Deficiency in non-classical major histocompatibility class II-like molecule, H2-O confers protection against Staphylococcus aureus in mice

Staphylococcus aureus is a human-adapted pathogen that replicates by asymptomatically colonizing its host. S. aureus is also the causative agent of purulent skin and soft tissue infections as well as bloodstream infections that result in the metastatic seeding of abscess lesions in all organ tissues. Prolonged colonization, infection, disease relapse, and recurrence point to the versatile capacity of S. aureus to bypass innate and adaptive immune defenses as well as the notion that some hosts fail to generate protective immune responses. Here, we find a genetic trait that provides protection against this pathogen. Mice lacking functional H2-O, the equivalent of human HLA-DO, inoculated with a mouse-adapted strain of S. aureus, efficiently decolonize the pathogen. Further, these decolonized animals resist subsequent bloodstream challenge with methicillin-resistant S. aureus. A genetic approach demonstrates that T-cell dependent B cell responses are required to control S. aureus colonization and infection in H2-O-deficient mice. Reduced bacterial burdens in these animals correlate with increased titers and enhanced phagocytic activity of S. aureus-specific antibodies. H2-O negatively regulates the loading of high affinity peptides on major histocompatibility class II (MHC-II) molecules. Thus, we hypothesize that immune responses against S. aureus are derepressed in mice lacking H2-O because more high affinity peptides are presented by MHC-II. We speculate that loss-of-function HLA-DO alleles may similarly control S. aureus replication in humans.


Introduction
Pathogens evolve successful strategies to evade defense mechanisms of their hosts.Staphylococcus aureus is a prime example of such a pathogen: it colonizes the human nares and skin, causes frequent skin, soft tissue, and bloodstream infections [1,2], and is adaptable to other hosts [3,4].While colonization is innocuous and occurs at birth [5,6], it represents the highest risk for infection as disease isolates are typically the same as those cultured from the skin and nares of the host [7][8][9][10].Despite the elicitation of S. aureus-specific antibodies, colonization persists and most studies point to the notion that these antibodies do not protect against invasive disease [8,[10][11][12][13][14].A number of reasons may account for this lack of protection: (i) S. aureus has evolved many mechanisms to escape the activity of antibodies [2]; (ii) the constant exposure to S. aureus may promote isotype switching and changes in glycosylation patterns of antibodies that preempt their engagement in adequate effector activities [14,15]; (iii) constant exposure may select for the presentation of non-protective epitopes of immunodominant antigens [16][17][18]; (iv) B and T cell superantigens (SAgs) secreted by S. aureus may skew the specificity of antibodies and T cells [19][20][21].
Several observations suggest that susceptibility to S. aureus is influenced by the genetic make-up of the host.For example, the major histocompatibility complex class II (MHC-II) locus has been proposed to act as a genetic determinant of susceptibility to S. aureus infection in humans [22,23].This may in part be accounted for by the function of T cell-specific SAgs, which have preferential binding for specific MHC-II alleles as nasal carriers inherit specific HLA class II alleles [24].However, there are over 100 genes mapped within the MHC locus in addition to MHC-II genes and all are polymorphic with half participating in immune responses.Because of the low frequency of recombination within the MHC region [25], the direct identification of genes that may modulate host-S.aureus interactions is challenging.
Previously we found that mice from the I/LnJ strain produce neutralizing antibodies against mouse mammary tumor virus (MMTV) [26,27].These neutralizing antibodies, generated in a CD4 T cell-dependent fashion [28], coat viral particles rendering MMTV non-infectious [26,27].This mechanism of resistance has been explained by the loss of H2-Ob (Ob) function and is not a peculiarity of I/LnJ animals, as MMTV-susceptible C57BL/6J (B6J) mice rendered Ob-deficient also produced virus-neutralizing antibodies [29].Ob encodes the beta subunit (Oβ) of the non-classical MHC-II-like αβ heterodimer (H2-O in mice and HLA-DO or DO in humans) and is expressed in B cells, dendritic and thymic epithelial cells [29][30][31][32][33].In humans, the non-classical αβ heterodimer is encoded by genes HLA-DOA (DOA) and HLA-DOB (DOB).Specific alleles of DOA and DOB with reduced-or loss-of-functions have been linked to resolution of infection with hepatitis B and C viruses [29,34] which require the early presence of broadly neutralizing antibodies [35][36][37].The elicitation of neutralizing antibodies in the absence of H2-O has been explained by the uninhibited function of H2-M [29].H2-M (HLA/DM or DM) catalyze peptide loading of MHC-II molecules in late endosomes and lysosomes of antigen presenting cells by replacing the MHC-II-associated invariant chain peptides with high-affinity, pathogen-derived peptides [38][39][40][41][42][43][44][45].As H2-O inhibits H2-M function, its expression leads to a small but significant decrease of high affinity peptides presented by MHC-II molecules [43].Conversely, the lack of H2-O/DO-mediated inhibition of H2-M/DM results in increased numbers of high affinity peptides presented by MHC-II molecules [43].
Since both the Oα and Oβ chains of H2-O/DO are encoded within the MHC-II locus, we wondered whether the loss of functional alleles influences the susceptibility or resistance toward S. aureus.Using a mouse model of colonization, we observed that animals with deficiency in H2-O cleared S. aureus.Successful decolonization required an intact T cell-dependent B cell response and directly correlated with increased titers of S. aureus-specific antibodies.Furthermore, H2-O-deficient mice also cleared the mouse common commensal, Staphylococcus xylosus.

Rapid S. aureus decolonization is observed in H2-O-deficient animals
Recently, we developed a model of S. aureus nasopharyngeal colonization in mice to better understand bacterial and host factors that contribute to persistent host-bacterial interactions [3].This model takes advantage of the mouse-adapted WU1 variant isolated from a mouse colony at Washington University [3].When the variant is inoculated intranasally, wild type B6J mice remain colonized as monitored by plating nasopharyngeal swabs and fecal material on mannitol salt agar in weekly intervals [3].This model is unlike other models that require pretreatment of animals with antibiotics for the transient colonization with human nosocomial or clinical isolates of S. aureus.B6J mice remain persistently colonized with WU1 provided that the bacterium produces Staphylococcal protein A (SpA), a factor that alters B cell responses [3,19,20].
To test whether the MHC-II antigen presentation pathway and specifically H2-O play a role in controlling S. aureus, B6J.Ob +/+ and B6J.Ob -/-mice were colonized with WU1 and bacterial loads in throat and feces were recorded weekly over the course of 7 weeks.Mice were scored as decolonized when both nasopharyngeal swabs and fecal matters were free of S. aureus following plating on mannitol salt agar.Both B6J.Ob +/+ and B6J.Ob -/-animals carried similar loads of bacteria one week post inoculation; however, while B6J.Ob +/+ mice remained colonized, Ob -/- animals cleared the bacteria within 5 weeks (Fig 1A ).A rifampicin-resistant WU1 variant (WU1 Rif ) that carries a point mutation in rpoB, encoding the β subunit of RNA polymerase [46], was also used for colonization (Fig 1B).This was performed because plating on mannitol salt agar revealed the presence of both pigmented S. aureus and a non-pigmented Gram-positive organism which was subsequently identified as S. xylosus (see below).As with strain WU1, B6J.Ob -/-cleared WU1 Rif in approximately 5 weeks while B6J.Ob +/+ animals remained colonized (Fig 1B).To examine whether loss of H2-O promotes decolonization of S. aureus independently of the mouse genetic background, BALB/cJ H2-O-deficient (BALB.Ob -/-) animals were also used.BALB.Ob +/+ mice were stably colonized with WU1 Rif throughout the duration of the experiment (Fig 2).Similar to B6J.Ob -/-mice, more than 70% of BALB.Ob -/-animals cleared S. aureus within 5 weeks post colonization (Fig 2).Therefore, we concluded that H2-O deficiency correlated with S. aureus decolonization in animals irrespectively of the genetic background.The notion that, when observed, clearance occurred approximately 5 weeks post colonization agrees with a role of H2-O within the MHC-II antigen presentation pathway specifically, the activation of adaptive immune responses.

PLOS PATHOGENS
S. aureus protective responses in absence of H2-O

H2-O-deficient animals clear S. xylosus
As noted above, plating on mannitol salt agar revealed the presence of non-pigmented colonies that acidified the medium as a result of metabolic activity.To rigorously establish the presence and identity of this bacterium, fecal samples of approximately 60-day old naïve B6J and BALB/ c mice were plated on mannitol salt agar (Fig 3A and 3C).Colonies were observed in fecal samples of all wild type animals but only in two of fifteen B6J.Ob -/-and none of ten BALB.Ob -/- mice (Fig 3A and 3C).All colonies looked identical on plate.Forty-five colonies were selected from 4 different cages per mouse strain and PCR products were obtained using 16S RNA specific-primers [47].DNA sequencing identified a single species, S. xylosus.
To ensure the H2-O-deficient mice could efficiently decolonize S. xylosus when they are exposed to the microbe, we foster-nursed newborn Ob -/-mice by Ob +/+ mothers of the same background (Fig 3B and 3D).Fecal contents of fostered animals were monitored for the presence of S. xylosus.Bacterial counts decreased significantly in Ob -/-mice of both backgrounds but not in Ob +/+ mice over time (Fig 3B and 3D).Decolonization became apparent at approximately 7 weeks of age.Thus, H2-O-deficient mice became colonized with both S. aureus and S. xylosus but successfully cleared these species in a manner coinciding with the development of T-cell-dependent Ab responses in mice [48,49].

Clearance of S. aureus in Ob -/-mice requires B and T cell functions
Since H2-O functions within the MHC-II antigen presentation pathway, we reasoned that control of staphylococcal species in H2-O-deficient mice is mediated via the adaptive immune response.To address this possibility, Ob -/-and Ob +/+ mice that either lack T cells or are monoclonal for the B cell receptor (BCR) were generated (Fig 4A).Briefly, to produce T-cell deficient mice, crosses were performed between B6.Tcrβ -/-, B6.Tcrδ -/-, B6.Ob -/-and wild type (B6.Ob +/+ ) animals to obtain B6.Tcrβδ -/-(which lack both αβ and γδ T cells) with and without Ob.The absence of T cells does not affect H2-O function as H2-O is expressed only in B, dendritic and thymic epithelial cells [30][31][32][33].To produce BCR monoclonal Ob -/-and Ob +/+ , the B6.MD4t transgenic mice, which carry a transgene for the rearranged heavy and light chains of a BCR recognizing hen egg lysozyme [50], were crossed to Ob -/-mice and subsequently to mice deficient for the IgH J segment locus (J H ) [51].The resulting B6.J H -/-MD4tOb -/-and B6.

Anti-S. aureus responses in H2-O-deficient mice are broadly protectives
Next, we sought to determine whether WU1 decolonization in Ob -/-mice correlated with increased pathogen-specific antibody titers.Sera from WU1-colonized B6J.Ob +/+ and B6J.Ob -/-mice from the experiment shown in Fig 1A , were analyzed for anti-S.aureus IgGs in ELISA using cellular extracts of a strain lacking spa and sbi (ΔspaΔsbi) to avoid non-specific binding with antibodies [52].This analysis revealed a significant increase in S. aureus-specific IgGs in colonized B6J.Ob -/-compared to B6J.Ob +/+ mice (Fig 5A).
To further assess the protective attribute of S. aureus-specific IgG responses, S. aureus-colonized B6J.Ob -/-and Ob +/+ animals (shown in Fig 1 ) were subjected to a sublethal intravenous challenge with 5×10 6 CFU of S. aureus strain USA300, a methicillin-resistant S. aureus (MRSA) clone and prevalent cause of community-associated skin and soft tissue and bloodstream infections [53,54].Similar to humans, bloodstream infection of mice with S. aureus results in either death or the dissemination to organ tissues and the development of abscess lesions that persist for weeks [1,55,56].Bloodstream dissemination is measured as bacterial loads in kidneys by plating tissues fifteen days after challenge and recording colony forming units (CFUs) per kidney [55,56].Bacterial burdens were reduced by approximately 5×10 3 times in WU1-colonized B6J.Ob -/-mice as compared to B6J.Ob +/+ (Fig 5B ).
BALB.Ob -/-and BALB.Ob +/+ mice (shown in Fig 2 ) were also challenged 9 weeks post colonization with what turned out to be a slightly higher dose (1×10 7 CFU of strain USA300) resulting in the development of acute disease (Fig 5C).None of the Ob +/+ animals survived the intravenous challenge, while ~60% of the Ob -/-mice did.Further, all surviving animals had recovered a normal health status by day 11 post infection (Fig 5C).
To rule out the possibility that the loss of Ob alone might confer resistance to bloodstream dissemination, groups of naïve B6J.Ob +/+ and B6J.Ob The data presented so far suggest that the S. aureus-specific antibodies generated in colonized Ob -/-mice have improved functional activity compared to S. aureus-specific antibodies produced by colonized Ob +/+ mice.Complement-and Fc receptor-mediated phagocytosis are two major mechanisms known to control S. aureus in an antibody-dependent manner [57].The presence of opsonophagocytic antibodies in sera can be tested using a whole blood killing assay [57].Thus, sera from colonized BALB.Ob -/-and BALB.Ob +/+ (shown in Fig 2) as well as sera from naïve BALB.Ob -/-and BALB.Ob +/+ (control) were added to anti-coagulated, freshly drawn blood of naïve C57BL/6.μMTanimals that lack mature B cells and thus, most immunoglobulins [58,59] and then incubated with S. aureus for 30 min.Next, samples were treated with saponin and streptokinase to lyze host cells as well as liberate extracellular bacteria that agglutinate in fibrin to escape opsnophagpcytic killing [60] and plated to enumerate S. aureus CFUs.A statisitcal reduction in bacterial CFUs was observed in samples incubated with sera from both colonized BALB.Ob +/+ and BALB.Ob -/-as compared to the isogenic naïve animals (Fig 5D).However, sera of colonized BALB.Ob -/-had superior killing activity as compared to all other sera.Bacterial killing was not observed when blood had been pretreated with cytochalasin D, a cell-permeable inhibitor of actin polymerization (Fig 5D).Thus, exposure to S. aureus elicits opsonophagocytic antibodies more effectively in Ob-deficient compared to Obsufficient animals.

Discussion
All humans are exposed to S. aureus at birth and most remain colonized through life while some may be intermittently colonized [5,6,61].Healthy carriage represents a major cause for infection through self-inoculation [7][8][9][10].S. aureus is a highly pathogenic organism and its wide distribution in human populations suggest that not all individuals may be similarly susceptible to infection.Because strains of S. aureus are adaptable to other species, inquiries on susceptibility can be experimentally addressed in animals.For example, mice and sheep from different genetic backgrounds were found to display variable susceptibilities to S. aureus infection [62][63][64].In humans, correlations may be primarily drawn between genetic conditions and increased susceptibility.It has long been recognized that primary immunodeficiencies such as chronic granulomatous disease (CGD) of childhood predispose the host to severe and recurrent bacterial infections [65][66][67].Polymorphonuclear neutrophils play a key role in defending against S. aureus [68,69] and the defective NADH oxidase accounts for the impaired bactericidal activity of CGD neutrophils [65,66].Other single-gene inborn errors of immunity have also been found to predispose individuals to S. aureus infections and include congenital neutropenia, leukocyte adhesion deficiency, and germline mutations of the TLR and IL-1R pathways [70].Inborn errors of immunity have also been found to exacerbate disease.For example, OTULIN haploinsufficiency has been shown to enhance the cytotoxicity of the staphylococcal virulence factor α-toxin [71].However, as these mutations are very rare, they do not account for the large number of staphylococcal infections that affect otherwise healthy individuals [72].Single nucleotide polymorphism analyses and genome wide association studies have also been applied to identify genetic markers that may influence S. aureus colonization focusing on host adhesive properties, recognition (mannose-binding lectin, TLRs, NLRs and PRRs) or eradication (antimicrobial peptides, cytokines) of the pathogen [73].Genetic variations of mannosebinding lectin, TLR2 and TLR4 were found to correlate with increased nasopharyngeal colonization in infants [74].It has also been reported that healthy S. aureus carriers produce more LL-37, a cathelicidin antimicrobial peptide, than non-carriers but a genetic basis was not revealed [75].Thus, known and unknown genetic variations could render a host more resistant to S. aureus colonization.
The contribution of H2-O, an inhibitor of H2-M that limits the number of high affinity peptides presented by MHC-II [43], was examined in a mouse model of S. aureus colonization.We find that mouse lacking H2-O become decolonized in a B-and T-cell-dependent manner.Further, animals lacking H2-O that had been colonized with S. aureus were significantly better protected against subsequent bloodstream challenge with S. aureus as compared to colonized wild type animals or naïve H2-O-sufficient and -deficient animals.This protection correlated with increased anti-S.aureus serum responses and enhanced opsonophagocytic activity of S. aureus-specific antibodies as revealed in a whole blood killing assay.Thus, loss of H2-O represents a gain of function with respect to control of the pathogen.Increased resistance to S. aureus in absence of H2-O was observed in both B6J and BALB/cJ mice with MHC haplotypes H2 b and H2 d , respectively, suggesting that protection may be achieved upon recognition of distinct antigens.Antibody-mediated decolonization promoted in H2-O animals mirrors decolonization observed with S. aureus lacking Staphylococcal protein A (WU1Δspa) [3].SpA is a B cell superantigen that induces the production of non-specific IgGs that dilute the activity of protective antibodies [19,20,76].Thus, lack of H2-O in the host or lack of SpA in the pathogen result in protective antibody responses that are key to clearing this mucosal pathogen.Deconvoluting the nature of such responses may prove informative for the identification of optimal epitopes for vaccine designs.
We also observe that H2-O-deficient animals clear S. xylosus.S. xylosus generally inhabits the skin and mucous membranes of birds and mammals including laboratory mice [77,78].Like S. aureus, S. xylosus has been shown to cause invasive disease in otherwise healthy carrier mice albeit that the mechanisms promoting colonization or infection have not been investigated [79].We surmise that loss of S. xylosus colonization in H2-O-deficient animals is also the result of increased antibody responses.

Ethics statement
Animal research was performed in accordance with institutional guidelines following experimental protocol review, approval, and supervision by the Institutional Animal Care and Use Committee at The University of Chicago.Experiments with S. aureus were performed in Biosafety Level 2 containment upon review by The University of Chicago Institutional Biosafety Committee.

Bacterial strains and growth conditions
S. aureus strains WU1, WU1 Rif , USA300 (LAC) and the ΔspaΔsbi variant were from our laboratory collection and propagated in tryptic soy broth or tryptic soy agar at 37˚C.Animal swabs and fecal samples were plated on mannitol salt agar at 37˚C.For animals colonized with WU1 Rif or infected with USA300, test samplings were plated on tryptic soy agar containing 100 μg/ml rifampicin or 50 μg/ml kanamycin, respectively.

Preparation of bacterial cultures and extracts
Bacterial inocula for animal colonization and infection were prepared as follows.For each experiment, an overnight culture grown from a single colony in tryptic soy broth was diluted 100 times and grown to absorbance at 600 nm (A 600 ) of approximately ~1.Cultures were spun to sediment bacteria and cells were washed by resuspension in an equal volume of PBS.After a second sedimentation, cells were resuspended in PBS at ~10 10 CFU/ml and 10 μl (10 8 CFU) were used for intranasal inoculation.For intravenous challenge, bacteria were resuspended at a concentration of ~5-10×10 7 CFU/ml and 100 μl of this suspension was injected in animals.
For whole blood killing assays, bacteria were resuspended at a concentration of ~1×10 6 CFU/ ml.For ELISA experiments, cultures grown to A 600 of ~0.5 were subjected to 20 μg/mL lysostaphin treatment for 30 min at 37˚C, followed by 10% trichloroacetic acid precipitation on ice for 1 h.Samples were centrifuged for 10 min at 20,000 × g, and precipitates washed with icecold acetone, allowed to dry, and suspended in PBS.

Animal experiments
Following intranasal inoculation with bacteria, mice were monitored daily and throat swabs and fecal pellets were obtained in weekly intervals.Throat swabs were performed with CON-STIX swabs (pointed 0.08x0.30in, cat# SC-4).Throats of mice were swabbed to a depth of approximately 17mm and streaked across agar plates.Fecal pellets were collected into sterile pre-weighed tubes, weighed, and homogenized in 500 μl of PBS.Following intravenous infection with bacteria, animals were monitored for signs of illness for up to 15 days.To measure bacterial loads in tissues, mice were sacrificed and kidneys were homogenized in PBS with 0.1% Triton X-100 and serially diluted prior to plating.

Whole blood killing assay
Tubes containing 300 μl of freshly drawn B6J.μMT mouse blood anticoagulated with heparin (10 unit/ml) were pre-incubated for 10 min with 20μM cytochalasin D in DMSO or vehicle

PLOS PATHOGENS
S. aureus protective responses in absence of H2-O control (DMSO).Each tube received 15 μl of sera pooled from either naïve or colonized BALB.Ob -/-and BALB.Ob +/+ mice.Reactions were started upon addition of 15 μl of PBS containing 1.5×10 4 CFU of S. aureus and proceeded for 30 min at 37˚C with rotation.300 μl SK buffer containing 2% saponin, 200U/ml streptokinase, 1mg/ml trypsin, 20 μg/ml DNase, 100 μg/ml RNase A was added to each sample for 10 min at 37˚C prior to plating for CFU enumeration.This step was performed to liberate bacteria from fibrin agglutinates [60].The experiment was repeated an additional time and all assays were performed in duplicate.

Enzyme-linked immunosorbent assay (ELISA)
Microtiter plates (NUNC MaxiSorp) were coated with 1 μg/ml of Δspa/Δsbi bacterial extracts in 0.1 M carbonate buffer (pH 9.5) at 4˚C overnight.Wells were blocked, incubated with serial dilutions of mouse sera prior to incubation with 1 mg/ml horseradish peroxidase (HRP)-conjugated goat anti-mouse antibodies (Fisher Scientific) and developed using OptEIA reagent (BD Biosciences).Experiments were performed in triplicate to calculate averages and standard error of the mean and repeated for reproducibility.

Statistical analyses
All experiments were performed at least twice.For experiments with repeated measures, data was plotted as median ± 95% confidence interval and analyzed using two-way analysis of variance (ANOVA) with multiple-comparison tests (GraphPad Software

Fig 1 .
Fig 1. H2-O-deficient B6J mice decolonize S. aureus.Eight-week-old B6J.Ob +/+ and B6J.Ob -/-mice were colonized intranasally with 10 8 CFUs of WU1 (A) or 10 8 CFUs of WU1 Rif (B) and monitored for colonization weekly by plating throat swabs (A and B, top panels) and fecal matter (A and B, bottom panels) on mannitol salt agar (A) or on rifampicin containing tryptic soy agar (B).Mice with different genotypes were co-housed in the same cages.Data are presented as median ± 95% confidence interval.Significance was calculated using two-way ANOVA tests with multiple comparisons.n, number of mice used.Each dot represents a mouse.Males and females were used at 50:50 ratio.CFU, colony forming unit.Fecal and throat cultures performed on mice prior to inoculation of WU1 were negative for S. aureus species.https://doi.org/10.1371/journal.ppat.1012306.g001

Fig 2 .
Fig 2. H2-O-deficient BALB/cJ mice decolonize S. aureus.Eight-week-old Ob +/+ and Ob -/-BALB/cJ mice were colonized intranasally with 10 8 CFUs of WU1 Rif and monitored for colonization weekly by plating throat swabs (top panel) and fecal matter (bottom panel).Data are presented as median ± 95% confidence interval.Significance was calculated using two-way ANOVA tests with multiple comparisons.n, number of mice used.Each dot represents a mouse.Males and females were used at 50:50 ratio.CFU, colony forming units.Fecal and throat cultures performed on mice prior to inoculation of WU1 were negative for S. aureus species.https://doi.org/10.1371/journal.ppat.1012306.g002 + animals lacked all endogenous B cells and had only B cells of the transgenic origin (Fig 4A; S1A and S1B Fig).T-cell deficient and BCR monoclonal mice with and without Ob were colonized with WU1 Rif and monitored for the presence of bacteria over time (Fig 4B and 4C).B6.Ob +/+ and B6.Ob -/-mice were also colonized as control (S2 Fig).As reported in Fig 1, B6.Ob -/-animals gradually decolonized S. aureus (S2 Fig) while B6.Tcrβδ -/-Ob -/-(Fig 4B)

Fig 3 .
Fig 3. H2-O-deficient mice decolonize a commensal bacterium, S. xylosus.Colonies of approximately 60-day old Ob -/-and Ob +/+ B6J (A) and BALB/cJ (C) mice born at the University of Chicago were screened by plating fecal matter on mannitol salt agar plates.All colonies looked identical on plate and the species was identified as S. xylosus using 16S RNA-specific primers.(B) B6J.Ob -/-and B6J.Ob +/ + were foster-nursed by S. xylosus-positive B6J.Ob +/+ mothers and (C) BALB/c.Ob -/-and BALB/c.Ob +/+ were foster-nursed by S. xylosuspositive BALB.Ob +/+ mothers.All animals were monitored for colonization weekly by plating fecal matter on mannitol salt agar but data are shown starting at weeks 6 (B) and 5 (D) after birth.Significance was determined using unpaired t test (A and C) and two-way ANOVA tests with multiple comparisons (B and D).Data are presented as mean ± SEM (A and C) and as median ± 95% confidence interval (B and D).Males and females were used at 50:50 ratio.CFU, colony forming unit.https://doi.org/10.1371/journal.ppat.1012306.g003
).For experiments show in Figs 3A, 3C, 5B and 5D with only two groups and a single time point, data was analyzed with unpaired t test.The Mantel-Cox test was used to analyze data in Fig 5A and 5C.