Over 90% of the
Staphylococcus aureus strains isolated from human infections are pigmented (
18). Staphyloxanthin (STX) is an orange-red triterpenoid, membrane-bound carotenoid which plays a role in the environmental fitness of
S. aureus (
4,
25). Membrane pigments have also been hypothesized to be virulence factors in
S. aureus, potentially by detoxifying reactive oxygen species produced by phagocytes (
19). Carotenoids may also stabilize the
S. aureus membrane during infection and pathogenesis. For example, Rohmer et al. (
26) postulated that polar carotenoids regulate membrane properties of prokaryotes in a manner similar to that observed for cholesterol in eukaryotes. In this regard, Liu et al. (
19,
20) recently observed that STX synthesis in staphylococcal cells is associated with resistance to phagocyte-mediated killing
in vitro and staphylococcal persistence in target organs in relevant
in vivo animal models.
DISCUSSION
Prior studies have indicated that carotenoids may afford
S. aureus protection against oxidative host defense mechanisms (
4,
19,
20,
25). Such antioxidant potential of carotenoids may be related to their specific orientation, localization, and organization within the membrane (
29). Nonpolar and polar carotenoids exert distinct effects on lipid membrane structure and physiology (
29). On the basis of several recent seminal studies (
19,
20), it appears that STX production plays a role in the ability of
S. aureus to resist clearance by the oxidative limb of the host immune system (i.e., the NADPH oxidase or “respiratory burst” pathway).
The STX pigment consists of a C
30-polyene backbone with alternating single and double bonds, capable of quenching oxidation by reactive oxygen species (
6,
21). Deletion of the gene encoding the upstream STX biosynthesis enzyme,
crtM, renders the bacterium more susceptible to killing by human and mouse polymorphonuclear leukocytes (PMNs) or whole blood (
4,
20). In addition, knockout mice with homozygotic deletions of the gp91
phox gene (encoding a component of the NADPH oxidase enzyme complex involved in oxidative immune defense) exhibited
in vitro defects in killing carotenoid-producing
S. aureus strains and decreased the ability to clear such organisms from systemic infections (
20). In a complementary manner, compared with its wild-type parental strain, a
S. aureus crtM mutant strain was much more susceptible to killing by hydrogen peroxide, superoxide radicals, hydroxyl radicals, hypochloride, and singlet oxygen species (
4,
20). Finally, loss of STX pigmentation translated to a significant reduction in
S. aureus virulence in murine skin abscess as well as systemic infection models. Despite these compelling investigations corroborating the protective role of carotenoid pigments in
S. aureus resistance to host oxidative defenses, the potential function of such pigments in nonoxidative innate host defense pathways is less well studied.
As carotenoids may significantly impact the CM biophysical properties of
S. aureus, and most mammalian host defense peptides initially target the microbial electronegative CM, examining the correlations among pigment production, CM biophysics, and peptide susceptibilities is important. In this regard, Liu et al. (
20) showed that carotenoid-deficient and wild-type
S. aureus strains had equivalent susceptibilities to murine CRAMP, an 18-amino-acid cathelicidin host defense peptide (CAP) that carries a charge of +5. Of interest, Katzif et al. have shown that pigment production in
S. aureus is dependent on the function of at least two stress response gene-dependent mechanisms (sigma factor B and cold shock protein A [
sigB-
cspA]) (
15). Moreover, they observed that
cspA is involved in the susceptibility of
S. aureus to at least one human CAP (cathepsin G) (
14). To investigate and clarify CAP-carotenoid interactions further, the current studies compared a panel of host defense peptides of different charges, structures, and sources regarding carotenoid influence on antistaphylococcal efficacy. A well-characterized isogenic MSSA strain set was studied, including a clinically derived parental strain (with an intact
crtOPQMN operon), its
crtM mutant, and a
crtMN-complemented and carotenoid-overproducing variant (
20).
Several interesting observations emerged from our studies. Fluidity characteristics of the CM are essential for bacterial viability. For example, CM-bound carotenoids stabilize intra- and extracellular leaflets of the lipid bilayer and increase CM rigidity by ordering the alkyl chains of CM lipids (
26,
28,
29,
31). In the present study, the amount of pigment production was directly correlated with the relative fluidity-rigidity CM profiles in our strain set. Thus, the CMs of the carotenoid-overproducing strain exhibited substantially more-rigid CMs than the parental and carotenoid-deficient strains. Of interest, we and others have shown a prominent correlation between the relative state of CM order in
S. aureus and the organism's profiles of susceptibility to selected host defense CAPs. Therefore, irrespective of the pathway or mechanisms involved, extremes of CM fluidity-rigidity can be associated with an altered susceptibility to certain CAPs (
2). For example, the presence of a CM-spanning quaternary ammonium transporter (
qacA) both alters CM fluidity and reduces killing by tPMPs in an efflux pump-independent manner (
2). Similarly, clinical
S. aureus strains which acquired DAP resistance
in vivo have been documented to coevolve both enhanced CM fluidity and reduced killing by mammalian host defense CAPs (
12,
34). In contrast,
S. aureus strains emerging as DAP resistant by serial
in vitro DAP passage became concomitantly more resistant to killing by certain host defense CAPs, while exhibiting more-rigid CMs (
23). Collectively, such findings have led to the hypothesis that CM fluidity differentially influences distinct binding of peptide to or subsequent perturbations of the target CM. A corollary to this paradigm is that extremes of CM order (i.e., very fluid or very rigid CMs) can be associated with reduced killing by selected CAPs, and there likely exists a CM order optimum for each bacterial CM-peptide interaction (the so-called “sweet spot”) (
1,
23,
28). This notion is underscored in our investigation by the finding that tipping CM order toward increased fluidity (e.g., in the
crtM knockout) did not impact CAP susceptibility profiles but that tipping CM order toward enhanced rigidity (i.e., in the carotenoid overproducer) exerted a major impact in this context. The current findings are also consistent with our recent nuclear magnetic resonance (NMR)-defined interaction studies between RP-1 and model prokaryotic lipid membrane systems (
3).
In the present study, the carotenoid-overproducing strain was less susceptible to
in vitro killing by many of the study peptides than the parental and/or
crtM knockout strain. It was important to confirm that the differences we observed in these susceptibility profiles were specifically related to extremes in carotenoid content and not an epiphenomena to the influences of other factors impacting CM or surface envelope characteristics (e.g., CM lipid composition, net surface charge, etc.). We were particularly interested in CM fatty acid profiles, since carotenogenesis modulation can lead to a buildup of C
30 precursor species that can then be shuttled into menaquinone-fatty acid oxidation pathways (
13). However, fatty acid profiling showed no differences in our strain set in any parameter known to impact CM order, i.e., chain lengths, unsaturation indices, or branched-chain (iso and anteiso species) and straight-chain profiles (
12).
The above-mentioned findings underscore the hypothesis that a net carotenoid homeostasis influences S. aureus survival in the face of nonoxidative host defenses. Thus, although the ΔcrtM knockout strain produced negligible amounts of STX, this strain did not substantially differ in CAP profiling from the parental strain. In contrast, excess STX production was sufficient to substantially impact such susceptibility profiles (albeit only statistically significant for the PMN CAP, hNP-1). A parallel theme emphasizes the idea that the ability of STX overexpression (and its attendant enhancement of CM rigidity) to impair CAP-mediated killing of S. aureus likely represents a relatively “CAP-nonspecific” mechanism. For example, these trends for STX overexpression to be associated with reduced CAP susceptibility were observed for an entire range of peptide secondary structures and cationicity, including RP-1 (α-helix), hNP-1 (β-sheet), PMB (cyclic peptide), and DAP (cyclic lipopeptide).
We recognize that our study has certain limitations: (i) only a single S. aureus strain set was evaluated; (ii) the parental strain was clinically derived, raising the possibility that host exposures and/or genetic pathways outside crtOPQMN might have influenced these results; (iii) a relatively narrow range of host defense CAPs was investigated, leaving open the potential that CAPs of different sources, structures, or mechanisms of action might exhibit distinct effects on their activity by carotenoid-related CM perturbations; and (iv) host defense CAPs were tested individually, in austere buffer systems, and at sublethal concentrations in vitro. These facts limit the potential for in vivo translatability of our findings. Nonetheless, the results from these investigations support the hypothesis that S. aureus employs adaptive mechanisms by which it may utilize carotenoids to modify their CMs in order to subvert innate host defenses. Current studies are ongoing to further explore these concepts.