White adipose tissue (WAT), long regarded as a tissue with only one purpose (storing energy), is now known to exert immune function [1
] and regulate inflammation [3
]. In concordance with its immune role, a heterogeneous population of innate and adaptive immune cells resides in WAT, including different subsets of macrophages, T-cells, natural killer (NK) cells and B-cells [5
The role of WAT in the immune response to viral infections is of interest because this tissue is ubiquitously distributed and makes up a large percentage of the body weight of both the average person [6
] and in vivo animal models such as the mouse [8
]. Innate receptors involved in antiviral responses such as Toll-like receptor 3, a pattern recognition receptor (PRR) that recognizes viral dsRNA and its synthetic homologue polyinosinic:polycytidylic acid (poly(I:C)), are expressed in human adipocytes [9
]. Interestingly, viruses such as the human immunodeficiency virus have developed mechanisms to survive within WAT when they are threatened, for example, by antiretroviral therapy [10
]. Likewise, adenovirus infections have been consistently linked to obesity and the expansion of WAT [11
Despite these recent advances, the immune role of WAT during viral infections is poorly understood. An interesting trait of WAT immune responses is that its endogenous cytokine production is skewed towards generating anti-inflammatory molecules such as interleukin-10 (IL-10) [12
] or IL-1 receptor antagonist (IL-1Ra) [13
]. Accordingly, WAT harbors a relatively abundant population of regulatory B-cells (Bregs), characterized by their ability to produce IL-10 [14
]. This WAT anti-inflammatory biased immune function could be particularly important in the presence of viruses of the family Herpesviridae
. The human cytomegalovirus (HCMV) and murine cytomegalovirus (MCMV) are both herpesviruses and have evolved to induce the expression of IL-10 in their respective hosts. Furthermore, HCMV encodes its own functional IL-10-like molecule [15
]. Interleukin-10 is able to suppress the immune response enabling both HCMV and MCMV to establish life-long infections in their hosts [16
In this study, we hypothesized that the ability of mouse WAT to produce IL-10 is strongly modulated in response to viral stimuli, and perhaps, cells specialized in production of IL-10 present in WAT could be involved. To test our hypothesis, we conducted in vivo and ex vivo experiments to measure production of IL-10 and other cytokines in gonadal white adipose tissue (GWAT), located next to the uterus and ovaries in female mice, and compared it to other organs. Our results show that, in comparison to spleen, liver, and uterus, GWAT depots in female C57Bl/6J mice produced large amounts of IL-10 and IL-21 in response to poly(I:C) in vivo via a TLR3-dependent pathway. In ex vivo GWAT explant cultures, IL-10-producing Bregs increased production of IL-10 in response to poly(I:C). In contrast, infection with MCMV down-regulated the expression of IL-10 and increased production of inflammatory cytokines IL-6 and TNFα. This change in IL-10 expression correlated with a reduction in the total number of B-cells, as well as the Breg subpopulation. To our knowledge, this is the first report that links viral pathogen-associated molecular patterns (PAMPs) and MCMV with the regulation of IL-10 in WAT, potentially via Bregs.
2. Materials and Methods
Female C57Bl/6J and TLR3 KO (6–8 weeks) mice were bred in house at the Reid Animal Facility, University of South Australia, and were housed in individually ventilated cages under standard specific-pathogen free conditions with food and water provided ad libitum. All animal experiments were carried out with dual approval from the University of South Australia and The University of Adelaide Animal Ethics Committees and conducted in accordance with National and Institutional ethical and regulatory guidelines.
Mouse fibroblast cell line 3T3-Swiss albino (ATCC, Manassas, VA, USA, CCL-92) was used to propagate and titrate murine cytomegalovirus. During the experiments 3T3-cells were passaged every three days and maintained in Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma, Castle Hill, Australia, D5671) supplemented with penicillin streptomycin solution (GIBCO, Waltham MA, USA, 15140-122) 10 mM of HEPES (Sigma, H0887), 2 mM of L-Glutamine (GIBCO, 25030-081) and 10% of heat inactivated (56 °C for 30 min) fetal bovine serum (FBS; GIBCO, 26140-079).
The K181 strain of murine cytomegalovirus was propagated by infecting confluent monolayers of 3T3-cells in DMEM medium supplemented with 2% heat inactivated FBS. Three days post-infection (dpi) the media was collected and frozen, and fresh media was added to infected flasks. On day five after infection the media was collected and pooled with media from 3 dpi. The remaining adhered cells were scraped from the bottom of the flasks, resuspended in DMEM and sonicated (30 W twice for 5 s) to release intracellular mature MCMV virus particles. The sonicated homogenate and pooled infection media were then mixed and centrifuged (5000× g for 20 min at 4 °C) to remove cell debris. The clarified medium was then mixed with 2.3% w/v of NaCl and 5% w/v of Polyethylene glycol Mw 6000 (PEG; SIGMA, 81255) with constant agitation for one hour at 4 °C. The virus was then allowed to precipitate with PEG overnight at 4 °C. To collect the virus the medium was centrifuged at 15,000× g for 20 min at 4 °C and the PEG-virus pellets were resuspended in TES buffer (0.01 M Tris-HCI, 0.002 M EDTA and 0.15 M NaCI) at pH 7–7.2. Excess PEG was removed by centrifugation and the virus kept at −80 °C until used.
For virus titrations 2× MEM culture medium was prepared from MEM powder (Sigma M0268) as indicated by the manufacturer. Ninety percent confluent 3T3-cells in 24-well plates were infected with 100 µL of 10-fold serial dilutions (in complete DMEM with 2% FBS) of infected supernatants or tissue homogenates for 1 h at 37 °C rocking the plates intermittently. Then, the inocula were removed and a 1:1 mixture of 2× MEM with 4% FBS and a sterile 1% agar solution was added to the wells. Five days after infection a 10% buffered formalin solution was added to the wells and allowed to slowly diffuse and fix the infected cell monolayers over a 24 h period. The medium was then removed and cells were stained with crystal violet to facilitate virus plaque count.
2.4. In Vivo Poly(I:C) Injections and Organ Collection for Cytokine Evaluation
Female C57Bl/6J or TLR3 KO mice were injected with 20 mg/kg of poly(I:C) (Sigma, Castle Hill, Australia P1530). At 12 h after the injection all mice were exposed to isoflurane until unconscious. Then while still under the anesthetic the thoracic cage was open to expose the beating heart. A cut was performed in the right atrium and 15 mL of ice-cold PBS was injected slowly through the left ventricle to replace circulating blood with the saline solution. After perfusion all tissues were collected and placed in 1 mL of tissue lysis buffer (150 mM of NaCl, 50 mM of Tris-HCl at pH 8, 1% triton x100, and a freshly added protease inhibitor cocktail; Roche, Sydney, Australia 11873580001). All samples were minced with sterile scissors and incubated in the lysis buffer for 30–40 min with 10–20 s vortexing every ten minutes. Then, the resulting tissue homogenates were centrifuged at 14,000× g for 10 min and supernatants were collected and kept at −80 °C for protein analysis. Throughout the process all tissues were kept at 4–6 °C.
2.5. Adipose Tissue Explant Cultures and MCMV Infections
The estrous cycle phase (i.e., metestrus, diestrus, proestrus, or estrus) for each mouse was determined by observation of cell smears from vaginal lavages. Mice were euthanized with CO2 only if in the estrus phase. Immediately after, the mice abdomens were open in sterile conditions. The gonadal adipose tissue pads (GWAT), attached to the uterine neck and lining the uterine horns were excised and placed in complete DMEM culture medium at room temperature. All GWAT pads were thoroughly examined to ensure that they were free of lymph nodes. GWAT from each mouse was then cut into approximately 2 mm pieces and placed in sterile PBS with 5% FBS. Then PBS was poured through a cell strainer to collect the GWAT. This process was repeated three times to wash the tissue explants. Between four to six explants were placed in a single well in 6-well plates containing 4 mL of complete DMEM medium. The explant cultures were allowed to rest in the culture media for at least 24 h before any treatment or infection was performed on them.
For MCMV infections, explants were incubated for 90 min in the presence of 100,000 PFU/mL of MCMV (K181) or culture media (complete DMEM +10% ΔFBS, control) in a total volume of 4 mL well. During incubation the plates were gently rocked every 15 min to enhance virus contact with the explants. At the end of the incubation period media were removed and all wells were washed with 2 mL of DMEM +10% ΔFBS. Then 4 mL of fresh culture medium was added. At 1, 2, and 3 days post-infection 400 µL of the medium was collected from each well for cytokine evaluation and replaced with same amount of fresh medium.
2.6. Cytokine Evaluation
Different LEGENDplexTM bead-array kits (BioLegend, San Diego, CA, USA) were used to detect mouse cytokines based on sandwich ELISA principles. Cytokine concentration was determined using standard curves obtained using recombinant cytokine standards provided in each kit. Bead signal readouts were obtained using the flow cytometer Becton DickinsonFACSAria™ Fusion and analyzed using LEGENDplex data analysis software.
2.7. Isolation of GWAT-Associated Stromal Vascular Fraction (SVF), Cell Staining, and Flow Cytometric Analysis
Cells of the SVF, containing GWAT-resident immune cells, were harvested from GWAT explants via enzymatic dissociation. To that end, fat pads were collected from the treatment wells and transferred into a 30 mm petri plate. After mincing the fat pads into smaller portions, the tissue was digested in 1 mL of the enzymatic mix composed of 900 μL of 2% FBS Roswell Park Memorial Institute (RPMI) cell culture medium, 100 μL of collagenase type I (10 mg/mL; Gibco Invitrogen, Carlsbad, CA, USA), and 1 μL of DNase I (4mg/mL; Roche, Basel, Switzerland). The petri dish was sealed and placed in an orbital shaker at 37 °C for 30 min with gentle rocking. Post digestion, the dissociated tissues were filtered through a 70 μm cell strainer to remove cell and tissue clumps then washed twice with PBS. The pelleted cells were then resuspended in 1 mL of Ammonium-Chloride-Potassium (ACK) lysis buffer (150 mM NH4Cl, 10 mM KHCO3, 0.1 mM Na2EDTA, pH 7.2) and incubated at room temperature for 5 min to eliminate red blood cells. Then, cells were washed once with PBS, counted, and dispensed at 1 × 106 cells/well in a V-bottom 96-well plate (Corning Inc, NY, USA).
Multi-parameter flow cytometry was used to identify B- and T-cell populations in SVF and to measure the relative intracellular production of IL-10. The following panel of conjugated monoclonal antibodies (mAb) specific for the following markers were used: B220-BV650 Clone RA3-6B2 (BioLegend, San Diego, CA, USA), CD3e-APCCy7 Clone 145-2C11 (BD Biosciences, NJ, USA), IL-10-PE Clone JES5-16E3 (BioLegend), and Fixable Viability Dye eFluor™ 660 (eBioscience, San Diego, CA, USA). The cells were first incubated with anti-mouse CD16/CD32 (eBioscience) for Fc receptor blockade then stained with the cocktail of mAb in staining buffer (PBS with 2% FBS) for 40 min in the dark on ice. After staining, cells were fixed and permeabilized using the BD Cytofix/Cytoperm Kit (BD Biosciences). The permeabilized cells were then stained with IL-10-PE or the isotype for 30 min at room temperature. Cells were washed and transferred to FACS tubes for flow cytometry analysis. Unstained, single color, fluorescence minus one controls and isotype controls were used to set gates and identify positive populations. All data were analyzed using FlowJo software (Treestar, San Carlos, CA, USA).
2.8. Normalisation of Cytokine and Breg Data
In all in vivo and ex vivo experiments for spleen, liver, uterus, and GWAT tissue samples were weighed using precision laboratory-grade scales. Cytokine levels and cell count (from immune staining assays) readouts were then normalized against tissue weight measured at the analysis time point and expressed as pg of cytokine X/100 mg of tissue or cells/100 mg of tissue to allow comparison between experimental conditions and treatments.
Statistical analysis and number of animals, experiments, or replicates are indicated for each experiment in their corresponding figure legends.
Adipose tissue is one of the most abundant tissues in humans, the majority of it is WAT. The average body of an American man and woman contains ~28% and 40% of adipose tissue, respectively, [6
] and in laboratory mouse strains the percentage of WAT ranges from ~15% to 40% [8
]. Yet, despite its abundance and being ubiquitously distributed, WAT has not received as much attention as other organs when studying immune responses to pathogens such as viruses. We observed that poly(I:C) evoked strong, TLR3-dependent cytokine responses in GWAT depots both in vivo and ex vivo. Others have shown that murine pre-adipocytes and adipocytes mount antiviral-like, pro-inflammatory responses to poly(I:C) [29
] in concordance with our in vivo results. Although other intracellular innate immune cytosolic receptors such as melanoma differentiation-associated 5 (MDA5) and retinoic acid-induced 1 (RIG-1) expressed in WAT cells can participate in sensing poly(I:C) [29
], in vivo systemic exposure to poly(I:C) is more likely to be sensed by TLR3, expressed on the outer cell membrane, as evidenced by the in vivo cytokine expression in TLR3 KO mice.
In vivo, the response of GWAT to poly(I:C) resulted in high expression of IL-21 and IL-10. Interleukin 21 negatively regulates regulatory T-cells in GWAT in male mice favoring chronic inflammation in obese mice [30
]. Investigating the role of IL-21 in modulating regulatory T-cells or other immune cell subsets in WAT and its overall effect on viral replication could be worthwhile in future studies. White adipose tissue responses to bacterial PAMPs have been shown to be skewed towards an anti-inflammatory response characterized by expression of IL-10 [12
] and IL-1Ra [13
] in similarity to our in vivo results for IL-10 upon exposure to poly(I:C). A closer analysis of the cells within the SVF of GWAT explants revealed that a subset of B-cells responded to poly(I:C) increasing intracellular expression of IL-10. In mice, WAT depots are relatively rich in regulatory B-cells that express IL-10 which play an important role in regulating inflammation in obesity [14
]. Others have found that IL-10-expressing cells localized in the peritoneum can be critical for pregnancy success in female mice [32
]. We recently showed that uterine Bregs number are increased in the embryo implantation period in the mouse downregulating potentially harmful T-cells subsets [33
]. Whether Bregs in GWAT depots (in close contact with the uterus) could have a role in pregnancy controlling inflammation during viral infections is worth investigating.
In humans, CMV causes intrauterine infections associated with life-long, severe sequelae in 0.2–2.6% of all babies [34
]. Part of the current understanding of MCMV and HCMV biology is that IL-10 plays a role in the establishment of life-long persistent infections in their respective hosts [16
]. Surprisingly, in our explant model MCMV down-regulated IL-10 expression and secretion in GWAT explants while increasing pro-inflammatory cytokines (TNFα and IL-6). Interleukin-10 expression peaked three days post-infection in vitro, while others have reported that in vivo, murine serum IL-10 peaked on day five post-infection during MCMV infections [17
]. Murine CMV was able to infect GWAT explants in culture. In vivo, MCMV can infect different depots of both white [26
] and brown [27
] adipose tissue depending on the route of infection, including epidydimal adipose tissue, the male homologue of GWAT in female mice [28
]. B-cell number was drastically reduced in MCMV-infected GWAT cultures. While, partly, MCMV could cause cell death during the course of its replicative cycle, the number of Bregs (IL-10 producing B-cells) was reduced further due to infection. Whether MCMV could preferentially infect these cells or whether it simply down-regulates the expression of IL-10 was not determined in this study and it deserves special attention in the future. In MCMV infected white adipose tissue, MCMV-specific CD8 T-cells can infiltrate the tissue and cause inflammation measurable long after infection as recently shown by Contreras et al. [28
]. In their study, the specific role of B-cells within the SVF was not assessed, but their data, similar to ours, suggests that MCMV benefits from debilitating the anti-inflammatory role of WAT.
In summary, this study highlights that WAT and in particular GWAT depots are an abundant source of IL-10 and IL-21 in the context of a viral infection. Moreover, WAT-resident B-cells are likely to play an important role regulating how much IL-10 is produced when exposed to viral PAMPs or whole virus particles. Murine CMV may have repressed the ability of GWAT to produce and release IL-10, highlighting the importance of future studies to investigate the interaction between WAT depots and other tissues during viral infections with MCMV and other viruses.