordan Fernández-Pereira 2, Uwe Groß 1, Michael Weig 1, Guillermo Quindós <sup>3</sup> , Piet W. J. de Groot 2,\* and Oliver Bader 1,\***


**Abstract:** *Candida parapsilosis* is a frequent cause of fungal bloodstream infections, especially in critically ill neonates or immunocompromised patients. Due to the formation of biofilms, the use of indwelling catheters and other medical devices increases the risk of infection and complicates treatment, as cells embedded in biofilms display reduced drug susceptibility. Therefore, biofilm formation may be a significant clinical parameter, guiding downstream therapeutic choices. Here, we phenotypically characterized 120 selected isolates out of a prospective collection of 215 clinical *C. parapsilosis* isolates, determining biofilm formation, major emerging colony morphotype, and antifungal drug susceptibility of the isolates and their biofilms. In our isolate set, increased biofilm formation capacity was independent of body site of isolation and not predictable using standard or modified European Committee on Antimicrobial Susceptibility Testing (EUCAST) drug susceptibility testing protocols. In contrast, biofilm formation was strongly correlated with the appearance of non-smooth colony morphotypes and invasiveness into agar plates. Our data suggest that the observation of non-smooth colony morphotypes in cultures of *C. parapsilosis* may help as an indicator to consider the initiation of anti-biofilm-active therapy, such as the switch from azole- to echinocandinor polyene-based strategies, especially in case of infections by potent biofilm-forming strains.

**Keywords:** *Candida parapsilosis*; biofilm; colony morphology; drug susceptibility

#### **1. Introduction**

*Candida parapsilosis* was first described as a non-pathogenic yeast with no clinical relevance [1]. However, increased use of medical devices, parenteral nutrition, and nosocomial infections [2] has made *C. parapsilosis* one of the most critical fungal species causing blood stream infections (BSI) [3], which are of particular relevance in critically ill neonates [4–6] and immunocompromised patients [5,7].

The high infection rate with *C. parapsilosis* among neonates is likely due to the frequent requirement of parenteral nutrition [8] and the concomitant ability of *C. parapsilosis* to utilize fats and fatty acids as major energy sources [9]. In addition, the immature or compromised immune system may favor infections with this species [10].

Another risk factor for acquiring *C. parapsilosis* infections is the use of indwelling catheters and other medical devices onto which *C. parapsilosis* may form biofilms in conjunction with other *Candida* species or bacteria [4]. Primarily, this is attributed to its capacity to attach to the different materials of which medical devices are made [5,11,12]. This feature is highly variable among individual clinical isolates [13]. In *C. parapsilosis*, the formation

**Citation:** Gómez-Molero, E.; De-la-Pinta, I.; Fernández-Pereira, J.; Groß, U.; Weig, M.; Quindós, G.; de Groot, P.W.J.; Bader, O. *Candida parapsilosis* Colony Morphotype Forecasts Biofilm Formation of Clinical Isolates. *J. Fungi* **2021**, *7*, 33. https://doi.org/10.3390/jof7010033

Received: 8 December 2020 Accepted: 5 January 2021 Published: 7 January 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/).

of biofilms is associated with the ability to form pseudohyphae [14–16] as well as the concomitant change in expression levels of cell wall-localized adhesins such as Als1-7 [17–19] or Rbt1 [20]. Importantly, susceptibility to commonly used azole-based antifungal agents in fungal biofilms on medical devices may be strongly reduced [21,22].

Similar to *C. albicans* (reviewed in [23]), *C. parapsilosis* can have different colony morphotypes on diagnostic agar plates (Figure 1). While mixed-morphology culture plates can be the result of infection with multiple strains [24] in diagnostic procedures, also, the morphologic switching of some strains is a well-described phenomenon [25]. In addition to their role in biofilm formation, pseudohyphae formation and adhesin expression are also key to the visual appearance of fungal colonies on solid agars [26–28], which in turn may well be correlated to the capacity to form biofilms in the host, the incorporation of cell wall proteins (CWP), and, consequently, virulence [15].

**Figure 1.** Colony morphotypes formed by *C. parapsilosis*. (**A**) Mixed morphotypes observed on a routine diagnostic plate. (**B**) Cells with distinct morphological colony phenotypes can be subcultured, but some strains (here strain PEU582: smooth-glossy vs. crepe, see below) sometimes undergo switching with strain-dependent frequencies. Smooth colonies are composed mainly of yeast-form cells, whereas non-smooth colonies are composed of pseudohyphal cells or mixtures of both morphologies.

Here, we phenotypically characterized a large collection of clinical *C. parapsilosis* isolates, including the description of novel intermediate morphotypes. We determined if early colony morphology was a potential predictor of biofilm production and pseudohyphal growth and as such might reveal the need to direct the antifungal therapy against biofilms containing *C. parapsilosis*.

#### **2. Materials and Methods**

#### *2.1. Clinical Routine Diagnostic and Strain Maintenance*

*C. parapsilosis* clinical isolates were routinely identified using MALDI-TOF (MALDI Biotyper, Bruker Daltonics, Bremen, Germany) according to the protocol described [29]. Mixed cultures were differentiated on YEPD agar (1% yeast extract, 2% peptone, 2% glucose, 2% agar) supplemented with 5 mg/mL phloxine B. On this medium, most tested colonies developed a final morphotype within 48 h of incubation at 30 ◦C (Figure 1, Supplementary Figures S1 and S2). Cells from colonies with stable morphotypes were transferred onto Sabouraud dextrose agar (SDA, Oxoid, Munich, Germany), regrown, and stored at −70 ◦C in cryovials (Mast Diagnostica, Reinfeld, Germany) for further analyses.

#### *2.2. Biofilm Quantitation*

Biofilm quantification assays on polystyrol were performed as described previously [11,30,31], with adaptations to suit *C. parapsilosis*. Briefly, isolates were grown on phloxine B-containing YEPD agar plates. Inoculum was prepared from single colonies grown to stationary phase in YEPD broth at 30 ◦C overnight at 220 rpm. A cell suspension adjusted to a cell density of McFarland = 2 was prepared using sterile saline, and 100 μL YEPD medium plus 50 μL aliquots of the cell suspensions were mixed in 96-well polystyrol microtiter plates (Greiner Bio-one) and incubated for 24 h at 37 ◦C. After removal of the medium by aspiration, the attached biofilms were washed once with 200 μL of distilled water. Cells were stained for 30 min in 100 μL of 0.1% aqueous crystal violet (CV). Excess CV was removed, and the biofilm was washed once with 200 μL of distilled water. To release CV from the cells, 200 μL of 1% SDS in 50% ethanol was added, and the cellular material was resuspended by pipetting. CV absorbance was quantified at 490 nm using a microtiter plate reader (MRX TC Revelation). Data shown are the average of three independent biological experiments, each including four technical repeats, using reference strain CDC317 as inter-experiment quality control.

#### *2.3. Antifungal Drug Susceptibility Testing*

Susceptibility testing was performed according to EUCAST e.def 7.3.1 standards [32]. Fluconazole (FLZ), voriconazole (VRZ), posaconazole (POS), and amphotericin B (AMB) substances were purchased from Discovery fine Chemicals Ltd. (Bournemouth, UK), micafungin (MFG) was kindly provided by Astellas, and caspofungin (CAS) was provided by Merck Sharp & Dohme Corp (MSD). Sequencing of the ERG11 and MRR1 genes was performed as previously described [33].

Preformed biofilms reduction

Cells were pre-grown on Sabouraud dextrose agar (SDA) for 96 h at 30 ◦C, and one colony was sub-cultured in 5 mL of YEPD broth overnight at 37 ◦C in an orbital incubator at 200 rpm. Cells were harvested by centrifugation and resuspended in Phosphate-buffered saline (PBS). Upon counting cells in a Neubauer chamber, the suspensions were adjusted to 1 × <sup>10</sup><sup>6</sup> cells/mL in both RPMI (2 g/L glucose, pH 7) and YEPD (20 g/L glucose, pH 6.7). One hundred μL of the inoculum was pipetted into each well of a Bioscreen plate (Labsystem, Helsinki, Finland), and the plates were incubated for 24 h at 37 ◦C to allow biofilm formation. Next, planktonic cells were removed, and the plates were washed twice in PBS leaving just the biofilm in the wells. Two-fold serial dilutions of four antifungal drugs were prepared in RPMI or YEPD ranging from 0.25 to 32 μg/mL for POS, VRZ, and MFG, and from 0.0125 to 16 μg/mL for AMB. Subsequently, 100 μl of each dilution was added to the corresponding wells with biofilm in triplicate, and the plates were incubated again at 37 ◦C for 24 h. Finally, plates were washed twice with PBS, and reduction of the biofilm metabolic activity was determined by measuring the absorbance at 492 nm with the XTT (2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H- tetrazolium-5-carboxanilide) colorimetric method [34].

#### *2.4. Morphotype Development and Agar Invasion*

Selected *C. parapsilosis* isolates were cultured overnight in YEPD liquid media in an orbital shaker at 220 rpm at 30 ◦C. Cell density was adjusted to 2 × <sup>10</sup><sup>2</sup> cells/mL, after which 100 μL was plated onto YEPD agar plates and incubated at 30 ◦C for ten successive days. Starting after 48 h, the morphotype development of colonies was captured every 24 h over ten days using a stereoscopic binocular loupe (SZM-1, OPTIKA®, Ponteranica, Italy) mounted with a digital camera. Morphotypes and agar invasion were classified according to the references given in Supplementary Figure S1.

For analysis of agar invasion, colonies with different morphotypes were plated onto YEPD agar plates supplemented with 5 mg/mL of phloxine B [25], and morphotype development was followed as described above. Agar invasion was scored from day 4 onwards by scraping selected colonies with an inoculation loop and eventually washing off the cells under running water on the last day. Agar invasion was classified as low (1), low-medium (2), medium (3), medium-high (4), high (5), and finally as very high (6) when cells could not be removed by rinsing (see scoring reference in Supplementary Figure S1).

#### *2.5. Statistical Analyses*

For statistical analyses of biofilms and antifungal drug susceptibility, unpaired twosamples Student's *t*-tests were used. All data used were the average of three independent analyses, and *p* values < 0.05 were considered statistically significant.

To detect potential correlations between biofilm formation capacity, colony morphotype, and/or agar invasion capacity, regression analyses were performed, and Pearson's correlation coefficient r was used as a predictor for correlation. All data were analyzed using IMB SPSS 22 statistics software.

#### **3. Results**

#### *3.1. Biofilm Formation Capacity Is Independent of Body Site of Isolation*

Over the course of two years, we collected 215 *C. parapsilosis* clinical isolates from our routine diagnostics (Figure 2). Isolates were classified according to nine different categories depending on the body site of isolation. *C. parapsilosis* is known to frequently occur in the nape region, reaching up to the ear. Consequently, most isolates stemmed from ear infections; however, a substantial number of isolates from invasive infections at other body sites as well as from indwelling devices such as central venous or urine catheters were included in the study. *C. parapsilosis* was only infrequently isolated from locations of the GI (2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H- tetrazolium-5-carboxanilide) tract, the oral cavity, or the skin.

Isolates obtained were systematically screened for their capacity to form biofilms in standard biofilm formation tests in polystyrol microtiter plates. No body site, including those isolate groups obtained from plastic materials, could be identified to be significantly associated with elevated numbers of biofilm-forming isolates (*p* = 0.371, Figure 2A). When stratified by quantitative measurement values, we observed a near even distribution across the study group; only a tentative cut-off for low biofilm formation capacity was observed at OD (optical density) measurement values of approximately 0.1 (Figure 2B, intersection of black lines). Microscopical imaging of cells in biofilms from some representative isolates confirmed the already established idea that the capacity to form biofilms is correlated with pseudohyphal development (Supplementary Figure S3) [25–28].

**Figure 2.** *C. parapsilosis* collection and selection of isolates for downstream experiments. (**A**) Distribution of collection isolates stratified according to site of isolation and biofilm production on polystyrol. Category "invasive" includes, e.g., blood culture, biopsies, or intraoperative swabs. Orange diamonds: six isolates used in pre-formed biofilm experiments, see text. Red and green lines: mean and two-fold mean values. (**B**) Biofilm formation capacity; isolates sorted by value from low to high. Intersection of black lines: approximated cut-off. Red boxes indicate strain selection of representative low (LBF, left box), intermediate (IBF, middle box) and high (HBF, right box) biofilm formers for subsequent experiments.

#### *3.2. Effect of Biofilm Formation on Antifungal Drug Susceptibility*

In order to estimate the correlation of the lead phenotype (biofilm formation on polystyrol) with drug susceptibility, we selected 40 isolates each of low (LBF), intermediate (IBF), and high (HBF) biofilm formation capacity (Figure 2B, red boxes) from our collection including two non-adherent control strains (CDC317 and ATCC22019). The strains were tested for susceptibility to selected azoles (FLZ, POS, VRZ), echinocandins (CAS, MFG), and one polyene (AMB) after one (young colonies) and eight days (matured colonies) of growth on phloxine B agar plates.

Four IBF isolates (PEU651, PEU768, PEU941, and PEU950), and reference strain CDC317 showed elevated (minimum inhibitory concentration) MIC (values of 4–16 μg/mL for FLC. To exclude potential biases through resistance mutations (e.g., Y132F [35]), we sequenced the ERG11 and MRR1 genes in these isolates. A non-synonymous ERG11 point mutation was found only in CDC317, which was heterozygous with respect to the Y132F amino acid exchange. Y132F is known to confer resistance to fluconazole [36,37]. MRR1 only contained non-synonymous SNPs (Single nucleotide polymorphisms) in PEU651. Since we could not exclude a potential influence of these mutations, data for PEU651

as well as those of the two reference strains were excluded from the drug susceptibility analysis, leaving a total of 117 isolates in three groups of 39 isolates each.

We did not observe large-scale differences in drug susceptibility between experiments undertaken with young or mature colonies except for CAS (LBF *p* = 0.017 and HBF *p* = 0.028, Figure 3A white vs. gray boxes).

**Figure 3.** Drug susceptibility. (**A**) Biofilm formation-phenotype dependent susceptibility testing where inoculum was prepared from cells after 1 day of growth on Sabouraud dextrose agar (SDA) (gray boxes) and after 8 days of growth (white boxes) on the same plates, when colonies had fully developed morphotypes. For each substance tested, the values for 1 and 8-day inoculum are depicted for each group (LBF, IBF, and HBF). Red lines: EUCAST clinical breakpoint (R>); green lines, susceptible cut-off (S ≤). \*: statistical significance (**B**,**C**) Effect of antifungal drugs on cell viability in pre-formed biofilms of selected biofilm-forming isolates tested in (**B**) RPMI (Roswell Park Memorial Institute )or (**C**) YEPD(yeast extract peptone dextrose) media.

Although statistically significant differences between HBF versus IBF or LBF isolate groups were clearly evident for some antifungal drugs (POS: LBF vs. HBF with young colonies *p* = 0.014, mature colonies *p* = 0.072, CAS: LBF vs. HBF with young colonies *p* = 0.015), the observed mean MIC differences did not result in a major change in classification of either susceptible (S) or resistant (R) according to EUCAST breakpoints. Differences in mean susceptibility values were 1-2 log2-fold decreases for VRZ, POS (mature colonies only), CAS, and MFG in the HBF group, as compared to the LBF group. No apparent differences for either FLZ or AMB were seen.

#### *3.3. Susceptibility of Biofilms to Antifungal Drugs*

Next, for six selected intermediate to high biofilm-forming isolates, we analyzed to which degree pre-formed *C. parapsilosis* biofilms resisted antifungal drug treatment in two different media, RPMI and YEPD. RPMI is considered the reference medium for antifungal drug susceptibility testing and, as mentioned above, it was also used for MIC determination according to the EUCAST protocol. Likewise, YEPD is a glucose-rich medium that is widely used in assays with yeasts due to the large amount of peptone and dextrose extremely necessary for yeast growth and biofilm formation. The ability to form biofilms was evaluated using this culture medium. Since both media are widely used in the literature, we decided to test the drug susceptibility of preformed biofilms in both of them, observing that some strains are more prone to form biofilms in one media and not in the other, as well as behaving differently when interacting with antifungal drugs. More specifically, one of the six isolates (PEU651) was not able to form biofilm in RPMI (Figure 3B), whereas another (PEU582) grew only at reduced rates. Both isolates also showed a reduction in biofilm development in YEPD (Figure 3C) but were kept in these assays as the results were qualitatively in agreement with the other strains used.

Antifungal drugs had different quantitative effects when the assay was carried out in RPMI (Figure 3B) or in YEPD (Figure 3C). In YEPD, sub-inhibitory drug concentrations caused increases in metabolic activity, as measured by XTT reduction. This may be a stress-response effect countering drugs at these levels, and it was considered an artifact for the purpose of this study. Higher concentrations of azoles (POS, VRZ) reduced biofilm metabolic activity by only 30–50% as compared to the drug-free control. There was no azole drug concentration in the measurement range (up to 32 μg/mL) that led to a full reduction in biofilm metabolic activity (all *p* > 0.05). In contrast, both MFG and AMB did achieve a strong reduction of metabolic activity, although under different conditions: in RPMI, an AMB concentration of 0.5 μg/mL was sufficient for 70% reduction, while 16 μg/mL were required in YEPD. For MFG, this was the opposite: in YEPD, a clear effect was seen at 2 μg/mL with only residual metabolic activity apparent up to the upper assay boundaries (32 μg/mL).

#### *3.4. Biofilm Formation Capacity on Polystyrol Correlates with Colony Morphotype and Agar Invasiveness*

On culture plates, individual *C. parapsilosis* isolates showed specific, stable morphotypes. Only a minority of isolates (18%) were also able to switch between such morphotypes upon re-plating (Figure 1 and scoring reference shown in Supplementary Figure S1).

In a preliminary analysis on a selected subset (Supplementary Figure S2), non-smooth colony morphologies were already apparent at the start of the observation period (20%) and reliably appeared after 72 h of growth. Across the entire collection, agar invasion (Supplementary Figure S1A) and colony morphotype (Supplementary Figures S1B and S2) were therefore scored over a course of four to ten days (Figure 4). Most LBF isolates (90%) showed smooth morphotypes, and only a small proportion (10%), for instance PEU944, developed wrinkled or crepe phenotypes. On day 10, about 20% had developed nonsmooth morphotypes as their major morphology (Figure 4A). With increasing biofilm formation capacity, also the frequency of non-smooth colony morphotypes increased. Of the IBF isolates, 20% had developed non-smooth morphotypes at day 4, and 55% had developed non-smooth morphotypes on day 10. HBF isolates mainly, but not exclusively, produced non-smooth phenotypes (66% on day 4, 83% on day 10), which in most cases were already distinguishable at day 2. Some isolates presented two independent stable

morphotypes (e.g., PEU525: non-smooth (cr) and smooth (s)), which were distinguishable from day 2 until day 10. Along with the increased formation of non-smooth colony morphotypes in IBF and HBF strains (r = 0.832, *p* < 0.001), also agar invasiveness increased from day 4 to day 10 (r = 0.969, \*\*\* *p* < 0.001 (Figure 4B, Table 1).

**Figure 4.** Emergence of colony morphotype over time. (**A**) Development of colony morphotype and (**B**) agar invasiveness scoring the same plates consecutively from 4 days to 10 days after inoculation. Isolates with low biofilm formation capacity (left panels), intermediate biofilm formation capacity (middle panels), and high biofilm formation capacity (right panels) were scored for the most frequent colony morphotype visible. Of note, HBF isolates rated "smooth" still developed minor frequencies of non-smooth colonies. See Supplementary Figure S1 for scoring references.


**Table 1.** Agar invasiveness and colony morphology.

<sup>a</sup> see Supplementary Figure S1 for classification of agar invasion and morphotypes.

#### **4. Discussion**

*C. parapsilosis* is frequently found as a cause of pathologies due to biofilm formation on medical devices in long-term hospitalized patients suffering from endocarditis, peritonitis, arthritis, or general sepsis [22,38–40]. We hypothesized that the capacity to form biofilm might be related to the origin of the clinical specimen, that is, infections at different body sites or fungus growing on medical devices such as indwelling catheters. However, when the 215 clinical isolates in our collection were scored for their biofilm-forming capacity on rich medium (YEPD), a strong inducer of biofilms in *C. parapsilosis* [41], no clear distribution of low (LBF) versus high (HBF) biofilm formation depending on the site of isolation was detected. Nevertheless, a high percentage of catheter-associated isolates belonged to the IBF and HBF groups, which is consistent with the notion that *C. parapsilosis* infections often start from infected indwelling devices [42,43].

Then, we raised the question of whether there might be a possible link between the biofilm-forming capacity of clinical isolates and their drug susceptibility [21,22], which could be useful for decision making about treatment strategies in the clinic. However, tests with the three groups of isolates (LBF, IBF, and HBF), reflecting a wide range—from negligible to high quantities of biomass—of biofilm-formation capacity on polystyrol, and six commonly used azole-, echinocandin-, and polyene antifungal agents did not reveal any such correlation. Observed MIC values were similar to those reported previously by others [44–47], and also, no remarkable differences were observed between inocula prepared from young or matured cells (as found in biofilms). Therefore, we conclude that the drug susceptibility data obtained with the standard EUCAST protocol do not seem to generate predictive information toward the biofilm-formation capacity of clinical isolates.

Nevertheless, fungal cells embedded within biofilms, including those formed by *C. parapsilosis*, are considered to be less susceptible to antifungal compounds [21,22]. This notion was confirmed here by studying the antifungal drug susceptibility of six biofilmforming strains embedded in preformed biofilms, which showed significantly reduced sensitivity to azoles. This experiment was executed in both YEPD (high glucose) and RPMI (lower glucose) medium. As expected [41], biofilm formation in YEPD was higher than in RPMI, and media-dependent differences in echinocandin and polyene sensitivity were also noted. These observations support earlier reported overall quantitative dependences of biofilm formation and drug susceptibility on glucose levels in the media [41,48,49].

Another important aspect of our study was the question of whether *C. parapsilosis* colony morphotypes might perhaps serve to forecast the biofilm-forming capacity of clinical isolates. We hypothesized that high biofilm formers would show more non-smooth morphotypes and pseudohyphae than poor biofilm formers and have an increased tendency to invade agar [50]. When following colony morphotype development in the three isolate groups (LBF, IBF, and HBF) over a course of ten days, we indeed observed a strong correlation of the HBF group with non-smooth colony morphotypes. In contrast, the appearance of non-smooth colonies occurred only in a minority of the isolates in the LBF group. As surface adherence represents the first step in biofilm formation, our data are supported by studies showing that non-smooth colonies are generally more adherent to plastic than smooth colonies [51].

Interestingly, in most cases, non-smooth morphotypes could already be observed within 48 h of growth, and only little change was observed after 96 h, indicating that prolonged incubation beyond this time point is not needed for morphotype determination. In addition, all the morphotypes appeared stable, as they were reproduced upon repeating the experiment, which is consistent with the idea that switching is not a common or frequent event. Finally, biofilm-forming capacity and the appearance of non-smooth morphotypes and pseudohyphae were positively correlated to agar invasion. However, a large proportion of smooth colony types also displayed medium-strength agar invasion, indicating pseudohyphae formation at least at the base of the colony [52,53].

In summary, our experiments show that there is a strong correlation between colony morphotype and biofilm formation capacity in isolates from clinical samples and that this is not reflected in the results from standard antifungal drug susceptibility testing. Our data are the first to indicate that the observation of non-smooth colony morphotypes of *C. parapsilosis* may help to consider the initiation of anti-biofilm-active therapy. This may include antifungal lock therapy and shifting treatment from azole-based to echinocandinor polyene-based strategies to eliminate biofilms from catheters [39]. However, due to the inherent ability of some strains to switch between morphotypes [25–28,54], the absence of such colonies should not be taken as an absolute indicator that biofilms do *not* exist in the patient.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2309-6 08X/7/1/33/s1, Supplementary Figure S1. Morphological classification of *C. parapsilosis* colony morphotypes. (A) Semi-quantitative classification of agar invasion on phloxine B-containing SDA. (B) Colony morphotypes observed in our isolate collection. In picture pairs, left pictures represent colony morphotype after 10 d incubation, right pictures show agar imprint left on phloxine B agar plates after flushing off colonies with running water. Scale bars = 0.25 cm; Supplementary Figure S2. Colony morphotype development during a ten-day time-lapse experiment. Morphotype development of colonies was followed during ten days of growth on YEPD agar at 30 ◦C. (A) Five selected low biofilm-forming (LBF) clinical isolates, (B) five intermediate biofilm-forming (IBF) isolates, and (C) five high-biofilm-forming (HBF) isolates. d, derby; cn, concentric; cr, crater; s, smooth; wr, wrinkled; Supplementary Figure S3. Morphology of *C. parapsilosis* cells in biofilms onto polystyrol. Representative strains with low (LBF, row 1), intermediate (IBF, row2), and high (HBF, rows 3 and 4) biofilm formation capacity are shown. Biofilms were let to develop for 24 h in YEPD as described in materials and methods. Unbound cells were removed by washing with PBS. Remaining biofilm cells were observed with a Leica DM1000 microscope mounted with a HC PL 100×/1.32 objective and MC170 HD digital camera.

**Author Contributions:** Conceptualization, E.G.-M., P.W.J.d.G. and O.B.; Formal analysis, E.G.-M., I.D.-l.-P., J.F.-P. and P.W.J.d.G.; Funding acquisition, U.G., G.Q., P.W.J.d.G. and O.B.; Investigation, E.G.-M., I.D.-l.-P. and J.F.-P.; Methodology, E.G.-M., M.W. and G.Q.; Project administration, O.B.; Resources, U.G.and M.W.; Supervision, G.Q., P.W.J.d.G. and O.B.; Writing–original draft, E.G.-M., M.W., P.W.J.d.G. and O.B.; Writing–review & editing, G.Q. and O.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was funded in part by grants or scholarships from the ZabaldUz program (Universidad del País Vasco/Euskal Herriko Unibertsitatea) to IDlP, the Consejería de Educación, Universidades e Investigación (GIC15/78 IT-990-16) of Gobierno Vasco-Eusko Jaurlaritza to GQ, the Ministerio de Economía y Competitividad (grants SAF2013-47570-P and SAF2017-86188-P, the latter co-financed by FEDER) of the Spanish government to P.G. and G.Q., and the FP7-PEOPLE-2013- ITN—Marie-Curie Action: "Initial Training Networks": Molecular Mechanisms of Human Fungal Pathogen Host Interaction, ImResFun, MC-ITN-606786, to O.B. and U.G.

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available upon reasonable request from the corresponding author.

**Acknowledgments:** Agnieszka Goretzki and Yvonne Laukat are thanked for performing susceptibility testing.

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

#### **References**


## *Article* **The Membranotropic Peptide** *gH625* **to Combat Mixed** *Candida albicans***/***Klebsiella pneumoniae* **Biofilm: Correlation between In Vitro Anti-Biofilm Activity and In Vivo Antimicrobial Protection**

**Angela Maione 1, Elisabetta de Alteriis 1, Federica Carraturo <sup>1</sup> , Stefania Galdiero <sup>2</sup> , Annarita Falanga <sup>3</sup> , Marco Guida <sup>1</sup> , Anna Di Cosmo 1, Valeria Maselli 1,\* and Emilia Galdiero 1,\***


**Abstract:** The antibiofilm activity of a gH625 analogue was investigated to determine the in vitro inhibition and eradication of a dual-species biofilm of *Candida albicans* and *Klebsiella pneumoniae*, two leading opportunistic pathogens responsible for several resistant infections. The possibility of effectively exploiting this peptide as an alternative anti-biofilm strategy in vivo was assessed by the investigation of its efficacy on the *Galleria mellonella* larvae model. Results on larvae survival demonstrate a prophylactic efficacy of the peptide towards the infection of each single microorganism but mainly towards the co-infection. The expression of biofilm-related genes in vivo showed a possible synergy in virulence when these two species co-exist in the host, which was effectively prevented by the peptide. These findings provide novel insights into the treatment of medically relevant bacterial–fungal interaction.

**Keywords:** *Galleria mellonella*; polymicrobial biofilm; experimental method in vivo

#### **1. Introduction**

Microorganisms rarely exist as single-species planktonic forms, and the biofilm mode of growth is the most common lifestyle adopted. Biofilms can be defined as biotic and abiotic surface-associated, structured microorganism communities embedded in an extracellular polysaccharide matrix. Living in a biofilm provides protection in a stressful environment where mechanical stress, desiccation, and biocides are frequent threats.

Progress in biofilm research has highlighted that these communities are rarely composed of a single-species microorganism, but mainly exist as complex, diverse, and heterogeneous structures. In fact, multiple species (fungi, bacteria, and viruses) frequently exist together in complex polymicrobial biofilm communities attached to sites where they compete for space and nutrients [1–3]. Moreover, polymicrobial biofilms are likely to influence disease severity by promoting intensified pathogenic phenotypes, including increased resistance to both host defense and antimicrobial therapies. Despite their clinical significance, polymicrobial biofilm infections continue to be largely understudied [4].

*Candida albicans* and *Klebsiella pneumoniae* are important pathogens causing a wide variety of infections. They possess the ability to co-exist as complex polymicrobial biofilms within the human host [5]. *C. albicans* is the predominant fungus frequently present in hospital infections with significant morbidity and mortality rates; unfortunately, it

**Citation:** Maione, A.; de Alteriis, E.; Carraturo, F.; Galdiero, S.; Falanga, A.; Guida, M.; Di Cosmo, A.; Maselli, V.; Galdiero, E. The Membranotropic Peptide *gH625* to Combat Mixed *Candida albicans*/*Klebsiella pneumoniae* Biofilm: Correlation between In Vitro Anti-Biofilm Activity and In Vivo Antimicrobial Protection. *J. Fungi* **2021**, *7*, 26. https://doi.org/10.3390/ jof7010026

Received: 29 October 2020 Accepted: 31 December 2020 Published: 5 January 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/).

is difficult to prevent, diagnose and treat. It is an opportunistic pathogen, which is a major cause of invasive fungal disease, principally in immunocompromised individuals, such as patients with organ transplantation and HIV infection or patients undergoing chemotherapy. The superficial mucosal and dermal infections caused by *C. albicans* can be disseminated to bloodstream infections with a mortality rate higher than 40% [6].

*Klebsiella pneumoniae*, the most common carbapenem-resistant member of the *Enterobacteriaceae* family, has emerged as an important opportunistic pathogen, mostly causing nosocomial infections associated with mortality rates up to 50% [7]. *K. pneumoniae* is responsible of infections both in human gastrointestinal tract and lung environments, and has a high propensity to form mono- and polymicrobial biofilms with consequent treatment difficulties [8].

Mixed biofilms of *Candida* with different bacteria, including *K. pneumoniae*, are present on implanted devices such as intravascular or urinary catheters, as well as in the oral environment [2].

Notwithstanding its clinical importance, information on mixed biofilms of *Candida/Klebsiella* is relatively scarce [9].

The potential use of antimicrobial peptides (AMPs) as a valid alternative to conventional antibiotics has been acknowledged and widely studied. In fact, their fast and strong antimicrobial activity, their antibiofilm action, and their reduced induction of resistance compared to conventional antibiotics make AMPs relevant compounds for controlling infections due to multi-drug resistant microorganisms embedded in a biofilm [10–12]. Among AMPs, there is a particularly relevant class of peptides, known as membranotropic peptides, which, apart from their eventual antimicrobial activity, present a high ability to disrupt the biofilm and thus may have an action both in the inhibition and in the eradication of the biofilm [13].

In this study, we evaluated the anti-biofilm activity of an analogue of the membranotropic peptide gH625, namely gH625-M, on a dual-species *C. albicans*/*K. pneumoniae* biofilm. Peptide gH625-M (HGLASTLTRWAHYNALIRAFGGGKKKK) is derived from gH625 (HGLASTLTRWAHYNALIRAF) and presents a C-terminal lysine sequence conjugated to the gH625 peptide by a glycine linker; the lysine functions to enhance the interaction with the negatively charged surfaces of bacterial biofilms and to enhance solubility. The glycine linker between the gH625 sequence and the lysine residues provides conformational flexibility to the peptide. gH625-M was shown to have activity on polymicrobial biofilms of *Candida tropicalis* and *Serratia marcescens* or *Staphylococcus aureus* [14] and on biofilms derived from *C. albicans* persister cells [15]. Therefore, the peptide was selected as a good candidate to evaluate the antibiofilm activity in vitro on a static biofilm of *C. albicans*/*K. pneumoniae*.

Since in vivo studies are crucial for the evaluation of the antimicrobial activities of new therapeutic agents and their modulatory effects on immune response, the larvae of the wax moth *G. mellonella* were used. In particular, *G. mellonella* is frequently exploited as an alternative to the murine model for studying microbial infections, because it is a simple, cheap and fast method, and implies fewer ethical concerns compared to the use of vertebrate models. As a matter of fact, several other studies have recently used *G. mellonella* to investigate the in vivo activity of antimicrobial agents against pathogenic microorganisms, including bacteria and fungi [16].

Here, we investigated for the first time, the correlation between the susceptibility profile shown by gH625-M in vitro and its antimicrobial efficacy in vivo. Therefore, the infection in the *G. mellonella* larva model was evaluated through the impact of gH625-M on the survival rate and on the immune response (galiomycin) of the larvae [17]. Furthermore, the expression of biofilm related genes of *C. albicans* (*HWP1*, *ALS3*) [18], and *K. pneumoniae* (*luxS*, *mrkA*) in *G. mellonella* larvae was investigated [19,20].

#### **2. Materials and Methods**

#### *2.1. Peptide Synthesis*

Peptide gH625-M (HGLASTLTRWAHYNALIRAFGGGKKKK) was synthetized by the Fmoc-solid-phase method [21]. Briefly, all amino acids were protected at their amino terminus with the Fmoc (9-fluorenylmethoxycarbonyl) group and coupled to the growing chain after activation of the carboxylic acid group. Consecutive cycles of amino deprotection (30% piperidine in dimethylformamide, for 10 min, twice) and coupling were performed to obtain the desired sequence. In particular, the first coupling was performed with four equivalents (equiv) of amino acid and four equiv DIC (Diisopropyl carbodiimide); while from the second coupling, we used four equiv amino acid, four equiv HATU (O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluroniumhexafluorophosphate) and eight equiv DIPEA (diisopropylethylamine): the synthesis was performed using a rink amide resin MBHA (4-methyl-benzhydrylamine-resin, 0.44 mmol/g). Side chain deprotection and cleavage of the peptide from the resin was achieved using an acid solution (95% *v/v* of TFA, trifluoroacetic acid). The peptide was precipitated in cold ethylic ether and the crude peptide was analyzed by HPLC–MS using a gradient of acetonitrile (0.1% TFA) in water (0.1% TFA) from 20 to 80% in 15 min. The purified peptide was obtained with a good yield (approximatively 60%) and its identity was confirmed using a LTQ-XL Thermo Scientific linear ion trap mass spectrometer.

#### *2.2. Microorganisms Culture and Biofilm Formation*

*Candida albicans* ATCC 90028 and *Klebsiella pneumoniae* ATCC 10031 were selected as pathogen representative of fungus and Gram-negative bacteria forming biofilm in hospital environments. *C. albicans* ATCC 90028 was subcultured into Tryptone Soya Broth (TSB) (Oxoid) medium with 1% of glucose and propagated in Sabouraud dextrose agar (1% yeast extract, 1% peptone, 4% glucose, 1% agar) as described previously [14]. *K. pneumoniae* ATCC 10031 was grown in Tryptone Soya Broth (TSB and maintained in Tryptone Soya Agar (TSA) at −80 ◦C.

Biofilm formation was carried out on microtiter plates, as previously described [14], with minor modifications. Briefly, for single species biofilms. *C. albicans* and *K. pneumoniae* suspensions were adjusted to 106 colony-forming units (CFU) mL−<sup>1</sup> in fresh TSB with 0.1% glucose and TSB, respectively. In the case of the dual species biofilm, microorganism suspensions were mixed in TSB with 0.1% glucose (1:1). Aliquots (100 μL) of these suspensions (single or mixed) were added to the wells of sterile flat-bottom 96-well microtiter test plates to allow single or mixed biofilm formation and incubated for 24 or 48 h at 37 ◦C. Then, the wells were washed with PBS in order to remove planktonic and sessile cells weakly attached to the surface. Quantification of biofilm biomass was carried out with crystal violet (CV) staining [22].

Experimental conditions were run in triplicate. Results are presented as mean values from at least four independent experiments.

#### *2.3. Minimum Inhibition Concentration (MIC)*

The concentration of the peptide that inhibited 80% microbial growth (MIC80) was determined by the microdilution method according to Clinical Laboratory Standards Institute guidelines [23]. Briefly, 100 μL of TSB with 1% glucose and TSB containing strain of *C. albicans* or *K. pneumoniae*, respectively (1 × 106 CFU/mL), was introduced into each well of 96-well microplate with different concentrations of gH625-M (2.5 μM–50 μM) and incubated for 24 h at 37 ◦C. The growth of each strain was measured at 590 nm wavelength with a plate reader (SYNERGY H4 BioTek).

#### *2.4. Inhibition and Eradication Biofilm Assays*

The inhibition activity of gH625-M on mono- and polymicrobial biofilm formation and the eradication activity against preformed biofilm were evaluated by using sub-MIC concentrations of the peptide ranging from 2.5 to 50 μM.

Briefly, for the inhibition assay, peptide was added together with the standardized inoculum in each well of a 96-well polystyrene microtiter plate and incubated at 37 ◦C for 24 h. Non-adherent microorganisms were removed by washing twice with 200 μL sterile PBS and adherent cells were fixed by incubation for 1 h at 60 ◦C and stained for 5 min at room temperature with 100 μL 1% crystal violet solution. Wells were then rinsed with distilled water and dried at 37 ◦C for 30 min. Biofilms were de-stained by treatment with 100 μL 33% glacial acetic acid for 15 min and OD570 measured.

The eradication activity of the peptide was evaluated exposing 24 h mono- and polymicrobial biofilms for additional 24 h to different sub-MIC concentrations of peptide and quantified the biomass by crystal violet assay as previously described. The percentages of inhibition or eradication were calculated as biofilm reduction % = OD570 control − OD570 sample/OD570 control × 100, where OD570 control and sample were the biomass formed in the absence and in the presence of the peptide, respectively. All tests were performed in triplicate in three independent experiments.

#### *2.5. Galleria Mellonella Survival Assay*

To determinate the in vivo effects of gH625-M, a *G. mellonella* survival assay was performed as described previously [24,25]. In brief, larvae of 250–300 mg each were used for each treatment (20 for each group). They were chosen to have clear color and a lack of spots and/or dark pigments on their cuticle. The experiments were performed in triplicate.

Larvae were cleaned by an alcohol swab prior to injection. Larvae were injected directly into the hemocoel with 10 μL *C. albicans* and/or *K. pneumoniae* suspensions prepared in PBS at a concentration of 1 × 106 CFU/larvae/pathogen (1 × <sup>10</sup><sup>6</sup> CFU/larvae total for co-infection), using a 50 μL microsyringe via the last left proleg. An aliquot of 10 μL of 50 μM gH625-M was delivered behind the last proleg on the opposite side of the pathogen injection site either 2 h pre-infection (for prevention experiments) or 2 h post-infection (for treatment experiments). One group of untreated larvae served as a blank control group (intact larvae), one group received 10 μL of PBS solution per leg and one group was injected with 10 μL of 50 μM gH625-M in one leg and 10 μL PBS in the other, in order to assess peptide toxicity.

Larvae were then incubated at 35 ◦C in plastic containers provided with a perforated lid and monitored daily for survival for 4 days. A larva was considered dead when it displayed no response to touch.

#### *2.6. Fungal/Bacterial Burden*

Larvae were inoculated with *C. albicans* or *K. pneumoniae* or co-infected with the two, at concentration of 1 × 106 CFU/larvae/pathogen or 1 × 106 CFU/larvae total for coinfection. The peptide was administered before or after infection/co-infection, as described in the previous paragraph. The infected models were incubated at 30 ◦C, and after 24 h of infection, two larvae were randomly selected and washed in 70% ethanol. Larvae were cut into small pieces using a sterile scalpel and added to falcon tubes containing 1 mL of PBS vortexed and 100 μL of each sample was collected and serially diluted. The dilutions were plated and incubated 48 h at 30 ◦C and colony forming units (CFU) of *C. albicans* and *K. pneumoniae* were counted on Sabouraud dextrose agar plus 20 μg mL−<sup>1</sup> cloramphenicol and TSA plus 1 μg mL−<sup>1</sup> caspofungin, respectively. The experiments were performed in triplicate.

#### *2.7. RNA Extraction and Gene Expression Analysis*

Larvae were infected, and RNA was extracted at 4 and 24 h post-treatment.

Therefore, three live larvae from each experimental group (4 and 24 h post-treatment) were snap-frozen in liquid nitrogen and ground to a powder by mortar and pestle in TRIzol (Invitrogen, Paisley, UK). The samples were further homogenized using a TissueLyser II (Qiagen, Valencia, CA, USA) and steal beads of 5 mm diameter (Qiagen, Valencia, CA, USA). RNA was extracted with RNeasy minikit (Qiagen, Valencia, CA, USA), following

the manufacturer's protocol. The quality and amount of purified RNA were analyzed spectrophotometrically with Nanodrop2000 (Thermo Scientific Inc., Waltham, MA, USA). 1000 ng of RNA was reverse transcribed with the QuantiTect Reverse Transcription Kit (Qiagen, Valencia, CA, USA), used as described by the manufacturer. Afterwards, Real-Time PCR was performed using the QuantiTect SYBR Green PCR Kit (Qiagen, Valencia, CA, USA) in a final volume of 25 μL, with 100 ng of cDNA, 1 μM of each primer, 12.5 μL of QuantiFast SYBR Green PCR Master Mix (2×). PCR cycling profile consisted of a cycle at 95 ◦C for 5 min, 40 two-step cycles at 95 ◦C for 15 s, at 60 ◦C for 60 s. Quantitative RT-PCR analysis was conducted using the 2(−ΔΔC(T)) method [26]. RT-PCR was performed in a Rotor-Gene Q cycler (Qiagen, Valencia, CA, USA). All primers used for quantitative PCR (qPCR) studies are shown in Table 1. At the end of each test, a melting curve analysis was done (plate read every 0.5 ◦C from 55 to 95 ◦C) to determine the formation of the specific products. Each sample was tested and run in duplicate. No-template controls were included.



mRNA levels in the different treatments were compared by ANCOVA (analysis of covariance). The control and the treatment groups in various assays were compared and analyzed using a Wilcoxon two group test and data with *p*-values < 0.05 were considered statistically significant [27].

Transcriptional activation is represented by the RNA fold change of the expression; for the *galiomycin* gene evaluation in *G*. *mellonella* actin was used as housekeeping gene, for the *luxS* and *mrkA* genes in *K. pneumoniae* 16S was used as housekeeping gene, and for *HWP1* and *ALS3* genes in *C. albicans* actin was used as housekeeping gene.

#### *2.8. Scanning Electron Microscopy (SEM)*

The 48 h polymicrobial biofilm was formed in multi-well plates as described above. The slides were prepared for scanning electron microscopy (SEM) using a previously published protocol [14]. Briefly, the slides were placed in 3% glutaraldehyde at 4 ◦C, then washed with PBS and post-fixed in 1% aqueous solution of osmium for 90 min at room temperature. Then, samples were dehydrated in a series of graded alcohols, dried to the critical drying point, and finally coated with gold. Specimens were evaluated with a scanning electron microscope (QUANTA 200 ESEM FEI Europe Company, Eindhoven, The Netherlands).

#### *2.9. Statistical Analysis*

Statistical analyses were performed using Microsoft® Excel 2016/XLSTAT©-Pro (version 7.2, Addinsoft, Inc., Brooklyn, NY, USA). Error bars in the graphs represent standard error of the mean (SEM and gene expression analysis) or standard deviations (SD, for biomass in mono- and polymicrobial biofilms, inhibition and eradication biofilm assay, CFU assay).

In the *G. mellonella* model of infection, the survival curves were plotted using Kaplan– Meier method and log-rank test. In all other assays, Tukey test was used to compare the means within the same set of experiments and ANOVA to consider the differences among the groups. A *p*-value of <0.05 was considered statistically significant.

#### **3. Results**

Both the examined strains were able to form single- and dual-species biofilm in vitro (Figure 1) under the experimental conditions adopted. In particular, the single biofilm production was weak for both species and moderate for dual species. The biomass of the dual species biofilm was even higher than the sum of the single species biofilm biomass, which likely indicated a synergism between the two species forming the biofilm. The analysis of the SEM images clearly showed the strict interconnection between the two species in the mixed biofilm (Figure 2). The co-existence of *C. albicans* and *K. pneumoniae* was clearly shown in the SEM image (Figure 2) and also confirmed by cell count: 4.1 × <sup>10</sup><sup>8</sup> ± 0.2 (SD) and 4.6 × <sup>10</sup><sup>8</sup> ± 0.4 (SD), respectively.

**Figure 1.** Biomass in mono- *(K. pneumoniae* and *C. albicans)* and polymicrobial biofilms (*n* = 3, mean ± SD) quantified by crystal violet staining and expressed as OD570. The dotted line corresponds to ODc, that is the cut-off value, defined as three standard deviations above the mean OD of the negative control. Negative (OD ≤ ODc), weak (ODc ≤ OD ≤ 2 ODc), moderate (2 ODc < OD ≤ 4 ODc), and strong biofilm production (4 ODc < OD), according to Stepanovich [22].

**Figure 2.** SEM observation of the 48 h dual species biofilm of *C. albicans* and *K. pneumoniae*. Scale bar = 10 μm.

The peptide gH625-M is an analogue of gH625, a peptide, which proved to be very effective in crossing membrane bilayers [15,28]. *In vitro*, gH625-M showed weak anticandidal activity with a MIC80 > 50 μM, which was in agreement with our previous studies [15]. Similarly, the MIC80 value of *K. pneumoniae* was higher than 50 μM, indicating a relatively scarce antibacterial activity of the peptide.

Interestingly, gH625-M was able to inhibit the formation of the biofilm, as well as to eradicate it (Figure 3A,B). After treatment with sub-MIC doses of gH625-M, the formation of mono- and polymicrobial biofilm was inhibited significantly (Figure 3A).

**Figure 3.** Action of increasing gH625-M concentrations on inhibition (**A**) and eradication (**B**) of mono- and polymicrobial biofilms. Quantification of the residual biofilm biomass was performed by crystal violet staining. Data with different letters (a–c; w–z) are significantly different (Tukey's, *p* < 0.05).

Figure 3B showed the eradication effect of gH625-M on preformed mono- and polymicrobial biofilm (48 h old). At 50 μM, gH625-M reduced the mono-preformed biofilms of *C. albicans* and *K. pneumoniae* of 80% and 50%, respectively, and the mixed biofilm of 50%. The in vivo antimicrobial activity of gH625-M was evaluated using *G. mellonella* larvae infected with *C. albicans* and *K. pneumoniae* isolates alone or mixed as shown in Figure 4.

**Figure 4.** Kaplan–Meier plots of survival curves of *G. mellonella* larvae infected with *C. albicans* (**A**), *K. pneumoniae* (**B**), and co-infected with *C. albicans* + *K. pneumoniae* (**C**). In all panels, survival curves of larvae are shown. All groups were treated with 50 μM gH625-M before or after infection/co-infection are reported. All groups were compared with control (infected or co-infected larvae). In all panels, survival curves of intact larvae, larvae injected with PBS, larvae injected with gH625 alone are reported. \* represents *p*-value < 0.001 (log-rank test).

Larval survival assay indicated that groups of larvae injected with PBS alone and gH625-M alone presented about 80% survival up to 96 h of observation with respect to intact larvae, indicating that gH625-M was not toxic for the larvae. Survival of larvae infected with *C. albicans* or *K. pneumoniae* (Figure 4A,B) was only 20% and 40% after 24 h, with 100% mortality observed after at 72 and 96 h, respectively.

To determine whether gH625-M had an effect in vivo, *G. mellonella* larvae were treated with gH625-M at 50 μM before or after the infection with each of the two species (Figure 4A,B) and before or after co-infection with the two (Figure 4C).

Both pre- and post-infection treatments showed significantly higher survival rates. In particular, the survival of larvae treated with gH625-M before infection with *C. albicans* was about 70%, and for *K. pneumoniae* was 80% after 24 h, preserving about 50% and 70% survival respectively at the end of the observation period. Moreover, the survival of the larvae treated with gH625-M after the infection with *C. albicans* or *K. pneumoniae* was significant, resulting in higher than 60% at 24 h, and 40% at 72 h of the experiment for both the microorganisms.

These results indicate that when gH625-M was administered before the infection, it was more effective compared to when administered after infection for both microorganisms. However, the increased survival rate of infected *Galleria* further confirms the significant activity of gH625-M.

In Figure 4C, survival of larvae co-infected with both *C. albicans* and *K. pneumoniae* is reported. Compared to single infection, mortality was enhanced, being 90% and 100% after 24 and 72 h. For co-infections, the administration of gH625-M greatly improved larvae survival both when given before and after the co-infection, with a slightly better prophylactic effect.

To detect the effect of gH625-M on the fungal and bacterial burden of infected larvae, a burden analysis was performed (Figure 5).

**Figure 5.** Effect pre- and post-treatment with gH625-M on bacterial/fungal burden in *G. mellonella* infected with *C. albicans* or *K. pneumoniae* and co-infected with *C. albicans* + *K. pneumoniae*. The peptide was administered before or after infection/co-infection. \* indicates that the differences vs. larvae injected with microorganisms alone or together is statistically significant (*t*-test; *p* < 0.05). Error bars represent the SD.

There was a significant decrease in the microbial burden for the gH625-M pre-treated groups and only a slight decrease for the gH625-M post-treated groups. These results corroborated the protective action of gH625-M towards infection and co-infection, as already seen in the analysis of survival curves.

The expression of the galiomycin peptide-encoding gene is associated with the immune response of *G*. *mellonella*. Galiomycin is an antimicrobial peptide playing a major role in innate immunity, showing broad-spectrum microbicidal activity and specificity to *G. mellonella*. To evaluate the insect humoral response after infection with and co-infection each of the two microorganisms and co-infection, as well as the role of gH625-M on infection and co-infection we evaluated the expression of the *galiomycin* peptide-encoding gene.

Figure 6 reports expression levels of *galiomycin* gene in *G. mellonella* larvae after 4 and 24 h after infection with each species, after co-infection with the two species and after treatment with the peptide. Analysis by real-time quantitative PCR showed that the levels of galiomycin were higher in insects infected after 4 h with *C. albicans* alone (*p* < 0.05) and co-infected with *C. albicans* and with *K. pneumoniae* (*p* < 0.05) than those found in insects infected with *K. pneumoniae* alone. Instead, the expression levels of the same gene after 24 h from the infection did not significantly differ among the insects infected with both microorganisms. In this context, it could be reasonable to explore some other markers able to evidence the biofilm-associated damage in infected larvae, or lack thereof in gH625M-treated larvae.

**Figure 6.** Relative mRNA expression levels of galiomycin gene in *G. mellonella* larvae infected with *C. albicans* and *K. pneumoniae* alone or together and pre-treated or post-treated with gH625-M, at 4 and 24 h, measured using real-time PCR analysis and calculated by the 2(−ΔΔC(T)) method. Actin gene was used as housekeeping. Each sample was tested and run in duplicate. No-template controls were included. \* asterisk indicates that the difference vs. intact larvae expression is statistically significant (Wilcoxon two group test, *p* < 0.05). Error bars represent the SEM.

Analysis of the expression of the selected genes revealed that the use of gH625- M before and after the infection, and the co-infection, significantly affects expression, decreasing the level of *galiomycin* compared to the corresponding intact and not treated samples. Data also show that the treatment with gH625-M produces an inhibition of galiomycin gene expression.

In Figure 7, levels of hyphal-specific and biofilm-related genes in *C. albicans* in the absence and presence of gH625-M at 24 h were quantified by real-time PCR. Hyphal specific gene, *HWP1* was downregulated in all cases showing hyphal structure formation failure. Biofilm-related gene *ALS3* was significantly downregulated in the presence of gH625-M 2 h after infection with *C. albicans* and in polymicrobial treated with gH625-M two hours before the infection. It was upregulated in larvae with only *C. albicans* infection pre-treated and in polymicrobial biofilm untreated or post-treated with gH625-M.

The relative expressions of *luxS* and *mrkA* genes were evaluated in *G. mellonella* infected with *K. pneumoniae* alone or in combination with *C. albicans* (Figure 8). The *luxS* gene encodes the AI-2 proteins of quorum sensing that play an important role in biofilm formation in Gram-negative bacteria such as *K. pneumoniae*, while *mrkA* gene (type 3 fimbriae) is a virulence -related gene detected in *K. pneumoniae;* both have critical roles in biofilm formation and antibiotic resistance.

It is interesting to observe that in co-infections, when the peptide was administered before, both *mrkA* and *luxS* genes were significantly down-regulated compared to untreated co-infections. The effect of pre-treatment was also observed in the case of *K. pneumoniae* infection. In contrast, both *mrkA* and *luxS* genes were upregulated when gH625-M was administered after infection and co-infection.

**Figure 7.** Relative mRNA expression levels of *C. albicans* virulence genes (*HWP1* and *ALS3*) in *G. mellonella* larvae at 24 h, measured using real-time PCR analysis and calculated by the 2(−ΔΔC(T)) method. Actin gene was used as housekeeping. Each sample was tested and run in duplicate. No-template controls were included. \* asterisk indicates that the difference vs. expression of larvae treated with *C. albicans* only is statistically significant (Wilcoxon two group test, *p* < 0.05). Error bars represent the SEM.

**Figure 8.** Relative mRNA expression levels of *K. pneumoniae* virulence genes (*luxS* and *mrkA*) in *G. mellonella* larvae at 24 h, measured using real-time PCR analysis and calculated by the 2(−ΔΔC(T)) method. 16S gene was used as housekeeping. Each sample was tested and run in duplicate. No-template controls were included. \* asterisk indicates that the difference vs. expression of larvae treated with *K. pneumoniae* only is statistically significant (Wilcoxon two group test *p* < 0.05). Error bars represent the SEM.

#### **4. Discussion**

Infections associated with polymicrobial fungal/bacterial biofilms represent a huge challenge due to intrinsic heterogeneity of these consortia, the low susceptibility to traditional drugs, as well as the high toxicity of many common antifungals. In this context, the development of novel strategies to combat polymicrobial biofilms of pathogen species is of great relevance.

In this paper, we focused our attention on the dual-species biofilms formed by *C. albicans* and *K. pneumoniae*, with *C. albicans* being the most common fungal opportunistic pathogen and *K. pneumoniae* being recognized in the group of 'ESKAPE' (*Enterococcus faecium*, *Staphylococcus aureus*, *Klebsiella pneumoniae*, *Acinetobacter baumannii*, *Pseudomonas aeruginosa*, *Enterobacter* spp.) pathogens, which 'escape' from the action of several antibiotics [29].

Previous work has demonstrated that the membranotropic peptide gH625-M was very effective against *C. albicans* biofilm, also when the biofilm was developed from persister cells [15]. In the present paper, we explored the efficacy of the peptide towards the dualspecies *Candida*/*Klebsiella* biofilm, showing its ability of both inhibiting and eradicating in vitro the biofilm at sub-MIC concentrations. The low antimicrobial activity of the peptide used in this work towards *C. albicans* and *K. pneumoniae* was very low and was similar to that of other previously tested membranotropic peptides [14,15]; nevertheless, these characteristics may represent an advantageous property in the prevention of the possible development of resistances in the entire microbial consortium, a feature typical of several compounds proposed as anti-biofilms molecules [11]. The mechanism of action of gH625-M against *C. albicans* monospecies biofilm, as shown by a CSLM analysis previously reported [15], is initially directed towards the esopolymeric matrix of the biofilm and then penetrates across cell membranes, producing a local and temporary destabilization of membranes, which nonetheless has scarce effects on cell viability. The results obtained in the present paper, with the scarce antimicrobial activity (MIC values > 50 uM) exhibited in the case of both fungal (*Candida*) and bacterial (*Klebsiella*) cells, further support the hypothesis that the de-structuring property is presumably at the basis of the inhibiting and eradicating efficacy of the peptide also in the case of the *Candida*/*Klebsiella* biofilm. Interestingly, the dual-species biofilm is characterized by a strict interconnection established between the two species, thus the ability to disrupt the structure of the biofilm is of even greater relevance.

*Candida/Klebsiella* biofilm-associated infections have been reported in mammalian hosts; nonetheless, the description of the interactions between *Candida* spp. and *K. pneumoniae* in mixed biofilms has been limited to few observations, derived from in vitro models [9]. Our results show an increase in the total biomass of the polymicrobial biofilm in vitro compared to the sum of the single species biofilm, suggesting a possible synergy between the two species.

In vivo studies are crucial for the evaluation of the efficacy of new therapeutic agents and their modulatory effects on the host immune response, which makes it necessary to evaluate outcomes in animal models. In this study, we used the larvae of the wax moth *G. mellonella* as an in vivo model, for its well-known advantages as an alternative to vertebrate models. Although *G. mellonella* does not replace the vertebrate model, it can be exploited as a screening step between in vitro and in vivo evaluations [16].

One of the main outcomes was derived from the analysis of the survival curves after fungal/bacterial infection in the larvae [30]. Our results clearly show that the killing of *G. mellonella* larvae by infection of the two pathogens together is greater than the sum of killing with each pathogen alone; this pattern suggests a synergistic pathogen–pathogen interaction or a change in host–pathogen interactions that is characterized by increased host susceptibility to one or both of the pathogens.

The analysis of the survival curves in experiments with gH625-M, showed that the peptide functions in vivo both as a prophylactic and therapeutic agent towards both single infections and co-infections, with a higher efficacy in prophylaxis as revealed also with the fungal burden analysis.

The actual interaction established in vivo between the two species examined and the host, as well as the action of the peptide, was further investigated in the larvae model through the analysis of the transcriptional profiles of both biofilm-associated genes and gene associated with the larvae innate immune response. Following infection, the expression of the biofilm-associated genes of the two species examined was enhanced in the larvae, suggesting the occurrence of a biofilm-like interaction in the animal, which was concomitant to the overexpression of the *galiomycin* gene indicating an activation of the larvae immune response, as expected. The reduction in the expression of the *galiomycin* gene following administration of the peptide seems to suggest that gH625-M could have an anti-inflammatory effect which may result in the protection of the infected larvae.

Interestingly, gH625-M treatment before the infection, significantly reduced the biofilmassociated gene expression of *HWP1* and *ALS3* particularly in co-culture, supporting the efficacy of the peptide already observed in vitro. It was previously reported that *HWP1* mutants produce a thin biofilm with less hyphae in vitro, but display serious biofilm defects in vivo, only forming yeast microcolonies, while *ALS3* mutants are able to form hyphae, but exhibit defects in biofilm formation [31,32]. Thus, the observed down-regulation of hyphal specific gene, *HWP1*, could determine a loss of physical scaffolds for yeast cell adhesion and aggregation, producing a decreased biofilm strength, integrity, and maturation. The results obtained confirm previous hypothesis on the ability of gH625-M in regulating the initial adhesion of yeast cells to surfaces, which is essential for all stages of biofilm development.

For *K. pneumoniae*, too, genes involved in biofilm formation and virulence indicated that only the pre-treatment with gH625-M before infection has a significant effect in decreasing gene activity, confirming that this peptide was able to reduce infection and biofilm formation. However, future experiments relative to differential gene expression after 4 h and 12 h, when % of viability is higher than 50%, could clarify the action of these genes involved in *G. mellonella* immune response and biofilm formation in *C. albicans* and *K. pneumoniae.*

The results obtained in this study confirm the importance of developing new strategies for dealing with polymicrobial biofilms and we foresee a novel role for membranotropic peptides such as gH625 for the inhibition of biofilm formation thanks to their physicochemical properties. It is likely that conformational flexibility and ability to destabilize hydrophobic domains typically present in membrane bilayers makes them able to disrupt the structure of the biofilm (eradication) or interfere with biofilm formation (inhibition). In conclusion, membranotropic peptides represent an appealing strategy to further evaluate for the development of innovative therapies meant to address problems such as biofilm inhibition/eradication and resistance.

**Author Contributions:** Conceptualization, E.d.A. and E.G.; Data curation, A.F.; Formal analysis, V.M.; Funding acquisition, M.G.; Investigation, A.M.; Project administration, F.C.; Resources, M.G.; Software, A.M.; Supervision, S.G.; Validation, E.d.A. and A.D.C.; Visualization, E.G.; Writing original draft, V.M.; Writing—review & editing, S.G. and E.G., All authors have read and agreed to the published version of the manuscript.

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

**Institutional Review Board Statement:** Ethical review and approval were not required for this animal species.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

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

#### **References**


## *Article* **A Screen for Small Molecules to Target** *Candida albicans* **Biofilms**

**Matthew B. Lohse 1,2, Craig L. Ennis 3,4 , Nairi Hartooni 1,† , Alexander D. Johnson 1,5,\* and Clarissa J. Nobile 4,6,\***


**Abstract:** The human fungal pathogen *Candida albicans* can form biofilms on biotic and abiotic surfaces, which are inherently resistant to antifungal drugs. We screened the Chembridge Small Molecule Diversity library containing 30,000 "drug-like" small molecules and identified 45 compounds that inhibited biofilm formation. These 45 compounds were then tested for their abilities to disrupt mature biofilms and for combinatorial interactions with fluconazole, amphotericin B, and caspofungin, the three antifungal drugs most commonly prescribed to treat *Candida* infections. In the end, we identified one compound that moderately disrupted biofilm formation on its own and four compounds that moderately inhibited biofilm formation and/or moderately disrupted mature biofilms only in combination with either caspofungin or fluconazole. No combinatorial interactions were observed between the compounds and amphotericin B. As members of a diversity library, the identified compounds contain "drug-like" chemical backbones, thus even seemingly "weak hits" could represent promising chemical starting points for the development and the optimization of new classes of therapeutics designed to target *Candida* biofilms.

**Keywords:** high-throughput screens; biofilms; biofilm inhibition; biofilm disruption; *Candida albicans*; antimicrobial resistance; therapeutics; Chembridge Small Molecule Diversity library

**1. Introduction**

*Candida albicans* is a normal commensal of the human microbiota that asymptomatically colonizes the skin, the mouth, and the gastrointestinal tract of healthy humans [1–4]. *C. albicans* is also one of the most common fungal pathogens of humans, typically causing superficial mucosal infections in healthy individuals [1,5–11]. When a host's immune system is compromised (e.g., in patients with AIDS), *C. albicans* can give rise to disseminated bloodstream infections with mortality rates exceeding 40% [1,12–15].

A notable virulence trait of *C. albicans* is its ability to form biofilms, multilayered, structured communities of cells that can grow on biotic and abiotic surfaces, such as mucosal surfaces and implanted medical devices (e.g., catheters, dentures, and heart valves) [1,2,10,16–21]. These biofilms are often resistant to antifungal drugs at concentrations that are normally effective against planktonic (free-floating) cells [20–25]. The drug-resistant nature of *C. albicans* biofilms frequently makes removal of biofilm-infected medical devices the only effective option to mitigate a biofilm-based infection, which can be especially problematic if patients are already critically ill or when device removal requires

**Citation:** Lohse, M.B.; Ennis, C.L.; Hartooni, N.; Johnson, A.D.; Nobile, C.J. A Screen for Small Molecules to Target *Candida albicans* Biofilms. *J. Fungi* **2021**, *7*, 9. https://dx.doi.org/ 10.3390/jof7010009

Received: 30 October 2020 Accepted: 23 December 2020 Published: 27 December 2020

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

**Copyright:** © 2020 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/).

surgical procedures (e.g., heart valve or prosthetic replacement) [20,26,27]. Since there are no biofilm-specific therapeutics available on the market today and only three major classes of antifungal drugs used to treat fungal infections in humans, the development of new therapeutics effective against *C. albicans* biofilms is an important and unmet medical need.

The search for new antibiofilm therapeutics has encompassed a wide range of approaches, many of which focus on compounds that have combinatorial effects with known antifungal drugs rather than (or in addition to) compounds that affect *Candida* biofilms by themselves [28–43]. One approach has focused on screening libraries of existing drugs and/or pharmacologically active compounds that would make promising candidates for repurposing [38–40,43,44]. A second approach has focused on targeted classes of compounds (e.g., compounds with known effects on signaling pathways [28,29], compounds with known effects on cell–cell communication [34], and secreted aspartyl protease inhibitors [42]) that influence specific aspects of *Candida* biology. In addition, compounds that might affect *Candida* that are produced by other organisms (e.g., antimicrobial peptides [35] and chemicals produced by plants [36,37]) are also included in this targeted approach. A third approach has focused on screening large, chemically diverse compound libraries to identify pharmacophores that inhibit and/or disrupt biofilms through novel mechanisms [41,45]. Examples of this third approach taken to identify compounds with effects against *C. albicans* biofilms include a screen of a 20,000 compound Chembridge NOVACore library [45] and a screen of a 120,000 compound National Institutes of Health Molecular Libraries Small Molecule Repository library [41]. Here, we report a screen of a 30,000 compound Chembridge Small Molecule Diversity library (a library which, we note, has few compounds that overlap with the NOVACore library from the same commercial vendor) for the ability of the compounds to inhibit biofilm formation and/or disrupt mature biofilms by themselves or in combination with the known antifungal drugs fluconazole, amphotericin B, and caspofungin.

#### **2. Materials and Methods**

#### *2.1. Media and Strains*

Media were prepared in accordance with previously reported biofilm protocols [46,47]. Yeast extract peptone dextrose (YEPD) liquid media contains 2% BactoTM peptone (Difco #211677 (Becton, Dickinson and Company, Franklin Lakes, NJ, USA)), 2% dextrose, and 1% yeast extract (Difco #212750 (Becton, Dickinson and Company, Franklin Lakes, NJ, USA)). YEPD plates also contain 2% agar. Biofilm assays were performed in Roswell Park Memorial Institute (RPMI)-1640 media (containing L-glutamine and lacking sodium biocarbonate, MP Biomedicals #0910601 (MP Biomedicals, Santa Ana, CA, USA)) supplemented with 34.5 g/L 3-(*N*-morpholino)propanesulfonic acid (MOPS) (Sigma #M3183 (Sigma Aldrich, St. Louis, MO, USA)) and adjusted to pH 7.0 with sodium hydroxide before sterilizing using a 0.22 μm filter. All biofilm assays used the previously reported SC5314-derived strain SN425, a commonly used prototrophic **a**/α *C. albicans* standard strain, which was created by introducing *HIS1, LEU2*, and *ARG4* markers back into the SN152 **a**/α *his1 leu2 arg4* strain [48]. Cells were recovered from glycerol stocks for two days at 30 ◦C on yeast extract peptone dextrose (YEPD) plates. Overnight cultures for assays were grown approximately 16 h at 30 ◦C in YEPD media.

#### *2.2. Reagents*

The Chembridge Small Molecule Diversity library, which consists of 30,000 "drug-like" compounds, including diverse and target-directed compounds, was obtained by the University of California – San Francisco's (UCSF's) Small Molecule Discovery Center (SMDC) from commercial vendors and proprietary sources. Stocks of candidate compounds (as well as the three positive control compounds (PC12, 2-[(1,5-dimethyl-1*H*-pyrazol-4-yl)methyl]-7- (4-isopropylbenzyl)-2,7-diazaspiro[4.5]decane, Chembridge Catalog #17159859; PC26, 7-(4 isopropylbenzyl)-2-(tetrahydro-2H-thiopyran-4-yl)-2,7-diazaspiro[4.5]decan-6-one, Chembridge Catalog #80527891; PC27, 7-(4-isopropylbenzyl)-2-[(2-methyl-5-pyrimidinyl)methyl]-

2,7-diazaspiro[4.5]decan-6-one, Chembridge Catalog #61894700) from the Chembridge NOVACore library that were hits from another high-throughput biofilm screen of a different Chembridge compound library [45]) were obtained directly from Chembridge (http://www.hit2lead.com/index.asp) for follow-up testing. Working stocks of the compounds were made at 20 mM in dimethyl sulfoxide (DMSO).

#### *2.3. Biofilm Assays*

The adherence inhibition, the sustained inhibition, and the disruption optical density biofilm assays followed previously reported 384-well format standard protocols [46,47,49,50]. In brief, for the biofilm inhibition assays, compounds were added during the 90 min adherence step (for the adherence and the sustained inhibition optical density biofilm assays) and/or at the 24 h growth step (for the sustained inhibition optical density biofilm assay). For the disruption optical density biofilm assay, a biofilm was grown for 24 h, after which the biofilm was incubated for an additional 24 h in the presence of the compound of interest. At the end of each assay, the media were removed from each well, and the OD600 of each well was measured using a Tecan Infinite M1000 Pro or a Tecan M200 plate reader (Tecan Group Ltd., Männedorf, Switzerland), taking the average of five reads per well.

The high-throughput adherence inhibition optical density biofilm assay screen of the Chembridge Small Molecule Diversity library was robotically conducted at UCSF's SMDC. Compounds were included at a final concentration of 10 μM per well in 384-well plates in the high-throughput adherence inhibition optical density biofilm assay. Compounds were included at a final concentration of 40 μM per well in 384-well plates in the sustained inhibition optical density biofilm and the disruption optical density biofilm assays. Candidate compounds were tested at 12.5 μM in the combination sustained inhibition optical density biofilm and the disruption optical density biofilm assays. In line with previously reported studies [42,43], the combination sustained inhibition optical density biofilm assays used 1 μg/mL amphotericin B, 0.125 μg/mL caspofungin, or 256 μg/mL fluconazole. The combination disruption optical density biofilm assays used 2 μg/mL amphotericin B, 0.5 μg/mL caspofungin, or 256 μg/mL fluconazole. These amphotericin B, caspofungin, and fluconazole concentrations were chosen to be close to but below the effective concentrations (as measured by OD600) in the respective assays in order to leave a dynamic range for observing any combinatorial interactions. The sensitivity of SN425 to the antifungal drugs amphotericin B, caspofungin, and fluconazole in the sustained inhibition optical density biofilm and the disruption optical density biofilm assays is included in File S3.

#### *2.4. Candidate Compound Selection*

Candidate compounds based on the results of the adherence inhibition optical density biofilm assay high throughput screen of the 30,000 compound Chembridge Small Molecule Diversity library were selected as follows. Separate lists of candidate compounds were developed for those compounds with an absorbance at least two standard deviations below that of the DMSO only controls and for those compounds with a B-score of less than −4 [51,52]. Other factors were also included in our selection criteria prioritizations, such as the selection of compounds with most favorable chemistries for optimizations as well as the selection of compounds that are available for purchase from Chembridge (http: //www.hit2lead.com/index.asp) (some compounds became unavailable for commercial purchase during this study). In total, we selected 64 compounds, 28 from the standard deviation list and 36 from the B-score list. Nineteen of these compounds were on both lists for a total of 45 candidate compounds (File S2). We selected three additional compounds available from Chembridge (PC12, Chembridge Catalog #17159859; PC26, Chembridge Catalog #80527891; PC27, Chembridge Catalog #61894700) that were not in the Chembridge Small Molecule Diversity library but were previously reported by Pierce and colleagues to inhibit biofilm formation in a screen of a different Chembridge libraries (the Chembridge NOVACore library) [45] to serve as positive controls. Data from our adherence inhibition

optical density biofilm assay screen of the 30,000 compound Chembridge Small Molecule Diversity library can be found in File S1. A list of the 45 selected candidate compounds can be found in File S2.

#### *2.5. Statistical Analysis and "Hit" Calling for the Biofilm Assays*

Statistical analyses and "hit" calling for the Biofilm Assays followed previously reported protocols [42,43]. For the stand-alone sustained inhibition and disruption optical density biofilm assays, individual repeats of candidate compounds (and controls) were performed in groups of eight wells. Between two and four repeats (16–32 total wells) were performed for each candidate compound. Each plate had seven sets of control wells (56 total wells) containing equivalent volumes of DMSO to the experimental wells spread throughout the plate to reduce positional effects. For each experimental set of eight wells, significance was evaluated relative to all of the control wells from the same plate using Welch's *t*-tests (two-tailed, assuming unequal variance). In order to correct for the multiple comparisons performed, we then applied the Bonferroni correction with α = 0.05. All of the comparisons for a given type of assay (e.g., all of the stand-alone sustained inhibition optical density biofilm assays) were pooled for the multiple comparisons correction step, giving a number of hypotheses, m, of 146 for the sustained inhibition optical density biofilm assay and of 105 for the disruption optical density biofilm assay (for final thresholds of 3.42 × <sup>10</sup>−<sup>4</sup> and 4.76 × <sup>10</sup><sup>−</sup>4, respectively). We then determined whether each experimental repeat (1) had an average absorbance less than the average of the control wells and (2) was significant after the multiple comparisons correction. To be considered a validated "hit", a compound had to satisfy both of these criteria.

For the combination sustained inhibition and disruption optical density biofilm assays, compounds (and controls) were again tested in groups of eight wells, and two distinct groups of controls were included on each plate. The first set of controls were wells where the candidate compound but no known antifungal drug was included. The second set of controls were wells where the antifungal drug but no candidate compound was included. In both cases, we used the same concentration of candidate compound or antifungal drug as was used in the experimental wells. Controls were included for all candidate compounds and antifungal drugs being tested on a given plate. In general, a single set of eight wells was included for each experimental or control condition on a given plate. Statistical analysis was performed using Welch's *t*-test and the Bonferroni correction as described above with the following modifications. Each experimental condition was compared to both the relevant antifungal drug control and the relevant candidate control (e.g., a compound CB01 plus caspofungin experiment was compared to the CB01 only control and the caspofungin only control from the same plate). All of the same comparisons for a given assay (e.g., all of the antifungal drug comparisons for the combination sustained inhibition optical density biofilm assay) were pooled for the multiple comparisons correction step, giving a number of hypotheses, m, of 144 for both the antifungal drug and the candidate compound comparisons in both the sustained inhibition and the disruption optical density biofilm assays (for a final threshold of 3.47 × <sup>10</sup>−4). To be considered a validated combination hit, a given experimental condition had to have (1) an average absorbance less than the averages of both sets of relevant control wells and (2) remain significant after the multiple comparisons correction for both sets of comparisons.

Data and statistics for the stand-alone and the combination biofilm assays are compiled in File S3. The chemical properties of the "hit" compounds (including molecular weights, polar surface area, logP, logSW, the number of rotatable bonds, and the numbers of Hbond acceptors and donors) that were available at the ChemBridge Online Chemical Store (www.hit2lead.com) are also included in File S3.

#### **3. Results**

The Chembridge Small Molecule Diversity library of 30,000 "drug-like" compounds covering a wide range of chemical scaffolds, diverse chemical backbones, chemotypes, and pharmacophores was robotically screened for compounds that inhibit *C. albicans* biofilm formation. This screen used the adherence inhibition optical density biofilm assay [46,47] (Figure 1a), where the compound of interest was added during the 90 min initial step of biofilm formation and then washed out (along with unadhered cells). The biofilm was then allowed to develop for 24 h in the absence of the compound. In total, 45 candidate compounds were then selected for further evaluation in secondary assays (Figure 1b, Files S1 and S2).

The 45 candidate Chembridge compounds (as well as the three positive control Chembridge compounds previously reported to inhibit biofilm formation that were not present in the 30,000 compound Small Molecule Diversity library [45]) were then evaluated for antibiofilm activity in the sustained inhibition optical density biofilm assay and the disruption optical density biofilm assay [46,47]. In the sustained inhibition optical density biofilm Assay, the compounds were added to the media during both the 90 min adherence step and the 24 h growth step of biofilm formation (Figure 1a). In the disruption optical density biofilm assay, a biofilm was grown for 24 h, after which the biofilm was incubated for an additional 24 h in the presence of the compound (Figure 1a). Other than the three positive controls (PC12, 2-[(1,5-dimethyl-1H-pyrazol-4-yl)methyl]-7- (4-isopropylbenzyl)-2,7-diazaspiro[4.5]decane, Chembridge Catalog #17159859; PC26, 7-(4 isopropylbenzyl)-2-(tetrahydro-2*H*-thiopyran-4-yl)-2,7-diazaspiro[4.5]decan-6-one, Chembridge Catalog #80527891; PC27, 7-(4-isopropylbenzyl)-2-[(2-methyl-5-pyrimidinyl)methyl]- 2,7-diazaspiro[4.5]decan-6-one, Chembridge Catalog #61894700) [45], none of the compounds tested inhibited biofilm formation throughout the duration of biofilm development (Figure 1c and Figure S1a). We do not fully understand why some compounds showed significant inhibition in the adherence inhibition optical density biofilm assay but not in the sustained inhibition optical density biofilm assay, but these different assays may be sensitive to different compound parameters such as solubility, stability, and pH dependence. Given the lack of a biofilm inhibition phenotype in the sustained inhibition optical density biofilm assay, we were surprised to find that one of the 45 compounds (CB17, 1-[2-(2-methylphenoxy)-3-pyridinyl]-*N*-(3-pyridinylmethyl)methanamine, Chembridge Catalog #80338143) disrupted mature *C. albicans* biofilms on its own at the same concentration (Figure 1d,e and Figure S1b). See File S3 for names and chemical properties of this compound.

Given the previous reports suggesting antibiofilm synergies between known antifungal drugs and certain drug classes, we next tested our initial 45 candidate compounds for their abilities to inhibit biofilm formation (using the sustained inhibition optical density biofilm assay and/or to disrupt mature biofilms (using the disruption optical density biofilm assay) when combined with sub-inhibitory concentrations of amphotericin B, caspofungin, or fluconazole (Figure 2 and Figures S2 and S3). Three compounds disrupted mature biofilms in the presence of caspofungin (CB14, 2,2- -({[2-(ethylsulfonyl)-1-(3-phenylpropyl)- 1*H*-imidazol-5-yl]methyl}imino)diethanol, Chembridge Catalog #10068182; CB36, *N*-[2-({2- [3-(1-azocanyl)-2-hydroxypropoxy]-4-methoxybenzyl}amino)ethyl]acetamide, Chembridge Catalog #29059737; CB40, 1-{3-[5-(1,3-benzodioxol-5-yl)-1,3,4-oxadiazol-2-yl]propanoyl}-4- (2-ethoxyphenyl)piperazine, Chembridge Catalog #35558198) (Figure 2a and Figure S2a). One of these compounds (CB36) also inhibited biofilm formation in the presence of caspofungin (Figure 2b and Figure S3a). In addition, a fourth compound (CB06, *N*-(2,3-dihydro-1,4-benzodioxin-6-yl)-1-[3-(1*H*-pyrazol-4-yl)propanoyl]-3-piperidinamine, Chembridge Catalog #22164746) inhibited biofilm formation in the presence of fluconazole (Figure 2c and Figure S3b). As noted above, none of these compounds had effects on biofilms on their own in this assay. Chemical properties of these compounds can be found in File S3. We also note that the positive control compounds PC12, PC26, and PC27 all disrupted mature biofilms in the presence of caspofungin, PC12 disrupted mature biofilms in the presence of fluconazole, and PC26 inhibited biofilm formation in the presence of fluconazole (Figure 2 and Figures S2 and S3).

**Figure 1.** Screen of the Chembridge 30,000 "drug-like" member library for compounds with the ability to inhibit *C. albicans* biofilm formation. (**a**) Overview of the adherence inhibition, the sustained inhibition, and the disruption optical density biofilm assays. (**b**) Comparisons of the differences from the mean (in units of standard deviation, *x*-axis) and the B-score (*y*-axis) for the entire library screened at a concentration of 10 μM in the adherence inhibition optical density biofilm assay. The 45 candidate hits that were pursued further are indicated in red, and all other compounds are indicated in black. (**c**,**d**) Statistically significant hits, positive controls, and additional selected candidates from the (**c**) stand-alone sustained inhibition optical density biofilm assay and the (**d**) stand-alone disruption optical density biofilm assay; compounds were included at concentrations of 40 μM. In both panels, the mean OD600 readings with standard deviations are shown. Significant differences from the DMSO solvent control, as determined by Welch's *t*-test (two-tailed, assuming unequal variance) with the Bonferroni correction, are indicated for α = 0.05 (\*) or mixed results (&). In the cases of PC12 (2-[(1,5-dimethyl-1*H*-pyrazol-4-yl)methyl]-7-(4-isopropylbenzyl)-2,7-diazaspiro[4.5]decane, Chembridge Catalog #17159859) and PC27 (7-(4-isopropylbenzyl)-2-[(2-methyl-5-pyrimidinyl)methyl]-2,7 diazaspiro[4.5]decan-6-one, Chembridge Catalog #61894700) in the disruption optical density biofilm assay, only one of the two repeats performed met the significance threshold. Data within a chart were taken from the same plate. (**e**) Structure of compound CB17 (1-[2-(2-methylphenoxy)-3-pyridinyl]- *N*-(3-pyridinylmethyl)methanamine, Chembridge Catalog #80338143) disrupted mature *C. albicans* biofilms on its own at a concentration of 40 μM.

**Figure 2.** Combination screening of candidate compounds with the antifungal drugs caspofungin and fluconazole. (**a**) combination disruption optical density biofilm assay and (**b**) combination sustained inhibition optical density biofilm assay with caspofungin. For each compound, wells with caspofungin (+ caspofungin) are indicated in yellow, and wells without caspofungin (− caspofungin) are indicated in red. (**c**) Combination sustained inhibition optical density biofilm assay with fluconazole. For each compound, wells with fluconazole (+ fluconazole) are indicated in grey and wells without fluconazole − fluconazole) are indicated in blue. Mean OD600 readings with standard deviations are shown, significant differences from the compound without antifungal drug controls (e.g., PC12, − caspofungin), as determined by Welch's *t*-test (two-tailed, assuming unequal variance) with the Bonferroni correction, are indicated for α = 0.05 (\*). Significant differences from the antifungal drug without compound control (e.g., DMSO, + caspofungin), determined by the same statistical analysis, are indicated for α = 0.05 (#). Data from different plates are separated by two vertical lines on the *x*-axis, and DMSO solvent controls are shown for each plate. Candidate compounds were included at concentrations of 12.5 μM in each of these assays. (**d**) Structures of compounds CB06 (*N*-(2,3-dihydro-1,4-benzodioxin-6-yl)-1-[3-(1*H*-pyrazol-4-yl)propanoyl]-3-piperidinamine, Chembridge Catalog #22164746), CB14 (2,2- -({[2-(ethylsulfonyl)-1-(3-phenylpropyl)-1Himidazol-5-yl]methyl}imino)diethanol, Chembridge Catalog #10068182), CB36 (*N*-[2-({2-[3-(1-azocanyl)-2-hydroxypropoxy]- 4-methoxybenzyl}amino)ethyl]acetamide, Chembridge Catalog #29059737), and CB40 (1-{3-[5-(1,3-benzodioxol-5-yl)-1,3,4 oxadiazol-2-yl]propanoyl}-4-(2-ethoxyphenyl)piperazine, Chembridge Catalog #35558198) inhibited and/or disrupted *C. albicans* biofilms in combination with at least one of the known antifungal drugs tested.

#### **4. Discussion**

Starting from an initial screen of a 30,000 compound diversity library and following standard high-throughput screening procedures for hit identification [53], we identified four compounds capable of inhibiting biofilm formation and/or disrupting mature biofilms in combination with caspofungin or fluconazole and a fifth compound capable of disrupting mature *C. albicans* biofilms on its own. As members of a diversity library, the identified compounds contain "drug-like" chemical backbones that represent promising chemical

starting points for the development and the optimization of new classes of therapeutics designed to target *Candida* biofilms. For example, all compounds within this library have low molecular weights, low polar surface areas, and are predicted to be soluble and capable of crossing membranes. Given the distinct structures of our specific individual and combination hits, these compounds are likely to display broad ranges of biological activities and should provide multiple amenable opportunities for structural elaboration. Thus, even seemingly "weak" hits have the potential to become potent hits upon chemical optimizations [53–55]. Therefore, even compounds we identified with relatively minor yet significant antibiofilm effects on their own (e.g., CB17) have promise. In addition, our combination results indicate potent effects for certain compounds (e.g., CB06, CB14, CB36, CB40) in combination with fluconazole and caspofungin, suggesting that these compounds are a priority for future chemical optimizations.

In addition to identifying several promising antibiofilm compounds, our results illustrate the degree to which the experimental setup for biofilm formation can affect compound efficacy. One example is our identification of several compounds with efficacy in combination with known antifungal drugs, where the combined effect is dependent on the assay conditions. A second example is our identification of compounds that disrupt mature biofilms but that do not inhibit biofilm formation (either on their own or in combination with known antifungal drugs). Given these findings, drug efficacy testing that focuses solely on one aspect of biofilm formation (e.g., inhibition of initial biofilm formation) may overlook promising compounds that may be broadly effective against mature biofilms, and vice versa. Thus, multiple testing parameters of compounds against different stages of biofilm formation are useful in identifying the most promising compounds for therapeutic development.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2309-608 X/7/1/9/s1. File S1, Screen of the Chembridge 30,000 "drug-like" member library for compounds with the ability to inhibit *C. albicans* biofilm formation in the adherence inhibition optical density biofilm assay. Differences from the mean (in units of standard deviation) and the B-score for the entire library screened at a concentration of 10 μM are provided. File S2, Identities of the 45 candidate compounds selected based on the initial adherence inhibition optical density biofilm assay as well as the three positive controls. Differences from the mean (in units of standard deviation) and the B-score are indicated for these compounds. File S3, Compiled data and statistics from the standalone and combination sustained inhibition and disruption optical density biofilm assays. For each compound, the average OD600, average OD600 of relevant control(s), and value(s) for Welch's *t*-test versus the relevant control(s) are provided. Whether the average OD600 was below the average OD600 of the relevant control(s) and whether the difference from the relevant control(s) remains significant following the Bonferroni Correction (α = 0.05) are indicated. The chemical properties of the "hit" compounds (including molecular weights, polar surface area, logP, logSW, the number of rotatable bonds, and the numbers of H-bond acceptors and donors) that were available at the ChemBridge Online Chemical Store (www.hit2lead.com) are also included. Figure S1, Additional results from the (**a**) stand-alone sustained inhibition and the (**b**) stand-alone disruption optical density biofilm assays. Mean OD600 readings with standard deviations are shown, significant differences from the DMSO solvent control, as determined by Welch's *t*-test (two-tailed, assuming unequal variance) with the Bonferroni Correction, are indicated for α=0.05 (\*) or mixed results (&). In the cases of CB36 and CB40 in the sustained inhibition optical density biofilm assay, only one of the two repeats performed met the significance threshold. In the case of CB40 in the disruption optical density biofilm assay, only two of the four repeats performed met the significance threshold. Data within a chart are all taken from the same plate on the same day. Figure S2, Additional results from the disruption optical density biofilm assay combination screening of candidate compounds with the antifungal agents caspofungin, fluconazole, and amphotericin B. Combination disruption biofilm assays with (**a**) caspofungin, (**b**) fluconazole, and (**c**) amphotericin B. In panel **a**, wells with caspofungin (+ caspofungin) are indicated in yellow and wells without caspofungin (− caspofungin) are indicated in red. In panel **b**, wells with fluconazole (+ fluconazole) are indicated in grey and wells without fluconazole (−fluconazole) are indicated in blue. In panel **c**, wells with amphotericin B (+ amphotericin B) are indicated in orange and wells without amphotericin B (− amphotericin B) are indicated in green. Mean OD600 readings with standard deviations are shown, significant differences from the compound without antifungal

agent controls (e.g., CB6, − caspofungin), as determined by Welch's *t*-test (two-tailed, assuming unequal variance) with the Bonferroni Correction, are indicated for α = 0.05 (\*). Significant differences from the antifungal agent without compound control (e.g., DMSO, + caspofungin), determined by the same statistical testing, are indicated for α = 0.05 (#). Candidate compounds were included at a concentration of 12.5 μM. Data from different plates are separated by two vertical lines on the x-axis, DMSO solvent controls are shown for each plate. Figure S3, Additional results from the sustained inhibition optical density biofilm assay combination screening of candidate compounds with the antifungal agents caspofungin, fluconazole, and amphotericin B. Combination sustained inhibition assays with (**a**) caspofungin, (**b**) fluconazole, and (**c**) amphotericin B. In panel **a**, wells with caspofungin (+ caspofungin) are indicated in yellow and wells without caspofungin (− caspofungin) are indicated in red. In panel **b**, wells with fluconazole (+ fluconazole) are indicated in grey and wells without fluconazole (− fluconazole) are indicated in blue. In panel **c,** wells with amphotericin B (+ amphotericin B) are indicated in orange and wells without amphotericin B (− amphotericin B) are indicated in green. Mean OD600 readings with standard deviations are shown, significant differences from the compound without antifungal agent controls (e.g., CB6, − caspofungin), as determined by Welch's *t*-test (two-tailed, assuming unequal variance) with the Bonferroni Correction, are indicated for α=0.05 (\*). Significant differences from the antifungal agent without compound control (e.g. DMSO, + caspofungin), determined by the same statistical tests, are indicated for α=0.05 (#). Candidate compounds were included at a concentration of 12.5 μM. Data from different plates are separated by two vertical lines on the x-axis, DMSO solvent controls are shown for each plate.

**Author Contributions:** Conceptualization, A.D.J. and C.J.N.; data curation, M.B.L. and C.L.E.; formal analysis, M.B.L. and C.L.E.; funding acquisition, A.D.J. and C.J.N.; investigation, M.B.L., C.L.E. and N.H.; methodology, M.B.L., C.L.E. and N.H.; project administration, C.J.N.; resources, A.D.J. and C.J.N.; supervision, C.J.N.; writing—original draft, M.B.L. and C.J.N.; writing—reviewing and editing, M.B.L, C.L.E, N.H., A.D.J., and C.J.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by National Institutes of Health (NIH) grants R43AI131710 (to M.B.L.), R01AI083311 (to A.D.J.), R35GM124594 (to C.J.N.), and R41AI112038 (to C.J.N.). C.L.E. was supported by NIH fellowship F31DE028488. This work was also supported by the Kamangar family in the form of an endowed chair (to C.J.N.). The content is the sole responsibility of the authors and does not represent the views of the NIH. The NIH had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

**Acknowledgments:** We thank the staff at UCSF's Small Molecule Discovery Center, especially Kenny Ang, for assistance with the high-throughput screens.

**Conflicts of Interest:** The authors declare the following competing interests. Clarissa J. Nobile and Alexander D. Johnson are cofounders of BioSynesis, Inc., a company developing inhibitors and diagnostics of *C. albicans* biofilms. Matthew Lohse was formerly an employee and currently is a consultant for BioSynesis, Inc.

#### **References**


*Article*
