*2.1. Inhibition of Bacterial Growth by CST*

The group of Metz-Boutigue first demonstrated the antibacterial activity of CST. Her group used bCST344–358 (coining the term cateslytin to describe this antimicrobial effect) to reveal the inhibition of growth of the Gram-positive and Gram-negative bacteria [21]. The minimal inhibitory concentrations (MICs) of CST (bCgA344–358, hCgA352–372, Gly364Ser-CST and Pro370Leu-CST) for Gram-positive bacteria (*Micrococcus luteus*, *Bacillus megaterium*, Group A *Streptococcus*, *S. aureus* ATCC 25923, *S. aureus* ATCC 49775, *S. aureus* S1 MRSA, *S. aureus* S1 MSSA, and *S. aureus* DmprF) range from 0.8 μM to >100 μM (Figure 1) [21,24]. The minimal concentration with 100% inhibition (MIC100) for Gram-positive bacteria range from 2 μM to >100 μM. The MIC of CST was higher (8 μM to 50 μM) for Gram-negative bacteria (*Escherichia coli* D22, *E. coli* 029, and *Pseudomonas aeruginosa*) compared to Grampositive bacteria (Figure 1). Likewise, the MIC100 of CST was higher (15 μM to 150 μM) for Gram-negative bacteria compared to Gram-positive bacteria (Figure 1). The higher MIC and MIC100 values of CST for Gram-negative bacteria are consistent with the presence of extra outer membrane containing lipopolysaccharide (LPS) [25,26]. Beyond the extra-thick cell membrane, Gram-negative bacteria also release exotoxins such as tetanus [27] and cholera toxins [28] that worsen prognosis.


**Figure 1.** Effects of wild-type (WT)-CST and natural human variants of CST (Gly364Ser and Pro370Leu) on the growth of Gram-positive and Gram-negative bacteria showing minimal inhibitory concentration (MIC) and lethal concentration (MIC100) of CST.

D-bCST1–15 was reported to exert more effective antimicrobial effects against various bacterial strains than L-bCST1–15 [23]. In addition to its antimicrobial effects, D-bCST1–15 was reported to potentiate (additive/synergistic) the antibacterial effects of cefotaxime,

amoxicillin, and methicillin [23]. Furthermore, it has been shown that D-bCST1–15 neither triggered bacterial resistance nor elicited cytokine release [23]. In addition, D-bCST1–15 was reported to be more resistant to degradation by secreted bacterial protease than LbCST1–15 [23]. Thus, it was suggested that D-bCST1–15 can be used as a monotherapy or as a combination therapy with currently prescribed antibiotics to counteract various diseases associated with bacterial infection [23].

#### *2.2. Composition of Bacterial Membranes*

While antibiotics target specific cellular activities (e.g., synthesis of DNA, protein, or cell wall), AMPs target the LPS layer of the cell membrane. Extensive studies have been conducted to learn the composition of the bacterial membrane. The bacterial cytoplasmic membrane consists of zwitterionic phospholipids (phosphatidylcholine, phosphatidylethanolamine, sphingomyelin, etc.) and anionic phospholipids (phosphatidyl serine, phosphatidyl glycerol, etc.), providing them with a negative charge [29–32]. In contrast, besides the cytoplasmic membrane, Gram-negative bacteria contain an additional strong electronegative LPS-containing thick outer membrane [25,26]. Furthermore, the peptidoglycan layer on the outer side of the cytoplasmic membrane is much thicker in Grampositive bacteria compared to Gram-negative bacteria (20–80 nm versus ~10 nm) [33,34]. The peptidoglycan layer in Gram-positive bacteria is connected by electronegative wall lipoteichoic acids and anchored on the phospholipid bilayer by electronegative lipoteichoic acids [35]. In contrast, in Gram-negative bacteria, the LPS forms the major lipid component of the outer leaflet of the outer membrane [35].

#### *2.3. Secondary Structure of CST Explains the Antibacterial Effects of CST*

Based on their secondary structure, AMPs are generally categorized into four groups: (i) α-helical AMPs, (ii) β-sheet AMPs, (iii) extended AMPs, and cationic loop AMPs [36]. Homology modeling followed by molecular dynamics simulation of bovine CST (bCgA342–370) performed in a water shell led to a β-strand-loop-β-strand structure. Molecular dynamics and computer simulations of human CST1–21 revealed the following: R10, A11, and Y12 contribute to a 310 helix [37]. In contrast, F7, R8, A9, F14, R15, G16, P17, and G18 contribute to the antiparallel β-sheet [37]. The mechanism of the antibacterial action of CST1–21 could start by interacting with negatively charged moieties such as LPS in the outer membranes of Gram-negative bacteria and lipoteichoic acid in the wall of Gram-positive bacteria. The primary structure of CST reveals that CST contains cationic and hydrophobic residues and adopt a β-sheet secondary structure via intermolecular forces [38]. This folding structure would facilitate CST to fold into an amphiphilic conformation with positively charged (polar) and hydrophobic (nonpolar) faces (Figure 2). The presence of a great number of positively charged residues (5 in bCST and 4 in hCST) will allow CST to interact preferentially with negatively charged bacterial membranes [1,39]. Since the hydrophilic and hydrophobic amino acids of CST are structurally segregated, it will provide solubility of CST in both aqueous and lipid-rich environments, as suggested for other AMPs [40]. In addition, positively charged amino acids in CST formed amphipathic structures, as evidenced by separated hydrophobic and hydrophilic surface domains [39,41] (Figure 2). When the concentration of CST would exceed a certain critical concentration, the cell membrane would form pores, leading to content leakage, cell lysis, and finally death. Since cyclization of peptide has been reported to induce high antimicrobial activity [39,41], it is reasonable to assume that cyclization of CST would markedly improve the antibacterial activity of CST.

**Figure 2.** Hydropathy profiles of (**A**) Consensus CST, (**B**) Human CST, and (**C**) Bovine CST. The values are plotted based on the parameters used from Kyte and Doolittle, 1982. The values above zero represents the hydrophobic property of the amino acids that might contribute to the hydrophobic core of the peptide. The values below zero represent the hydrophilic property of the amino acids, which are instrumental in interaction with other protein factors.

Metz-Boutigue's group has shown that bCgA344–358 is unstructured in solution but is converted to an antiparallel β-structure and forms aggregates at the surface of negatively charged bacterial membranes [42]. As for catecholamine secretion [15], arginine residues were found to be crucial for binding to negatively charged lipids [42,43]. They proposed that the phase boundary defects caused by zones of different rigidity and thickness lead to permeability induction and peptide crossing through the bacterial membrane [42]. The fact that CST penetrates through the bacterial wall was shown by measuring the optical density of the released β-galactosidase from ML-35p [24]. Electron microscopical studies of *E. coli* ML-35p confirmed that CST rapidly disrupts the *E. coli* membrane, with visible membrane blebbing compared to untreated cells within 10 min [24].
