Figures
Abstract
Linezolid (L), a potent antibiotic for Methicillin Resistant Staphylococcus aureus (MRSA), inhibits bacterial protein synthesis. By contrast, vancomycin (V) is a cell wall active agent. Here, we used a murine sepsis model to test the hypothesis that L treatment is associated with differences in bacterial and host characteristics as compared to V. Mice were injected with S. aureus USA300, and then intravenously treated with 25 mg/kg of either L or V at 2 hours post infection (hpi). In vivo alpha-hemolysin production was reduced in both L and V-treated mice compared to untreated mice but the reduction did not reach the statistical significance [P = 0.12 for L; P = 0.70 for V). PVL was significantly reduced in L-treated mice compared to untreated mice (P = 0.02). However the reduction of in vivo PVL did not reach the statistical significance in V- treated mice compared to untreated mice (P = 0.27). Both antibiotics significantly reduced IL-1β production [P = 0.001 for L; P = 0.006 for V]. IL-6 was significantly reduced with L but not V antibiotic treatment [P<0.001 for L; P = 0.11 for V]. Neither treatment significantly reduced production of TNF-α. Whole-blood gene expression profiling showed no significant effect of L and V on uninfected mice. In S. aureus-infected mice, L altered the expression of a greater number of genes than V (95 vs. 42; P = 0.001). Pathway analysis for the differentially expressed genes identified toll-like receptor signaling pathway to be common to each S. aureus-infected comparison. Expression of immunomodulatory genes like Cxcl9, Cxcl10, Il1r2, Cd14 and Nfkbia was different among the treatment groups. Glycerolipid metabolism pathway was uniquely associated with L treatment in S. aureus infection. This study demonstrates that, as compared to V, treatment with L is associated with reduced levels of toxin production, differences in host inflammatory response, and distinct host gene expression characteristics in MRSA sepsis.
Citation: Sharma-Kuinkel BK, Zhang Y, Yan Q, Ahn SH, Fowler VG Jr (2013) Host Gene Expression Profiling and In Vivo Cytokine Studies to Characterize the Role of Linezolid and Vancomycin in Methicillin-Resistant Staphylococcus aureus (MRSA) Murine Sepsis Model. PLoS ONE 8(4): e60463. https://doi.org/10.1371/journal.pone.0060463
Editor: Karsten Becker, University Hospital Münster, Germany
Received: December 6, 2012; Accepted: February 26, 2013; Published: April 2, 2013
Copyright: © 2013 Sharma-Kuinkel et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported primarily by a grant from Pfizer Inc., WS499060 to S.H.A. V.G.F. was supported in part by K24 AI093969 from the United States National Institutes of Health. No additional external funding was received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have read the journal’s policy and have the following conflicts. V.G.F. has received grants or research support from Astellas Pharma, Inc., Cubist Pharmaceuticals, Merck & Co., Inc., Theravance, Inc., MedImmune, Cerexa, Inc., Pfizer, Advanced Liquid Logic, MedImmune and the United States National Institutes of Health. He is a consultant for Astellas Pharma, Inc., Biosynexus, Novartis, Cubist Pharmaceuticals, Inimex, Merck & Co., Inc., Galderma, Johnson & Johnson, Medicines Company, Novartis and is on the speaker’s bureau and advisory committee for Cubist Pharmaceuticals. He has received honoraria from Achaogen, Arpida, Astellas Pharma, Inc., Cubist Pharmaceuticals, Durata, Inhibitex, Leo Pharma, Merck & Co., Inc., Pfizer, Targanta, Theravance, Inc., and Ortho-McNeil. This study uses Linezolid, a product produced by Pfizer, Inc. The co-author V.G.F. is a PLOS ONE Editorial Board member. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials. B.K.S.K, Y.Z., Q.Y. and S.H.A. have no competing interests to declare.
Introduction
Staphylococcus aureus is a leading cause of nosocomial and community-acquired infections [1], [2]. An emerging body of evidence suggests that specific bacterial toxins are involved in the pathogenesis of a variety of MRSA infections [3]–[6]. For example, alpha-hemolysin is a potent cell lysing agent, thought to induce the production of IL-1β, IL-6, and IL-8, and TNF-α [3]. In addition, the Panton-Valentine Leukocidin (PVL), a phage-associated bicomponent leukocidin, is a well-recognized virulence factor in S. aureus. PVL lyses neutrophils and induces the release of proinflammatory cytokines from these cells [7], [8]. Its presence in S. aureus has been associated with a number of clinical syndromes, including necrotizing S. aureus pneumonia [9]–[12], necrotizing fasciitis [13], osteomyelitis [14], and induces the release of proinflammatory cytokines from neutrophils [7], [8].
Significant evidence suggests that the interruption of bacterial toxin synthesis by protein-synthesis inhibitors may reduce toxin production and potentially improve clinical outcome [15]–[18]. It has been shown that linezolid (L), but not vancomycin (V), is a potent inhibitor of protein synthesis [18]–[20]. Thus, evidence supports the hypothesis that linezolid’s inhibition of bacterial protein synthesis can influence the clinical course of toxin-mediated infections.
In contrast to cell-wall active agents, the majority of which induce bacterial toxin production [18], [21], [22], L specifically down-regulates the production of a variety of virulence toxin genes, including PVL and alpha hemolysin, even in its sub-inhibitory concentrations [20], [22], [23]. These findings suggest that these two classes of antibiotics may have significantly different effects on the expression of bacterial toxin genes. However, despite previous efforts [24]–[26], the full impact of L on the modulation of the host-pathogen interaction in MRSA sepsis is still unknown. The current study is designed to address this gap by simultaneously evaluating the influence of two broadly different classes of antibiotics (L and V) on both the host and the pathogen. Although the host gene expression profiling has been previously used to study the host pathogen interaction in infections [27], [28] and in non-infectious conditions like cancer [29]–[31], none of the studies have used this technique to compare and contrast the effect of antibiotics on host gene expression. In the current study, we test the hypothesis that L treatment elicits a different gene expression profile compared to V treatment in MRSA sepsis. To do this, we evaluated and compared the impact of L and V on both host and pathogen gene expression using a well-characterized strain of CA-MRSA in a murine sepsis model of infection.
Materials and Methods
S. aureus Strain and Culture
S. aureus USA300 was obtained from NARSA (Network on Antimicrobial Resistance in Staphylococcus aureus). For preparation of S. aureus for injection, an overnight bacterial culture of S. aureus was diluted with fresh tryptic soy broth (TSB) (Becton Dickinson, Sparks, MD) and incubated (37°C) with aeration to log-phase growth [optical density at wavelength 600 nm (OD 600) of 1.0] [32]. S. aureus was harvested by centrifugation, rinsed, and re-suspended in PBS.
Antimicrobial Agents
Linezolid was obtained from Pfizer Inc. (Groton, CT). Vancomycin was purchased from Sigma-Aldrich Chemical (St. Louis, MO).
Animals
A/J mice (Jackson Laboratory, Bar Harbor, Maine), 8 to 10 weeks old and weighing approximately 18 g (16 to 20 g) were used in this study. Mice were housed in sterile microisolator cages (5 per cage) with sterile bedding, feed, and acidified water. All animal experiments were carried out in strict accordance with the recommendations of NIH guidelines, the Animal Welfare Act, and US federal law. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Duke University (IACUC number: #1890907) which has been accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International.
Inoculum
Mice were challenged with S. aureus via intraperitoneal (i.p.) administration with 6×106 CFU/g of S. aureus USA300 (0.2 ml), then intravenously injected with 25 mg/kg L or V dissolved in 0.2 ml of pyrogen free distilled water 2 hours post-infection (hpi) [33], [34]. To mimic the natural course of S. aureus infection in humans, which typically arises from a primary focus of infection and disseminates to other sites, we employed an i.p. route of infection in our animal model [35]. The 2 hpi time point was chosen to reflect the fact that bacteria are demonstrable in murine bloodstream within 2 hour post i.p. infection [35].
Quantitative Culture
Blood and kidney samples were obtained from infected mice (n = 5 in each group) 24 hpi, which were serially diluted and plated on TSA agar plates at 37°C to quantify CFU. Kidneys collected from euthanized animals were homogenized in 1X PBS and diluted 10 fold serially. Blood was collected via cardiac puncture. Tissue homogenates and blood were serially 10-fold diluted in sterile 1X PBS. The serial dilutions were plated in Tryptic Soy Agar (TSA) plates and incubated (37°C, overnight) to count the number of colony forming units (CFU) of S. aureus. Finally the number of bacteria in blood and kidney were expressed as CFU/ml and CFU/g respectively. Statistical significance of the comparison was tested by using F-test.
Production of Antibody to PVL
The gene encoding the PVL toxin, lukF, was cloned into an overexpression his-tag fusion vector (pET100/D-TOP; Invitrogen Co.), as described elsewhere [36], and purified according to the manufacturer’s instructions. Two rabbits (Rockland Immunochemicals, Inc., PA, USA) were used for antibody production by three times boosting with the LukF recombinant protein mixed in an equal volume of incomplete Freund’s adjuvant 1 week apart. Serum was obtained 45 days after the initial dose. Polyclonal antibody efficiency against PVL has been successfully tested by Western blotting (data not shown).
Enzyme Linked Immunosorbent Assay (ELISA)
In vivo levels of S. aureus toxins (alpha-hemolysin and PVL) and cytokines in mouse serum were determined by ELISA. A 96-well plate (BD Falcon, Franklin, NJ) was coated with 10 µg/ml polyclonal antibody against rabbit in 1X PBS overnight at 4°C, after which the wells were washed and blocked with 5% bovine serum albumin in PBS. Serum was added to the wells and incubated at RT for 1 hour. Primary antibodies specific for alpha-hemolysin (Sigma-Aldrich, St. Louis, MO) or PVL (Rockland Immunochemicals, Inc.) were added to the supernatant, followed by anti-rabbit IgG horseradish-peroxidase conjugate (Sigma-Aldrich, St. Louis, MO), 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid). After 20 min, absorbance was read at 405 nm. For the measurement of cytokine, serum was separated from blood obtained by intracardiac puncture. Protein concentrations were determined by bicinchoninic acid method kit (Pierce, Rockford, IL). Equal amounts of protein for each sample were used to measure the levels of interleukin-1β, IL-6, and tumor necrosis factor alpha (TNF-α) using ELISA (Duo kit; Invitrogen) following the manufacturer’s instructions. The serum levels of cytokines were expressed as pg/ml of serum. The statistical significance between the experimental groups was studied using unpaired t-test.
Microarray Using Whole Blood RNAs
Total RNA was prepared from mouse blood using the Mouse RiboPure Blood RNA isolation kit (Ambion, Austin, TX) following the manufacturer’s instruction. Globin mRNA was removed from whole blood RNA samples using the Globinclear kit (Ambion, Austin, TX). A total of 35 samples passed the quality criteria of the Agilent Bioanalyzer and were used for microarray analysis. Since the total RNA yield of many samples was low, one round of linear amplification was performed for all samples using the MessageAmp Premier kit (Ambion, Austin, TX). RNA integrity numbers were calculated for all samples and found to be within tolerance limits. Affymetrix GeneChip® Mouse Genome 430 2.0 Arrays were used (Affymetrix, Santa Clara, CA). The biotin-labeled cDNA was hybridized to the arrays for 16 hours at 45°C following the manufacturer’s instruction. The arrays were then washed and labeled with streptavidinphycoerythrin (strep-PE), and the signal was amplified using biotinylated antistreptavidin followed by another round of staining with strep-PE. Washing and staining were performed on the Affymetrix fluidics station according to the recommended fluidics protocol. Amplification and microarray hybridization were performed at the Duke University Microarray Core. Labeled gene chips were scanned using an Affymetrix Genechip Scanner 7G (Affymetrix, Santa Clara, CA). This array contains 45,101 probe sets to analyze the expression level of over 39,000 transcripts and variants from over 34,000 well characterized mouse genes. The microarray data have been deposited in the NCBI GEO and are accessible through GEO series accession no GSE38531.
Microarray Data Analysis
Gene expression data were imported into Partek Genomics Suite 6.5 (Partek, St Louis, Mo) as CEL files using default parameters. Raw data were preprocessed, including background correction, normalization, and summarization using robust multiarray average analysis, and expression data were log2 transformed. Differential expression analysis for the whole blood cells was performed using 1-way analysis of variance (either infection status or drug treatment alone). Gene lists were created using a cutoff of P<0.05, 2-fold change, although second-level analysis using a false-discovery rate of <0.05, 2-fold change was also performed. Hierarchical clustering (heat map image) was also performed using Partek Genomics suite.
Pathway Analysis
Pathway analysis for functional annotation of the genes was performed by GATHER KEGG pathway analysis (http://gather.genome.duke.edu/) [37]. The significance of the association between the data set and the KEGG pathway was measured in two ways: 1) Fischer’s exact test to calculate a P-value and 2) calculating the ratio of the pathway-associated genes in the experimental data to the total number of genes in that pathway.
Results
Effect of Antibiotics on Bacterial Load in Kidney and Blood
To characterize the bacteriological effect of L and V, we measured the tissue burden of S. aureus USA300 in A/J mice at 24 hours post-infection (hpi). Compared to untreated mice, bacterial loads in kidney (Figure 1A) and blood (Figure 1B) were significantly reduced in mice treated with either L or V. Bacterial counts were similar from the kidney in L and V-treated mice (19.93 × 106 cfu/gm vs. 39.38 × 106 cfu/gm; P = 0.394), but significantly lower from blood in V-treated mice (308 × 103 cfu/ml vs. 7.778 103 cfu/ml; P = 0.0003) (Figure 1).
The bacterial loads in the (A) Kidney and (B) Blood from A/J mice after intraperitoneal infection with S. aureus (6×106 CFU/g, strain USA300), and treatment with linezolid (L) or vancomycin (V) (25 mg/kg). Infected mice were sacrificed as follows: 2 hpi S. aureus; 24 hpi S. aureus; 24 hpi S. aureus with linezolid or vancomycin. Each symbol represents one mouse. P-values were obtained by F-test. The error bars correspond to the standard error of mean (SEM) and the dashed line correspond to the mean value. Values of P<0.05 were considered significant.
Effect of Antibiotic on In Vivo S. aureus Toxins and Cytokines Production in Serum
To study the effect of toxin production suppression in staphylococcal sepsis, we evaluated the effect of L and V on in vivo concentration of PVL and alpha-hemolysin in mouse serum. Compared to untreated controls, L but not V was associated with significant reduction in PVL production (P = 0.02 in L vs. P = 0.27 in V) (Figure 2A). Mice treated with L and V did not differ significantly in reduction of PVL production (P = 0.15). Neither L nor V treatment resulted in a statistically significant difference on alpha-hemolysin production compared to the untreated controls (P = 0.12 in L and P = 0.70 in V). (Figure 2B). Despite the fact that V treatment significantly lowered the bacterial counts in blood compared to L treatment (Figure 1B), the level of both PVL and alpha-hemolysin was higher in V treated serum compared to that of L treated.
In vivo level of (A) Panton Valentine Leukocidin (PVL, µg) (B) and Alpha-toxin (ng) in the serum of A/J mice after 24 h of intraperitoneal (ip) infection with S. aureus (6×106 CFU/g, strain USA300), and treatment with linezolid or vancomycin (25 mg/kg) at 2 h was measured using ELISA. P-value was obtained by unpaired t-test. The error bar corresponds to the standard error of mean (SEM). Values of P<0.05 were considered significant.
To study the relative immunomodulatory effects of L and V treatment, we next measured the cytokine levels in mouse serum at 24 hpi using ELISA. Compared to uninfected controls, cytokine production was significantly increased at 24 hours following S. aureus infection compared to uninfected controls (P<0.001). Both antibiotics significantly reduced IL-1β production compared to untreated group [P = 0.001 in L and P = 0.006 in V] and there was no statistically significant difference between the L and V treatment groups (P = 0.78) (Figure 3A)]. Although treatment with L, but not V, resulted in a statistically significant reduction of IL-6 as compared to untreated controls [P<0.001 in L and P = 0.11 in V], the differences in IL6 production between the two treatment groups did not differ significantly (P = 0.11) (Figure 3B). Neither L nor V treatment resulted in statistically significant reduction in production of TNF-α (Figure 3C).
Level of (A) IL-1beta (pg/ml), (B) IL-6 (pg/ml), and (C)TNF-alpha (pg/ml) from the serum of A/J mice without infection, with 2 hours post infection (hpi) with S. aureus (6×106 CFU/g, strain USA300), and 24 hpi with and without antibiotic treatment as labeled was measured using ELISA. P-value was obtained by unpaired t-test. The error bar corresponds to the standard error of mean (SEM). Values of P<0.05 were considered significant.
Effect of Antibiotic on Host Gene Expression Pattern
To ascertain that the different treatment groups (e.g. uninfected, S. aureus infected, S. aureus infected with L treatment, and S. aureus infected with V treatment) can be separated into several distinct clusters based on their gene expression profiles, we next conducted an unsupervised hierarchical clustering using the gene expression data from 35 microarrays. As seen in the Figure 4 heat map, distinct clustering is clearly evident based on infection status (Red squares: uninfected, Blue squares: S. aureus infected) and the infection time-course [Green squares: 0 hpi, Purple squares: 2 hpi, and Orange squares: 24 hpi (with or without antibiotic treatment)].
Each colored row in the heat map represents the gene expression value for a probe and each column represents a sample. The color conventions are as follows: red indicates over-expressed transcripts, blue represents under-expressed transcripts. Time of analysis post-infection are marked as follows: green rectangles indicate 0 hpi, purple rectangles indicate 2 hpi, and orange rectangles indicate 24 hpi (with and without antibiotic treatment).
To address the antibiotic-specific differences in gene expression, we identified the number of genes uniquely expressed in response to S. aureus infection, infection with L treatment, and infection with V treatment (Figure 5). When the drug-alone treatment group (L or V) was compared with uninfected/untreated group (controls), only five genes were differentially expressed solely in each group. This finding suggests that treatment with either L or V have a minimal impact on host gene expression itself and each drug shows a unique expression profile without any overlap when administered in an uninfected host (Figure 5A). Within the S. aureus-infected host, L treatment significantly affected more genes (95 genes) compared to V treatment (42 genes) (P = 0.0013). An additional 22 genes were affected by both antibiotics, indicating some overlap in the expression response induced by L or V in S. aureus-infected host (Figure 5B). A comparison of S. aureus infection dynamics shows a considerably higher number of genes differentially expressed at 24 hpi compared to 2 hpi (1691 at 24 hpi vs. 599 at 2 hpi). Around 74% of genes (447 of 599) differentially expressed at 2 hpi were also differentially expressed at 24 hpi (compared to uninfected controls) (Figure 5C). The details of all the genes differentially expressed in each comparison groups are summarized in Supplemental Table S1.
(A) Effect of linezolid and vancomycin on host, (B) Effect of linezolid and vancomycin on host with S. aureus infection, and (C) Effect of S. aureus infection on host immune response. Venn diagrams indicate the number of genes unique to each comparison pair and the number of genes common to each comparison pair (overlapped).
Effect of Antibiotic on Pathways of Differentially Expressed Genes
To correlate the significance of differentially expressed genes in terms of its biological relevance, we next studied pathway associations for the differentially expressed genes in each comparison group in mice using Kyoto Encyclopedia of Genes and Genomes (KEGG). Pathways significantly associated with the differentially expressed genes are summarized in Table 1. Toll-like receptor signaling pathway (Atm, Ccl3, Cd14, Cxcl10, Cxcl9, Ikbkb, Il6, Jun, Lbp, Ly96, Mapk11, Mapk13, Myd88, Nfkbia, Stat 1, Tirap, Tlr2, and Tlr4) was the most frequently identified pathway common to all comparison groups studied other than Control L vs. Control V group. MAPK signaling pathway (Casp1, Casp11, Casp3, Casp7, Cd14, Dusp1, Dusp14, Fas, Fgf13, Gadd45g, Gng12, Ikbkb, Il1r2, Jun, Map3k1, Map3k4, Map3k8, Map4k1, Map4k2, Mapk11, Mapk13, Mef2c, Pak1, Ppp3cc, Prkacb, Prkcb1, Rapgef4, Rasgrp1, Rasgrp4, Rps6ka5, Rras2, Tgfbr2, and Tnfrsf1a) was common between Control vs. 24 h S. aureus infection and 24 h S. aureus infection vs. 24 h S. aureus L comparisons.
When looked into further details for the genes involved in immune modulation, genes like Cxcl9, Cxcl10 and Il1r2 were differentially expressed between 24 h S. aureus L vs. 24 h S. aureus V. Cxcl9 and Cxcl10 were up-regulated in 24 h S. aureus L and Il1r2 was down-regulated. Also Cd14, Nfkbia, and Il1r2 were differentially expressed between 24 h S. aureus vs. 24 h S. aureus L and all three genes were down-regulated in 24 h S. aureus L. In addition, Nfkbia was down-regulated in 24 h S. aureus V with a significantly different expression between 24 h S. aureus and 24 h S. aureus V (Table 1 and Supplemental Table S1).
Glycerolipid metabolism pathway was significantly associated with 24 h S. aureus infection vs. 24 h S. aureus L comparison group while Huntington’s disease pathway was only associated with 24 h S. aureus infection vs. 24 h S. aureus V comparison. There was no overlap in pathways within the Control L vs. Control V groups. Similarly, comparisons involving S. aureus infection (24 h S. aureus L vs. 24 h S. aureus V, 24 h S. aureus vs. 24 h S. aureus L, 24 h S. aureus vs. 24 h S. aureus V and Control vs. 24 h S. aureus) also yielded no common gene expression pathways.
Discussion
Despite several recent studies of antibacterial agents possessing immunomodulatory effects [20], [23], [38]–[45], the precise mechanism and effects of these antibiotics in host gene expression and immunomodulation in MRSA infection is unknown. This present study sought to understand the impact of L on the host-pathogen interaction during MRSA infection in comparison with V. Using a murine model of S. aureus infection, we determined that L and V induced differential production of bacterial toxins and host cytokines, differences in host gene expression, and differences in immunomodulators during MRSA bloodstream infection.
Various studies have shown that L suppresses toxin production, whereas the cell wall active antibiotics had either no effect or enhanced the production of extracellular virulence factors [18], [21]–[23], [46]–[48]. For example, Stevens et al [49] showed that L significantly reduced production of TSST even in the setting of high bacterial counts. The results of our in vivo study substantiate these previous findings showing more pronounced effect of L in in vivo production of PVL and alpha hemolysin compared to V. Collectively, these data provide the evidence that protein-synthesis inhibitors such as L that act on a ribosomal level are more effective toxin suppressors and could be more effective in treating MRSA sepsis.
Antibiotic including L and V are known to have immunomodulatory effects during S. aureus infection [20], [23], [38]–[45]. Consistent with previously reported findings [45], [50]–[54], treatment with both L and V reduced proinflammatory cytokine production in the serum of S. aureus infected mice, albeit by different mechanisms (Figure 3). Evidence from previous studies suggests that the network of inflammatory cytokines and chemokines play a major role in mediating, amplifying, and perpetuating inflammation [55], [56]. Findings from two recent studies have suggested that PVL treatment of neutrophils results in elevated gene expression and protein production of proinflammatory cytokines including IL-6, IL-8 and TNF-a [7], [8]. It has also been shown that L inhibits MRSA growth and suppresses the production of toxins in vitro [18], [19] and potentiates the susceptibility of S. aureus to human neutrophils [20]. Thus the pronounced effect of L on mice could be due to its two-pronged effect –1) direct effect of L in protein synthesis of bacterial toxins including PVL; and 2) indirect effect on toxin-mediated cytokine production during S. aureus infection. Although the role of PVL in S. aureus infections is highly controversial, there is evidence that PVL is associated with severe disease in community acquired pneumonia (CAP) caused by S. aureus both in clinical reports [9], [10] and in some [11], [12], but not all [57]–[60], in vivo model systems. Taken together, these findings suggest that L’s effect on S. aureus infected patients is complex and due at least in part to its effect on both bacterial protein and host cytokine production.
The murine model of host gene expression profiling has been widely used to assess the host pathogen interaction [27], [28] as well as in non-infectious conditions like radiation exposure and breast cancer [29]–[31]. In this study, we have described the drug-specific gene expression profiles in mice by comparing the response of S. aureus infection with and without L or V treatment. In consistent with previously reported findings [61], [62], we found that cytokine-cytokine receptor interactions is an important pathway in the murine response to S. aureus infection (Table 1). Similar gene expression-based analyses of the human response to bacterial infection have also revealed the importance of cytokine-cytokine receptor interactions; TLR signaling; MAPK signaling; Jak-STAT signaling; focal adhesion; and complement and coagulation cascades [63]–[65], three of which (Cytokine-cytokine receptor interactions, TLR signaling and MAPK signaling) were significantly associated with S. aureus infection in this study.
We have noticed that Nfkbia, a hallmark of the inflammatory response [66] was down-regulated in both treatment groups (24 h S. aureus L and 24 h S. aureus V) compared to untreated controls (Table 1 and Supplemental Table S1). This is in agreement with the fact that both L and V treatment reduce inflammation compared to untreated controls. Interestingly, in 24 h S. aureus L treatment, Cxcl9 and Cxcl10 were up-regulated whereas Il1r2 was down-regulated compared to 24 h S. aureus V. Moreover, Cd14 and Il1r2 were down-regulated in 24 h S. aureus L compared to untreated controls (Table 1 and Supplemental Table S1).The chemokines CXCL9 and CXCL10 are known to promote protective immune response during infections [67]. Recently, CD14 inhibition is known to efficiently attenuate early inflammatory response in bacterial sepsis [68]. IL1R2 is the well-known receptor of pro-inflammatory cytokines IL-1 alpha and IL-1 beta [69]. Taken together, as additional genes involved in inflammatory cascade (e.g. Cxcl9, Cxcl10, Il1r2 and Cd14 as discussed above) are differentially expressed with 24 h S. aureus L treatment, this could have a significant effect in reducing inflammation by L treatment in comparison to V treatment and untreated controls.
The glycerolipid metabolism pathway was strikingly reported only with L treatment comparison. Recent studies have shown that the disturbances in lipid metabolism, particularly those involving the components of GL/FFA cycling, are strongly associated with diseases related to the metabolic syndrome, inflammation and the pathogenesis of some cancers [70]–[72]. Also, many agents interfering with glycerolipid metabolism have been successfully used as additives or to treat several infections including toxic shock syndrome caused by S. aureus [73]–[76]. Thus, the involvement of glycerolipid pathway only with L treatment could have significance with its unique mode of action in treating staphylococcal infections. However, the exact significance of L treatment associated with glycerolipid pathway requires further elucidation.
The current study increases our understanding of the interplay between bacterial virulence and compensatory host response, and the impact of antibiotic treatment on this interplay. The in vivo evaluation of bacterial pathogenesis and host response is novel. The findings of this study help to unravel the complex relationship between host and pathogen in MRSA infection. Limitations include the fact that this study looked only at alpha hemolysin and PVL. We also studied only few cytokines based on the reported involvement of L and V on these cytokines [44], [45]. However, several other cytokines, both pro- and anti-inflammatory cytokines, might be associated with blood stream infection and we were not able to include all these cytokines in this study. For example, the serum level of an anti-inflammatory cytokine IL-10 is shown to be associated with mortality in severe sepsis [77] and recently in S. aureus bacteremia patients [78]. This warrants for the expansion of the cytokine panel in similar future studies. Based on the fact that the bacteria are demonstrable in murine blood stream within 2 hour post i.p. infection [35], we administered antibiotics 2 hour post infection. However, we recognize the fact that changing the time of introducing the antibiotic therapy could have a significant effect on cytokine and gene expression profile as is shown for its effect in mortality in patients with sepsis by Kumar A et al [79].
In summary, this study has provided the evidence that protein synthesis inhibitors like L could be superior in treating MRSA blood stream infections than the cell wall acting antibiotics. The superiority of protein synthesis inhibitors could be partly attributed to its better effect on lowering down the in vivo level of important staphylococcal toxins like alpha hemolysin and PVL as well as to its better immunomodulatory effects.
Supporting Information
Table S1.
Genes differentially expressed in each comparison.
https://doi.org/10.1371/journal.pone.0060463.s001
(XLSX)
Acknowledgments
S. aureus USA300 isolate was obtained through the NARSA program supported under NIAID/NIH Contract # HHSN272200700055C. We would like to acknowledge Dr. Lawrence Park, Duke University Medical Center for his statistical review of this manuscript.
Author Contributions
Conceived and designed the experiments: BKSK SHA VGF. Performed the experiments: BKSK SHA. Analyzed the data: BKSK SHA. Contributed reagents/materials/analysis tools: BKSK YZ QY SHA. Wrote the paper: BKSK SHA VGF.
References
- 1. David MZ, Daum RS (2010) Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic. Clin Microbiol Rev 23: 616–687.
- 2. Lowy FD (1998) Staphylococcus aureus infections. N Engl J Med 339: 520–532.
- 3. Fournier B, Philpott DJ (2005) Recognition of Staphylococcus aureus by the innate immune system. Clin Microbiol Rev 18: 521–540.
- 4. Gould IM, David MZ, Esposito S, Garau J, Lina G, et al. (2012) New insights into meticillin-resistant Staphylococcus aureus (MRSA) pathogenesis, treatment and resistance. Int J Antimicrob Agents 39: 96–104.
- 5. Kobayashi SD, DeLeo FR (2009) An update on community-associated MRSA virulence. Curr Opin Pharmacol 9: 545–551.
- 6. Otto M (2010) Basis of virulence in community-associated methicillin-resistant Staphylococcus aureus. Annu Rev Microbiol 64: 143–162.
- 7. Ma X, Chang W, Zhang C, Zhou X, Yu F (2012) Staphylococcal panton-valentine leukocidin induces pro-inflammatory cytokine production and nuclear factor-kappa B activation in neutrophils. PLoS ONE 7: e34970.
- 8. Tseng CW, Kyme P, Low J, Rocha MA, Alsabeh R, et al. (2009) Staphylococcus aureus Panton-Valentine leukocidin contributes to inflammation and muscle tissue injury. PLoS ONE 4: e6387.
- 9. Gillet Y, Issartel B, Vanhems P, Fournet JC, Lina G, et al. (2002) Association between Staphylococcus aureus strains carrying gene for Panton-Valentine leukocidin and highly lethal necrotising pneumonia in young immunocompetent patients. Lancet 359: 753–759.
- 10. Francis JS, Doherty MC, Lopatin U, Johnston CP, Sinha G, et al. (2005) Severe community-onset pneumonia in healthy adults caused by methicillin-resistant Staphylococcus aureus carrying the Panton-Valentine leukocidin genes. Clin Infect Dis 40: 100–107.
- 11. Labandeira-Rey M, Couzon F, Boisset S, Brown EL, Bes M, et al. (2007) Staphylococcus aureus Panton-Valentine leukocidin causes necrotizing pneumonia. Science 315: 1130–1133.
- 12. Diep BA, Chan L, Tattevin P, Kajikawa O, Martin TR, et al. (2010) Polymorphonuclear leukocytes mediate Staphylococcus aureus Panton-Valentine leukocidin-induced lung inflammation and injury. Proc Natl Acad Sci U S A 107: 5587–5592.
- 13. Miller LG, Perdreau-Remington F, Rieg G, Mehdi S, Perlroth J, et al. (2005) Necrotizing fasciitis caused by community-associated methicillin-resistant Staphylococcus aureus in Los Angeles. N Engl J Med 352: 1445–1453.
- 14. Cremieux AC, Dumitrescu O, Lina G, Vallee C, Cote JF, et al. (2009) Panton-valentine leukocidin enhances the severity of community-associated methicillin-resistant Staphylococcus aureus rabbit osteomyelitis. PLoS ONE 4: e7204.
- 15. Stevens DL, Gibbons AE, Bergstrom R, Winn V (1988) The Eagle effect revisited: efficacy of clindamycin, erythromycin, and penicillin in the treatment of streptococcal myositis. J Infect Dis 158: 23–28.
- 16. Stevens DL, Maier KA, Laine BM, Mitten JE (1987) Comparison of clindamycin, rifampin, tetracycline, metronidazole, and penicillin for efficacy in prevention of experimental gas gangrene due to Clostridium perfringens. J Infect Dis 155: 220–228.
- 17. Zimbelman J, Palmer A, Todd J (1999) Improved outcome of clindamycin compared with beta-lactam antibiotic treatment for invasive Streptococcus pyogenes infection. Pediatr Infect Dis J 18: 1096–1100.
- 18. Stevens DL, Ma Y, Salmi DB, McIndoo E, Wallace RJ, et al. (2007) Impact of antibiotics on expression of virulence-associated exotoxin genes in methicillin-sensitive and methicillin-resistant Staphylococcus aureus. J Infect Dis 195: 202–211.
- 19. De Gascun C, Rajan L, O’Neill E, Smyth EG (2006) Linezolid use in sepsis due to methicillin-susceptible Staphylococcus aureus. J Antimicrob Chemother 57: 150–151.
- 20. Gemmell CG, Ford CW (2002) Virulence factor expression by Gram-positive cocci exposed to subinhibitory concentrations of linezolid. J Antimicrob Chemother 50: 665–672.
- 21. Doss SA, Tillotson GS, Amyes SG (1993) Effect of sub-inhibitory concentrations of antibiotics on the virulence of Staphylococcus aureus. J Appl Bacteriol 75: 123–128.
- 22. Dumitrescu O, Boisset S, Badiou C, Bes M, Benito Y, et al. (2007) Effect of antibiotics on Staphylococcus aureus producing Panton-Valentine leukocidin. Antimicrob Agents Chemother 51: 1515–1519.
- 23. Bernardo K, Pakulat N, Fleer S, Schnaith A, Utermohlen O, et al. (2004) Subinhibitory concentrations of linezolid reduce Staphylococcus aureus virulence factor expression. Antimicrob Agents Chemother 48: 546–555.
- 24. Gill CJ, Abruzzo GK, Flattery AM, Misura AS, Bartizal K, et al. (2007) In vivo efficacy of a novel oxazolidinone compound in two mouse models of infection. Antimicrob Agents Chemother 51: 3434–3436.
- 25. LaPlante KL, Leonard SN, Andes DR, Craig WA, Rybak MJ (2008) Activities of clindamycin, daptomycin, doxycycline, linezolid, trimethoprim-sulfamethoxazole, and vancomycin against community-associated methicillin-resistant Staphylococcus aureus with inducible clindamycin resistance in murine thigh infection and in vitro pharmacodynamic models. Antimicrob Agents Chemother 52: 2156–2162.
- 26. Pultz NJ, Stiefel U, Donskey CJ (2005) Effects of daptomycin, linezolid, and vancomycin on establishment of intestinal colonization with vancomycin-resistant enterococci and extended-spectrum-beta-lactamase-producing Klebsiella pneumoniae in mice. Antimicrob Agents Chemother 49: 3513–3516.
- 27.
Ahn SH, Deshmukh H, Johnson N, Cowell LG, Rude TH, et al.. (2010) Two Genes on A/J Chromosome 18 Are Associated with Susceptibility to Staphylococcus aureus Infection by Combined Microarray and QTL Analyses. Plos Pathogens 6.
- 28. von Kockritz-Blickwede M, Rohde M, Oehmcke S, Miller LS, Cheung AL, et al. (2008) Immunological mechanisms underlying the genetic predisposition to severe Staphylococcus aureus infection in the mouse model. Am J Pathol 173: 1657–1668.
- 29. Dressman HK, Muramoto GG, Chao NJ, Meadows S, Marshall D, et al. (2007) Gene expression signatures that predict radiation exposure in mice and humans. PLoS Med 4: e106.
- 30. He M, Mangiameli DP, Kachala S, Hunter K, Gillespie J, et al. (2010) Expression signature developed from a complex series of mouse models accurately predicts human breast cancer survival. Clin Cancer Res 16: 249–259.
- 31. LaBreche HG, Nevins JR, Huang E (2011) Integrating factor analysis and a transgenic mouse model to reveal a peripheral blood predictor of breast tumors. BMC Med Genomics 4: 61.
- 32. Rice KC, Firek BA, Nelson JB, Yang SJ, Patton TG, et al. (2003) The Staphylococcus aureus cidAB operon: evaluation of its role in regulation of murein hydrolase activity and penicillin tolerance. J Bacteriol 185: 2635–2643.
- 33. Miyazaki S, Fujikawa T, Kobayashi I, Matsumoto T, Tateda K, et al. (2002) The in vitro and in vivo antibacterial characterization of vancomycin and linezolid against vancomycin-susceptible and -resistant enterococci. J Antimicrob Chemother 50: 971–974.
- 34. Patel R, Piper KE, Rouse MS, Steckelberg JM (2000) Linezolid therapy of Staphylococcus aureus experimental osteomyelitis. Antimicrob Agents Chemother 44: 3438–3440.
- 35. Thakker M, Park JS, Carey V, Lee JC (1998) Staphylococcus aureus serotype 5 capsular polysaccharide is antiphagocytic and enhances bacterial virulence in a murine bacteremia model. Infect Immun 66: 5183–5189.
- 36. Ahn SH, Han JH, Lee JH, Park KJ, Kong IS (2005) Identification of an iron-regulated hemin-binding outer membrane protein, HupO, in Vibrio fluvialis: effects on hemolytic activity and the oxidative stress response. Infect Immun 73: 722–729.
- 37. Chang JT, Nevins JR (2006) GATHER: a systems approach to interpreting genomic signatures. Bioinformatics 22: 2926–2933.
- 38. Dalhoff A, Shalit I (2003) Immunomodulatory effects of quinolones. Lancet Infect Dis 3: 359–371.
- 39. Desaki M, Okazaki H, Sunazuka T, Omura S, Yamamoto K, et al. (2004) Molecular mechanisms of anti-inflammatory action of erythromycin in human bronchial epithelial cells: possible role in the signaling pathway that regulates nuclear factor-kappaB activation. Antimicrob Agents Chemother 48: 1581–1585.
- 40. Giamarellos-Bourboulis EJ (2008) Macrolides beyond the conventional antimicrobials: a class of potent immunomodulators. Int J Antimicrob Agents 31: 12–20.
- 41. Kanoh S, Rubin BK (2010) Mechanisms of action and clinical application of macrolides as immunomodulatory medications. Clin Microbiol Rev 23: 590–615.
- 42. Shinkai M, Henke MO, Rubin BK (2008) Macrolide antibiotics as immunomodulatory medications: proposed mechanisms of action. Pharmacol Ther 117: 393–405.
- 43. Tateda K, Ishii Y, Matsumoto T, Kobayashi T, Miyazaki S, et al. (2000) Potential of macrolide antibiotics to inhibit protein synthesis of Pseudomonas aeruginosa: suppression of virulence factors and stress response. J Infect Chemother 6: 1–7.
- 44. Coyle EA, Cha R, Rybak MJ (2003) Influences of linezolid, penicillin, and clindamycin, alone and in combination, on streptococcal pyrogenic exotoxin a release. Antimicrob Agents Chemother 47: 1752–1755.
- 45. Garcia-Roca P, Mancilla-Ramirez J, Santos-Segura A, Fernandez-Aviles M, Calderon-Jaimes E (2006) Linezolid diminishes inflammatory cytokine production from human peripheral blood mononuclear cells. Arch Med Res 37: 31–35.
- 46. Kernodle DS, McGraw PA, Barg NL, Menzies BE, Voladri RK, et al. (1995) Growth of Staphylococcus aureus with nafcillin in vitro induces alpha-toxin production and increases the lethal activity of sterile broth filtrates in a murine model. J Infect Dis 172: 410–419.
- 47. Libert N, Borne M, Janvier F, Batjom E, Cirodde A, et al. (2009) [Successful treatment of life-threatening Panton-Valentine leucocidin positive Staphylococcus aureus pneumonia with antibiotics and immunoglobulins targeting the toxin production]. Rev Med Interne 30: 907–910.
- 48. Ohlsen K, Ziebuhr W, Koller KP, Hell W, Wichelhaus TA, et al. (1998) Effects of subinhibitory concentrations of antibiotics on alpha-toxin (hla) gene expression of methicillin-sensitive and methicillin-resistant Staphylococcus aureus isolates. Antimicrob Agents Chemother 42: 2817–2823.
- 49. Stevens DL, Wallace RJ, Hamilton SM, Bryant AE (2006) Successful treatment of staphylococcal toxic shock syndrome with linezolid: a case report and in vitro evaluation of the production of toxic shock syndrome toxin type 1 in the presence of antibiotics. Clin Infect Dis 42: 729–730.
- 50. Danin J, Linder L, Lundqvist G, Wretlind B (2003) Cytokines in periradicular lesions: the effect of linezolid treatment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 96: 492–498.
- 51. Krehmeier U, Bardenheuer M, Voggenreiter G, Obertacke U, Schade FU, et al. (2002) Effects of antimicrobial agents on spontaneous and endotoxin-induced cytokine release of human peripheral blood mononuclear cells. J Infect Chemother 8: 194–197.
- 52. Laborada G, Cruz F, Nesin M (2005) Serial cytokine profiles in shunt-related ventriculitis treated with intraventricular vancomycin. Chemotherapy 51: 363–365.
- 53. Takahashi G, Sato N, Yaegashi Y, Kojika M, Matsumoto N, et al. (2010) Effect of linezolid on cytokine production capacity and plasma endotoxin levels in response to lipopolysaccharide stimulation of whole blood. J Infect Chemother 16: 94–99.
- 54. Yoshizawa S, Tateda K, Saga T, Ishii Y, Yamaguchi K (2012) Virulence-suppressing effects of linezolid on methicillin-resistant Staphylococcus aureus: possible contribution to early defervescence. Antimicrob Agents Chemother 56: 1744–1748.
- 55. Goodman RB, Pugin J, Lee JS, Matthay MA (2003) Cytokine-mediated inflammation in acute lung injury. Cytokine Growth Factor Rev 14: 523–535.
- 56. Ware LB, Matthay MA (2000) The acute respiratory distress syndrome. N Engl J Med 342: 1334–1349.
- 57. Bubeck Wardenburg J, Palazzolo-Ballance AM, Otto M, Schneewind O, DeLeo FR (2008) Panton-Valentine leukocidin is not a virulence determinant in murine models of community-associated methicillin-resistant Staphylococcus aureus disease. J Infect Dis 198: 1166–1170.
- 58. Kobayashi SD, Malachowa N, Whitney AR, Braughton KR, Gardner DJ, et al. (2011) Comparative Analysis of USA300 Virulence Determinants in a Rabbit Model of Skin and Soft Tissue Infection. J Infect Dis 204: 937–941.
- 59. Olsen RJ, Kobayashi SD, Ayeras AA, Ashraf M, Graves SF, et al. (2010) Lack of a major role of Staphylococcus aureus Panton-Valentine leukocidin in lower respiratory tract infection in nonhuman primates. Am J Pathol 176: 1346–1354.
- 60. Voyich JM, Otto M, Mathema B, Braughton KR, Whitney AR, et al. (2006) Is Panton-Valentine leukocidin the major virulence determinant in community-associated methicillin-resistant Staphylococcus aureus disease? J Infect Dis 194: 1761–1770.
- 61. Chin CY, Monack DM, Nathan S (2010) Genome wide transcriptome profiling of a murine acute melioidosis model reveals new insights into how Burkholderia pseudomallei overcomes host innate immunity. BMC Genomics 11: 672.
- 62. Yu SL, Chen HW, Yang PC, Peck K, Tsai MH, et al. (2004) Differential gene expression in gram-negative and gram-positive sepsis. Am J Respir Crit Care Med 169: 1135–1143.
- 63. Ardura MI, Banchereau R, Mejias A, Di Pucchio T, Glaser C, et al. (2009) Enhanced monocyte response and decreased central memory T cells in children with invasive Staphylococcus aureus infections. PLoS ONE 4: e5446.
- 64. Pankla R, Buddhisa S, Berry M, Blankenship DM, Bancroft GJ, et al. (2009) Genomic transcriptional profiling identifies a candidate blood biomarker signature for the diagnosis of septicemic melioidosis. Genome Biol 10: R127.
- 65. Ramilo O, Allman W, Chung W, Mejias A, Ardura M, et al. (2007) Gene expression patterns in blood leukocytes discriminate patients with acute infections. Blood 109: 2066–2077.
- 66. Karin M, Greten FR (2005) NF-kappaB: linking inflammation and immunity to cancer development and progression. Nat Rev Immunol 5: 749–759.
- 67. Hardison JL, Wrightsman RA, Carpenter PM, Lane TE, Manning JE (2006) The chemokines CXCL9 and CXCL10 promote a protective immune response but do not contribute to cardiac inflammation following infection with Trypanosoma cruzi. Infect Immun 74: 125–134.
- 68. Thorgersen EB, Hellerud BC, Nielsen EW, Barratt-Due A, Fure H, et al. (2010) CD14 inhibition efficiently attenuates early inflammatory and hemostatic responses in Escherichia coli sepsis in pigs. FASEB J 24: 712–722.
- 69. McMahan CJ, Slack JL, Mosley B, Cosman D, Lupton SD, et al. (1991) A novel IL-1 receptor, cloned from B cells by mammalian expression, is expressed in many cell types. Embo J 10: 2821–2832.
- 70. Muoio DM, Newgard CB (2006) Obesity-related derangements in metabolic regulation. Annu Rev Biochem 75: 367–401.
- 71. Prentki M, Madiraju SR (2008) Glycerolipid metabolism and signaling in health and disease. Endocr Rev 29: 647–676.
- 72. Swinnen JV, Brusselmans K, Verhoeven G (2006) Increased lipogenesis in cancer cells: new players, novel targets. Curr Opin Clin Nutr Metab Care 9: 358–365.
- 73. Brissette JL, Cabacungan EA, Pieringer RA (1986) Studies on the antibacterial activity of dodecylglycerol. Its limited metabolism and inhibition of glycerolipid and lipoteichoic acid biosynthesis in Streptococcus mutans BHT. J Biol Chem 261: 6338–6345.
- 74. Haynes MP, Buckley HR, Higgins ML, Pieringer RA (1994) Synergism between the antifungal agents amphotericin B and alkyl glycerol ethers. Antimicrob Agents Chemother 38: 1523–1529.
- 75. Lin YC, Schlievert PM, Anderson MJ, Fair CL, Schaefers MM, et al. (2009) Glycerol monolaurate and dodecylglycerol effects on Staphylococcus aureus and toxic shock syndrome toxin-1 in vitro and in vivo. PLoS ONE 4: e7499.
- 76. Ved HS, Gustow E, Mahadevan V, Pieringer RA (1984) Dodecylglycerol. A new type of antibacterial agent which stimulates autolysin activity in Streptococcus faecium ATCC 9790. J Biol Chem 259: 8115–8121.
- 77. Gogos CA, Drosou E, Bassaris HP, Skoutelis A (2000) Pro- versus anti-inflammatory cytokine profile in patients with severe sepsis: a marker for prognosis and future therapeutic options. J Infect Dis 181: 176–180.
- 78. Rose WE, Eickhoff JC, Shukla SK, Pantrangi M, Rooijakkers S, et al. (2012) Elevated serum interleukin-10 at time of hospital admission is predictive of mortality in patients with Staphylococcus aureus bacteremia. J Infect Dis 206: 1604–1611.
- 79. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, et al. (2006) Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 34: 1589–1596.