This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vylkova, S.
Right arrow Articles by Edgerton, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vylkova, S.
Right arrow Articles by Edgerton, M.

 Previous Article  |  Next Article 

Antimicrobial Agents and Chemotherapy, January 2006, p. 324-331, Vol. 50, No. 1
0066-4804/06/$08.00+0     doi:10.1128/AAC.50.1.324-331.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Distinct Antifungal Mechanisms: ß-Defensins Require Candida albicans Ssa1 Protein, while Trk1p Mediates Activity of Cysteine-Free Cationic Peptides

Slavena Vylkova,1 Xuewei S. Li,1 Jennifer C. Berner,1 and Mira Edgerton1,2*

Departments of Oral Biology,1 Restorative Dentistry, School of Dental Medicine, State University of New York at Buffalo, Buffalo, New York 142142

Received 15 August 2005/ Returned for modification 26 September 2005/ Accepted 8 October 2005


arrow
ABSTRACT
 
Salivary histatin 5 (Hst 5) kills the fungal pathogen Candida albicans via a multistep process which includes binding to Ssa1/2 proteins on the cell surface and requires the TRK1 potassium transporter. Hst 5-induced membrane permeability to propidium iodide (PI) was nearly abolished in strain CaTK1 (TRK1/trk1), suggesting that Hst 5-induced influx of PI is via Trk1p. To explore the functional role of Trk1p in the mechanism of other antifungal peptides, we evaluated candidacidal activity and PI uptake in wild-type strain CaTK2 (TRK1/TRK1) and strain CaTK1 following treatment with lactoferricin 11 (LFcn 11), bactenecin 16 (BN 16), and virion-associated protein VPR 12. Strain CaTK1 was resistant to killing with these peptides (VPR 12 > LFcn 11 > BN 16), showing the requirement of Trk1p for fungicidal activity. In contrast, human neutrophil defensin 1 (HNP-1), human ß-defensin 2 (hBD-2), and hBD-3 effects on viability of and membrane permeability to PI were not different between mutant and wild-type strains, clearly showing that their candidacidal mechanism does not involve Trk1p as a functional effector. To test whether defensins require binding to Candida surface Ssa1/2 proteins for their activity, we measured the killing effectiveness in SSA1/2 mutant strains. Both hBD-2 and hBD-3, but not HNP-1, exhibited reduced killing of ssa1{Delta} and ssa2{Delta} strains compared to the wild type, showing that Ssa1 and Ssa2 proteins are required for their fungicidal activity. These results demonstrate that (i) Trk1p mediates candidacidal activities of cysteine-free peptides, but not of defensins, and (ii) hBD-2 and hBD-3, but not HNP-1, require Ssa1/2p for antifungal activity.


arrow
INTRODUCTION
 
Human saliva contains antimicrobial proteins that provide a first line of defense against a wide spectrum of bacteria and fungi. In the oral cavity, {alpha}- and ß-defensins, lactoferricin, and histatins supply a significant source of antifungal activity against many Candida species and shield the oral mucosa from development of candidiasis. Among these antifungal proteins, perhaps the best studied in terms of mechanisms of action against Candida albicans are salivary histatins.

Histatin 5 (histatin31/24 or Hst 5) is a small (24 amino acid residues) basic protein derived from trypsin-like cleavage of the secreted parent salivary protein histatin 3 (3). Histatin 5 has the highest antifungal activity of the histatin family, having 50% lethal dose (LD50) values of 2 to 10 µM with most strains of C. albicans. The N terminus of histatin 5, comprised of only 14 amino acid residues, retains full candidacidal activity (21). Histatin 5 does not display a well-defined secondary structure in its native aqueous environment, and neither ordered secondary structure nor amphipathicity appears to be related to its killing activity (5, 10), suggesting that peptide interactions with yeast cell membranes do not play a role in the antimicrobial action of histatin 5.

Histatin 5 first binds to C. albicans cell wall proteins, which facilitate intracellular transport to other effector sites for its toxic activity. Initial binding by histatin 5 to specific cell wall proteins, identified as Ssa1p and Ssa2p (16), is obligatory prior to intracellular translocation. Hst 5 binding, translocation, and toxicity are closely related processes, and the loss of either the Ssa1 or Ssa2 protein results in diminished histatin 5 intracellular transport and cell killing (16). We previously found that human neutrophil defensin 1 (HNP-1) competes for C. albicans cell wall Hst 5 binding sites (7), now identified as Ssa1/2 proteins. With the availability of C. albicans SSA1 and SSA2 deletion mutants, we now have the ability to directly test the contribution of cell wall Ssa1/2p for fungicidal activity of other cationic antimicrobial peptides.

The hallmark of histatin 5 killing of C. albicans is the rapid efflux of cellular ATP and other small nucleotides and ions from the cell as well as concurrent intracellular uptake of propidium iodide (PI) (9, 14, 20, 27). These events do not reflect membrane pore formation since pretreatment of C. albicans cells with ion transport inhibitors such as 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (DIDS) prevents Hst 5 killing (1). Recently, we found that a deletion of theTRK1 gene encoding the major plasma membrane K+ uptake system in C. albicans nearly abolished Hst 5-initiated ATP release and cell killing (2). As Trk1p and its homologues are potassium uptake systems unique to prokaryotes, fungi, and plants, these membrane proteins have the potential to be specific effectors in these cell types. However, it is not known whether other cationic antifungal peptides also interact with Trk1p in the process of cell killing. Our C. albicans TRK1 mutants provide a means for testing such connections by assessment of the fungicidal activity of other cationic peptides in C. albicans upon the deletion of TRK1.

Lactoferrin is an abundant 77-kDa iron-binding glycoprotein found in saliva. The iron-binding domain of the lactoferrin molecule is located at the carboxy terminus, while the highly basic N-terminal region contains a 25-amino-acid microbicidal domain termed lactoferricin (LFcn). The N terminus of LFcn, consisting of only 11 amino acid residues (LFcn 11), retains complete bactericidal and antifungal properties (13, 17). LFcn 11 appears to have features in its antifungal activity that are similar to those of histatin 5, in that membrane disruption, extracellular release of ATP, and subsequent loss of viability occur in an energy-dependent manner following peptide treatment (4, 17). Although irreversible damage to the cell ultimately follows exposure to LFcn 11, the precise mechanism of its activity with C. albicans remains to be identified.

Mammalian defensins are a family of cationic peptides that are divided into {alpha}- and ß-defensin subfamilies on the basis of the connectivity of highly conserved cysteine residues. These proteins serve as one of the major arms of innate defense in the oral cavity, and {alpha}-defensins, or human neutrophil peptides (HNP-1 to HNP-4), and human ß-defensins exhibit high antifungal activity (11, 12). Although the spectrum of antimicrobial activity of these defensins has been characterized, very little is known about their mechanism of action with C. albicans and other pathogenic fungi. We have identified similarities in the overall effects of defensins with Hst 5 in the temporal release of intracellular ATP pools; however, it is not known whether this is a result of generalized membrane permeabilization or more specific effects with membrane proteins such as Trk1p. Thus, it is possible to examine the mechanism of toxic effects of a range of host defense peptides with C. albicans by employing strains with selective mutations in key proteins that are involved in the toxicity of histatin 5. In this manner, we can directly test whether binding and uptake proteins (Ssa1/2p) and/or transporters for potassium regulation (Trk1p) are implicated in the fungicidal activity of {alpha}- and ß-defensins and other cationic cysteine-free peptides in C. albicans. Here, we report that C. albicans Trk1p is required for the toxicity of other cationic peptides but that the fungicidal activity of defensins is independent of this protein. In addition, Ssa1/2 proteins mediate candidacidal activity of ß-defensins but not {alpha}-defensins, suggesting key differences in their modes of action.


arrow
MATERIALS AND METHODS
 
Synthetic peptides. Hst 5, LFcn 11 (corresponding to residues 12 to 21 of bovine lactoferrin), and the virus-based protein VPR (VPR 12) (representing amino acid residues 71 to 82 of the C terminus of virus protein R) were synthesized using standard solid-phase synthesis protocols and purified by reversed-phase high-performance liquid chromatography by Genemed Synthesis Inc., San Francisco, CA. Peptide purity was assessed by amino acid analysis and mass spectroscopy. The bactenecin 16 (BN 16) peptide from the N terminus of bovine bactenecin was synthesized as previously described (22). HNP-1 was purchased from Sigma. Human ß-defensin 2 (hBD-2) and hBD-3 were purchased from Peptides Int., Louisville, KY. Primary structures of these peptides are shown in Table 1.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Primary structures of antimicrobial peptides used in this study

Strains and media. To analyze whether the candidacidal activity of the peptides used in the study share similar features to Hst 5 cytotoxic action with respect to their binding to the Ssa1/2 protein on the C. albicans cell envelope, the following C. albicans strains were used: CAF4-2 as the wild-type strain, an ssa1{Delta} mutant strain with a deletion of both alleles of the SSA1 gene and one wild-type allele of the SSA2 gene (ssa1/ssa1 SSA2/ssa2), and an ssa2{Delta} mutant that has both alleles of the SSA2 gene and one wild-type allele of the SSA1 gene deleted (SSA1/ssa1 ssa2/ssa2). In all other experiments, C. albicans wild-type strain CaTK2 [CaTK2(wt)] (TRK1/TRK1), the hemizygous strain CaTK1 (trk1/TRK1) constructed by single-allele disruption of the TRK1 locus from the wild-type strain (2), and CaTK2 complementation of the hemideleted strain CaTK1 (trk1/TRK1/rp10::TRK1) obtained by transforming an additional copy of the wild-type gene into the Candida RP10 locus were used. Yeast nitrogen base (YNB) medium (Qbiogene), with or without uridine and supplemented with 2% glucose as a carbon source, was used to culture the strains. Solid media were prepared by the addition of 1.5% Difco agar.

Candidacidal assay. Antifungal activity of antifungal peptides was examined by microdilution plate assay (6) with the following modifications. Briefly, C. albicans cells were grown in YNB medium, washed twice with 10 mM sodium phosphate buffer (NaPB) (Na2HPO4/NaH2PO4, pH 7.5), and resuspended in sodium phosphate buffer at a concentration of 106 cells/ml. The cell suspensions were then mixed with Hst 5 (7.5 to 62 µM), LFcn 11 (1 to 15 µM), BN 16 (5 to 30 µM), VPR 12 (1 to 30 µM), HNP-1 (2.9 to 31.25 µM), hBD-2 (1-11.6 µM), or hBD-3 (0.1-10 µM) for 1 h at 37°C with shaking. Control cultures were incubated with 10 mM NaPB alone. Cell suspensions were diluted in 10 mM NaPB, and aliquots of 500 cells were spread onto YNB agar plates and incubated for 48 h at room temperature. Candidacidal assays were performed in triplicate. Cell survival was expressed as a percentage of the control, and the loss of viability was calculated as follows [1 – (colonies from peptide-treated cells/colonies from control cells)] x 100.

Flow cytometry. C. albicans cells (106) were mixed with Hst 5 (62 µM), HNP-1 (31.25 µM), hBD-3 (10 µM), LFcn 11 (15 µM), BN 16 (30 µM), or VPR 12 (30 µM) to reach a final volume of 100 µl and were incubated for 1 h at 37°C with shaking. The DNA-binding dye propidium iodide (final concentration, 50 mg/ml) was added to the cells and incubated for an additional 30 min. To remove excess PI, cells were harvested by centrifugation (5,500 x g, 2 min) and washed three times with 10 mM NaPB, and cell pellets were reconstituted in 1 ml buffer. Control cells were incubated with 10 mM phosphate buffer or PI only. Samples were analyzed on a FACScan flow cytometer (Becton Dickinson) by acquisition of 105 events using a 15-mW argon laser at 488-nm excitation for fluorescence detection of propidium iodide. Data were recorded as histograms of fluorescence (FL1) versus counted events.

Statistical analysis. A statistical comparison of candidacidal activities of cationic peptides on CaTK2(wt) and CaTK1 strains or on strain CAF4-2 and ssa1{Delta} or ssa2{Delta} strains was performed by using an unpaired two-tailed t test. A P value of ≤0.05 was considered to be statistically significant.


arrow
RESULTS
 
TRK1 is the channel for Hst 5-induced PI uptake and killing. To determine whether cells lacking one functional copy of the TRK1 gene exhibit elevated membrane permeability following Hst 5 treatment, measures for propidium iodide uptake were undertaken in C. albicans strains CaTK2(wt), CaTK1, and CaTK2. Strain CaTK2(wt) showed elevated PI uptake, as observed by the increased fluorescence intensity in about 77.5% of the cells (Fig. 1, black boldface line) following treatment with 62 µM Hst 5. This corresponded with the high susceptibility to Hst 5 killing in these cells, with about 83% of cells killed at this peptide concentration (Table 2). Cells alone exhibited little autofluorescence (Fig. 1, light gray line). In contrast to the wild type, strain CaTK1, possessing only one functional copy of the TRK1 gene treated with 62 µM Hst 5, exhibited significantly lowered PI uptake (Fig. 1B, thin black line), whereas only 22% of the cells remained permeable to propidium iodide (Table 2). These data closely parallel Hst 5 resistance of CaTK1 cells (Table 2). These effects were not due to lowered cellular uptake of Hst 5, since we found that CaTK2(wt) and CaTK1 had equivalent total cellular levels of Hst 5 following 60 min of incubation with fluorescein isothiocyanate-Hst 5, as assessed by FACScan analysis (data not shown). Complementation of the CaTK1 strain by insertion of a second wild-type TRK1 gene at the RP10 locus in strain CaTK2 fully restored the sensitivity to Hst 5 (Table 2) and resulted in an elevated PI uptake in about 73% of the cells (Fig. 1, dark gray line, and Table 2), thus verifying that the loss of susceptibility to Hst 5 observed in strain CaTK1 is not an artifact of genetic manipulation. In addition, wild-type cells pretreated with the anion channel inhibitor DIDS showed PI uptake similar to that of CaTK1 (data not shown), consistent with our previous reports on the inhibitory effect of DIDS on Trk1p function (2).



View larger version (38K):
[in this window]
[in a new window]
  		FIG. 1. Hst 5-induced membrane permeability to PI is inhibited in C. albicans TRK1 mutants. CaTK2(wt) (black boldface line), CaTK1 (thin black line), and CaTK2 (gray boldface line) cells were incubated for 30 min with 50 mg/ml PI (final concentration) following incubation with 62 µM Hst 5 for 1 h at 37°C. Cells were washed in 10 mM sodium phosphate buffer, and 105 cells were analyzed by FACScan. CaTK1 cells showed reduced PI uptake, which was restored upon complementation of the TRK1 gene (CaTK2).


View this table:
[in this window]
[in a new window]
  		TABLE 2. PI uptake corresponds to cell killing by Hst 5

Collectively, these results indicate that the loss of viability and uptake of the DNA-binding dye propidium iodide of Hst 5-treated C. albicans cells is specifically related to the gene status of the TRK1 potassium transporter and further suggest that Hst 5 could induce the influx of PI through the transporter itself. These data also show that PI uptake correlates quantitatively with Hst 5-induced cell death.

Other cationic antimicrobial peptides exert fungicidal activity mediated by Trk1p. Next, we evaluated whether other basic antifungal peptides without disulfide linkages exhibit similarities to Hst 5 candidacidal activity by using wild-type and TRK1 deletion strains. LFcn 11, BN 16, and the virus-based protein VPR (VPR 12) were selected, as they have high fungicidal activities (17, 18, 22). All tested peptides exhibited candidacidal activities in a dose-dependent manner with the wild-type strain (Fig. 2). Incubation of CaTK2(wt) with LFcn 11 resulted in nearly complete (~90%) killing at 15 µM (Fig. 2A, closed circles). However, killing by LFcn 11 was significantly lower (P < 0.01) with the CaTK1 strain than with the wild-type strain at all concentrations tested (Fig. 2A, open circles). Killing reached only ~50% of CaTK1 cells at 15 µM, and the LD50 was doubled from 7 µM in the wild type to 14.6 µM in CaTK1.



View larger version (11K):
[in this window]
[in a new window]
  		FIG. 2. Cysteine-free cationic peptides share similarities in their killing mechanisms with Hst 5. C. albicans strains CaTK2(wt) (closed circles) and CaTK1 (trk1/TRK1) (open circles) were incubated with (A) LFcn 11 (1 to 15 µM), (B) BN 16 (5 to 30 µM), or (C) VPR 12 (1 to 30 µM) in 10 mM sodium phosphate buffer (pH 7.4) for 1 h at 37°C, as described in Materials and Methods. The loss of cell viability is expressed as follows: [1 – (colonies after Hst 5 treatment/colonies after incubation with buffer only)] x 100. Each data point represents the mean ± standard deviation of at least three independent experiments. TRK1 mutants were protected from killing by each peptide tested.

Treatment of CaTK2(wt) and CaTK1 strains with BN 16 showed killing profiles with similarity to that of LFcn 11; however, nearly twice the concentration (30 µM) was required to induce cell death in ~90% of the wild-type cells (Fig. 2B, closed circles). At this peptide concentration, susceptibility to killing by BN 16 in the CaTK1 mutant strain was reduced to 52.8% (Fig. 2B, open circles). At reduced peptide concentrations (5 µM or 10 µM), the mutant strain also displayed significantly reduced (P = 0.03) BN 16 susceptibility compared with that of the wild-type strain. Overall, strain CaTK1 had about 1.5-fold-reduced susceptibility to BN 16 killing, with an LD50 concentration of 28.4 µM, compared with an LD50 of 18 µM in the wild type.

The fungicidal activity of VPR 12 on C. albicans wild-type and TRK1 deletion strains differed somewhat from that of LFcn 11 and BN 16. In contrast to the dose-dependent linear increase of killing with LFcn 11 and BN 16, treatment of strain CaTK2(wt) with VPR 12 reached a high level of killing (~80%) at only 10 µM, and increasing peptide doses up to 30 µM only increased this level of killing to 89% (Fig. 2C, closed circles). However, as for the other peptides tested, the killing of the CaTK1 mutant was reduced to 56.4% upon incubation with an LD90 (30 µM) of VPR 12 (Fig. 2C, open circles). The susceptibility of the TRK1 mutant strain was significantly reduced (P = 0.04) by about half compared with that of the wild-type strain at every concentration tested, except at the lowest tested concentration (1 µM). Overall, CaTK1 cells exhibited about sixfold-reduced susceptibility to VPR 12 killing, with an LD50 of 26.6 µM for the TRK1 mutant strain, compared with 4.6 µM for strain CaTK2(wt).

These results show that the reduced candidacidal activity of LFcn 11, VPR 12, and BN 16 on strain CaTK1, which has only one functional copy of the TRK1 gene, is dependent on TRK1 potassium transporter function. A comparison of the candidacidal activities of these small basic cysteine-free peptides examined here shows that VPR 12 killing has the highest dependence on Trk1p function, since strain CaTK1, having one functional copy of the TRK1 gene, was about six times less susceptible to VPR 12-induced killing compared to the 1.5- to 2-fold reduction in susceptibility with the other two peptides tested. However, Hst 5, having an eightfold reduction in killing of the TRK1 mutant strain at its LD90, is most sensitive to TRK1 function compared with LFcn 11, VPR 12, and BN 16.

In order to further delineate the importance of Trk1p in the candidacidal activity of LFcn 11, VPR 12, and BN 16, we examined PI uptake in strains CaTK2(wt) and CaTK1 following treatment with these cationic peptides. For this study, we selected LD90 concentrations of peptides (15 µM for LFcn 11 and 30 µM for BN 16 and VPR 12) at which CaTK1 killing was most clearly distinguishable from CaTK2(wt) cells. Incubation of the wild-type strain having two copies of the TRK1 gene with 15 µM LFcn 11 (Fig. 3A, green line), 30 µM BN 16 (Fig. 3A, magenta line), or 30 µM VPR 12 (Fig. 3A, blue line) induced a high level of PI uptake comparable to those observed with Hst 5 (Fig. 3A, solid yellow area). Interestingly, although VPR 12 is the most potent fungicidal peptide against the wild-type strain, it exhibited slightly reduced fluorescence intensity compared with the other peptides. CaTK1 cells treated with 15 µM LFcn 11 (Fig. 3B, green line), 30 µM BN 16 (Fig. 3B, magenta line), or 30 µM VPR 12 (Fig. 3B, blue line) showed a substantial number of cells (40 to 50%) without PI uptake, which correlated well with protection from killing (~50%) at these concentrations. Thus, these results indicate that while the cationic peptides exert potent antifungal activities with respect to strain CaTK2(wt), their candidacidal activities are diminished in the TRK1 deletion strain CaTK1, which suggests similarities in the killing mechanisms of these peptides.



View larger version (37K):
[in this window]
[in a new window]
  		FIG. 3. C. albicans TRK1 mutants treated with cysteine-free peptides, but not with defensins, exhibit reduced PI uptake. Strain CaTK2(wt) (A, C) or TRK1 hemizygous strain CaTK1 (B, D) was incubated for 1 h at 37°C with 31.25 µM Hst 5 (solid yellow area), 15 µM LFcn 11 (A, B) (green lines), 30 µM BN 16 (A, B) (magenta lines), 30 µM VPR 12 (A, B) (blue lines), 31.25 µM HNP-1 (C, D) (red lines), or 10 µM hBD-3 (C, D) (purple lines) followed by a 30-min incubation with 50 mg/ml PI. Cells were washed three times in 10 mM sodium phosphate buffer (pH 7.4) and analyzed by FACScan. Control cells (no PI treatment) are indicated with a black line. PI uptake in CaTK1 cells was reduced following incubation with LFcn 11, BN 16, and VPR 12 but not with defensins.

Comparison of candidacidal activities of Hst 5 and {alpha}- and ß-defensins in CaTK2(wt) and CaTK1. In previous studies, we have shown that the candidacidal activity of HNP-1 shared some similar features to Hst 5 cytotoxic action (7). Therefore, we questioned whether HNP-1 and other defensins require TRK1 potassium transporter function for their fungicidal activities. For this purpose, we tested the candidacidal activities and PI uptake of HNP-1 and human ß-defensins 2 and 3. Incubation of C. albicans strain CaTK2(wt) with HNP-1 (7.5 µM to 31.25 µM) for 60 min resulted in concentration-dependent killing compared to untreated cells, reaching 91.7% cell death with the highest concentration tested (Fig. 4A, closed circles). In contrast to the significant reduction in cell death of the TRK1 mutant strain following treatment with LFcn 11, BN 16, or VPR 12, incubation of CaTK1 cells with HNP-1 did not result in any significant reduction in killing compared to that of the wild-type strain, except at the highest concentration tested (31.25 µM; P = 0.03). No significant difference in LD50 values for HNP-1 was observed between these strains, with LD50s of 13.0 µM and 11.6 µM for strains CaTK2(wt) and CaTK1, respectively, demonstrating that HNP-1 effects are not mediated substantially by Trk1p.



View larger version (11K):
[in this window]
[in a new window]
  		FIG. 4. Trk1p is not required for {alpha}- and ß-defensin candidacidal action. C. allbicans strains CaTK2(wt) (closed circles) and CaTK1 (trk1/TRK1) (open circles) were incubated with (A) HNP-1 (7.5 to 31.25 µM), (B) hBD-2 (1 to 10 µM), or (C) hBD-3 (0.5 to 10 µM) in 10 mM sodium phosphate buffer (pH 7.4) for 1 h at 37°C. Cell suspensions were diluted, and aliquots (500 cells) were spread onto YNB agar plates. Cell survival is expressed as a percentage of the control, and the loss of viability is calculated as follows: [1 – (colonies from peptide-treated cells/colonies from control cells)] x 100. Strain CaTK1 has sensitivity to {alpha}- and ß-defensins equivalent to that of wild-type cells.

We next tested human ß-defensins 2 and 3 for similarity with HNP-1 in the killing of strains CaTK2(wt) and CaTK1. Both strains CaTK2(wt) (Fig. 4B, closed circles) and CaTK1 (Fig. 4B, open circles) showed equally high susceptibility to hBD-2, with LD50 values of 5.9 µM and 5.5 µM, respectively. Similarly, the TRK1 mutant strain (Fig. 4C, open circles) was not less sensitive to hBD-3 than wild-type strain CaTK2(wt) (Fig. 4C, closed circles), except at 1 µM (P = 0.01). However, hBD-3 was ~6 times more potent than hBD-2 in fungicidal activity, with LD50 concentrations of hBD-3 of 0.8 µM and 1.35 µM for wild-type and TRK1 mutant strains, respectively. These results indicate that hBD-2 and -3 are undistinguishable in their toxicity between wild-type and TRK1 potassium transporter-deficient strains.

In order to confirm the results from testing of the candidacidal effect of HNP-1, hBD-2, and hBD-3, we compared PI uptakes in strains CaTK2(wt) and CaTK1. Incubation of CaTK2(wt) cells for 1 h with either 62 µM Hst 5 (Fig. 3C, yellow area), 31.25 µM HNP-1 (Fig. 3C, red line), or 10 µM hBD-3 (Fig. 3C, purple line), followed by a 30-min incubation with propidium iodide (50 mg/ml), resulted in pronounced PI uptake in wild-type cells. In contrast to results with Hst 5 (Fig. 3D, solid yellow area), the mutant CaTK1 strain showed no reduction in the levels of PI permeability following treatment with HNP-1 (Fig. 3D, red line) or hBD-3 (Fig. 3D, purple line). Results from hBD-2-induced membrane permeability for PI were identical to those from hBD-3 (data not shown). Altogether, the effect of human {alpha}- and ß-defensins on viability (Fig. 4) and membrane permeability (Fig. 3C and D) in TRK1 deletion strain CaTK1 clearly indicate that the candidacidal action of these peptides does not involve the TRK1 potassium transporter as a functional effector.

Hst 5 and ß-defensins bind to heat shock protein Ssa1/2 on the Candida cell surface. Since we previously demonstrated by using an overlay that HNP-1 and Hst 5 could compete for binding to C. albicans histatin 5 binding protein, subsequently identified as Ssa1/2 proteins (16), we next investigated whether these defensins require Ssa1/2 proteins on the C. albicans cell surface for their candidacidal activity. For this purpose, we compared the candidacidal activity of {alpha}- and ß-defensins in wild-type CAF4-2 cells with those of ssa1{Delta} (expressing only Ssa2p) and ssa2{Delta} (expressing only Ssa1p) mutant strains. Incubation of wild-type strain CAF4-2 with HNP-1 (3.6 to 10.9 µM), hBD-2 (2.9 to 11.6 µM), or hBD-3 (0.2 to 0.8 µM) also resulted in a concentration-dependent loss of cell viability, reaching ~90% with the highest concentrations tested (Fig. 5, closed squares). However, ssa1{Delta} (Fig. 5A, open circles) and ssa2{Delta} (Fig. 5A, open triangles) mutant strains showed no differences in susceptibility to HNP-1 compared to the wild-type strain. LD50 concentrations were 5.45 µM for wild-type, 3.96 µM for ssa1{Delta}, and 3.62 µM for ssa2{Delta} strains, thus showing that HNP-1 does not require Ssa1p or Ssa2p for killing.



View larger version (13K):
[in this window]
[in a new window]
  		FIG. 5. Heat shock 70 protein Ssa1/2 is required for hBD-2 and hBD-3 killing of Candida albicans. CAF4-2 (closed squares), ssa1{Delta} (open circles), and ssa2{Delta} (open triangles) strain cells of C. allbicans (106 cells/ml) were incubated with (A) HNP-1 (3.6 to 10.9 µM), (B) hBD-2 (2.9 to 11.6 µM), or (C) hBD-3 (0.2 to 0.8 µM) in 10 mM sodium phosphate buffer (pH 7.4) for 1 h. Control groups were incubated with 10 mM phosphate buffer alone. The loss of cell viability is expressed as follows: [1 – (colonies after Hst 5 treatment/colonies after incubation with buffer only)] x 100. Each data point represents the mean ± standard deviation of at least three independent experiments. Both ssa1{Delta} and ssa2{Delta} strains have reduced susceptibility to ß-defensins but not to HNP-1.

In contrast to HNP-1, killing with both hBD-2 and hBD-3 was reduced in mutant cells lacking either Ssa1 or Ssa2p. Incubation of ssa1{Delta} (Fig. 5B, open circles) with hBD-2 resulted in significantly reduced susceptibility to killing compared to wild-type cells (P < 0.03), and the ssa2{Delta} (Fig. 5B, open triangles) strain also showed reduced killing of these strains compared to the wild type (Fig. 5B, closed squares). LD50 concentrations for hBD-2 were increased from 5.3 µM for the wild type to 8.54 µM for ssa1{Delta} and 8.77 µM for ssa2{Delta} strains. Thus, both Ssa1 and Ssa2 proteins appear to contribute to hBD-2-induced killing of C. albicans cells.

The candidacidal activity of hBD-3 against wild-type strain CAF4-2 was highly potent, achieving ~90% killing at only 0.8 µM (Fig. 5C, closed squares). This strain was about 14 times more sensitive to hBD-3 than to hBD-2 (Fig. 5B, closed squares). Incubation of the ssa2{Delta} strain with hBD-3 resulted in a ~30% reduction in sensitivity at lower concentrations (0.2 µM and 0.4 µM; P < 0.05) compared to that of the wild-type strain (Fig. 5C, open triangles). However, at higher concentrations (0.6 µM and 0.8 µM), the ssa2{Delta} strain showed sensitivity comparable to that of the wild-type strain. In contrast, the ssa1{Delta} strain was significantly protected from killing by hBD-3 at all concentrations tested (P < 0.001) (Fig. 5C, open circles). Overall, ssa2{Delta} cells exhibited a less-than-twofold reduction and ssa1{Delta} cells exhibited about a fourfold reduction in sensitivity to hBD-3 killing compared to wild-type cells. The LD50 concentration for ssa2{Delta} was 0.35 µM, while that ssa1{Delta} was 0.77 µM, compared with an LD50 of 0.19 µM for strain CAF4-2. Thus, hBD-3 fungicidal activity is highly dependent upon the presence of Ssa1p, although Ssa2p also appears to play a role in its killing function.


arrow
DISCUSSION
 
The mechanism of action of antimicrobial peptides differs from that of classical antifungal drugs; hence, they have the potential to be used therapeutically, perhaps to augment conventional antifungal drug treatment. Although many antimicrobial peptides have common features, such as net cationic charge, a unifying principle for their mechanism of antimicrobial action has not been demonstrated. While human defensins have been shown to interact with membranes in susceptible bacteria and fungi (15), other peptides such as histatins and lactoferricin have been shown to be transported intracellularly and to affect critical cytosolic components. Here, we provide evidence for fungicidal mechanisms of Hst 5 that overlap those of other small cationic peptides, including disulfide-linked human ß-defensin proteins.

Although the complete mechanism of Hst 5 candidacidal activity has not yet been elucidated, it is now clear that Hst 5 fungicidal activity requires TRK1 potassium transporter function (2) and is associated with the efflux of ATP (14) and total cell volume reduction (1). In this study, we show that Hst 5 induces membrane permeability for PI in the susceptible C. albicans wild-type strain but not in the TRK1 hemizygous strain. Hence, we conclude that histatin 5 may directly or indirectly alter Trk1p function, allowing the efflux of larger anions, including ATP, and the influx of small cationic dyes, such as propidium iodide.

When we compared killing and PI uptake induced by other small cationic peptides in wild-type and TRK1 mutant strains, similarities to Hst 5 were observed. The cationic peptides LFcn 11, BN 16, and VPR 12 showed potent antifungal activity against the wild-type strain but exhibited reduced killing potency in strain CaTK1. Similarly, treatment of CaTK1 with these peptides at lethal concentrations resulted in significantly lowered PI uptake. Therefore, it is likely that these small cationic peptides also ultimately involve Trk1p function to increase the membrane permeability to PI. However, the inhibition of PI uptake with the peptides tested was not as pronounced as the Hst 5-induced membrane permeability to PI in CaTK1 cells, indicating that the requirement of Trk1p for Hst 5 killing is more explicit or that these peptides may have additional or alternative targets. In this regard, it has been shown that the synthetic D-decapeptide BM2 (D-NH2-RRRFWWFRRR-CONH2), which has high fungicidal activity, acts as a specific inhibitor of the yeast plasma membrane proton pump Pma1p (19). Interestingly, this peptide was found to possess some similarity in primary structure to other cationic peptides, including the first 9 residues of LFcn 11, which was able to inhibit Pma1p activity as well (19). Our data suggest that cationic peptides are capable of modulating the function of Candida cell membrane transporters and possibly other proteins maintaining membrane integrity, thus eventually permitting the influx of PI.

Although cationic peptides may interact with cell membrane proteins, other cytoplasm-localized effectors are likely involved. It has been reported that bactenecin 7 kills bacteria in a two-stage process in which it first gains entry to cytoplasm and then inhibits intracellular targets (24). Similarly, it has been demonstrated that intracellularly expressed VPR causes cell death in Saccharomyces cerevisiae (18). Intracellular effectors may be required for candidacidal action of LFcn 11 as well, since LFcn 11-treated C. albicans cells show killing, PI uptake, and ATP release (17) similar to that induced by Hst 5. Thus, intracellular effector sites are likely to be involved in the candidacidal mechanism of BN 16, VPR 12, and LFcn 11.

In previous studies, we have demonstrated that salivary Hst 5 and HNP-1 kill C. albicans via shared pathways, since they exhibit similarities in their active concentrations, ATP efflux, and inhibitor profiles. In addition, HNP-1 was able to compete with Hst 5 for interaction with a cell envelope binding protein in Candida, which we subsequently identified as Ssa1/2 proteins (16). However, when we compared the killing profiles following HNP-1 treatment of C. albicans ssa1{Delta} and ssa2{Delta} mutants with that of the wild-type strain, no difference in susceptibility to HNP-1 among the strains was observed. These results are at odds with the binding we previously observed by using an overlay assay (6). It is possible that HNP-1 binding to C. albicans could be done with another homologous Candida cell envelope protein, perhaps a related heat shock protein such as HSP104. Studies to identify additional proteins that are candidate Hst 5-interacting partners on the yeast cell surface are ongoing in our laboratory. HNP-1 seems to differ from Hst 5 not only in the interaction with C. albicans surface proteins but also in the lack of a requirement for Trk1p, as shown by the equivalent susceptibility of the TRK1 mutant and CaTK2(wt) strains to HNP-1 killing. Thus, our data do not rule out a mechanism by which Hst 5 alters plasma membrane permeability of susceptible cells, perhaps assisted by other Candida cell envelope proteins. However, these results point to the involvement of cytoplasmic effectors distinct from those utilized by Hst 5. Altogether, our results show that HNP-1 does not require binding to Ssa1/2p, nor does it involve Trk1p, for cell killing.

Human ß-defensin 2 and the recently identified human ß-defensin 3 have been shown to possess potent antifungal activity (8, 11); however, the mechanism for their candidacidal activity is poorly understood. We have identified a more specific basis for these differences, as we found that killing profiles of C. albicans ssa1{Delta} and ssa2{Delta} strains were dissimilar between hBD-2 and hBD-3. Interestingly, while both ssa1{Delta} and ssa2{Delta} strains showed 15 to 20% reduction of killing following hBD-2 treatment compared to that shown by the wild type, hBD-3 killing was significantly inhibited, indicating that ß-defensins 2 and 3 seem to exert differential effects that are dependent upon Ssa1/2p. Moreover, hBD-3 may possess a higher affinity for Ssa1p, as the ssa1{Delta} strain with both alleles of SSA1 deleted was substantially protected from hBD-3 killing. Recent evidence that the radish antifungal defensin RsAFP2, which has high structural homology with mammalian ß-defensins, interacts directly with glucosylceramides on the plasma membrane of susceptible fungi suggests that ß-defensins are capable of specific interactions with cell envelope components (26). In addition, hBD-2-treated C. albicans cells showed morphological evidence of thinning and dissolution of the cell wall (8). However, differences in binding among ß-defensins seem to exist, since hBD-2 but not hBD-3 activity was mediated by binding to lipooligosaccharides on the Haemophilus influenzae plasma membrane (25). In addition, it has been shown that the bactericidal activity of hBD-2, but not of hBD-3, is affected by increased ionic strength (11). Altogether, our results suggest that hBD-3 may exhibit specific interactions with heat shock protein 70 family Ssa1p on the Candida cell surface that is involved in cell killing. As hBD-3 is highly potent against Candida spp. at low concentrations and is expressed in various oral tissues upon Candida challenge (23), knowledge about its mechanism of action is important for the prevention and treatment of oral candidiasis.


arrow
ACKNOWLEDGMENTS
 
This work was supported by NIH grant DE10641 from the National Institute of Dental and Craniofacial Research (to M.E.).


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: 310 Foster Hall, State University of New York at Buffalo, Main Street Campus, 3435 Main Street, Buffalo, NY 14214. Phone: (716) 829-3067. Fax: (716) 829-3942. E-mail: edgerto{at}buffalo.edu. Back


arrow
REFERENCES
 
    1
  1. Baev, D., X. S. Li, J. Dong, P. Keng, and M. Edgerton. 2002. Human salivary histatin 5 causes disordered volume regulation and cell cycle arrest in Candida albicans. Infect. Immun. 70:4777-4784.[Abstract/Free Full Text]
    2. 2
  2. Baev, D., A. Rivetta, S. Vylkova, J. N. Sun, G. F. Zeng, C. L. Slayman, and M. Edgerton. 2004. The TRK1 potassium transporter is the critical effector for killing of Candida albicans by the cationic protein, histatin 5. J. Biol. Chem. 279:55060-55072.[Abstract/Free Full Text]
    3. 3
  3. Castagnola, M., R. Inzitari, D. V. Rossetti, C. Olmi, T. Cabras, V. Piras, P. Nicolussi, M. T. Sanna, M. Pellegrini, B. Giardina, and I. Messana. 2004. A cascade of 24 histatins (histatin 3 fragments) in human saliva. Suggestions for a pre-secretory sequential cleavage pathway. J. Biol. Chem. 279:41436-41443.[Abstract/Free Full Text]
    4. 4
  4. Chapple, D. S., C. L. Joannou, D. J. Mason, J. K. Shergill, E. W. Odell, V. Gant, and R. W. Evans. 1998. A helical region of human lactoferrin. Its role in antibacterial pathogenesis. Adv. Exp. Med. Biol. 443:215-220.[Medline]
    5. 5
  5. Den Hertog, A. L., H. W. W. F. Sang, R. Kraayenhof, J. G. Bolscher, W. Van't Hof, E. C. Veerman, and A. V. N. Amerongen. 2004. Interactions of histatin 5 and histatin 5-derived peptides with liposome membranes: surface effects, translocation and permeabilization. Biochem. J. 379:665-672.[CrossRef][Medline]
    6. 6
  6. Edgerton, M., S. E. Koshlukova, T. E. Lo, B. G. Chrzan, R. M. Straubinger, and P. A. Raj. 1998. Candidacidal activity of salivary histatins. Identification of a histatin 5-binding protein on Candida albicans. J. Biol. Chem. 273:20438-20447.[Abstract/Free Full Text]
    7. 7
  7. Edgerton, M., S. E. Koshlukova, M. W. Araujo, R. C. Patel, J. Dong, and J. A. Bruenn. 2000. Salivary histatin 5 and human neutrophil defensin 1 kill Candida albicans via shared pathways. Antimicrob. Agents Chemother. 44:3310-3316.[Abstract/Free Full Text]
    8. 8
  8. Feng, Z., B. Jiang, J. Chandra, M. Ghannoum, S. Nelson, and A. Weinberg. 2005. Human beta-defensins: differential activity against candidal species and regulation by Candida albicans. J. Dent. Res. 84:445-450.[Abstract/Free Full Text]
    9. 9
  9. Helmerhorst, E. J., P. Breuwer, W. van't Hof, E. Walgreen-Weterings, L. C. Oomen, E. C. Veerman, A. V. Amerongen, and T. Abee. 1999. The cellular target of histatin 5 on Candida albicans is the energized mitochondrion. J. Biol. Chem. 274:7286-7291.[Abstract/Free Full Text]
    10. 10
  10. Helmerhorst, E. J., W. van't Hof, P. Breeuwer, E. C. Veerman, T. Abee, R. F. Troxler, A. V. Amerongen, and F. G. Oppenheim. 2001. Characterization of histatin 5 with respect to amphipathicity, hydrophobicity, and effects on cell and mitochondrial membrane integrity excludes a candidacidal mechanism of pore formation. J. Biol. Chem. 276:5643-5649.[Abstract/Free Full Text]
    11. 11
  11. Joly, S., C. Maze, P. B. McCray, and J. M. Guthmiller. 2004. Human ß-defensins 2 and 3 demonstrate strain-selective activity against oral microorganisms. J. Clin. Microbiol. 42:1024-1029.[Abstract/Free Full Text]
    12. 12
  12. Kagan, B. L., M. E. Selsted, T. Ganz, and R. I. Lehrer. 1990. Antimicrobial defensin peptides form voltage-dependent ion-permeable channels in planar lipid bilayer membranes. Proc. Natl. Acad. Sci. USA 87:210-214.[Abstract/Free Full Text]
    13. 13
  13. Kang, J. H., M. K. Lee, K. L. Kim, and K. S. Hahm. 1996. Structure-biological activity relationships of 11-residue highly basic peptide segment of bovine lactoferrin. Int. J. Pept. Protein Res. 48:357-363.[Medline]
    14. 14
  14. Koshlukova, S. E., T. L. Lloyd, M. W. Araujo, and M. Edgerton. 1999. Salivary histatin 5 induces non-lytic release of ATP from Candida albicans leading to cell death. J. Biol. Chem. 274:18872-18879.[Abstract/Free Full Text]
    15. 15
  15. Lehrer, R. I., and T. Ganz. 1996. Endogenous vertebrate antibiotics. Defensins, protegrins, and other cysteine-rich antimicrobial peptides. Ann. N. Y. Acad. Sci. 797:228-239.[Medline]
    16. 16
  16. Li, X. S., M. S. Reddy, D. Baev, and M. Edgerton. 2003. Candida albicans Ssa1/2p is the cell envelope binding protein for human salivary histatin 5. J. Biol. Chem. 278:28553-28561.[Abstract/Free Full Text]
    17. 17
  17. Lupetti, A., A. Paulusma-Annema, M. M. Welling, S. Senesi, J. T. van Dissel, and P. H. Nibbering. 2000. Candidacidal activities of human lactoferrin peptides derived from the N terminus. Antimicrob. Agents Chemother. 44:3257-3263.[Abstract/Free Full Text]
    18. 18
  18. Macreadie, I. G., D. R. Thorburn, D. M. Kirby, L. A. Castelli, N. L. de Rozario, and A. A. Azad. 1997. HIV-1 protein Vpr causes gross mitochondrial dysfunction in the yeast Saccharomyces cerevisiae. FEBS Lett. 410:145-149.[CrossRef][Medline]
    19. 19
  19. Monk, B. C., K. Niimi, S. Lin, A. Knight, T. B. Kardos, R. D. Cannon, R. Parshot, A. King, D. Lun, and D. R. Harding. 2005. Surface-active fungicidal D-peptide inhibitors of the plasma membrane proton pump that block azole resistance. Antimicrob. Agents Chemother. 49:57-70.[Abstract/Free Full Text]
    20. 20
  20. Pollock, J. J., L. Denepitiya, B. J. MacKay, and V. J. Iacono. 1984. Fungistatic and fungicidal activity of human parotid salivary histidine-rich polypeptides on Candida albicans. Infect. Immun. 44:702-707.[Abstract/Free Full Text]
    21. 21
  21. Raj, P. A., M. Edgerton, and M. J. Levine. 1990. Salivary histatin 5: dependence of sequence, chain length, and helical conformation for candidacidal activity. J. Biol. Chem. 265:3898-3905.[Abstract/Free Full Text]
    22. 22
  22. Raj, P. A., E. Marcus, and M. Edgerton. 1996. Delineation of an active fragment and poly(L-proline) II conformation for candidacidal activity of bactenecin 5. Biochemistry 35:4314-4325.[CrossRef][Medline]
    23. 23
  23. Schofield, D. A., C. Westwater, and E. Balish. 2004. Beta-defensin expression in immunocompetent and immunodeficient germ-free and Candida albicans-monoassociated mice. J. Infect. Dis. 190:1327-1334.[CrossRef][Medline]
    24. 24
  24. Skerlavaj, B., D. Romeo, and R. Gennaro. 1990. Rapid membrane permeabilization and inhibition of vital functions of gram-negative bacteria by bactenecins. Infect. Immun. 58:3724-3730.[Abstract/Free Full Text]
    25. 25
  25. Starner, T. D., W. E. Swords, M. A. Apicella, and P. B. McCray, Jr. 2002. Susceptibility of nontypeable Haemophilus influenzae to human ß-defensins is influenced by lipooligosaccharide acylation. Infect. Immun. 70:5287-5299.[Abstract/Free Full Text]
    26. 26
  26. Thevissen, K., D. C. Warnecke, I. E. Francois, M. Leipelt, E. Heinz, C. Ott, U. Zahrihger, B. P. Thomma, K. K. Ferket, and B. P. Cammue. 2004. Defensins from insects and plants interact with fungal glucosylceramides. J. Biol. Chem. 279:3900-3905.[Abstract/Free Full Text]
    27. 27
  27. Xu, Y., I. Ambudkar, H. Yamagishi, W. Swaim, T. J. Walsh, and B. C. O'Connell. 1999. Histatin 3-mediated killing of Candida albicans: effect of extracellular salt concentration on binding and internalization. Antimicrob. Agents Chemother. 43:2256-2262.[Abstract/Free Full Text]

Antimicrobial Agents and Chemotherapy, January 2006, p. 324-331, Vol. 50, No. 1
0066-4804/06/$08.00+0     doi:10.1128/AAC.50.1.324-331.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Conti, H. R., Shen, F., Nayyar, N., Stocum, E., Sun, J. N., Lindemann, M. J., Ho, A. W., Hai, J. H., Yu, J. J., Jung, J. W., Filler, S. G., Masso-Welch, P., Edgerton, M., Gaffen, S. L. (2009). Th17 cells and IL-17 receptor signaling are essential for mucosal host defense against oral candidiasis. JEM 206: 299-311 [Abstract] [Full Text]  
  • Chaffin, W. L. (2008). Candida albicans Cell Wall Proteins. Microbiol. Mol. Biol. Rev. 72: 495-544 [Abstract] [Full Text]  
  • Gank, K. D., Yeaman, M. R., Kojima, S., Yount, N. Y., Park, H., Edwards, J. E. Jr., Filler, S. G., Fu, Y. (2008). SSD1 Is Integral to Host Defense Peptide Resistance in Candida albicans. Eukaryot Cell 7: 1318-1327 [Abstract] [Full Text]  
  • Luque-Ortega, J. R., Hof, W. v., Veerman, E. C. I., Saugar, J. M., Rivas, L. (2008). Human antimicrobial peptide histatin 5 is a cell-penetrating peptide targeting mitochondrial ATP synthesis in Leishmania. FASEB J. 22: 1817-1828 [Abstract] [Full Text]  
  • van der Weerden, N. L., Lay, F. T., Anderson, M. A. (2008). The Plant Defensin, NaD1, Enters the Cytoplasm of Fusarium Oxysporum Hyphae. J. Biol. Chem. 283: 14445-14452 [Abstract] [Full Text]  
  • Vylkova, S., Jang, W. S., Li, W., Nayyar, N., Edgerton, M. (2007). Histatin 5 Initiates Osmotic Stress Response in Candida albicans via Activation of the Hog1 Mitogen-Activated Protein Kinase Pathway. Eukaryot Cell 6: 1876-1888 [Abstract] [Full Text]  
  • Vylkova, S., Nayyar, N., Li, W., Edgerton, M. (2007). Human {beta}-Defensins Kill Candida albicans in an Energy-Dependent and Salt-Sensitive Manner without Causing Membrane Disruption. Antimicrob. Agents Chemother. 51: 154-161 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vylkova, S.
Right arrow Articles by Edgerton, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vylkova, S.
Right arrow Articles by Edgerton, M.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»