**1. Introduction**

Cyanobacteria are among the first living beings to exist on Earth. The oldest fossil cyanobacteria registries date back 3.8 billion years ago. Their presence was crucial to the creation of an aerobic atmosphere, resulting in the emergence of an enormous species variety [1]. They are defined as prokaryotic oxygen photosynthetic microorganisms and are mainly known for their ability to synthesize structurally diverse and biologically active natural products [2]. In addition, similar to other bacteria, these microorganisms are nucleus-free and have an immense morphological diversity. The various structural shapes encountered in these species are the result of their ability to alter their morphology according to the environment allowing for higher energy accumulation and growth [3,4].

These microorganisms are at mercy of various stress situations found in diverse types of environments, including water-based and land-based. The ability to thrive in these heterogenous environments can be attributed to an enormous secondary metabolite repertory, which has intrigued numerous scientists for its rarity and richness [5,6]. Peptides generated by ribosomal synthesis and produced by large multi-domain enzymes called nonribosomal peptide synthetases (NRPS) are among these metabolites [7,8]. The macrolides present in these photosynthetic species derive from an enzyme complex called polyketide synthase, which is also modular in nature, similar to animal fatty acid synthase. Some molecules are synthesized from the combination of these two metabolic pathways, such as toxin nodularin and microcystin. Products from these two pathways constitute the majority of the secondary metabolites described in cyanobacteria [9].

The ribosomal peptide pathway forms a group very diverse and complex of products, and it is present in all three domains of life. The building blocks used by this pathway

are usually limited to 20 proteinogenic amino acids. The enormous structural diversity of these proteinaceous substances can be enriched by post-translational modifications, which are also responsible for the functional diversity contained in this category. Such modifications occur in the side chains and can lead to different forms of macrocyclization [10,11]. The precursor peptide is mainly formed by a leader peptide (LP) and core peptides (CP), which act as recognition and modification sites, respectively. This identification assists post-translational enzymes to focus a biosynthetic effort on a particular precursor peptide. The different types of post-translational modifications (PTMs) are used to differentiate the subfamilies of this group and can enhance the stability of the peptide and its activities [12,13].

Microviridins are among the most promising peptides found in cyanobacteria. These molecules are potent inhibitors of protease found in an enormous variety of cyanobacteria, mainly those of the genus *Microcystis*, *Planktothrix*, *Anabaena*, *Nostoc* and *Nodularia* [14]. An in silico analysis revealed that the occurrence of microviridins in bacteria belonged to other phyla [15,16]. Here, we present a review of the microviridins produced by cyanobacteria and their biotechnological and ecological relevance.

## **2. Microviridin**

Microviridins are one of the most known and largest oligopeptides formed by cyanobacteria. They are ribosomally produced, classified as depsipeptides. Their size can vary from 12 to 20 amino acids, where the N-terminal residue is typically acetylated [17–19]. By posttranslational modifications, the side chains of some of these amino acids lead to ω-ester and an ω-amide linkage, which result in distinct ring formations. When completely cyclic, microviridins typically exhibit two ester bonds between the Thr-Asp/Glu and Ser-Asp/Glu side chains and an amide bond formed between the Lys side chain at position 9 and Glu or Asp at position 2. The formation of amide and ester bonds are catalyzed by ATP-grasp enzymes. Mono- and bicyclical structures may also be formed, possibly due to the lack of one of the PTM enzymes or further modification of the tricyclic microviridin [14,15,20]. These oligopeptides are capable of inhibiting the hydrolytic activity of several serine protease, including elastase, trypsin, thrombin and chymotrypsin, as well as tyrosinase. Hence, they have cogitated as promising agent in the treatment of several metabolic disorders [21,22]. Their selectivity can be related to their amino acid sequence, especially that occupying the fifth position from the C-terminal. All known microviridins normally share the TxKxPSD motif and possess Asp, Thr, Ser and Lys residues (Figure 1) [20].

Microviridins have been identified in different cyanobacterial genera, mostly isolated from freshwater. The screening of environmental samples and isolated strains showed a wide distribution and diversity of this oligopeptide [14]. The majority of reports focused mainly on the strains of *Microcystis* and *Planktothrix*, as these genera are bloom-forming and are usually found in the eutrophic ambient. Over the last few years, more microviridin variants have been discovered in phyla other than cyanobacteria [15,16].


**Figure 1.** Diversity of microviridin sequences and the conserved KYPSD motif. Multiple alignment was obtained by Clustal Omega (https://www.ebi.ac.uk) and visualized using JalView software (https://www.jalview.org), and the consensus sequence was generated by WebLogo (https://weblogo.berkeley.edu).

#### **3. Microviridin Structure**

Microviridin was firstly described in the toxic *Microcystis viridis* (NIES-102), which was isolated from a bloom on Kasumigaura Lake, by Ishitsuka et al. (1990) [21]. Its amino acid sequence was defined as Ac-Tyr (I)-Gly (I)-Gly (I)-Thr-Phe-Lys-Tyr (II)-Pro-Ser-Asp-Trp-Glu (I)-Glu (II)-Tyr-OH, where Lys is bound to Glu (II) through its ε-NH with γ-CO of Glu (II). Thr and Ser amino acids are esterified and form ester bonds with the γ and δ carboxylic moieties of Asp and Glu (I), respectively (Figure 2). After the discovery of microviridin A, Okino et al. (1995) [23] identified a further two novel microviridins in the freshwater cyanobacterium *M. aeruginosa* (NIES-298). They were named microviridin B and C, the former exhibiting high similarity to microviridin A. They differ solely by three amino acid residues: Phe, Thr and Leu, which occupy the same position of Tyr (I), Gly (I) and Phe in microviridin A. The microviridin B amino acid composition was defined as Ac-Phe-Gly-Thr-(I)-Thr (II)-Leu-Lys-Tyr-Pro-Ser-Asp-Trp-Glu-(I)-Glu (II)-Tyr-OH. Microviridin C is closely related to microviridin B, exhibiting the same amino acid composition but containing a methoxy group in the γ carboxylic acid of Glu (I) and one additional hydroxyl group correlated to Ser. In this oligopeptide, neither Ser nor Glu are esterified. The slight difference between anti-elastase activity exhibited by both inhibitors was important to demonstrate that the ester bond between Ser and Glu(I) is not included in the reactive site.

belonginggroup

**Figure2.** *Cont*.

One year later, Shin et al. (1996) [24] revealed the presence of three novel microviridins in *Planktothrix agardhii* (NIES-2014), known as microviridins D, E and F. Microviridin D is a bicyclic peptide, the N-terminal of which is occupied by an acetylated Tyr. Similar to microviridin A, this metabolite also possesses a ester bond formed between the side chains of the Thr and Asp residues. Differing from the former, microviridin D has Asn and Met residues instead of Gly and Phe, respectively. Furthermore, the ester bond between the γ-carboxyl of the Glu and the Ser hydroxyl group is missing in microviridin D, since γ-carboxyl of the Glu existed as a methyl ester. Microviridin E was the first microviridin composed of 13 amino acids described. In microviridin E, three Phe residues replaced two Tyr and one Trp residues of microviridin D. Unlike the other microviridins mentioned above, which have Glu occupying the second position from the C-terminal, this oligopeptide presents the residue of Asp in this position. Microviridin F seems to be a hydrolyzed microviridin E product with the same amino acid sequence. The absence of an ester bond between Thr and Asp is the main difference compared to other microviridins mentioned above. *Nostoc minutum* (NIES-26) was uncovered in 1997 as a source of two novels microviridins (G and H). Microviridin G is structurally related to microviridins A and B, while microviridin H has its structure closely related to microviridin C. These newly identified peptides have the same amino acid compositions. However, microviridin H does not have an ester bond between the Ser and Glu amino acid residues [25].

Microviridin I was firstly identified in the nontoxic *P. agardhii* strains 2 and 18. This oligopeptide exhibits high similarity to microviridins A, B and G. They share the Lys-Tyr (2)-Pro (2)-Ser-Asp (1)-Trp-Glu amino acid sequence, as can be seen in Figure 1 [26]. Microviridin J was firstly described in *M. aeruginosa* strain UWOCC MRC, being composed of 13 amino acids organized in three rings and two linear side chains. Unlike the previous microviridins, this peptide has arginine residues between Thr and Lys, which confer a special arrangement with the hydrophobic regions formed between the side chain of this residue and other amino acid residues. This novel structure conferred by the Arg residue occupying the fifth position provides ring stabilization and may be associated with a strong inhibition of trypsin, which has been identified solely in this microviridin [27]. The N-acetyl group of microviridin J also contributes to a marginal increase in the inhibition of trypsin by hydrogen bond formation [28]. The greatest amount of this toxin was obtained by utilizing MeOH at a concentration between 40–80%. The lowest yield was achieved by utilizing absolute methanol [27].

Reshef and Carmeli (2006) [29] isolated, for the first time, three microviridins with the nonproteinogenic amino acid β-hydroxyaspartic acid (Has) bound to lysine through an amide bond. These oligopeptides received the names of microviridin SD1684, SD1634 and SD1652 and were isolated from the extract of *M. aeruginosa* (IL-215). All these microviridins exhibit the same amino acid compositions. However, they differ regarding the number of ester bonds. SD1684 has no ester bonds (solely the amide bond), while SD1634 possesses the two-ester bonds and SD1552 contains only one ester bond, Ser-Glu.

Vegman and Carmeli (2014) [30] isolated from the extract of a yellow-brown bloom material composed of *Microcystis* spp. (TAU IL-376) the microviridin LH1667, whose amino acid sequence was defined as Ac-Tyr (I)-Ser(I)-Thr-Leu-Lys-Tyr (II)-Pro-Ser (II)-Asp-Trp-Glu(I)-Glu (II)-Tyr (III), with a Lys side chain amine and Glu (II) side chain carboxylic acid connected via a lactam, Ser (II) side chain hydroxyl and Glu (I) side chain carboxylic acid connected via a lactone and a side chain of Thr forming a lactone ring with a side chain carboxylic acid of Asp [30].

The increased number of genome sequences belonging to cyanobacteria opened the doors to a deeper knowledge about microviridins, allowing the discovery and engineering of new variants. The structure of microviridin K was determined by Philmus et al (2008) [15] in *P. agardhii* CYA126/8. Its amino acid composition is similar to microviridin D. However, the residue of Glu12 is not methylated. This oligopeptide thus contains two rings of lactone. Microviridin L, detected in cyanobacterium *M. Aeruginosa* (NIES843), was one of the first cyanobacterial oligopeptides to be characterized with the assistance of genomic data. The

gene cluster of this metabolite was inserted into a fosmid and subsequently expressed in *Escherichia coli* [31].

Microviridins N3−N9 were identified in the model strain *N. punctiforme* PCC73102 via a genomic approach. These unusual microviridins contain between 15 and 20 amino acid residues and are not acetylated. The name was given to highlight the difference between the number of N-terminal amino acids, which can range from three to nine [19].

Two new microviridins have recently been discovered in strain *M. aeruginosa* EAWAG 127A: microviridin 1777 and microviridin O [32]. The former is the most potent chymotrypsin inhibitor of the microviridin class, while the latter was not detected in the extract, although the precursor peptide gene was contained in the genome (EZJ55 03525). An antiSMASH analysis allowed the identification of its gene cluster. This oligopeptide exhibits high similarly with microviridins A, B, G and J. They share the Lys-Tyr (2)-Pro (2)-Ser-Asp (1)-Trp-Glu amino acid sequence. Its peptide sequence is AC-Tyr-Asn-Val-Thr-Leu-Lys-Tyr-Pro-Ser-Asp-Trp-Glu-Glu-Phe.

Based on the number and structure of the ester bonds, microviridins can be classified into four classes. The amide bond is conserved in all of them. Group I consists of microviridins with two ester bonds. The second and third groups have only one ester bond between Thr1-Asp7 and Ser6-Glu9, respectively. In the fourth, microviridins are present with only the amide bond conserved (Figures 2–5).

**Figure 3.** Microviridin structures belonging to group II.

**Figure 4.** Microviridin structure belonging to group III.

**Figure 5.** Microviridin structure belonging to group IV.
