In this study, we tested the in vitro susceptibility of Staphylococcus aureus to an elapid snake-derived cathelicidin, NA-CATH, as well as related novel, synthetic peptides and compared the performance of these peptides to that of the human cathelicidin LL-37. We demonstrated that LL-37 has similar potency in vitro against S. aureus to NA-CATH, as opposed to our earlier findings for E. coli and other gram-negative bacteria where we determined NA-CATH to be more potent than LL-37 [25, 26]. The EC50 values were converted from μg/ml to μM to reflect the number of molecules of peptide and to accommodate the different molecular weights of the peptides. Therefore, on a molar basis, LL-37 was slightly (2.4-fold) more effective against S. aureus than the NA-CATH, but the difference was not statistically significant. The EC50 for the D-enantiomer, D-LL-37, was found to be ~10 fold higher than for LL-37, suggesting that it is less effective as an antimicrobial peptide under these conditions for S. aureus.
Three 11-residue peptides based on the ATRA motifs of the NA-CATH sequence (ATRA-1, ATRA-2, and ATRA-1A) were compared. The three ATRA peptides all had a nominal charge of +8 at pH 7, and their sequences differed only by the residues at the 3rd (F/A) and 10th position (L/P). On a molar basis, ATRA-1 is significantly more potent against S. aureus than ATRA-2, by ~10-fold. We demonstrated in prior work that the presence of alanine and proline at the 3rd and 10th positions, respectively, in the sequence of ATRA-2 (KRAKKFFKKPK) resulted in a significant alteration to the predicted hydrophobic face of the peptide and disrupted peptide helicity . The peptide ATRA-1A (KRAKKFFKKLK) was synthesized as a variation on the ATRA-1 peptide sequence (KRFKKFFKKLK) in order to determine the degree to which the Ala->Phe substitution at the 3rd position contributed to the reduced potency ATRA-2 exhibited against S. aureus. ATRA-1A is ~25 times more effective against S. aureus than is ATRA-2. However, comparing ATRA-1A to ATRA-1, the alanine substitution did not statistically change its activity against the gram-positive S. aureus (1.4 fold, p > 0.05), in contrast to the significantly improved activity against gram-negative bacteria . The side chain of alanine is smaller than phenylalanine, which could affect the peptide's hydrophobic face. The proline residue tends to make the peptide structure destabilized and disrupts the helical structure of peptides. This may impact the ability of the ATRA-2 to achieve a stable and well-defined helical conformation when interacting with bacterial membranes. We conclude that the substitution of alanine in ATRA-1A does not significantly contribute to the antimicrobial activity of the ATRA motif against S. aureus. Thus, the presence of the proline residue is likely to be the major contributor to the decreased anti-microbial activity of ATRA-2 peptide , and potentially also contributing to the overall anti-microbial activity of NA-CATH.
In earlier work, we demonstrated that ATRA-1 exhibited significant helical character in 60 mM SDS, while ATRA-2 showed no substantial helical character under these conditions. This behavior parallels their anti-microbial potencies. In this study, we found that NA-CATH:ATRA1-ATRA1 had significantly greater helical character in both 50% TFE and 60 mM SDS than did wild-type NA-CATH. In fact, the CD spectrum for NA-CATH:ATRA1-ATRA1 in 60 mM SDS suggests that the peptide has greater helical character under these conditions than the parental NA-CATH does in 50% TFE, a strongly helix-promoting environment. The anionic SDS is frequently used as a model system in studying the interaction between CAMPs and bacterial membranes [36, 37]. Accordingly, the increased helical nature/propensity of NA-CATH:ATRA1-ATRA1 could be a significant factor in its ~6 times (p < 0.05) greater anti-microbial potency against S. aureus than the parental NA-CATH. Accordingly, the increased helical nature/propensity of NA-CATH:ATRA1-ATRA1 could be a significant factor in its ~6 fold (p < 0.05) greater anti-microbial potency against S. aureus relative to the parental NA-CATH.
The range of effective concentrations displayed by these novel AMPs against S. aureus varied from 0.51 to 2.85 μg/ml (excluding peptides that proved ineffective). At these concentrations, these peptides have shown no significant hemolytic activity against erythrocytes, implying that the peptides can specifically target the cell membranes of bacteria at their effective EC50 without lysing mammalian cells.
In addition to assessing their anti-microbial activities, the capabilities of the peptides to inhibit S. aureus biofilm formation were tested. Biofilm formation by S. aureus is clinically relevant because biofilm formation allows pathogens to adhere to and accumulate on scabs or in-dwelling medical devices, such as catheters. Furthermore, in addressing wound infections, biofilm-embedded bacteria are often more difficult to combat than bacteria in planktonic form. This difficulty applies to both antibiotic regimes and the host immune response [38, 39]. Thus, it would be beneficial to prevent biofilm production as part of wound treatment. NA-CATH:ATRA1-ATRA1 proved effective at inhibiting biofilm formation at concentrations much lower than is required to reduce bacterial growth under high salt conditions. These findings are important, as there are few reports of AMPs or other antimicrobials exerting anti-biofilm activity against S. aureus at sub-anti-microbial concentrations. This suggests that these peptides may act internally on the bacteria, affecting the expression of genes that are essential for the development of biofilm [15, 32]. For example, in S. aureus, production of PNAG polysaccharide, which is a major component of the biofilm matrix, is regulated by genes of the agr locus  (in response to an autoinducer peptide, AIP) and the ica locus . In addition, a critical role for Bap (biofilm-associated protein) has been demonstrated for biofilm formation by this bacterium, with Bap and genomic DNA (or eDNA) contributing to the strength of the biofilm. In Pseudomonas aeruginosa, the human cathelicidin LL-37 alters the expression of biofilm related genes such as Type IV pili, Rhamnolipid and Las quorum sensing system at sub-antimicrobial levels . Staphylococcus aureus lacks these genes, and the molecular and genetic targets of LL-37 against S. aureus remain undefined.
By performing biofilm attachment experiments against S. aureus, we were able to determine that NA-CATH:ATRA1-ATRA1 and its parent peptide, NA-CATH, inhibit biofilm but not by inhibiting attachment. D- and L-LL-37 peptides are capable of inhibiting initial biofilm attachment (58-62%), suggesting a potential interaction of these peptides with bacterial adhesins may be part of their mechanism.
We have not yet determined the bacterial target of NA-CATH:ATRA1-ATRA1 or the D- and L-LL-37 peptides in S. aureus, but we intend to investigate this further in future work. One mechanism could be by directly promoting biofilm dispersal (as has been observed for some cationic detergents such as cetylpyridinium chloride ) or by inhibiting attachment. It is unlikely that the mechanism involves killing the bacteria, since we have observed that bacterial growth under high-salt conditions is not affected by these peptides. Moreover, anti-biofilm activity was observed for peptides associated with poor anti-microbial effect such as D-LL-37. In addition, we have shown specificity of response, as the scrambled LL-37 peptide does not inhibit biofilm production nor attachment in S. aureus.
The in vivo relevance of the host cathelicidin response to S. aureus infection is not fully established. It has been demonstrated that exposing keratinocytes to live S. aureus induces production of beta-defensin peptides, hBD1 and 3, but does not induce expression of hBD2 or LL-37. In addition, intracellular S. aureus did not induce LL-37 expression. However, heat-killed S. aureus or lipotechoic acid (LTA), a component of S. aureus cell wall, were able to induce LL-37 expression in keratinocytes . These studies indicate that the presence of this bacterium in or on the human host may induce the expression of LL-37 in vivo under the appropriate circumstances. Finally, in addition to direct effects on the bacteria, these peptides can also exert direct effects on host cells (although they do not appear to lyse host cells at these concentrations). LL-37 may have wound-healing properties . The host targets of LL-37 in human cells were found to include GAPDH , EGFR [45, 46] and the P2X7 receptor . D-LL-37 has been reported to exhibit powerful immuno-stimulatory activity on the host (more effectively than the L-peptide), such as the induction of IL-8 in keratinocytes and promoting fibroblast proliferation , which suggests that it could promote wound healing as an added effect. The bacterial and host-cell targets of these peptides will be the focus of our continued studies.