**4. Discussion**

A notable number of AMPs including some bacteriocins exhibit potent and broadspectrum antimicrobial activities. Some of them are cationic and perturb the permeability of the bacterial membrane bilayers. The presence of positive charges and hydrophobic residues constitutes a common trait of the AMPs which interact with the bacterial membrane, causing pores and membrane depolarization or altering the microbial metabolic pathways or inhibiting DNA/RNA/protein synthesis [40,48–50]. Bacteriocins can be bacteriostatic or bactericidal, induce a rapid killing effect and are thought to have a lower propensity to develop resistance than conventional antibiotics [10,20,49,51]. Overall, the modes of action of LAB-bacteriocins acting against Gram-positive bacteria have been well

studied [8,28], but those used by LAB-bacteriocins acting against Gram-negative bacteria are much less documented [25,26,29], and remain to be fully explored.

Recently, we have isolated the *L. paracasei* CNCM I-5369 strain, and shown its potential to produce five distinct class II bacteriocins and inhibit Gram-negative bacteria, including strains of *E. coli* resistant to colistin [29]. Here, we focused on lacticaseicin 30, one of these bacteriocins, which was then used as a model to understand this original activity. Of note, its antibacterial activity may result from the combined action between the peptide itself and an acidic pH (pH 5) [29,30]. According to the classification proposed by Alvarez-Sieiro et al. [22], lacticaseicin 30 appeared to have the characteristics of class IIc bacteriocins. Indeed, this bacteriocin is synthesized without a leader sequence and does not undergo post-translational modifications, suggesting it could be a novel leaderless bacteriocin. Pérez-Ramos et al. [21] reported that leaderless bacteriocins disrupt the cell membrane of target bacteria and most of them do not require any docking molecule for their antimicrobial activity.

As with a few AMPs already reported [52,53], lacticaseicin 30 contains a higher number of negatively charged residues (Asp and Glu) than of positively charged ones (Arg and Lys), with an isoelectric point (p*I*) of 6.05, and many hydrophobic residues. At a neutral pH (pH > p*I*), the peptide is anionic while at an acidic pH (pH < p*I*), it is cationic. As previously mentioned, hydrophobic and cationic residues are one of the main characteristics of AMPs, and the presence of cationic residues can mediate interactions with negatively charged bacterial lipids, while the hydrophobic residues could contribute to the membrane perturbation [54].

The lacticaseicin 30 amino acid sequence is structurally organized into five distinct helices (H1 to H5) (Figure 2), based on the AlphaFold2 predictions and the circular dichroism spectroscopy data obtained in the absence or presence of SDS micelles at pH 5 or 7 (Figure 3).

As indicated in Table 1, the activity against Gram-negative bacteria appeared to be exerted in a strain-dependent manner. On the other hand, we observed an increase of the MIC values with the shortened peptide carrying the C-terminal region of lacticaseicin 30 (C-ter-lacticaseicin 30), in comparison with the native peptide (Table 1). Accordingly, the antibacterial susceptibility has decreased 4-fold against *E. coli* ATCC 8739 and 2-fold against *Salmonella enterica* serovar Newport ATCC 6962 and *Proteus vulgaris* ATCC 33420. Remarkably, these shortened analogs displayed similar activity against *Pseudomonas aeruginosa* ATCC 27853. It is worthy of note that antibacterial assays performed on different *E. coli* strains carrying or not carrying modifications on their LPS were conducted with novel bacteriocins including thereof the class II lacticaseicin 30 [29], and the results obtained did not suggest LPS as the main target of these novel class II bacteriocins. Nonetheless, the activity against *Listeria innocua* CIP 80.11, used as the Gram-positive target strain, remained unchanged. These results delineate the dual role of the N-terminal region composed of helix 1 and helix 2, while the C-terminal region and more particularly helices 4 and 5 could modulate the antibacterial activity according to the target bacteria, suggesting another mechanism of action that remains to be determined. This study is in line with that of Van Kraaij et al. [55], who showed the role of the C-terminal region of nisin across the membrane. Similarly, Johnsen et al. [56] and Rihakova et al. [57] revealed the involvement of the C-terminal region of pediocin-like bacteriocins (class IIa bacteriocins) in determining their antibacterial spectrum.

To gain further insights on the antibacterial activity of lacticaseicin 30, another analog named N-ter-H1-lacticaseicin 30, consisting of a shortened N-terminal region carrying the 1 to 20 amino acids and including only the first predicted α-helix, was designed, expressed and produced. Interestingly, the antibacterial activity remained unchanged against the Gram-positive target strain, whereas it was completely abolished against the Gram-negative target bacteria (Table 1). This result argues that at least two helices located in the N-terminal region are required for activity against Gram-negative bacteria. Furthermore, substitutions of selected amino acids have been performed by site-directed mutagenesis with the aim

to understand their roles in the global antibacterial activity. Site-directed mutation refers to the redesign of natural antimicrobial peptides by adding, deleting or replacing one or several amino acid residues [58]. Therefore, two types of key mutations have been created. The first one consisted of replacing glutamic acid, aspartic acid or tyrosine by glycine or serine, and the second consisted of replacing threonine or alanine by proline. These mutations have been created inside each predicted α-helix. Then, the peptide variants E6G, T7P, D57G, A74P, Y78S, Y93S, A97P, E32G, T33P and T52P were tested for their activities against the target Gram-negative bacteria. Overall, they exhibited a loss in the antibacterial activity except for E32G, T33P and T52P, for which anti-Gram-negative activity remained unchanged (Table 2). Proline is a non-polar amino acid and proline-rich AMPs act differently from other AMPs. Indeed, some proline-rich AMPs have been shown to enter the bacterial cytoplasm through the inner membrane transporter SbmA instead of killing bacteria through membrane disruption [59]. Once in the cytoplasm, some proline-rich AMPs target ribosomes and block the binding of aminoacyl-tRNA to the peptidyltransferase center and interfere with protein synthesis [60], while others bind and inhibit DnaK [61]. In the case of lacticaseicin 30, which is devoid of proline residues, the substitution of T7, T52, A74 or A97 with prolines decreases the activity of the corresponding peptides, indicating that the mechanism implied in the anti-Gram-negative activity of lacticaseicin 30 is perturbed by the presence of proline. On the other hand, glycine is generally classified as a non-polar amino acid that induces flexibility of the peptide backbones [62,63].

#### **5. Conclusions**

To sum up, lacticaseicin 30 is predicted to adopt a secondary structure characterized by five helices. The generation of peptide variants carrying single point mutations, or truncated sequences, enabled some amino acids that play a major role in the structure of lacticaseicin 30 and its activity against Gram-negative bacteria, including strains of *E. coli* resistant to colistin. Moreover, this study permitted us to establish that at least two helices located in the N-terminal region are required for the anti-Gram-negative activity, while the C-terminal region would serve as a modulator of the activity, conferring its selectivity. These promising achievements open a new avenue in the characterization of LAB-bacteriocins endowed with activity against Gram-negative bacteria. Further experiments consisting of designing novel variants with enhanced antibacterial activity directed especially against resistant strains constitute our next goal.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/pharmaceutics14091921/s1, Table S1: Bacterial strains and plasmids used in this study; Table S2: Sequences of oligonucleotide primers used in this study; Figure S1: Structural alignment of the predicted structures of native lacticaseicin 30 (brown) and its truncated derivatives N-ter-lacticaseicin 30 (transparent red, RMSD between 37 pruned atom pairs 0.381 angstroms; across all 39 pairs: 1.535 angstroms), N-ter-H1- lacticaseicin 30 (transparent green, RMSD between 20 pruned atom pairs 0.797 angstroms; across all 20 pairs: 0.797 angstroms) and C-ter-lacticaseicin 30 (transparent blue, RMSD between 31 pruned atom pairs 0.888 angstroms; across all 73 pairs 8.364 angstroms). Substituted amino acids in the variant peptides are visible as balls and sticks in their position in the native peptide.

**Author Contributions:** All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by D.M.-M., B.D., R.T., Y.L. and M.M. The first draft of the manuscript was written by D.M.-M. and all authors commented on previous versions of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by la Région des Hauts-de-France, through ALIBIOTECH CPER/FEDER 2016/2021.

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** All data generated or analyzed during this study are included in this published article.

**Acknowledgments:** The authors are grateful to Yanath Belguesmia and Marc Maresca for the critical reading of the manuscript. The MALDI-TOF/MS experiments were performed on the REALCAT platform funded by the French government grant managed by the French National Research Agency (ANR) as part of the "Investments for the Future" program (ANR-11-EQPX-0037). We also acknowledge the Hauts-de-France region, the ERDF, the Ecole Centrale de Lille, and the Foundation Centrale Initiatives for financial support with the acquisition of the REALCAT platform equipment.

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