Exacerbated Staphylococcus Aureus Foot

Exacerbated Staphylococcus aureus Foot Infections in Obese/Diabetic Mice Are Associated with Impaired Germinal Center Reactions, Ig Class Switching, and Humoral This information is current as Immunity of October 1, 2021. Christopher W. Farnsworth, Eric M. Schott, Abigail Benvie, Stephen L. Kates, Edward M. Schwarz, Steven R. Gill, Michael J. Zuscik and Robert A. Mooney

J Immunol 2018; 201:560-572; Prepublished online 1 June Downloaded from 2018; doi: 10.4049/jimmunol.1800253 http://www.jimmunol.org/content/201/2/560 http://www.jimmunol.org/ Supplementary http://www.jimmunol.org/content/suppl/2018/05/31/jimmunol.180025 Material 3.DCSupplemental References This article cites 46 articles, 20 of which you can access for free at: http://www.jimmunol.org/content/201/2/560.full#ref-list-1

Why The JI? Submit online. by guest on October 1, 2021

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2018 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Exacerbated Staphylococcus aureus Foot Infections in Obese/Diabetic Mice Are Associated with Impaired Germinal Center Reactions, Ig Class Switching, and Humoral Immunity

Christopher W. Farnsworth,*,† Eric M. Schott,*,† Abigail Benvie,*,† Stephen L. Kates,‡ Edward M. Schwarz,† Steven R. Gill,x Michael J. Zuscik,† and Robert A. Mooney*,†

Obese patients with type 2 diabetes (T2D) are at an increased risk of foot infection, with impaired immune function believed to be a critical factor in the infectious process. In this study, we test the hypothesis that humoral immune defects contribute to exacerbated foot infection in a murine model of obesity/T2D. C57BL/6J mice were rendered obese and T2D by a high-fat diet for 3 mo and were compared with controls receiving a low-fat diet. Following injection of Staphylococcus aureus into the footpad, obese/T2D mice had greater foot swelling and reduced S. aureus clearance than controls. Obese/T2D mice also had impaired humoral immune responses as indicated by lower total IgG levels and lower anti–S. aureus Ab production. Within the draining popliteal lymph Downloaded from nodes of obese/T2D mice, germinal center formation was reduced, and the percentage of germinal center T and B cells was decreased by 40–50%. Activation of both T and B lymphocytes was similarly suppressed in obese/T2D mice. Impaired humoral immunity in obesity/T2D was independent of active S. aureus infection, as a similarly impaired humoral immune response was demonstrated when mice were administered an S. aureus digest. Isolated splenic B cells from obese/T2D mice activated normally but had markedly suppressed expression of Aicda, with diminished IgG and IgE responses. These results demonstrate impaired humoral immune responses in obesity/T2D, including B cell–speciﬁc defects in Ab production and class-switch recombination. http://www.jimmunol.org/ Together, the defects in humoral immunity may contribute to the increased risk of foot infection in obese/T2D patients. The Journal of Immunology, 2018, 201: 560–572.

besity and type 2 diabetes (T2D) are among the greatest index ,30) and avoiding skin ulceration. There is currently no risk factors for infection (1). Obese T2D patients have a known medical treatment that proactively improves immune re- O 25% lifetime risk of developing a foot ulcer (2), and sponses and lessens infection risk in this population. Importantly, ∼56% of these become infected (3, 4). Although infections are the consequences of a foot infection are catastrophic, causing by guest on October 1, 2021 common among those with obesity/T2D, the only preventative substantial physical, emotional, and financial stress. Twenty percent treatments are reducing body weight to a healthy range (body mass of foot infections result in amputation, the most feared consequence of a foot ulcer (5). Because obese/T2D patients are disproportion- ally affected by foot infection, new therapies are needed to proac- *Department of Pathology and Laboratory Medicine, University of Rochester, tively reduce the risk of infection in this population. Rochester, NY 14642; †Center for Musculoskeletal Research, University of Roches- ter, Rochester, NY 14642; ‡Department of Orthopaedic Surgery, Virginia Common- It is generally accepted that diabetes increases the risk of foot x wealth University, Richmond, VA 23298; and Department of Microbiology and ulcers via two primary mechanisms: impaired vascular flow and Immunology, University of Rochester, Rochester, NY 14642 peripheral neuropathy (2). However, these factors alone do not ORCIDs: 0000-0001-7169-3850 (C.W.F.); 0000-0002-0160-1890 (A.B.); 0000-0002- fully explain the increased risk of infection in this population. 4854-9698 (E.M. Schwarz); 0000-0002-6651-4747 (R.A.M.). Multiple studies have shown obesity to be a risk factor for in- Received for publication February 22, 2018. Accepted for publication May 7, 2018. fection in surgical sites (1, 6). Our studies have shown that obesity, This work was supported by AOTrauma Research (Clinical Priority Program: Bone Infection) and by National Institutes of Health Grants P30 AR069655, P50 but not hyperglycemia, impairs humoral immune responses to AR072000, UL1 TR000042, and T32 AR053459. orthopedic Staphylococcus aureus infection (7). Other studies All authors contributed to the study design. C.W.F., E.M. Schott, and A.B. contrib- have demonstrated impaired innate (8, 9) and adaptive immune uted to the acquisition of data. C.W.F., E.M. Schwarz, S.L.K., S.R.G., M.J.Z., and responses in obesity (7, 10). However, the underlying mechanisms R.A.M. were responsible for drafting and editing the manuscript. All authors read and approved the submitted version. that alter immune function in obesity remain elusive. Much effort has focused on macrophages and neutrophils, demonstrating im- The RNA sequencing data presented in this article have been submitted to the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE113075) paired phagocytosis and bacterial killing despite increased cell under accession number GEO113075. recruitment (8, 9, 11). We have previously demonstrated impaired Address correspondence and reprint requests to Dr. Robert A. Mooney, Department Ab production in obese/T2D mice in a model of orthopedic im- of Pathology and Laboratory Medicine, University of Rochester, Box 626, 601 plant–associated infection, a result that was confirmed in patients Elmwood Avenue, Rochester, NY 14642. E-mail address: robert_mooney@urmc. rochester.edu infected with S. aureus (7). Recent studies have demonstrated The online version of this article contains supplemental material. reduced T cell response upon challenge in obese patients despite Abbreviations used in this article: AID, activation-induced cytidine deaminase; CSR, increased prechallenge activation (12). Others have demonstrated class-switch recombination; GC, germinal center; miR-155, microRNA 155; PLN, reduced Ab titers following vaccination in obese individuals (13, popliteal lymph node; PNA, peanut agglutinin; RNAseq, RNA sequencing; T2D, type 14), an effect that results in increased mortality following infec- 2 diabetes/diabetic; TfH, T follicular helper. tion in mice (15). However, the effect of obesity/T2D on humoral Copyright Ó 2018 by The American Association of Immunologists, Inc. 0022-1767/18/$35.00 immune responses is largely understudied. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1800253 The Journal of Immunology 561

In the absence of infection, the immune system of the obese host overnight at 4˚C. Anti-fluorescein conjugated to Alexa Fluor 488 (Thermo is in a state of chronic inflammation (16, 17). Importantly, Fisher Scientific) or mouse IgG kappa binding protein conjugated to B lymphocytes are known to participate in this chronic inflam- CruzFluor 594 (mIgGk BP-CFL 594) (Santa Cruz Biotechnology, Dallas, TX) were used at 1:200 for 1 h in TBS, followed by DAPI staining (Thermo matory state. B lymphocytes have been shown to produce IgG Fisher Scientific) and fixation using ProLong Gold Antifade Mounting autoantibodies (16) and secrete proinflammatory cytokines (18) Reagent (Life Technologies, Eugene, OR). All slides were visualized using within the adipose tissue of high fat–fed mice, propagating in- an Olympus VS120 Virtual Slide Scanning Microscope with Olympus flammation and contributing to insulin resistance. How chronic OlyVIA software. Quantification was as follows: germinal centers (GCs) were quantitated by a blinded observer by counting the number of PNA+ activation of the immune system in obesity/T2D, particularly in- B220+ areas within the lymph node. GC size and anti-AID staining were volving B cells, impacts normal B lymphocyte function during an determined using Visiopharm software (Brookfield, CO). infection is largely unknown. Together, these previous studies suggest that obesity increases Flow cytometry the risk of foot infection in obese/T2D patients by altering immune Lymph nodes were harvested and passed through a 70-mM nylon strainer function. In this article, we expand on our previous finding of (Corning, Corning, NY) in ice-cold PBS with 1% FBS. Cells were then impaired humoral immunity in obese/T2D patients and test the pelleted and stained with fixable viability stain v450 (BD Biosciences) for 40 min. Following washing, Fc block (rat anti-mouse CD16/CD32) was hypothesis that impaired Ab production in obesity is driven, at least performed for 5 min prior to staining. The following Abs were used for in part, by impaired B lymphocyte function. phenotyping of cells: BV711 rat anti-mouse IL-10, BV650 rat anti-mouse CD8a, BV605 rat anti-mouse IL-4, PerCP-Cy5.5 rat anti-mouse IFN-g, Materials and Methods APC-Cy 7 rat anti-mouse CD19, FITC hamster anti-mouse CD69, PerCP- Animals Cy5.5 rat anti-mouse Ly-6G and Ly-6C, PE rat anti-mouse GL-1, Alexa

Fluor 700 rat anti-mouse CD3, PerCP-Cy5.5 hamster anti-mouse CD69, Downloaded from All handling of mice and associated experimental procedures were reviewed BV711 rat anti-mouse F4/80, BB515 rat anti-mouse GL7, Alexa Fluor 700 and approved by the University Committee on Animal Resources at the rat anti-mouse CD4, and Alexa Fluor 647 mouse anti-mouse RoR g T (all University of Rochester Medical Center. Male C57BL/6J mice purchased from BD Biosciences) and PE-Cy7 anti-mouse CD11c and PE anti-mouse from The Jackson Laboratory (Bar Harbor, ME) were housed five per cage CD11b (both from eBioscience, San Diego, CA). in one-way housing on a 12-h light/dark cycle at the University of Superfolding variant GFP UAMS-1 S. aureus was used for cytometric Rochester. To model obesity and T2D, mice were placed on a high-fat (60% analyses on infected feet. Intracellular staining was performed using a cell kcal D12492) or low-fat (10% kcal D12450J) diet at 5 wk of age (Open- fixation and permeabilization kit (BD Biosciences). All analysis was Source Diets, Research Diets, New Brunswick, NJ). Prior to infection, performed on a BD LSR II 18-color flow cytometer (BD Biosciences). http://www.jimmunol.org/ fasting blood glucose levels were measured using OneTouch glucose meters Compensation was completed using anti-mouse or anti-rat/anti-hamster (LifeScan, Milpitas, CA) after an overnight fast. Glucose tolerance testing compensation control beads (BD Biosciences). Fluorescence minus one was also performed on mice fasted overnight. Lean and high fat–fed mice controls were used to determine positive gating for all Abs except for had an average fasting blood glucose of 120 and 190 mg/dl, respectively. CD19, CD3, CD4, and CD11b. All fluorescence minus one controls were performed on pooled samples from both control and obese/T2D mice. Infection model Gating strategies are outlined in Supplemental Fig. 1. Following feeding, chronic foot infections were established. S. aureus Serum analysis (UAMS-1) was cultured overnight, washed, and diluted 1:10 in sterile saline. Mice were given 60 mg/kg ketamine and 4 mg/kg xylazine as well Total IgG and IgM assays. Two micrograms per milliliter goat anti-mouse as a preoperative dose of buprenorphine. Mice were then injected in the IgG and IgM (Southern Biotech, Birmingham, AL) were coated onto by guest on October 1, 2021 foot with 60 ml of the diluted S. aureus. Immediately before and following 96-well plates (Nunc 442404). After blocking in 3% BSA for 1 h, serum at the injection, the diameter of the footpad was measured using digital 1:10,000 was added to wells (100 ml) and incubated at 4˚C for 1 h. After calipers (Mitutoyo, Japan). All feet returned to the original size within 2 h five washes in PBS with Tween 20, secondary Ab (anti-IgG or anti-IgM; of injection. Feet were also measured at indicated time points. S. aureus Southern Biotech) conjugated to HRP was added at a dilution of 1:4000 digest (UAMS-1 DSpa) was digested at 37˚C in the presence of lysozyme and incubated for 1 h at 4˚C. KPL SureBlue peroxidase was used as a and lysostaphin (Sigma-Aldrich, St. Louis, MO) for 3 h in PBS. The digest substrate (52-00-01; Gaithersburg, MD). For measuring in vitro Ab pro- was then washed prior to combination with alum (Thermo Fisher Scien- duction, the same method was used with 100 ml of undiluted conditioned tific, Waltham, MA) per the manufacturer’s instructions. Following sus- media after 48 h of stimulation. pension with alum, 50 ml of the S. aureus digest was immediately injected into the footpad of control or obese/T2D mice. Immunogenicity of the Anti–S. aureus ELISA assay digest was confirmed prior to suspension in alum via an ELISA using S. aureus extracts were prepared by incubating 1.0 ml of an overnight serum from previously infected mice. Complete S. aureus killing in the culture of DSpa UAMS-1 S. aureus with 20 mg of lysozyme (Sigma- digest was confirmed prior to alum suspension by plating on tryptic soy Aldrich) and 1 mg lysostaphin (Sigma-Aldrich) in sterile water. Extract broth agar plates overnight at 37˚C. was then used as a coating Ag in an ELISA at 1:2000 in PBS. After Colony-forming units blocking and washing in PBS with Tween 20, serum was diluted 1:10 in PBS, followed by incubation for 1 h. IgG or IgM secondary conjugated to S. aureus was isolated and CFU quantitated from infected footpads at HRP was next incubated for 1 h, followed by incubation with KPL sacrifice. Necrotic soft tissue surrounding the infection site was placed in SureBlue. Absorbance was measured on a plate reader at 450 nm. Results 2 ml of PBS. Tissues were then homogenized using an IKA T-10 Handheld are reported as arbitrary units by dividing the absorbance of each sample Homogenizer (Wilmington, NC). Serial dilutions were then prepared in by the mean absorbance of lean mice. PBS, and 100-ml aliquots were plated on tryptic soy agar. Plates were then placed at 37˚C for 24 h before counting colonies. RNA extraction and sequencing Histological analysis RNA sequencing (RNAseq) was performed on naive lymph nodes. Fol- lowing sacrifice, lymph nodes were harvested, immediately flash-frozen, Lymph nodes were harvested after sacrifice and fixed overnight in 4% neutral and stored at 280˚C. Prior to extraction of RNA, lymph nodes were buffered formalin. Samples were then washed, processed, and embedded in pulverized at liquid nitrogen temperatures in Lysing Matrix B tubes (MP paraffin, and 5-mm transverse sections were prepared. Immunofluorescence Biomedicals). RNA was then extracted by mechanical lysis in ice-cold staining included the following: sodium citrate (Dako, Carpinteria, CA) acid phenol using the FastPrep-24 instrument (MP Biomedicals) at a was diluted per the manufacturer’s instructions, and slides were incubated setting of 5.5 for 40 s. Extracted RNA was purified using RNeasy Mini at 95˚C for 40 min. Following Ag retrieval, blocking was performed with Columns (Life Technologies). Contaminating genomic DNA was re- 10% donkey serum in PBS for 2 h at room temperature. Primary Abs in- moved using TURBO DNase (Ambion). Illumina library construction for cluded the following: APC rat anti-mouse CD45R/B220 (BD Biosciences, RNAseq was as follows: libraries were prepared from rRNA-depleted San Jose, CA), lectin from Arachis hypogaea (peanut agglutinin [PNA]), RNA (Ribo-Zero; Epicentre) using the ScriptSeq version 2 RNAseq FITC (Sigma-Aldrich), and mouse anti–activation-induced cytidine deam- Library Preparation Kit (Epicentre) according to the manufacturer’s in- inase (AID) (Thermo Fisher Scientific) at 1:200 in PBS were incubated structions. Libraries were sequenced on an Illumina HiSeq 2500 platform 562 IMPAIRED HUMORAL IMMUNITY IN OBESITY with four to five libraries multiplexed per lane and ∼50 million 100-nt To track the fate of S. aureus in the obese and lean mice, single-direction reads for each sample. Initial sequence data from the footpads were infected with superfolder GFP UAMS-1 S. aureus, HiSeq 2500 were evaluated for quality. The low-quality sequence reads a GFP-expressing isolate. On day 7 postinfection, footpad tissue were removed prior to final analysis using SeqClean (http://sourceforge. net/projects/seqclean/). The remaining high-quality processed reads were was harvested, pooled, stained, and analyzed by flow cytometry. then mapped or aligned to reference Mus musculus genomes with Consistent with increased CFU, there was a considerably greater SHRiMP version 2.2.3 (19) to match sequence reads to specific mouse proportion of fluorescence associated with GFP+ S. aureus in genes. Differential expression levels of all genes were determined using samples from obese/T2D mice (19%) compared with lean-fed Cufflinks (cuffdiff2)version 2.0.2 (20) and a false discovery rate of 0.05, which calculates gene expression as the number of sequence reads that controls (6%) (Fig. 1E). Interestingly, a substantial portion of + maptoeachgene.Agreaterthanlog2-foldincreaseordecreaseinex- GFP S. aureus was bound or internalized by host cells in obesity/ pression level was used to identify genes that were significantly different T2D, as indicated by association with high forward light scatter between control and obese/T2D mice. Three mice were used per group (cell size). Further analysis of the GFP+ S. aureus population in for RNAseq experiments. All RNAseq data (raw sequence reads obese/T2D mice revealed that the large majority of bacteria was and primary analysis) have been deposited in the National Center for Biotechnology Information’s Gene Expression Omnibus and are acces- observed in nonviable cells (69%). In contrast, only half of sible through Gene Expression Omnibus Series accession number remaining GFP+ S. aureus were within nonviable cells of lean GEO113075 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? mice (Fig. 1F). Despite the greater S. aureus burden in obese/T2D acc=GSE113075). Enrichment analysis was performed using Ingenuity mice, the largest proportion (47% of total cells) of noninfected Pathway Analysis software (Qiagen, Redwood City, CA) and Ensembl + + annotation software (21). live cells was CD11b Gr-1 granulocytes (Fig. 1G). In contrast, only 12% of the live cell fraction in control mice was composed of Real-time quantitative PCR granulocytes (Fig. 1G). Together, the data suggest that obesity/ Downloaded from RNA samples were DNase treated (Invitrogen, Carlsbad, CA), and 15 mlof T2D is associated with impaired innate immune responses to RNA was then added to iScript cDNA Synthesis per the manufacturer’s S. aureus footpad infection. instructions (Bio-Rad, Hercules CA). 1.5 ml of cDNA was then added to 7 ml water, 10 ml of SYBR Green (Quanta Biosciences, Gaithersburg, MD) Impaired humoral immune response and GC formation in and 1.5 ml of forward and reverse primer mix. PCR was carried out on a obesity/T2D Rotor-Gene Q (Qiagen, Hilden, Germany) real-time PCR machine. The primers used were the following: Aicda forward: 5-9AAGAATGCAC- We previously demonstrated an impaired humoral immune re- GATCGCCTCT-39, reverse: 5-9CTTTGGATCTGGGCGCTTTG-39; BCl6 sponse in obese/T2D mice to an S. aureus orthopedic implant http://www.jimmunol.org/ forward: 5-9CACACCCGTCCATCATTGAA-39, BCL6 reverse: 5-9TGTCC- infection (7). In the footpad infection model, obese/T2D mice TCACGGTGCCTTTTT-39; Rpl32 forward: AAGCGAAACTGGCGGAA- again displayed impaired total IgG, increased total IgM, and re- AC-39,reverse:5-9TCAGGATCTGGCCCTTGAACC-39; and Actin forward: 5-9GCTACAGCTTCACCACCACA-39, reverse: 5-9GGGGTGT- duced anti–S. aureus IgG levels (Fig. 2A–C). Consistent with TGAAGGTCTCAAA-39. reduced anti–S. aureus IgG levels, immunofluorescence staining (B220+,PNA+) revealed reduced GC formation in popliteal lymph Splenic B cell isolation and stimulation nodes (PLN) from obese/T2D mice at day 14 postinfection Following 12 wk on diet, spleens were harvested and passed through a (Fig. 2D, 2E). Diminished GC formation in obesity/T2D led to 70-mM nylon filter in sterile, ice-cold PBS. B lymphocytes were isolated further examination of cell populations within the draining PLN using an EasySep Negative Selection Mouse B Cell Enrichment Kit (Stem by flow cytometry. Approximately 100-fold more cells were iso- by guest on October 1, 2021 Cell Technologies, Vancouver, Canada) per the manufacturer’s instructions. B lymphocytes were then plated at 1 3 106 cells/ml in 24-well plates lated from the draining PLN of both control and obese mice on and cultured with RPMI 1640 (Life Technologies, Grand Island, NE) days 7 and 14 postinfection compared with noninfected mice supplemented with gentamicin and 10% FBS. Cells were then stimulated (Fig. 2F), with marked increases in B and T lymphocytes with 20 ng/ml S. aureus peptidoglycan (Sigma-Aldrich), 1 mg/ml CD40 (Figs. 3B, 4B). However, PLN from obese/T2D mice consis- ligand + 20 ng/ml IL-4 + anti-HA, or anti-mouse IgM. tently yielded larger cell numbers than controls. Thus, the Statistics reduction in Ab production and GC formation in PLN from in- Multiple analyte comparisons were measured using two-way ANOVA and fected obese/T2D mice was associated with increased cellularity. Bonferroni post hoc test. Unpaired t test was used when two groups were Despite increased total cell numbers, the proportion of macro- compared, including area under the curve measurements. Mann–Whitney phages (F4/80+ CD11b+) in the lymph nodes was lower in obese/ U test was used for nonnormally distributed data. All statistics were an- T2D mice (Fig. 2G, 2H). There was no difference in the alyzed using GraphPad Prism. percentage of neutrophils (data not shown) or dendritic cells within the draining PLN in obese/T2D mice compared with Results controls (Fig. 2I). Thus, obesity/T2D was associated with im- Mice fed a high-fat diet for 3 mo developed the hallmarks of obesity paired GC formation and Ab production, a phenotype that is and T2D (Fig. 1A, 1B) (7, 22). A foot infection was then initiated linked with increased lymph node cellularity and impaired mac- in the obese mice and lean controls by injecting S. aureus into the rophage infiltration. mouse footpad. On day 5 postinfection, there was a trend toward impaired clearance of S. aureus from the footpads of obese/T2D Obesity/T2D is associated with impaired T cell activation, GC mice (Fig. 1C). By day 12, infections persisted in the footpads of T cell recruitment, and T cell polarization obese/T2D mice but were cleared in all but one control mouse Because obesity was associated with reduced GC formation, (Fig. 1C). Footpad swelling was measured throughout the infec- we next sought to determine the effect of obesity/T2D on tion as a marker of inflammation and immune cell migration to the T lymphocytes within the draining PLN. As with total cells infection site. The high fat–fed obese mice had increased footpad (Fig. 2F), there was a robust increase in CD3+ T cells in the swelling as early as day 5 postinfection (Fig. 1D). By day 12, draining PLN from day 0 to day 7, persisting to day 14 postin- significant foot swelling remained in obese/T2D mice, but fection (Fig. 3A, 3B) in both control and obese/T2D mice. In swelling had resolved in control mice (Fig. 1D). These results are contrast to the difference in total cellularity, no differences in total similar to a study in a genetic model of obesity that reported in- T cells were noted between obese/T2D and control mice at any of creased immune cell infiltration into infected footpads of obese the time points (Fig. 3B). Moreover, the number of T cells per- mice and impaired bacterial clearance (11). sisted from day 7 to 14 in both cohorts. Interestingly, despite the The Journal of Immunology 563 Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 1. Impaired S. aureus clearance and increased inflammatory cell infiltration in obesity/T2D. C57BL6/J mice were fed a diet consisting of 60% kcal from fat or 10% kcal from fat for 3 mo. Mice were then (A) weighed and (B) nonfasting blood glucose was measured prior to footpad infection with S. aureus.(C) Infected mice were harvested for CFU quantitation on day 5 and 12 (n = 5 for each group). (D) Footpad swelling was measured as a readout for immune cell infiltration. Control and obese/T2D mice were infected with superfolder GFP S. aureus, harvested on day 7, and pooled for flow cytometry. (E) Gating on GFP+ S. aureus indicated an increased proportion of S. aureus associated with host cells in infected footpads of obese/T2D mice. (F) Further gating on the GFP+ S. aureus population with a mammalian-specific viability dye revealed an increased proportion of S. aureus in dead cells of obese/T2D mice. (G) Proportion of noninfected live cells (gated on forward and side light scatter and viability) that were granulocytes. (A–C) Unpaired Student t test. (D) Two-way ANOVA and Bonferroni post hoc analysis. Each point represents one mouse, and error bars represent SEM. (E)–(G) are pooled specimens from six mice and are representative of two independent experiments. *p , 0.05, **p , 0.01, ***p , 0.001. same number of total T cells, obese/T2D mice had a reduced Consistent with this reasoning, PLN from obese/T2D mice had proportion of CD3+ T cells on day 0 and 14 postinfection 50% fewer CD4+ GL7+ GC T cells compared with controls 14 d (Fig. 3C) compared with control mice. Twenty-five percent of the postinfection (Fig. 3D, 3E). We also noted a trend toward de- total cells were composed of T lymphocytes on day 14 in lean creased GC T cells on day 7 postinfection in obese/T2D mice controls as opposed to only 12% in the obese/T2D. This coupled (Fig. 3E). We next sought to establish if the reduction in GC with the same number of total T cells in each group indicates T cells in obesity/T2D was a result of impaired T cell activation increased infiltration or proliferation of other cell types into the and Ag presentation. Consistent with this, obese/T2D mice had a PLN in obese/T2D mice. 50% decrease in activated T cells compared with lean-fed controls A reduction in GCs in obesity/T2D (Fig. 2D, 2E) led us to by day 7 postinfection (Fig. 3F). This effect was transient, how- speculate that GC T cells would also be reduced in this population. ever, as the percentage of activated T cells in obese/T2D mice 564 IMPAIRED HUMORAL IMMUNITY IN OBESITY Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 2. Impaired humoral immune responses and GC formation in obese/T2D mice. Mice (n = 6–10) were infected in the footpad with S. aureus. On day 14, (A) total IgG, (B) total IgM, and (C) anti–S. aureus IgG were measured by ELISA. Mice (n = 4) were sacrificed on day 14, and lymph nodes were isolated for histological staining. (D) Representative fluorescence micrographs obtained at original magnifications 32 and 320 of medial sections of lymph nodes from control and obese/T2D mice (n = 4) stained with DAPI (blue, cells), B220 (pink, B cells), and PNA (green, GCs). (E) is quantitation of GC numbers per lymph node. (F) PLN (n = 4–5 mice) were isolated for flow cytometry, and total cells were quantitated prior to infection and 7 and 14 d postinfection. (G) Representative surface staining for macrophages in the draining PLN 7 d postinfection. (H) Quantitation of (G). (I) Quantitation of total dendritic cells in draining PLN. Unpaired Student t test, except for (F) (two-way ANOVA and Bonferroni post hoc analysis). Quantification in (E)is from lymph nodes from four mice. Each data point represents an average of three sections per node, with all sections separated by 100–150 mm. *p , 0.05, **p , 0.01. normalized to that of the lean control by day 14. Together, these speculated that distorted T cell polarization due to a high-fat diet results associate impaired T cell activation with reduced GC and obesity would be further propagated following S. aureus in- T cells in the PLN of obese/T2D mice following footpad S. aureus fection. To this end, we further examined Th cell subsets within infection. the draining PLN 14 d postinfection. Interestingly, obesity/T2D Previous studies have established enhanced T cell polarization was associated with an increase in CD4+ T helper cells compared and recruitment to the adipose tissue in obesity. Adipose tissue with control mice (Fig. 3G, 3H). Having established reduced GC inflammation and insulin resistance have been implicated with Th1 T cells in obesity/T2D, we next examined T follicular helper and Th17 cell polarization and infiltration (23, 24). Therefore, we (TfH) cell subsets. There was a 50% reduction in BCL6+ TfH cells The Journal of Immunology 565 Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 3. Impaired T cell activation and polarization in obese/T2D mice. Mice (n = 4–7 per group) were infected in the footpad with S. aureus. Draining PLN were harvested, stained, and analyzed by ﬂow cytometry. (A) Representative plots 14 d postinfection of CD19+ B cells and CD3+ T cells. (B) Quantitation of CD3+ T cells from (A). (C) Percentage of total T cells from draining PLN in (A). (D) Representative images of gating strategy for GC T cells (CD3+, CD4+, GL7+). (E) Quantitation of GCs preinfection and 7 and 14 d postinfection. (F) Quantitation of CD69+-activated T cells. (G) Representative surface staining for CD4, demonstrating increased CD4+ T cells in obese/T2D mice. (H) Quantitation of (G). Intracellular cytokine and transcription factor staining of T cells was quantitated from PLN of infected mice in (G). BCL6+ staining of TfH cells (I), Th17 cells (J), Th1 cells (K), and T regulatory cells (L) (all gates on CD3+ CD4+ cells). *p , 0.05, ** p , 0.01, unpaired Student t test (H–L) (others were analyzed by two-way ANOVA and Bonferroni post hoc analysis). in obese/T2D mice compared with controls (Fig. 3I), a result activation, reduced recruitment of TfH cells to GCs, and alter- consistent with impaired GC formation in this population. Con- ations in T cell polarization. versely, obesity/T2D was associated with an increase in Th17 cells in the draining PLN compared with lean-fed mice (Fig. 3J). Obesity/T2D is associated with reduced GC B cells and T regulatory cell populations and Th1 populations were also aberrant B cell activation present but showed no difference between obese/T2D mice and Because of reduced GCs and impaired Ab production in obesity/ control mice (Fig. 3K, 3L). Negligible numbers of Th2 cells were T2D, B cells were assessed in the draining PLN. Interestingly, found in either group (,0.01% cells, data not shown). Together, signiﬁcantly more B cells were located in the PLN from obese/T2D we conclude that obesity/T2D is associated with impaired T cell mice than control mice prior to infection (Fig. 4A, 4B). On day 7 566 IMPAIRED HUMORAL IMMUNITY IN OBESITY Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 4. Impaired B cell activation and reduced GC B cells in obese/T2D mice. Mice (n = 5–7 per group) were infected in the footpad with S. aureus. Draining PLN were harvested, stained, and analyzed by ﬂow cytometry. (A) Representative plots 14 d postinfection of B cells (CD19+). (B) Quantitation of B cells in (A). (C) Percentage of total B cells from (A). (D) Representative images of gating strategy for activated B cells (CD19+, GL7+). (E) Quantitation of activated B cells from (D). (F) Representative plots of GC (Gl7+) B cells (gated on CD19). (G) Quantitation of (F) on uninfected and 7 and 14 d post- infected. *p , 0.05, **p , 0.01, ***p , 0.001, two-way ANOVA and Bonferroni post hoc analysis. postinfection, there was a .100-fold increase in the total number obese/T2D mice, which remained elevated and was signiﬁcantly of B cells in both control and obese/T2D mice compared with greater than the number of B lymphocytes in PLN of lean-fed naive lymph nodes. By day 14, the absolute number of control mice (Fig. 4B). Moreover, B lymphocytes were an in- B lymphocytes was reduced in lean-fed control mice (Fig. 4B). creased percentage of the total cell fraction in obese/T2D This was in marked contrast to the number of total B cells from mice compared with controls (Fig. 4C). These data implicate The Journal of Immunology 567

B lymphocytes as the primary driver of the increased cell attained when B cells were stimulated with peptidoglycan and infiltration/proliferation (Fig. 2F) in the lymph nodes of obese/ anti-mouse IgM for 48 h (Fig. 6B). Consistent with this, CFSE T2D mice following S. aureus infection. staining to measure proliferation revealed no differences between Consistent with impaired Ag presentation (25, 26), the per- obese/T2D mice and controls 2 d poststimulation (data not centage of activated B cells was reduced in obesity/T2D on day 7 shown). Together, these results reveal no defect in B lymphocyte and 14 postinfection (Fig. 4D, 4E). Moreover, the 50% reduction activation in obesity/T2D. in activated B cells in obesity/T2D correlated with a similar re- Isolated B lymphocytes were subsequently stimulated for 4 d duction in GC B cells in this population by day 14 postinfection with CD40L + IL4 to induce CSR and to observe Ab production. (Fig. 4F, 4G). We also noted a trend toward reduced GC B cells Consistent with our in vivo findings, B lymphocytes from obese/ 7 d postinfection in obese/T2D mice. These results associate T2D mice produced markedly less IgG and IgE in vitro com- obesity/T2D with reduced B cell activation and reduced GC pared with controls (Fig. 6C, 6D). No difference was noted in IgM B cells. production (Fig. 6E). The finding of impaired CSR in obesity/T2D despite normal activation led us to further investigate AID. Con- Impaired Ab response in obesity/T2D independent of sistent with in vivo staining for AID, Aicda expression was sig- bacterial killing nificantly reduced in B cells from obese/T2D mice stimulated To determine if reductions in Ab production, GCs, and GC B cells in vitro for 2 d (Fig. 6F). To validate these results, a second cohort reflected a decreased ability of obese/T2D mice to kill live of mice were infected for 7 d, and lymph nodes were harvested for S. aureus as part of Ag presentation, mice were challenged with RNA. Again, a marked reduction in Aicda expression was ob- S. aureus that was heat killed, treated with lysozyme and lysos- served in obese/T2D mice (Fig. 6G). Additionally, expression of taphin, and prepared in alum. Interestingly, obese/T2D mice dis- Bcl6, a gene that is crucial for CSR in GCs, was almost com- Downloaded from played an increase in footpad swelling over the duration of the pletely suppressed in lymph nodes from obese/T2D mice com- 14 d compared with controls, indicating that impaired bacterial pared with control mice (Fig. 6H). Together, these results reveal killing was not necessary for elevated inflammation in the diabetic impaired CSR in B lymphocytes of obese/T2D mice. foot (Fig. 5A). Consistent with results in mice infected with live It is logical to assume that high-fat diet–induced obesity/T2D S. aureus, obese/T2D mice had impaired anti–S. aureus IgG elicited changes in naive lymph nodes and B lymphocytes prior to production (Fig. 5B). Moreover, there was an increase in anti– S. aureus infection, which contributed to the subsequent impaired http://www.jimmunol.org/ S. aureus IgM production in obese/T2D mice (Fig. 5C). This is humoral immune response to infection. In support of this, whole- consistent with our previous results in bone infection that indi- transcriptome sequencing of naive lymph nodes from obese/T2D cated impaired class-switch recombination (CSR) in obesity/T2D and control mice revealed 1802 differentially expressed genes: (7). Quantitation by flow cytometry of the draining PLN revealed 1512 were upregulated in obesity and 290 were downregulated no difference in the total number of cells when comparing control (Fig. 7A, 7B). Although many genes were differentially expressed, and obese/T2D mice on day 14 postinjection (Fig. 5D). Similar to their differences were relatively small, as only two genes were 10- obese/T2D mice infected with live S. aureus, we noted an in- fold increased in obese/T2D PLN compared with controls. Nota- creased proportion of B lymphocytes in this population (Fig. 5E). bly, Il17f was 16-fold more highly expressed, and Mki67 was by guest on October 1, 2021 Despite reduced anti–S. aureus IgG levels in obesity/T2D, the 9-fold higher in obese/T2D mice compared with controls total proportion of GC B cells, macrophages, and dendritic cells (Supplemental Table I). Pathway analysis revealed an enrichment was equivalent to that of controls 14 d postchallenge (Fig. 5F–H). in terms associated with BCR signaling, TCR signaling, cell mi- Moreover, histology revealed no differences in GCs or total GC gration, chemokine signaling, and transcription in obese/T2D area in obesity/T2D compared with lean mice 14 d postinjection. mice (Fig. 7A). Moreover, cell type enrichment revealed enrich- (Fig. 5I, 5J). ment in B cell subsets, particularly follicular B cells in obese/T2D The surprising observation of impaired Ab production despite mice compared with controls (Fig. 7B). These results indicate normal GC formation led us to hypothesize that CSR is impaired in increased B cell infiltration and accumulation in the naive PLN of obese/T2D mice independent of Ag presentation. To examine this obese/T2D mice. possibility, immunofluorescence staining for AID was performed Histological analysis of naive PLN revealed increased lymph on PLN of mice challenged for 14 d with killed S. aureus. AID is an node size and B cell infiltration/proliferation in obese/T2D mice enzyme that is crucial for CSR and is upregulated in GC B cells (Fig. 7C), confirming RNAseq results. No generation of GCs was (27, 28). Consistent with the indication of impaired CSR in noted in obesity/T2D despite an increase in B cells (Fig. 7C). Flow + obesity/T2D, there was a significant reduction in AID immuno- cytometry on naive PLN from uninfected obese/T2D mice has staining in GCs in this population compared with lean-fed controls demonstrated increased total cellularity (Fig. 2F) and increased (Fig. 5K). This finding translated statistically, as obese/T2D mice total B cells compared with lean-fed controls (Fig. 4C). Further + had reduced total AID staining within GCs compared with con- analysis demonstrated reciprocal changes in B cell and T cell trols despite normal GC formation (Fig. 5L). Together, these data percentages (Fig. 7D, 7E) in obese/T2D mice, a result that was imply that diminished Ab production in obesity/T2D is due in part primarily driven by an increase in the absolute number of to impaired CSR and reduced AID in GCs. B lymphocytes in obesity/T2D. Further analysis of naive PLN Reduced Ab production and Aicda expression are associated revealed an increase in activated B cells in obesity/T2D (Fig. 7F). with increased inflammation in obesity/T2D Together, these data indicate that the naive PLN from obese/T2D mice have an influx of B lymphocytes and that these cells are in a To directly examine the influence of obesity/T2D on B cell acti- proinflammatory, activated state. These changes likely predispose vation and CSR, naive B cells were isolated from the spleen of the host to defective humoral immune response to infection. obese/T2D mice and control mice. Following stimulation with CD40L and IL4 for 48 h, B cell activation was assessed by quantitating CD69+ cells by flow cytometry. No differences were Discussion noted in the proportion of activated B cells in obese/T2D mice In this study, we demonstrate impaired S. aureus clearance by compared with controls (Fig. 6A, 6B). Comparable results were obese/T2D mice in a model of foot infection. Importantly, this 568 IMPAIRED HUMORAL IMMUNITY IN OBESITY Downloaded from

FIGURE 5. Impaired GC reactions and humoral immunity in obese/T2D mice occur in the presence of normal numbers of APCs. Mice (n = 8–9) were http://www.jimmunol.org/ challenged in the footpad with enzymatically digested S. aureus mixed in alum. (A) Footpad swelling was determined prior to harvest. (B) Anti–S. aureus IgG Ab production and (C) anti–S. aureus IgM were measured in serum 14 d postchallenge. Draining PLN were isolated for flow cytometry and (D) total cells, (E) proportion of CD19+ B cells, (F) GC B cells (GL7+ CD19+), (G) macrophages (F480+ CD11b+), and (H) dendritic cells (CD11b+ CD11c+) were quantitated. (I) Immunofluorescence staining was performed on PLN 14 d postinfection to determine GCs (DAPI, blue; B cells, pink). (J) Quantitation of GC area in (I). (K) Immunofluorescence staining for AID. Dashed white lines outline GCs. (L) Quantitation of (K). Each point represents one mouse, and error bars indicate SEM. For quantitation in (J) and (L), each point is the average of three sections, 100 mm apart, from one lymph node. *p , 0.05, **p , 0.01, ***p , 0.001, unpaired Student t test, except (A) (two-way ANOVA and Bonferroni post hoc analysis). Scale bar, 200 mm(I). by guest on October 1, 2021 result was associated with increased cellular infiltration at the mice and patients to infection, a result that was independent of infection site, impaired innate immune function, and reduced hyperglycemia (7). Therefore, we conclude that to reduce the humoral immune responses. With regards to infection, peripheral incidence of foot infection, it is likely insufficient to only treat the vascular disease and neuropathy associated with T2D account for underlying hyperglycemia in obesity/T2D. Rather, effective much of the increased risk of foot ulcers in diabetic patients. treatment to restore immune function and reduce infection risk However, these factors alone do not fully account for the impaired may require addressing the other elements of the obesity pheno- immune response that appears to be at least partially attributable type. To date, the most effective approach remains reducing body to obesity. In support of this, multiple studies have demonstrated mass index. an association between obesity and increased infection suscepti- Given the impaired innate immune function of obese/T2D in the bility following surgery (1, 6, 29). Moreover, obese patients are foot infection model, it is not surprising that defects were also noted also at increased risk of developing community-associated infec- within the draining PLN. To our knowledge, the effect of impaired tions, such as influenza (30), periodontitis (8), and skin infections innate immunity and reduced bacterial killing in obesity/T2D on (31). Consistent with other studies (32), we demonstrate that humoral immune function has not been established. In this report, obese/T2D mice have an increase in inflammatory cell infiltration the humoral immune response in obese/T2D mice infected with live to the infection site and impaired bacterial killing (Fig. 1) (8, 9, S. aureus was characterized by impaired Ab production, reduced 33). Specifically, the primary source of inflammatory cells within GC formation, impaired B and T cell activation, altered T cell the footpad of obese/T2D mice 7 d postinfection was CD11b+ polarization, and reduced GC B and T cells. Moreover, on day 7 GR1+ granulocytes. Despite increased phagocytes in obesity/T2D, postinfection, this was associated with reduced numbers of mac- we noted reduced clearance of S. aureus. Moreover, an increased rophages within the draining PLN. Interestingly, when enzymati- proportion of S. aureus was observed to be extracellular or re- cally digested S. aureus was administered, we noted a rescue in siding within dead host cells in the obese population. To our GC formation and macrophage number in the lymph nodes knowledge, this is the first study linking impaired immune func- of obese/T2D mice. This argues that impaired S. aureus killing tion with reduced bacterial clearance in a model of diet-induced in the obese/T2D foot instigates impaired Ag presentation obesity/T2D and foot infection. These results were similar to a and lymphocyte activation in the draining PLN. It is also worth footpad infection in genetically modified leptin-deficient mice, noting that the kinetics of inflammation within the PLN matched which also demonstrated impaired S. aureus killing in obesity that of inflammation in the footpad. In both control and obese/T2D (11). Additionally, our results reveal an impaired humoral im- mice, there was a robust increase in cells within the footpad by munity that may be a major driver of increased infection risk in day 7 and a subsequent increase in total cells in the PLN. By day the obese/T2D population. Moreover, we have previously dem- 14, most lean mice resolved the infection, and there was a sub- onstrated in a model of bone infection that obesity predisposes sequent decrease in total lymphocytes within the draining PLN. In The Journal of Immunology 569 Downloaded from http://www.jimmunol.org/

FIGURE 6. Obesity/T2D is associated with impaired CSR in vitro and in vivo. B lymphocytes isolated from the spleens of control or obese/T2D mice (n $ 4) were cultured with vehicle control or CD40L + IL-4, followed by immunostaining for activated B cells (CD69+). (A) Representative panels demonstrating in vitro activation 48 h poststimulation. (B) Quantitation of activated B cells from (A) or from stimulation with peptidoglycan or anti-mouse IgM. Fold change in (C)IgG,(D)IgE,and(E) IgM produced by splenic B cells activated with CD40L + IL-4 after 4 d. (F) Aicda expression was measured 4 d by guest on October 1, 2021 poststimulation of splenic B cells with CD40L + IL-4. (G) Aicda and (H) Bcl6 expression in PLN of mice infected for 7 d with S. aureus. (B) is pooled data from three separate experiments performed in triplicate. (C–F) are pooled data from four independent experiments, each performed in triplicate. *p , 0.05, **p , 0.01, unpaired Student t test. contrast, in obese/T2D mice, inflammation persisted in the foot in vivo and stimulated B cells from obese/T2D mice in vitro. AID concurrently with enhanced inflammation and cell recruitment/ is a critical enzyme in CSR, producing double-stranded breaks in proliferation in the draining PLN. Taken together, we demon- the DNA of B cells and ultimately instigating the production of strate in this study a clear association in obesity/T2D between class-switched, highly specific Abs (27, 35, 36). Our data argue impaired innate immune killing of S. aureus, increased lymph that reduced AID expression and impaired CSR in B lymphocytes node cellularity, aberrant lymphocyte activation, and diminished from obese/T2D mice is a source of dampened Ig production in humoral immunity. Further studies must be completed to con- this population. Consistent with these findings, clinical studies clusively show the cellular source and mechanism of impaired Ag have shown that obesity is associated with impaired humoral presentation in the lymph nodes in obesity/T2D. immunity to influenza vaccination (13) and reduced titers to the We also demonstrate impaired B lymphocyte function in obesity/ tetanus vaccine in obese children (14). This further implicates T2D independent of S. aureus killing. Supporting this conclusion, impaired humoral immune function with obesity and argues that obese/T2D mice challenged with killed S. aureus had impaired the defect is not specific to S. aureus. However, further studies anti–S. aureus Ab production, similar to those infected with live must be completed to conclusively determine the role of impaired S. aureus. Moreover, reduced anti–S. aureus IgG in obese/T2D class switching and Ab production in obesity/T2D in other in- was established despite correcting for APC, lymphocyte activa- fectious diseases. tion, and GC formation when challenged with killed S. aureus. It is tempting to link increased inflammation in the naive obese/ This implies that impaired humoral immune responses in obesity/ T2D lymph node with impaired humoral immune responses and Ab T2D are due, at least in part, to compromised B lymphocyte production upon immune challenge. It is generally accepted that function. This is consistent with studies demonstrating altered diet-induced obesity is associated with increased systemic in- B cell responses in ex vivo stimulated B cells from obese mice flammation (17, 37). Consistent with this, B lymphocytes are in- (34). Consistent with this finding, we noted impaired in vitro IgG strumental in the progression of diabetes, driving inflammation and IgE production by previously naive B lymphocytes in obese/ and, ultimately, insulin resistance (16, 18). How obesity-induced T2D mice stimulated with CD40L + IL-4 to induce CSR. More- inflammation and activation of B cells affect responses to infec- over, this phenomenon was associated with a reduction in Aicda tion has not been established. It was recently revealed that ele- (AID) gene expression in both PLN of infected obese/T2D mice vated TNF-a production by B cells in aged mice limits B cell 570 IMPAIRED HUMORAL IMMUNITY IN OBESITY Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 7. Increased activation and inflammation in naive lymph nodes from obese/T2D mice. PLN were isolated from naive obese/T2D and control mice (n = 3), and whole-transcriptome analysis was performed. (A) Gene ontology terms that were significantly enriched in obese/T2D mice using In- genuity Pathway Analysis are shown. (B) Cell type enrichment was performed using Ensembl, demonstrating increased follicular B cells in obesity/T2D. (C) Representative histological images demonstrating increased follicular B cells in obesity/T2D (DAPI, blue; B cells, pink; PNA GC, green). Flow cytometry was performed on naive lymph nodes from control and obese/T2D mice. (D) Representative flow cytometry panels demonstrating B cell and T cell populations in naive PLN of obese/T2D mice and control mice. (E) Proportion of CD19+ B cells or CD3+ T cells in naive PLN. (F) Quantification of activated B cells (CD69+) within the naive PLN. Each point represents one mouse or biological replicate. n . 6 for (C–F). Original magnification 35(C). Error bars indicate SEM. *p , 0.05, **p , 0.01, unpaired Student t test. responses, impairs CSR, and reduces AID production following and ex vivo B cells from obese mice had reduced CSR in the LPS stimulation (38). Importantly, diabetes and obesity are driven, absence of Th17 stimulus. Nonetheless, it is possible that altered at least in part, by elevated proinflammatory cytokines (37, 38). T cell polarization negatively affects Ig production in obese/T2D Moreover, we have previously shown that TNF-transgenic mice mice. accumulate CD23+CD21highCD1dhigh B cells in inflamed lymph In this study, we demonstrate inflammation within the naive nodes, which are nonactivated polyclonal Ag–capturing cells (25, lymph nodes in obese/T2D mice by whole-transcriptome se- 26). Thus, we observed that accumulation of these bystander cells quencing and further validated this by flow cytometry, demon- from obese/T2D chronic inflammation could also contribute to the strating increased activation of lymph node B cells in this increase in B cells and inefficient GC reactions. It is also possible population. One side effect of enhanced inflammatory cytokine that reduced Ig production by obese/T2D mice is a result of al- production in aging and obesity is an increase in microRNA 155 tered T cell polarization. In support of this, TfH cells were de- (miR-155) (42, 43). Importantly, miR-155 is a known AID and creased and Th17 cells were increased in lymph nodes of infected CSR regulator (28, 44). Therefore, chronic low-grade inflamma- obese/T2D mice (Fig. 3). Additionally, IL17f was highly tion associated with obesity may mediate downregulation of B cell expressed in the noninfected obese lymph node (Supplemental responses via an AID- and miR-155–dependent mechanism, ulti- Table I). IL17A and IL17F are implicated in the inflammation mately decreasing B lymphocyte Ab production. The source of obesity (39, 40), and Th17 cells have a known role in B cell of inflammation in B lymphocytes is currently unclear. Exoge- class switching to isotypes IgG2a and IgG3 (41). We did not nous cytokine production and inflammation are well-known observe any differences in Ig isotype production (data not shown), comorbidities of obesity (17). However, altered metabolism in The Journal of Immunology 571

B lymphocytes as a result of changes in substrate (i.e., increases in 16. Winer, D. A., S. Winer, L. Shen, P. P. Wadia, J. Yantha, G. Paltser, H. Tsui, P. Wu, M. G. Davidson, M. N. Alonso, et al. 2011. B cells promote insulin re- saturated fat) may also play a role in impaired B cell responses sistance through modulation of T cells and production of pathogenic IgG anti- and inflammation. Consistent with this, the humoral immune re- bodies. Nat. Med. 17: 610–617. sponse is known to be tightly regulated by metabolism (45), and 17. Gregor, M. F., and G. S. Hotamisligil. 2011. Inflammatory mechanisms in obesity. Annu. Rev. Immunol. 29: 415–445. studies have shown that B cell function in obesity can be modu- 18. DeFuria, J., A. C. Belkina, M. Jagannathan-Bogdan, J. Snyder-Cappione, lated by fatty acid supplementation (34, 46). Nonetheless, further J. D. Carr, Y. R. Nersesova, D. Markham, K. J. Strissel, A. A. Watkins, M. Zhu, studies must be completed to fully elucidate the mechanism of et al. 2013. B cells promote inflammation in obesity and type 2 diabetes through regulation of T-cell function and an inflammatory cytokine profile. Proc. Natl. impaired B lymphocyte class switching in obesity and T2D. Acad. Sci. USA 110: 5133–5138. Together, these results demonstrate impaired immunity in 19. David, M., M. Dzamba, D. Lister, L. Ilie, and M. Brudno. 2011. SHRiMP2: sensitive yet practical SHort read mapping. Bioinformatics 27: 1011–1012. obesity/T2D that contributes to impaired clearance of S. aureus in a 20. Trapnell, C., A. Roberts, L. Goff, G. Pertea, D. Kim, D. R. Kelley, H. Pimentel, model of foot infection. Moreover, we demonstrate impaired hu- S. L. Salzberg, J. L. Rinn, and L. Pachter. 2012. Differential gene and transcript moral and innate immune function in obesity/T2D, each of which expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat. Protoc. 7: 562–578. is instrumental in reduced Ab production. The defects in humoral 21. Birney, E., T. D. Andrews, P. Bevan, M. Caccamo, Y. Chen, L. Clarke, G. Coates, immunity likely contribute to the increased risk of infection in J. Cuff, V. Curwen, T. Cutts, et al. 2004. An overview of Ensembl. Genome Res. obese/T2D patients and warrant further study. 14: 925–928. 22. Farnsworth, C. W., E. M. Schott, S. E. Jensen, J. Zukoski, A. M. Benvie, M. A. Refaai, S. L. Kates, E. M. Schwarz, M. J. Zuscik, S. R. Gill, and Acknowledgments R. A. Mooney. 2017. Adaptive upregulation of clumping factor A (ClfA) by Staphylococcus aureus in the obese, type 2 diabetic host mediates increased We thank Ann Gill, Sarah Mack, and Kathy Maltby for very capable tech- virulence. Infect. Immun. 85: e01005-16. nical skills. The support of the Histology, Biochemistry, and Molecular Im- 23. McLaughlin, T., S. E. Ackerman, L. Shen, and E. Engleman. 2017. Role of Downloaded from aging Core in the Center for Musculoskeletal Research at the University innate and adaptive immunity in obesity-associated metabolic disease. J. Clin. Invest. 127: 5–13. of Rochester is greatly appreciated. RNAseq and transcriptome analysis 24. Cso´ka, B., P. Pacher, P. Bai, and G. Hasko´. 2015. New piece in the jigsaw puzzle: described in this study were completed by the University of Rochester adipose tissue-derived stem cells from obese subjects drive Th17 polarization. Genomics Research Center. Diabetes 64: 2341–2343. 25. Li, J., I. Kuzin, S. Moshkani, S. T. Proulx, L. Xing, D. Skrombolas, R. Dunn, I. Sanz, E. M. Schwarz, and A. Bottaro. 2010. Expanded CD23(+)/CD21(hi) Disclosures B cells in inflamed lymph nodes are associated with the onset of inflammatory- The authors have no financial conflicts of interest. erosive arthritis in TNF-transgenic mice and are targets of anti-CD20 therapy. http://www.jimmunol.org/ J. Immunol. 184: 6142–6150. 26. Moshkani, S., I. I. Kuzin, F. Adewale, J. Jansson, I. Sanz, E. M. Schwarz, and A. Bottaro. 2012. CD23+ CD21(high) CD1d(high) B cells in inflamed lymph References nodes are a locally differentiated population with increased antigen capture and 1. Dowsey, M. M., and P. F. Choong. 2009. Obese diabetic patients are at sub- activation potential. J. Immunol. 188: 5944–5953. stantial risk for deep infection after primary TKA. Clin. Orthop. Relat. Res. 467: 27. Pasqualucci, L., G. Bhagat, M. Jankovic, M. Compagno, P. Smith, 1577–1581. M. Muramatsu, T. Honjo, H. C. Morse, III, M. C. Nussenzweig, and R. Dalla- 2. Singh, N., D. G. Armstrong, and B. A. Lipsky. 2005. Preventing foot ulcers in Favera. 2008. AID is required for germinal center-derived lymphomagenesis. patients with diabetes. JAMA 293: 217–228. Nat. Genet. 40: 108–112. 3. Smith, D. M., M. Weinberger, and B. P. Katz. 1987. A controlled trial to increase 28. Zan, H., and P. Casali. 2015. Epigenetics of peripheral B-cell differentiation and office visits and reduce hospitalizations of diabetic patients. J. Gen. Intern. Med. the antibody response. Front. Immunol. 6: 631.

2: 232–238. 29. Ja¨msen, E., P. Nevalainen, A. Eskelinen, K. Huotari, J. Kalliovalkama, and by guest on October 1, 2021 4. Block, P. 1981. The diabetic foot ulcer: a complex problem with a simple T. Moilanen. 2012. Obesity, diabetes, and preoperative hyperglycemia as treatment approach. Mil. Med. 146: 644–646. predictors of periprosthetic joint infection: a single-center analysis of 7181 5. Levin, M. E. 2002. Management of the diabetic foot: preventing amputation. primary hip and knee replacements for osteoarthritis. J. Bone Joint Surg. Am. South. Med. J. 95: 10–20. 94: e101. 6. Winfield, R. D., S. Reese, K. Bochicchio, J. E. Mazuski, and G. V. Bochicchio. 30. Mertz, D., T. H. Kim, J. Johnstone, P. P. Lam, M. Science, S. P. Kuster, 2016. Obesity and the risk for surgical site infection in abdominal surgery. Am. S. A. Fadel, D. Tran, E. Fernandez, N. Bhatnagar, and M. Loeb. 2013. Pop- Surg. 82: 331–336. ulations at risk for severe or complicated influenza illness: systematic review and 7. Farnsworth, C. W., C. T. Shehatou, R. Maynard, K. Nishitani, S. L. Kates, meta-analysis. BMJ 347: f5061. M. J. Zuscik, E. M. Schwarz, J. L. Daiss, and R. A. Mooney. 2015. A humoral 31. Huttunen, R., and J. Syrjanen. 2013. Obesity and the risk and outcome of in- immune defect distinguishes the response to Staphylococcus aureus infections in fection. Int. J. Obes. (Lond.) 37: 333–340. mice with obesity and type 2 diabetes from that in mice with type 1 diabetes. 32. Brotfain, E., N. Hadad, Y. Shapira, E. Avinoah, A. Zlotnik, L. Raichel, and Infect. Immun. 83: 2264–2274. R. Levy. 2015. Neutrophil functions in morbidly obese subjects. Clin. Exp. 8. Amar, S., Q. Zhou, Y. Shaik-Dasthagirisaheb, and S. Leeman. 2007. Diet- Immunol. 181: 156–163. induced obesity in mice causes changes in immune responses and bone loss 33. Yano, H., M. Kinoshita, K. Fujino, M. Nakashima, Y. Yamamoto, H. Miyazaki, manifested by bacterial challenge. Proc. Natl. Acad. Sci. USA 104: 20466–20471. K. Hamada, S. Ono, K. Iwaya, D. Saitoh, et al. 2012. Insulin treatment directly 9. Zhou, Q., S. E. Leeman, and S. Amar. 2009. Signaling mechanisms involved in restores neutrophil phagocytosis and bactericidal activity in diabetic mice and altered function of macrophages from diet-induced obese mice affect immune thereby improves surgical site Staphylococcus aureus infection. Infect. Immun. responses. Proc. Natl. Acad. Sci. USA 106: 10740–10745. 80: 4409–4416. 10. Vallerskog, T., G. W. Martens, and H. Kornfeld. 2010. Diabetic mice display a 34. Kosaraju, R., W. Guesdon, M. J. Crouch, H. L. Teague, E. M. Sullivan, delayed adaptive immune response to Mycobacterium tuberculosis. J. Immunol. E. A. Karlsson, S. Schultz-Cherry, K. Gowdy, L. C. Bridges, L. R. Reese, et al 184: 6275–6282. 2017. B cell activity is impaired in human and mouse obesity and is responsive 11. Park, S., J. Rich, F. Hanses, and J. C. Lee. 2009. Defects in innate immunity to an essential fatty acid upon murine influenza infection. J. Immunol. 198: predispose C57BL/6J-Leprdb/Leprdb mice to infection by Staphylococcus au- 4738–4752. reus. Infect. Immun. 77: 1008–1014. 35. De Silva, N. S., and U. Klein. 2015. Dynamics of B cells in germinal centres. 12. Richard, C., M. Wadowski, S. Goruk, L. Cameron, A. M. Sharma, and Nat. Rev. Immunol. 15: 137–148. C. J. Field. 2017. Individuals with obesity and type 2 diabetes have additional 36. Mesin, L., J. Ersching, and G. D. Victora. 2016. Germinal center B cell dy- immune dysfunction compared with obese individuals who are metabolically namics. Immunity 45: 471–482. healthy. BMJ Open Diabetes Res. Care 5: e000379. 37. Hotamisligil, G. S., P. Arner, J. F. Caro, R. L. Atkinson, and B. M. Spiegelman. 13. Sheridan, P. A., H. A. Paich, J. Handy, E. A. Karlsson, M. G. Hudgens, 1995. Increased adipose tissue expression of tumor necrosis factor-alpha in A. B. Sammon, L. A. Holland, S. Weir, T. L. Noah, and M. A. Beck. 2012. human obesity and insulin resistance. J. Clin. Invest. 95: 2409–2415. Obesity is associated with impaired immune response to influenza vaccination in 38. Frasca, D., M. Romero, A. Diaz, S. Alter-Wolf, M. Ratliff, A. M. Landin, humans. Int. J. Obes. (Lond.) 36: 1072–1077. R. L. Riley, and B. B. Blomberg. 2012. A molecular mechanism for TNF- 14. Eliakim, A., C. Schwindt, F. Zaldivar, P. Casali, and D. M. Cooper. 2006. Re- a-mediated downregulation of B cell responses. J. Immunol. 188: 279–286. duced tetanus antibody titers in overweight children. [Published erratum appears 39. Ip, B., N. A. Cilfone, A. C. Belkina, J. DeFuria, M. Jagannathan-Bogdan, in 2006 Autoimmunity 39: 349.] Autoimmunity 39: 137–141. M. Zhu, R. Kuchibhatla, M. E. McDonnell, Q. Xiao, T. B. Kepler, et al. 2016. 15. Karlsson, E. A., V. A. Meliopoulos, N. C. van de Velde, L. A. van de Velde, Th17 cytokines differentiate obesity from obesity-associated type 2 diabetes and B. Mann, G. Gao, J. Rosch, E. Tuomanen, J. McCullers, P. Vogel, and S. Schultz- promote TNFa production. Obesity (Silver Spring) 24: 102–112. Cherry. 2017. A perfect storm: increased colonization and failure of vaccination 40. Eljaafari, A., M. Robert, M. Chehimi, S. Chanon, C. Durand, G. Vial, leads to severe secondary bacterial infection in influenza virus-infected obese N. Bendridi, A. M. Madec, E. Disse, M. Laville, et al. 2015. Adipose tissue- mice. MBio 8: e00889-17. derived stem cells from obese subjects contribute to inflammation and reduced 572 IMPAIRED HUMORAL IMMUNITY IN OBESITY

insulin response in adipocytes through differential regulation of the Th1/Th17 43. Frasca, D., A. Diaz, M. Romero, F. Ferracci, and B. B. Blomberg. 2015. balance and monocyte activation. Diabetes 64: 2477–2488. MicroRNAs miR-155 and miR-16 decrease AID and E47 in B cells from elderly 41. Mitsdoerffer, M., Y. Lee, A. Ja¨ger,H.J.Kim,T.Korn,J.K.Kolls, individuals. J. Immunol. 195: 2134–2140. H. Cantor, E. Bettelli, and V. K. Kuchroo. 2010. Proinﬂammatory T helper 44. Teng, G., P. Hakimpour, P. Landgraf, A. Rice, T. Tuschl, R. Casellas, and type 17 cells are effective B-cell helpers. Proc. Natl. Acad. Sci. USA 107: F. N. Papavasiliou. 2008. MicroRNA-155 is a negative regulator of activation- 14292–14297. induced cytidine deaminase. Immunity 28: 621–629. 42. Karkeni, E., J. Astier, F. Tourniaire, M. El Abed, B. Romier, E. Gouranton, 45. Boothby, M., and R. C. Rickert. 2017. Metabolic regulation of the immune L. Wan, P. Borel, J. Salles, S. Walrand, et al. 2016. Obesity-associated in- humoral response. Immunity 46: 743–755. ﬂammation induces microRNA-155 expression in adipocytes and adipose 46. Whelan, J., K. M. Gowdy, and S. R. Shaikh. 2016. N-3 polyunsaturated fatty tissue: outcome on adipocyte function. J. Clin. Endocrinol. Metab. 101: acids modulate B cell activity in pre-clinical models: implications for the im- 1615–1626. mune response to infections. Eur. J. Pharmacol. 785: 10–17. Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021 Viability Cells Singlets Samples FMO control Samples FMO control

Supplemental Figure 1 B and T Lymphocyte Gating Strategy Representative gating strategy used to distinguish activated B and T cells or germinal center T cells and B cells. Also shown are fluorescent minus one controls demonstrating how positive staining was confirmed. Top 200 upregulated genes Top 200 downregulated genes Rps17 -0.54 7.64E-03 Oas3 -0.77 4.97E-02 Gene log2(fold_change) q_value Gene log2(fold_change) q_value Gene log2(fold_change) q_value Gene log2(fold_change) q_value Il17f 4.01 1.20E-02 Saa3 1.51 4.98E-03 Laptm4a -0.55 5.51E-03 Gpr34 -0.78 9.32E-04 Mki67 3.14 0.00E+00 Ncoa6 1.51 8.60E-09 Rpl27a -0.55 1.01E-02 Gpx6 -0.78 3.95E-02 Oxtr 2.89 1.07E-05 Kif11 1.50 5.74E-06 Tspan3 -0.55 7.13E-03 Reln -0.78 1.80E-02 Gm4759 2.58 0.00E+00 Il21 1.50 4.07E-02 2010107E04Rik-0.55 1.36E-02 Hspb11 -0.79 2.45E-02 Plin4 2.46 0.00E+00 Rif1 1.50 4.50E-07 Mgst1 -0.55 1.40E-02 Ms4a8a -0.79 1.83E-02 Kif20b 2.44 9.77E-06 Mll2 1.50 1.16E-08 Sult1a1 -0.55 3.26E-02 Cd209d -0.79 2.76E-04 Slx4 2.24 6.55E-07 Myo9b 1.49 1.38E-06 Ier2 -0.55 2.91E-02 Mrgprx2 -0.79 2.59E-03 Ltf 2.15 1.49E-06 Top2a 1.49 5.95E-09 Rpl35a -0.56 8.16E-03 Sh3gl3 -0.80 1.40E-02 Lipf 2.14 3.59E-02 Itsn2 1.49 5.45E-05 Angptl4 -0.56 4.74E-02 Cfhr1 -0.81 1.98E-03 Uvssa 2.10 1.04E-05 Chd6 1.49 8.15E-08 Car3 -0.56 1.75E-02 Dnahc6 -0.81 2.20E-02 Kif15 2.07 6.02E-04 Mcm10 1.48 8.87E-03 Rplp1 -0.56 9.99E-03 Hscb -0.81 2.43E-03 Ankrd11 2.06 8.79E-12 BC018507 1.48 2.24E-08 Itm2a -0.56 1.93E-02 Cdo1 -0.82 9.24E-03 Gen1 2.06 7.11E-03 Szt2 1.48 4.75E-03 Pfdn5 -0.56 4.35E-03 Gpx3 -0.82 3.50E-05 Golga4 2.04 1.90E-11 Pik3c2b 1.47 3.13E-06 S100a16 -0.56 2.01E-02 Syngr1 -0.82 2.77E-02 Chd7 1.99 2.64E-10 Kif18b 1.47 1.54E-03 Ndufa6 -0.57 1.06E-02 Il15 -0.82 9.03E-03 Nin 1.99 8.77E-10 Chd4 1.47 5.51E-13 Bst2 -0.57 1.62E-02 Ppp1r3c -0.83 1.05E-02 Gm8979 1.99 0.00E+00 Akna 1.47 1.24E-11 Rps15a -0.57 8.74E-03 Ccl8 -0.83 1.01E-05 Pou2f2 1.96 7.06E-08 Cntln 1.47 1.52E-02 Churc1 -0.57 1.93E-02 Pck1 -0.84 3.95E-02 Fcamr 1.94 9.85E-03 Lmtk2 1.47 4.47E-08 Rpl37a -0.57 4.95E-03 Hes1 -0.85 2.59E-03 Ngp 1.94 3.60E-11 Apc 1.46 8.43E-09 Bpgm -0.57 2.58E-02 Gpm6a -0.85 4.74E-04 Id1 -0.85 4.56E-03 Rest 1.92 1.17E-07 Klf7 1.46 1.37E-02 Rgs18 -0.57 4.91E-02 Khdrbs3 -0.89 2.43E-02 Nipbl 1.91 1.63E-08 Nlrc3 1.46 1.21E-03 Rps24 -0.57 1.97E-02 Tpsab1 -0.89 2.25E-04 Camp 1.90 6.27E-08 Zfml 1.46 9.87E-07 Rps15a-ps6 -0.57 3.20E-02 Ccdc148 -0.90 6.81E-03 Eppk1 1.87 7.34E-08 Plcb4 1.46 6.91E-03 Rps27a -0.57 5.29E-03 Tpsb2 -0.90 1.20E-06 Kif4 1.87 5.96E-04 BC005561 1.46 1.18E-07 Sptssa -0.57 4.78E-03 Cd209a -0.91 2.51E-04 Trim56 1.87 3.06E-08 Zfp831 1.45 5.15E-08 Havcr1 -0.58 4.36E-02 Adh1 -0.91 8.67E-05 Ckap2l 1.85 4.19E-05 Tnrc6b 1.44 4.35E-06 Laptm4b -0.58 1.10E-02 Hoxd10 -0.91 1.14E-02 Hivep1 1.85 5.76E-11 Card11 1.44 4.03E-09 Mrpl52 -0.58 2.50E-02 Mrgprb1 -0.92 8.16E-05 Cep110 1.85 8.18E-05 Smg1 1.44 3.94E-10 Nenf -0.58 3.64E-02 Retn -0.92 1.25E-02 Zfp957 1.84 7.94E-07 Wbp7 1.44 2.92E-07 Rpl22 -0.58 3.46E-03 Hist3h2ba -0.92 4.11E-02 Gvin1 1.84 7.15E-10 Nol8 1.43 1.85E-03 Rpl38 -0.58 3.17E-03 Rtp4 -0.93 3.71E-04 Setd2 1.83 0.00E+00 Zfp608 1.43 5.52E-06 Rps15a-ps4 -0.58 2.52E-02 Zfp292 1.42 9.96E-09 Oosp1 -0.94 1.70E-02 Mll1 1.83 1.17E-12 Gm4832 -0.58 3.73E-02 Nnat -0.95 4.51E-02 Upf1 1.82 4.20E-09 Stk10 1.42 2.89E-11 Gm6525 -0.58 4.96E-03 Golgb1 1.42 2.91E-06 Inmt -0.95 5.74E-06 Mis18bp1 1.81 1.30E-03 Rpl22l1 -0.58 3.12E-03 Fxyd6 -0.96 6.86E-06 Lcn2 1.80 7.11E-06 Sptbn1 1.42 7.58E-12 Cd209b -0.58 4.52E-03 Trrap 1.42 1.99E-07 Upb1 -0.96 2.18E-02 Lyst 1.79 3.10E-11 Rps18 -0.58 6.48E-03 Ppp1r12a 1.42 1.98E-07 Slc6a4 -0.96 1.30E-03 Akap13 1.77 0.00E+00 Usp18 -0.58 3.86E-02 Gm4975 1.41 3.23E-02 Cldn5 -0.97 7.45E-07 Cep250 1.77 2.23E-07 Rpl36a -0.59 2.65E-03 Adap2 1.41 3.16E-02 Klrg1 -0.97 2.96E-02 Zbtb38 1.77 1.34E-06 Rps21 -0.59 2.85E-03 Xlr3b -0.98 1.07E-02 Sltm 1.41 6.74E-06 Bod1l 1.76 1.99E-08 Klf10 -0.59 2.44E-02 Rsad2 -0.98 2.31E-03 Fry 1.41 1.50E-05 Rrbp1 1.76 2.37E-09 Rps28 -0.59 2.29E-03 Xlr3c -0.98 1.13E-02 Macf1 1.40 4.81E-07 Mns1 1.76 6.38E-03 Gm15421 -0.59 2.83E-03 Mrgprb2 -0.99 1.20E-03 Upf2 1.39 2.07E-06 Ccdc88c 1.75 3.06E-08 Klk8 -0.60 1.07E-02 Il22ra2 -1.00 2.21E-07 Zcchc6 1.39 2.17E-07 Smc4 1.74 3.03E-13 Tmem160 -0.60 1.40E-02 Klk1b9 -1.05 2.03E-02 Cenpj 1.39 9.47E-03 Myh10 1.72 4.99E-12 Ube2l6 -0.60 5.49E-03 Slc40a1 -1.08 5.46E-08 Ylpm1 1.39 3.47E-07 Golga3 1.72 4.73E-08 Cd209c -0.60 1.79E-02 Enho -1.08 7.02E-03 Arid5b 1.38 6.94E-06 Srrm2 1.72 7.75E-10 Cd209e -0.60 2.44E-02 Fibin -1.08 2.46E-02 Nsd1 1.38 3.60E-10 Tcf20 1.71 2.92E-07 Lhfp -0.60 1.45E-02 Mylpf -1.09 1.36E-02 Rfc1 1.38 5.62E-07 Zfp646 1.70 1.19E-10 Sox18 -0.61 4.71E-02 Cldn11 -1.09 5.30E-05 Ankhd1 1.37 6.37E-09 Myh11 1.70 1.01E-12 Uqcrh -0.61 1.52E-03 Atp8b5 -1.11 8.97E-04 Atrx 1.37 4.03E-09 Zbtb16 1.69 5.91E-03 Rgs10 -0.61 2.59E-03 Rbp4 -1.12 6.49E-03 Bdp1 1.37 8.20E-06 Limch1 1.68 1.25E-02 Siglech -0.61 9.81E-03 Klk1b24 -1.14 1.32E-02 Eea1 1.37 3.65E-06 Cep350 1.68 1.59E-10 Bok -0.62 1.67E-02 Klk1b21 -1.14 8.94E-03 Gm13034 1.37 2.30E-06 Tnrc18 1.68 1.27E-06 Fmo1 -0.62 7.13E-03 Xlr3a -1.15 3.68E-03 Zfp609 1.36 9.84E-06 Gm608 1.67 4.75E-11 Tspan7 -0.62 2.09E-03 Sncb -1.15 2.53E-02 Zcchc11 1.36 1.23E-08 Chd5 1.67 1.08E-04 Bod1 -0.62 3.04E-02 Gabra3 -1.16 1.47E-02 Tmie 1.36 1.96E-02 Iqgap3 1.66 3.49E-04 Bola3 -0.62 2.78E-02 Slc13a3 -1.17 4.86E-02 Ankrd52 1.36 2.21E-07 Ankrd12 1.66 8.30E-08 Folr4 -0.62 2.30E-02 Clec4g -1.19 3.90E-04 Zfp335 1.36 1.76E-04 Dvl3 1.65 1.48E-04 Rps12 -0.62 1.47E-03 Hba-x -1.19 1.24E-02 Gm1995 1.36 7.89E-05 Brwd3 1.65 3.76E-05 Fam198b -0.62 3.33E-02 Tm4sf20 -1.20 9.29E-03 Arid1a 1.36 1.29E-10 Chd3 1.65 1.73E-12 Rpl39 -0.63 2.12E-03 Klk1b4 -1.20 7.68E-03 Acin1 1.36 3.31E-03 Ngb -1.21 4.91E-02 Pcnt 1.65 2.17E-07 Fbln2 1.36 8.50E-04 Cmpk2 -0.63 3.62E-02 Cbr2 -1.24 3.41E-07 Tcof1 1.64 6.26E-06 Znf512b 1.36 2.17E-02 Camk1 -0.63 4.42E-02 Plec 1.64 1.34E-07 Xaf1 -0.63 4.74E-02 Klk1b3 -1.25 2.84E-03 Nemf 1.35 4.16E-05 Ckm -1.26 2.37E-03 Gm7134 1.64 2.24E-02 Depdc5 1.35 7.23E-04 Nt5c3 -0.63 3.36E-02 Fancm 1.63 3.22E-04 Prg3 -0.63 2.74E-02 Tnni2 -1.27 2.47E-04 Cep164 1.35 2.72E-04 Timp4 -1.27 9.77E-06 2810474O19Rik1.63 4.03E-09 Gm5124 1.35 2.13E-06 Fcgr1 -0.63 4.09E-02 Tpr 1.62 3.08E-07 Tm4sf1 -0.65 6.08E-03 Klk1 -1.28 7.21E-04 Bcl9l 1.35 6.88E-07 Klk1b5 -1.28 1.19E-03 Prrc2c 1.62 8.61E-12 Paox -0.65 1.90E-02 Tln2 1.34 7.41E-05 Gm4956 -1.32 4.33E-03 Syne3 1.62 1.73E-05 Oas1g -0.65 2.73E-02 Hp 1.34 1.86E-08 Cyp2f2 -1.35 1.71E-07 5830417I10Rik1.62 9.96E-12 Oas1a -0.66 2.42E-02 Rb1cc1 1.34 1.33E-06 Hbb-bh2 -1.36 7.14E-06 Spen 1.61 3.23E-06 Gm13476 -0.66 4.43E-02 Tln1 1.34 2.50E-11 Hbb-y -1.38 1.70E-06 Mest 1.59 1.29E-06 Tmem98 -0.67 2.58E-02 Kdm6b 1.33 2.02E-05 C1ql2 -1.39 5.40E-03 Mll5 1.59 2.88E-08 Dusp1 -0.67 8.00E-03 Uaca 1.33 1.73E-04 Hbq1a -1.39 2.81E-03 Cep135 1.59 4.74E-04 Tal1 -0.67 4.10E-02 Herc1 1.33 7.53E-07 Mpz -1.40 1.57E-03 Akap9 1.58 2.69E-06 2610035D17Rik-0.68 1.39E-02 Tshz3 1.33 1.50E-02 Alas2 -1.40 3.06E-07 Cdk5rap2 1.58 3.46E-06 Mmrn1 -0.68 1.24E-02 Smarca4 1.33 3.06E-08 Ifi44 -1.41 2.42E-05 Phf3 1.57 1.00E-10 Cma1 -0.68 5.18E-04 Casp8ap2 1.33 2.79E-05 Beta-s -1.41 5.44E-09 Crebbp 1.57 6.95E-08 Wfdc18 -0.69 2.26E-02 Setd5 1.33 1.67E-08 Hbb-bh1 -1.42 2.45E-10 Prr14l 1.57 2.37E-09 Klhl4 -0.69 1.48E-02 Ino80 1.32 1.01E-05 Cd209g -1.42 4.73E-09 Adam12 1.57 1.15E-06 Gdf11 1.32 5.08E-03 Dab2 -0.69 7.66E-03 Cd209f -1.44 4.30E-10 Gcc2 1.56 8.77E-06 Trip11 1.32 8.01E-05 Cpa3 -0.70 5.44E-04 Hba-a1,Hba-a2-1.46 4.70E-08 Upf3b 1.56 8.18E-05 Prdm2 1.32 9.64E-08 Irf7 -0.70 2.24E-02 Hbq1b -1.46 1.32E-03 Ciita 1.56 1.60E-10 Clspn 1.32 1.11E-02 Wfdc17 -0.71 4.81E-03 Klk1b11 -1.48 6.47E-03 Med13l 1.56 5.44E-09 Bclaf1 1.31 1.71E-07 Gng11 -0.71 1.40E-03 Ifit3 -1.55 1.10E-09 Gon4l 1.55 2.44E-06 E2f8 1.31 4.14E-03 Ccl12 -0.72 2.18E-02 I830012O16Rik-1.56 1.07E-09 Kif21b 1.55 3.22E-12 Sfswap 1.31 2.18E-04 Jam3 -0.74 3.15E-04 Hba-a2 -1.56 3.32E-03 Myh9 1.54 0.00E+00 Atad2 1.31 1.06E-05 Mgp -0.75 2.38E-02 Snca -1.66 1.42E-02 Ckap5 1.54 1.99E-07 Hivep3 1.31 5.89E-04 Cfhr2 -0.76 3.17E-03 Isg15 -1.67 2.82E-09 Igfbp2 1.53 9.29E-03 Dnmt1 1.31 7.29E-06 Aldh1l1 -0.76 2.46E-02 Thrsp -1.76 2.34E-09 Tph2 1.53 2.79E-03 Igfbp5 1.31 4.41E-11 Efna1 -0.76 2.02E-02 Fos -1.78 0.00E+00 Nktr 1.53 1.92E-07 Zfp407 1.31 5.74E-05 Ninj1 -0.76 1.41E-04 Smpx -1.81 4.54E-02 Myh14 1.53 2.34E-07 Esf1 1.31 2.39E-04 Nrep -0.76 4.95E-02 Ppbp -1.81 2.65E-03 Atp10d 1.53 1.23E-04 Rexo1 1.31 6.78E-05 Fcer1a -0.76 7.11E-03 Hbb-b2 -2.31 2.28E-04 Cspp1 1.53 1.16E-03 Dot1l 1.31 2.09E-04 Sepp1 -0.77 1.76E-03 Myoz1 -2.54 3.86E-02 Clip1 1.53 1.50E-07 Atp2b4 1.30 4.19E-06 Hbb-b1,Hbb-b2-0.77 4.61E-02 Pvalb -2.67 4.86E-06 Dlgap5 1.52 5.58E-03 Dzip1 1.30 1.27E-02 Ces1d -0.77 1.49E-02 Cyp2e1 -2.68 0.00E+00 Kat6a 1.51 2.02E-12 Brpf3 1.30 3.73E-05 Oas3 -0.77 4.97E-02 Tnnc2 -4.20 1.04E-05 Supplementary Table 1 List of top upregulated and downregulated genes in obese lymph nodes Popliteal lymph nodes were isolated from naïve obese/T2D and control mice and whole transcriptome analysis performed (see Fig 7 for analysis). Shown are the top 200 significantly upregulated and downregulated genes in obese/T2D mice relative to lean-fed control mice. n= 3 mice per group.