Natural IgM Anti-Leukocyte Auto-Antibodies (IgM-ALA) Attenuate Excess Inflammation Mediated by Innate and Adaptive Immune Mechanisms Involving TH-17

Abstract

Little is known about the function of natural IgM auto-antibodies and especially IgM with anti-leukocyte reactivity (IgM-ALA). Natural IgM-ALA auto-antibodies are present at birth and characteristically increase during inflammatory and infective conditions. Our prior clinical observations and those of others showing less rejections in renal and cardiac allografts transplanted into recipients with high levels of IgM-ALA, led us to investigate if IgM-ALA regulate the inflammatory response. Here we show that IgM, in physiologic doses, inhibit pro-inflammatory cells from proliferating and producing IFN-γ and IL-17 in response to alloantigens (MLR), anti-CD3 and the glycolipid alpha-gal ceramide. We show in an IgMko murine model, with intact B cells and Tregs, that there is more severe inflammation and loss of function in absence of IgM after renal ischemia reperfusion injury (IRI) and cardiac allograft rejection. Replenishing IgM in IgMko or increasing the levels of IgM-ALA in WT-B6 mice significantly attenuated the inflammation in both these inflammatory models which involve IFN-γ and IL-17. The protective effect on renal IRI wasnot observed using IgM pre-adsorbed with leukocytes to remove IgM-ALA. We provide data to show that the anti-inflammatory effect of IgM is in part mediated by inhibiting TLR4 induced NF-kB translocation into the nucleus and inhibiting differentiation of activated T cells into TH-1 and TH-17 cells. These observations highlight the importance of IgM-ALA in regulating excess inflammation mediated by both innate and adaptive immune mechanisms and where the inflammatory response involves TH-17 cells that are not effectively regulated by Tregs.

INTRODUCTION

The physiologic relevance of natural IgM auto-antibodies and the IgM subset that bind to leukocyte receptors remains to be elucidated. Prior studies on natural IgM with binding reactivity to leukocytes (IgM-ALA) have been reviewed by us (1). Briefly, IgM-ALA were initially discovered because of their binding reactivity to lymphocytes. These IgM auto-antibodies and the B-1 lymphocytes that produce them can be found in the umbilical cord blood, prior to exposure to foreign antigens, and hence such antibodies are referred to as “naturally occurring” or “natural” IgM. Such auto-antibodies that bind to leukocyte receptors (IgM-ALA) are present at low levels in normal individuals and increase during inflammatory disorders and various infections, including HIV-1. Previous studies in our laboratory and those of others have demonstrated that IgM-ALA are a heterogeneous group of several different antibodies that are reactive to different receptors present on autologous and allogeneic leukocytes and other cells that express leukocyte receptors (1). IgM-ALA have been shown to bind to various undefined membrane receptors comprising glycoproteins, phospholipids and glycolipids (1). Such naturally occurring IgM auto-antibodies are encoded by minimally or non-mutated germline genes and hence are characteristically polyreactive with low binding affinity. Of particular importance, these IgM-ALA do not mediate cytolysis in the presence of complement at body temperature. Naturally occurring IgM differ from disease producing auto-antibodies, in that the latter are predominantly of the IgG isotype, bind with high affinity and specificity to the auto-antigen and mediate cytolysis at 37°C.

Human kidney and heart transplants performed in the subset of patients having high levels of IgM-ALA have been shown to have a lower incidence of acute rejections and of less severity, thus permitting better graft survival (1-6). This observation showing a strong co-relation between high levels of IgM-ALA and protection from allograft rejection together with the finding that IgM-ALA are non-cytolytic to leukocytes at body temperature and increase in various inflammatory and infective states, led us to investigate if IgM-ALA had a regulatory role in attenuating inflammation mediated by innate and adaptive immune mechanisms. We hypothesized that IgM-ALA could bind to different cell membrane receptors, i.e. receptors that initiate and activate the inflammatory process, as well as receptors that are important in enhancing chemokine production and facilitating chemotaxis. Several observations favored such a hypothesis. Firstly, our studies with human B cell clones derived from umbilical cord, clearly demonstrated that only 10 percent of IgM secreting clones had IgM-ALA reactivity and that IgM-ALA from these clones had different receptor specificities (1). Secondly, we showed that IgM isolated from human serum immuno-precipitated CD3 and CD4, inhibit T cell activation/proliferation, and inhibit leukocyte production of certain cytokines, e.g. TNF-α. Additionally, we showed that human IgM immunoprecipitated chemokine receptors, e.g. CXCR4 and CCR5, and inhibited the binding of chemokines (and HIV-1) to these receptors, as well as inhibited chemotaxis induced by chemokines (1,7).

These in-vitro findings with human leukocytes showing that IgM inhibits T cell function and leukocyte chemotaxis, as well as the clinical observations on human transplants, prompted us to study the inhibitory role of IgM in two murine models of acute inflammation, i.e. heart allograft rejection where inflammation is mediated by allopeptide activated T cells and kidney ischemia reperfusion injury (IRI) where inflammation is mediated by NK and NKT cells that are activated by endogenous ligands released after ischemia. In these studies, we show that C57/BL6 mice (referred to as B6) have IgM-ALA. In in-vitro studies, we show that purified murine serum IgM, but not IgM pre-absorbed with splenic leukocytes to remove IgM-ALA, inhibits the production and activation of pro-inflammatory cells involved in innate and adaptive immune responses as well as inhibits TLR-4, which initiates the inflammatory process. Secondly, we show that allograft rejection and kidney IRI is more severe in B6/S4-IgMko mice even though they possess competent regulatory T cells (i.e. Tregs) and administering physiologic doses of IgM to B6/S4-IgMko mice ameliorates this inflammatory response. More importantly, our studies show that WT-B6 mice can also be protected from renal IRI and cardiac allograft rejection when these mice are pretreated with IgM to augment their IgM-ALA levels. Finally, we show that IgM inhibits TH-17 cell differentiation even when added 48 hours after activation, and lack of such a mechanism in IgMko could explain the severity of the inflammatory process, especially involving TH-17 cells, even though these IgMko mice have adequate levels of functional Tregs.

MATERIALS AND METHODS

IgM and IgG purification from sera

IgM and IgG was purified by size exclusion column chromatography (Sephacryl S-300 HR) from heat-inactivated (56°C for 1 hour) WT-B6 murine sera (Innovative Research, MI) using previously described procedures and with modifications detailed below (1). IgM was not isolated by dialyzing sera in water or by ammonium chloride precipitation as both these techniques yielded IgM with impaired functional activity. Column purified IgM was absorbed with protein G and re-passaged through Sephacryl S-300 to remove other contaminating proteins. With this approach more than 92% of the protein fraction contained IgM as determined with protein electrophoresis and there was <3% IgG and IgA contamination as identified with ELISA. Purified IgM and IgG was concentrated to 1.3-1.5mg/ml (higher concentrations lend to IgM aggregation and precipitation), dialyzed against RPMI-1640 and then millipored prior to use in cultures and for in-vivo use. Purified IgM was stored at 4°C to prevent precipitation of IgM when frozen. The effect of IgM on in-vitro cultures was dose dependent and the maximum effect of IgM was observed using IgM at a physiological dose of 5 to 15μg/250 × 10³ cells/0.5 ml of culture.

Adsorption of IgM with splenic leukocytes

Aliquots of 1mg IgM in 4.0 ml RPMI were absorbed at 37°C for 45 minutes with 1×10¹² splenic leukocytes that were pre-activated for 24 hours with sol anti-CD3 and LPS. Pre-activation of leukocytes increased receptor expression while adsorption at 37°C prevented IgM mediated cell cytolysis which occurs at colder temperatures, e.g. room temperature. About 65% of IgM was recovered after the adsorption procedure. Adsorbed IgM was re-passaged through Sephacryl S-300 HR column to remove cytokines and other contaminants resulting from the absorption procedure. Absorbed IgM was redialyzed in RPMI 1640 prior to use.

Mice and surgical protocol for kidney IRI and heart transplantation

All experiments were performed in accordance with NIH and Institutional Animal Care and Use Guidelines. The Animal Research Committee of the University of Virginia approved all procedures and protocols. We obtained BALB/c mice, C57BL/6 mice (WT-B6) either expressing CD45.1 or CD45.2 on their leukocytes, and B6-bm12 mice (Strain B6(C)-H2-Ab1bm12/KhEgJ) incompatible at MHC-Class II (Ia) with WT-B6, were obtained from Jackson Labs (Bar Harbor, ME). B6/S4-IgMko mice were derived from a background of C57BL/6 and 129S4 mice (StrainB6; 129S4-Igh-6tm1Che/J) and obtained from Jackson Labs (8). There was no detectable endogenous IgM in the sera of B6/S4-IgMko using an ELISA technique nor could we detect the presence of IgM-ALA when B6/S4-IgMko sera was added to autologous or allogeneic leukocytes (data not shown). In experiments involving B6/S4-IgMko mice, we used as WT controls mice with the same C57BL/6 and 129S4 background (referred to as WT-B6/S4) which were obtained by breeding heterozygous pairs. Foxp3^EGFP mice on a C57BL/6 background, co-expressing EGFP and Foxp3 under the control of endogenous promoter, were obtained from Jackson Lab. All experiments were performed on 6-8 week old male mice, weighing 20g.

Kidney IRI was performed under anesthesia with bilateral flank incisions as we have previously described (9). Both kidney pedicles were exposed and cross-clamped for either 26 minutes (mild ischemia) or 32 minutes (severe ischemia) and then clamps were released and kidneys were reperfused for 24 hours. Body temperature was maintained at 35 to 36C° (rectal temp) during surgery with heating pads. Kidney pedicles were exposed, but not clamped in sham-operated mice. Heart transplantation was performed according to previously described techniques (10). Briefly, the donor heart was harvested after heparinization and preserved in saline at 4°C. Following a mid-line laparotomy, the recipient inferior vena cava and aorta were cross-clamped proximal and distal to the anastomotic site. Donor aorta was anastomosed end to side to the recipient aorta and donor pulmonary artery was anastomosed end to side to the abdominal inferior vena cava. Post operatively, mice were administered analgesia and maintained at 35 to 37°C prior to returning to the vivarium.

IgM and IgG administration to mice

IgM predialyzed in RPMI 1640 was warmed at 37°C and administered in the tail vein of mice that were also kept pre-warmed at 37°C so as to avoid IgM mediated cytolysis in-vivo at colder temperatures. In experiments, requiring repeated doses, IgM was given every 48 hours intra-peritonealy. Mice received 150 to 200μg IgM with each dose and at this dose the serum IgM increased by 150 to 200 μg/ml. Lower doses of IgM did not inhibit the inflammatory response induced by IRI or rejection. Normal levels of IgM in serum of WT-B6 varies from 200 to 420μg/ml.

Assessment of kidney function and histology

Plasma creatinine was determined using a colorimetric assay according to the manufacturer’s protocol (Sigma) (11). For histology, kidneys were fixed in 0.2% sodium periodate – 1.4% DL-lysine-4% paraformaldehyde in 0.1 MPBS, pH 7.4, 4% PLP and then embedded in paraffin. Kidney sections (4μm) were stained with H and E and we quantitated tubular injury using a scoring system described previously (12).

Characterization of leukocytes eluted from kidney

Flow cytometry was used to analyze kidney leukocyte content 24 hours post reperfusion (9). Briefly, kidneys were weighed, minced and digested with collagnase type 1A in EDTA. Cells were isolated by passage through a cotton column treated with 10% FCS. Cells were pretreated with anti-mouse CD16/32 (clone 2.4G2) to block for non-specific Fc binding and with 7-aminoactinomycin (7-AAD), to distinguish between live/dead cells. Leukocytes were identified by labeling cells with CD45 (30-F11) and cell suspension was mixed with Caltag Counting Beads to normalize for differences in cell recovery among kidney samples. Surface labeling of leukocyte subsets was performed using antibodies that we have previously described (9). Appropriate fluorochrome conjugated isotype-matched, irrelevant mAbs were used as negative controls. Seven color flow cytometry was used to quantitate leukocytes and their subsets. Flow cytometry data was analyzed with FlowJo software 9.1 (Tree Star Inc., Ashland, OR).

Immunofluoresence staining of heart and kidney sections and endothelial cells

Kidney and cardiac tissue was fixed and frozen as we have previously described in detail for kidney tissue (13). Briefly, these tissues were fixed in 1% PLP (1% paraformaldehyde, 1.4% DL-lysine, 0.2% sodium periodate in 0.1 MPBS, pH7.4) overnight, incubated in 30% sucrose for 48 hours at 4°C and embedded and frozen in OCT (Ted Pella Inc., Redding, CA). Frozen sections (5μm) of kidney or heart were permeabilized with 0.3% Triton X-100, and non-specific binding was blocked with 10% horse serum and anti-mouse CD16/32. Tissue sections were stained with the following conjugated antibodies: rat anti-neutrophil (clone 7/4, Cederlane Labs, Burlington, NC), rat anti Foxp3 (clone FJK-16s, eBioscience, San Diego, CA), rabbit anti-IL17A (polyclonal, Santa Cruz, Biotechnology Inc., Santa Cruz, CA), rat anti-CD31 (clone MEC 13 3, Biolegend, San Diego, CA), rat anti F4/80 (clone BM8, Invitrogen, Carlsbad, CA), rat anti TLR-4/MD-2 (clone MTS 510, eBioscience) rabbit anti-murine JE (MCP-1, polyclonal, PeproTech) and rabbit anti-murine KC (CXCL1, polyclonal, PeproTech, Rocky Hill, NJ). Nuclei were visualized using DAPI. Specimens were mounted with Prolong Gold Antifade Reagent (Molecular Probes, Eugene, OR) and examined using a Zeiss Axiovert 200 microscope with Apo Tome (Zeiss).

LPS activated adherent endothelial cells on glass cover slips were fixed in 4% formaldehyde and then permeabilized and blocked using the same reagents as described for kidney tissue. Intracellular staining was performed with conjugated rabbit anti-NF-kBp69 (clone C22B4, Cell Signaling, Technology Inc., Danvers, MA), phalloidin conjugated actin (Sigma) and antibodies to detect chemokines.

Real-time reverse transcriptase-PCR

Kidney sections were immediately transferred into RNA Later^® (Ambion, Austin, TX). RNA and cDNA were prepared from these reagents as we have described previously (13). Subsequently, real-time reverse transcriptase-PCR was performed using a MyiQTM Single Color Real-Time PCR detection system (BioRad, Hercules, CA). Primers were obtained from Integrated DNA Technologies (Coralville, IA) and primer sequences used to detect IFN-γ, TNF-α and chemokines have been previously described (13). Sample values in triplicate were calculated with normalization to glyceraldehyde -3 – phosphate dehydrogenase (GPDH).

Alloantigen activation of splenocytes (MLR)

We performed a one-way mixed lymphocyte reaction (MLR) where 2.5×10⁵ WT-B6 responder splenocytes were co-cultured with 7.5×10⁵ radiated (3,000 Rad) BALB/c splenocytes, which are fully MHC incompatible with WT-B6. MLR co-cultures were set up in a total of 0.5 ml RPMI culture media which for murine in-vitro cultures were supplemented with 10% FCS, 10mM HEPES buffer, 1mM sodium pyruvate, 2mM L-glutamine, and 50μM mercapto-ethanol. To evaluate the proliferative response, we used WT-B6 responder that were Carboxyfluorescein succinimidyl ester (CFSE) labeled and added to co-cultures. A well described technique was used to label cells with CFSE (See protocol B, Nature protocols, Vol. 2, 2051, 2007). Cells were co-cultured in 5% CO₂ at 37°C for 5 to 6 days prior to quantitating proliferation and intra-cellular cytokine production. TGFβ, IL-6, IL-21, IL-23 (2ng each) were added to MLR co-cultures to maximize TH-17 differentiation whereas IL-4 (2ng) was added to maximize TH-2 differentiation. Intracytoplasmic staining of cytokines was performed on cells that were incubated with ionomycin/PdBU (500μg/ml) in presence of brefeldin A for the last 4 hours of the co-culture incubation to inhibit cytokine secretion. Cells were then washed x2 in RPMI, stained for 30 minutes with LIVE/DEAD^® near IR stain according to the InVitrogen kit protocol, fixed with 4% paraformaldehyde, washed x2, permeabilized using Becton-Dickson (Franklin Lakes, NJ) permeabization kit and then stained using the following fluorochrome labeled antibodies, obtained from e-Bioscience: rat anti-mouse IFN-γ (clone XMG1 2), rat anti-IL-17A (clone 17B- 7), rat anti-IL4 (clone 11B11), rat anti-Foxp3 (clone FJK-16s), and mouse anti-CD45.2 (clone 104). A five-color flowcytometer (BD FACS Caliber, BD Biosciences) was used to acquire data, which was analyzed with Flow Jo software 9.1 (Tree Star, Inc.).

Anti-CD3 activation of splenocytes and T cells

Freshly isolated splenocytes or isolated T cells or T cell subsets (2.5 × 10⁵ in 0.5 ml RPMI culture media) were activated with 12μl insoluble anti-CD3/CD28 beads (InVitrogen-Dynal bead, T cell activators) and cells were cultured for 4 to 5 days in 5% CO₂ at 37°C.

In certain experiments, splenocytes (2.5 × 10⁵ in 0.5 ml RPMI culture media) were also activated with LPS (0.35μg/ml) and soluble anti-CD3 (1.0μg/ml) and with these conditions, cells were cultured for 3 to 4 days. We used the same cytokines as described in MLR to maximize TH-2 and TH-17 differentiation. Intracytoplasmic staining of cells and staining of leukocyte subsets was performed as described in the MLR assay.

Alpha-galactosyl-ceramide activation of splenocytes

Alpha-galactosyl-ceramide (100ng) was added to 400,000 splenic leukocytes in 0.5 ml RPMI culture media. Cells were cultured for 48 hours at 37°C in 5% CO₂. IFN-γ in supernatant was quantitated by an ELISA kit.

LPS activation of cultured endothelial cells

Murine glomerular endothelial cells, a kind gift from Dr. Michael Madaio (Medical College of Georgia) were cultured on glass cover slips 2.5 × 10⁵ cells/slip in a tissue culture well containing 1ml DMEM/HamF12 (ratio 3:1) culture media supplemented with 10% FCS and 2mM L-Glutamine (14). IgM (50μg) was added one hour prior to cell activation with LPS (1μg/ml of media), Sigma, 055.B5). After 1 hour, adherent cells on cover slips were fixed, permeabilized and stained with anti-NF-kB (1:50 dil, Cell Signaling). Intracellular chemokine production was evaluated after cells were cultured for 6 to 12 hours.

Isolation of Tregs and EGFP labelled Foxp3⁺ cells

CD4⁺CD25⁺Tregs were isolated from splenocytes using the Miltenyi Biotech Inc., (Auburn, CA) isolation kit which employs magnetic beads (Cat #130-091-041). Isolated cells were >93% CD4⁺CD25⁺ and 70 to 75% of these cells were Foxp3⁺. Magnetic bead isolated CD25⁺ Tregs from splenocytes of Foxp3^EGFP mice were sorted using a flow cytometer to obtain purified (99% pure) Foxp3^+EGFP cells.

Statistics

Graph Pad Instat 3 (GraphPad Inc., LaJolla, CA), SigmaPlot 11.0 (Systat Software Inc.), and Canvas X (ACD Systems of America Inc., Chicago, IL) were used to analyze and present the data. Data were analyzed, after transformation if needed to generate a normal distribution, by 2-tailed t test or 1-or 2-way ANOVA with post-hoc analysis as appropriate. P< 0.05 was used to indicate significance.

RESULTS

Murine serum IgM binds to non-FcIgM receptors expressed on the cell membrane

We wanted to confirm whether WT-B6 and WT-B6/S4 mice have IgM-ALA present in their sera. IgM was purified from serum using size-exclusion chromatography, and interacted with splenic derived pronase treated T cells for 45 minutes at 37°C prior to staining at 4°C for binding of IgM splenic cells (2×10⁶ cells/ml) were pronase digested (250 ug/ml at 37°C for 30 min) to remove FcIgM receptors (15). Data in Figure 1a and 1b clearly shows that purified polyclonal IgM from WT-B6 sera binds to T cells in a dose-dependent manner. IgM binding to macrophages was much more pronounced when compared to T and B cells (data not shown). Adsorbing polyclonal IgM with splenic leukocytes significantly decreased the binding of adsorbed IgM to T cells (Figure 1c). A murine monoclonal IgM (isotype control) failed to bind to leukocytes. Additionally, we could not detect IgM in the sera of B6/S4-IgMko using an ELISA technique nor could we detect the presence of IgM-ALA when B6/S4-IgMko sera was added to autologous or allogeneic leukocytes (data not shown).

Figure 1.

Polyclonal WT IgM bind to membrane receptors on leukocytes. Figure 1a depicts binding of IgM to pronase pretreated splenic CD3⁺T cells in a dose dependent manner. Maximum binding was observed with 10-15μg IgM per 1.5 × 10⁵ splenic leukocytes. Figure 1b depicts immunofluorescence microscopy images of IgM binding to cell membranes of pronased splenic T lymphocytes. Figure 1c compares binding of IgM, leukocyte absorbed IgM (Leu-Abs) and isotype IgM on CD3⁺WT-B6 pronase pretreated splenic leukocytes. Equal quantities (10μg) of IgM were added to 1.5 × 10⁵ splenic leukocytes. Histogram data in Figure 1a and 1c are representative examples from 3 to 4 separate experiments. Figure 1d is a representative example of a Western blot from two separate experiments demonstrating immunoprecipitation by WT-polyclonal IgM of biotinylated membrane proteins from the murine macrophage cell line J77. In this experiment, WT-polyclonal IgM is compared with an equal amount of isotype IgM that has no binding activity to leukocytes using flow cytometry.

IgM purified from murine serum is a polyclonal preparation. Hence, based on our studies with human monoclonal IgM, it became important to determine if polyclonal murine serum IgM, like human IgM, would bind and immunoprecipitate several different receptors expressed on the cell membrane (1). In these studies, we used the strategy of surface biotinylation of leukocyte membrane proteins, followed by immunoprecipitation and western blotting. Initially, we compared the binding of purified serum IgM with mouse cell lines: WEHI7.1 (T cell), Sp2/0 (B cell) and J774 (Macrophage). Based on flow cytometry analysis, maximum binding of IgM was observed with the macrophage cell line J774. Figure 1d shows that murine IgM, purified from serum, immunoprecipitates biotin labeled J774 cell membrane protein shaving different molecular weights. Two different isotype control IgM antibodies did not immunoprecipitate these membrane proteins.

Murine IgM inhibits production of pro-inflammatory cytokines induced by innate and adaptive immune mechanisms in-vitro

Endogenous glycolipids, released by ischemia injured cells initiates inflammation associated with IRI. Endothelial and dendritic cells present these antigens to NK and NKT cells which when activated, augment cytokine production (IFN-γ and IL-17) of infiltrating leukocytes, especially neutrophils (9,13). To test the in-vitro effect of IgM on this innate immune response, we activated splenic leukocytes in the presence of the glycolipid, α-gal-ceramide. As can be seen in Figure 2a, murine polyclonal IgM, but not isotype control IgM, significantly inhibited splenic leukocytes from producing IFN-γ, in response to α gal-ceramide. Inhibition of IFN-γ production was observed even when IgM was added one hour after cells were activated with α-gal-cermide thus indicating that IgM can inhibit post-activation intracellular processes (Figure 2a).

Figure 2.

Polyclonal WT IgM inhibits IFN-γ production and T cell proliferation and differentiation into TH-1 and TH-17 cells. In Figure 2a, 4 × 10⁵ splenic leukocytes were activated with 100ng α-gal-ceramide for 48 hours prior to quantitating IFN-γ in culture media with an ELISA. 10μg IgM was added either 0.5 hours before activation (IgM pre) or one hour post activation. Data from three separate experiments are combined in Figure 2a. In Figures 2b to 2d, CFSE labeled WT-B6 splenocytes (2.5×10⁵ in 0.5ml media) were activated either in a one way MLR (using 7.5×0⁵ BALB/c irradiated splenocytes) or LPS (350ng) and soluble anti-CD3. Cells were cultured for 4 to 5 days. IgM (10 to 15μg) was added at the initiation of culture unless otherwise indicated. In Figure 2d, the effect of Tregs was evaluated by co-culturing 2.5×10⁵ CD45.1 WT-B6 splenic leukocytes, containing 1.8% CD4⁺ Foxp3⁺ cells, with 0.5×10⁵ CD45.2 WT-B6 Tregs (76% Foxp3⁺) under cytokine conditions favoring TH-17 differentiation. Co-cultures were harvested on Day 4. In Figure 2f, flow sorted Foxp3^+EGFPTregs were cultured for 4 days with insoluble anti-CD3/CD28 antibody. 10μg IgM was added on Day 0 to 2×10⁵ cells. In Figure 2f, IgM was added on day 2 and on day 4 dead cells were evaluated with IR Live/Dead stain (InVitrogen). In figure 2g, 40 μL of heat inactivated (56C, 1 hr) and dialysed WT-B6 mouse serum was added to 0.5 mL cell culture on Day 0. Data are representative examples of 4 to 6 separate experiments.

Allograft rejection is mediated by alloantigen activated T cells which proliferate and differentiate into cytokine producing T cell subsets. TH-1 and TH-17 subsets produce pro-inflammatory cytokines (IFN-γ and IL-17), which activate the infiltrating leukocytes (primarily NK cells, macrophages and effector T cells) to mediate the allograft rejection. To test the in-vitro effect of IgM on this adaptive immune response, T cells were activated with either LPS and anti-CD3 antibody or with alloantigens using a one-way mixed MLR where CFSE labeled WT-B6 responder splenocytes were co-cultured with fully MHC incompatible BALB/c radiated splenocytes. As depicted in Figure 2b, purified IgM, significantly inhibited IFN-γ production of the T cell subset activated by alloantigens in the MLR. Additionally, IgM slowed down the proliferative response of the activated T cells. These inhibitory effects of IgM on TH-1 differentiation were not observed using leukocyte adsorbed IgM which is deficient in IgM-ALA (Figure 1c). Interestingly, IgM did not inhibit the compensatory increase in IL4 production that occurs in the setting of decreased IFN-γ production induced by IgM (see Figure 2c). These data would therefore indicate that the inhibitory effect of polyclonal IgM is mediated by the subset of IgM that binds to leukocyte receptors, i.e., IgM-ALA and that murine IgM, like human IgM, inhibits T-cell proliferation and production of certain pro-inflammatory cytokines (1).

We next wanted to determine if polyclonal IgM inhibited T cell differentiation towards TH-17 cells especially in the presence of cytokines that favor maximum TH-17 differentiation. Figure 2d clearly shows that physiological doses of IgM, inhibited activated T cells from differentiating into TH-17 cells. Since Foxp3⁺CD4⁺ Tregs can also differentiate into TH-17 cells (especially in presence of IL-6/21), it became important to determine if IgM can prevent Tregs from differentiating into TH-17 cells under these cytokine conditions. Accordingly, CD 45.1 WT-B6 responder splenocytes were co-cultured with CD 45.2 WT-B6 purified Tregs and then activated with LPS and soluble anti-CD3 in presence of cytokines favoring TH-17 differentiation. As depicted in Figure 2d, CD45.2 Tregs failed to inhibit CD45.1 splenocytes from differentiating into TH-17 cells. Instead, relatively more TH-17 differentiation occurred in CD45.2 Tregs than in CD45.1 splenocytes. Addition of IgM inhibited both CD45.1 splenocytes and CD 45.2 Tregs from differentiating into TH-17 cells and decreased the downregulation of Foxp3 that occurred in CD45.2 Tregs under TH-17 cytokine conditions (Figure 2d). To more conclusively determine the effect of IgM on differentiation of Foxp3⁺Tregcells into TH-17 cells, we obtained 99% purified Foxp3^+EGFP cells by cell sorting and activated these cells with insoluble anti-CD3/CD28 beads under TH-17 conditions. Data in Figure 2e depicts that IgM also inhibits TH-17 differentiation of CD4+Foxp3⁺Treg cells. Isotype IgM failed to inhibit TH-17 differentiation of both non-Tregs and Tregs (data not shown). These inhibitory effects of polyclonal IgM on T cell were not a consequence of isolating and purifying IgM as similar inhibitory effects were observed with WT-B6 sera containing similar quantities of purified IgM (Figure 2g).

Since IgM in all these in-vitro experiments was added at the initiation of cultures, it became important to determine if IgM could also inhibit intra-cellular processes induced after cell activation. Accordingly, IgM was added 48 hours after T cells were activated by alloantigens in an MLR. Data in Figure 2f clearly shows that IgM, when added at 48 hours, inhibits both cell proliferation and TH-17 differentiation of the T cell subset activated by alloantigens in the MLR. Reduction of cell proliferation, in presence of IgM, was also associated with less cell death (Figure 2f). The lack of apoptosis on T cells with physiological doses of IgM and the anti-proliferative effect of IgM on T cells was also observed with human IgM (1,16). IFN-γ production was similarly inhibited by IgM when added at 48 hours (data not shown). These data would therefore indicate that IgM can mediate its inhibitory effect by binding to cell membrane receptors designed to inhibit post activation intracellular processes. However, these data do not negate the possibility of IgM mediated inhibition of early events that occur during T cell activation.

IgM-ALA protects B6 mice from renal IRI

Since polyclonal IgM inhibited the in-vitro activation of NK and NKT cells by α-gal-ceramide, (Figure 2a) we proceeded to determine if IgM inhibited the in-vivo inflammatory response induced by kidney IRI. We employed two approaches to test the protective role of IgM on suppressing this innate inflammatory response. Firstly, we performed renal IRI in B6/S4-IgMko mice. In these studies, rectal temperatures of mice were maintained at 35 to 36°C with heating pad during renal ischemia. The IgM expressing B cells in these IgMko mice cannot secrete IgM, but can differentiate into plasma cells and secrete normal levels of other immunoglobulins. As can be seen from Figure 3a, 26 minutes of ischemia that was insufficient to cause renal IRI in WT-B6/S4 mice, led to severe IRI (as assessed by plasma creatinine and by histology) in B6/S4-IgMko mice. Importantly, intravenous administration of physiologic dose of purified IgM (250μg), given 24 hours prior to ischemia, protected B6/S4-IgMko mice from developing IRI (Figure 3a) thus indicating that it is the IgM deficiency that predisposes these mice to IRI after minimal ischemia.

Figure 3.

Figure 3a – B6/S4-IgMko mice are more sensitive to renal IRI when compared to their WT counterparts (WT-B6/S4). In Figure 3a, kidneys from B6/S4-IgMko mice and their WT counterparts (WT-B6/S4) were subjected to mild ischemia (26 minutes) and then reperfused. Data depict 24 hour plasma creatinine comparing (WT-B6/S4)WT mice with B6/S4-IgMko mice, and B6/S4-IgMko pretreated with 240μg IgM, 24 hours before ischemic injury. Representative data from one of four experiments is also presented comparing H&E staining of renal outer medulla between B6/S4-IgMko and WT-B6/S4 after 24 hours of reperfusion. Figures 3b and 3c -Here studies were done in the WT-B6 mice and not WT-B6/S4 mice. Polyclonal IgM, but not leukcocyte adsorbed IgM (Leu-Ads IgM), protects against renal IRI in WT-B6 mice. In these studies WT-B6 mice were pretreated with equal quantities (150μg in 0.75ml) of IgM or Leu-Ads IgM or IgG, 24 hours before subjecting the kidneys to severe ischemia (32 minutes). Kidneys were reperfused for 24 hours prior to determining plasma creatinine and obtaining kidneys for histology. Control mice were pretreated with 0.75 ml RPMI containing 150μg bovine albumin to exclude variables such as volume/colloid that can protect against ischemic injury. Five to seven mice were used for each group in Figure 3. A student’s t-test was used to calculate p values in Figure 3a while in Figure 3b, a two-way ANOVA test was used. Values are mean ± SEM.

In the second approach, we determined if increasing IgM levels would protect WT-B6 from developing IRI with more severe (32 minutes) renal ischemia. As can be seen from Figure 3b and 3c, intravenous administration of 150μg IgM, 24 hours before ischemic injury, protected WT-B6 mice from severe IRI as determined by plasma creatinine and tubular necrosis score (Figure 3b) and by histology (Figure 3c). This protective effect was mediated by the IgM-ALA subset of polyclonal IgM as leukocyte pre-absorbed IgM had no protective effect on IRI. Most of the renal injury seen after IRI is associated with infiltration of neutrophils and increased production of IFN-γ as seen in mice pretreated with Albumin and Leu-Ads IgM, (see Figure 4). Natural IgG, unlike natural IgM-ALA, binds to leukocytes via its Fc domain, i.e. to FcγR on the cell membrane. Hence we used IgG as another control to determine if other natural antibodies can inhibit inflammatory processes even if they bind to leukocytes via their Fc domain. Data in these experiments would indicate that IgG or FcγR do not inhibit the innate inflammatory processes that occur in renal IRI (see Figure 4).

Figure 4.

Protective effect of IgM, after renal ischemia is associated with lack of leukocyte infiltration into renal parenchyma and with suppression of ischemia induced cytokine/chemokine production. Data presented are obtained from all WT-B6 kidneys in Figure 3b. In Figure 4a, leukocytes eluted from kidneys were quantitated and subjected to flow cytometry to identify neutrophils, dendritic cells and macrophages. Figure 4b depicts immunofluorescence staining of neutrophils (7/4 green), macrophages (F4/80 (white)) and CXCL-1 (red) in outer medulla (x400 magnification). In Figure 4c renal parenchyma mRNA for IFN-γ, TNF-α and CXCL1 are quantitated with values that are relative to GPDH. In Figure 4d CD45⁺ leukocytes eluted from kidneys, were examined by flow cytometry for intracytoplasmic IFN-γ and IL-17. P values, using a two-way ANOVA test, compared albumin control, Leu-Ads IgM with IgM. *p=0.016 **p<0.01. Values are mean ± SEM.

Data in Figure 4 clearly demonstrates that the protective effect mediated by IgM is associated with minimal neutrophil infiltration into the ischemic renal parenchyma (Figure 4a and 4b), as well as with decreased production of CXCL1, which attracts neutrophils into the interstitial space. IgM also prevented IRI induced increase in IFN-γ and IL-17, which have an important role in promoting renal IRI (13). For example, IFN-γ production, as determined by mRNA levels in renal parenchyma was significantly suppressed and levels were lower than in “sham” treated mice (Figure 4c). Additionally, IgM suppressed intracytoplasmic INF-γ and IL-17 of infiltrating leukocytes (Figure 4d). Furthermore, the mRNA levels of TNF-α and CXCL-1 in IgM pretreated mice did not increase after renal IRI and levels were similar to that of “sham” treated mice (Figure 4c).

IgM protects B6 mice from cardiac allograft rejection

Since the in-vitro studies demonstrated that IgM inhibited alloantigen activated T cell proliferation and differentiation into TH-1 and TH-17, we performed studies aimed at determining if IgM could also inhibit allograft rejection which is an in-vivo model of inflammation mediated by alloantigen activated T cells. Two approaches were used. Firstly, cardiac transplants were performed in B6/S4-IgMko mice using B6-bm12 donor hearts which are mostly incompatible at the MHC-Class II locus (Ia). In this transplant model, there is a chronic from of cellular rejection and a vasculopathy that is initiated by a T cell-mediated inflammatory process not involving anti-MHC antibodies (17,18). As a result, rejection induced decrease in cardiac function is detected between Day 17 to 28 when with finger palpation one can clearly detect a diminution in cardiac contractility. This palpation technique has been found to be reliable for detecting significant rejection that impairs cardiac function and was initially described in detail by Corry et al., (10). However, cardiac allograft ceases having a heart beat at 2 to 3 months in this model (17). In Figure 5a, one observes that rejection (as defined by a decrease in cardiac contractility) occurs significantly earlier, i.e. between Day 10 to 18, when B6-bm12 donor hearts are transplanted into B6/S4-IgMko recipients. Furthermore, Figure 5a, clearly shows that by Day 10, the rejection induced inflammatory process in cardiac allografts is severe when transplanted into B6/S4-IgMko recipients but minimal in WT-B6/S4 recipients. Additionally, cardiac allografts in B6/S4-IgMko recipients cease having a heart beat in 2 to 3 weeks, which is significantly earlier when compared to their WT-B6/S4 counterpart, where cessation of heart beat occurs after more than 2 months.

Figure 5.

Figure 5a – Allograft rejection is more rapid and severe in B6/S4-IgMko. B6/S4-IgMko mice and their WT counterparts (WT-B6/S4) received cardiac allografts from B6-bm-12 donors that are only incompatible at the MHC-Ia locus. Figure 5a (left panel) depicts the post-transplant day when cardiac contractility was found to be decreased by finger palpation. Figure 5a (right panel) presents data on B6-bm12 cardiac allograft histology on Day 10 post-transplant. This is a representative example from 3 mice in each group that were sacrificed on Day 10 to 12 post-transplant. Figure 5b – Polyclonal IgM protects WT-B6 mice from allograft rejection. In these studies WT-B6 mice received cardiac allografts from BALB/c donors that are fully MHC incompatible. One group of mice received 175μg IgM on Day 1, 3 and 5 after the cardiac transplant. The control group received 3 doses of RPMI with 175μg bovine albumin. Mice were sacrificed on Day 6. H&E and immunofluorescence staining were performed on cardiac allografts. In the middle panel, sections were stained for neutrophils (7/4, green) and CD3⁺ T cells (white). In the bottom panel, capillary endothelial cells were labeled with CD31 (anti-PCAM, white) and CXCL-1 (red). Representative examples from one of the three mice in each group are depicted.

In the second approach, we wanted to determine if IgM, when administered to WT-B6 mice inhibited the severe and rapid rejection that occurs in the setting of fully MHC incompatible donor hearts, i.e. from BALB/c donors. In this model, rejection in WT-B6 recipients is detectable by Day 5 with finger palpation and the heart ceases having a heart beat by Day 7 to 9 (17). In these studies, 175μg IgM was administered 24 hours after ascertaining that cardiac surgery was successful and the dose of IgM was repeated on Day 3 and 5. Mice were euthanized on Day 6. Figure 5b clearly shows that IgM inhibited the severe inflammation in the cardiac allograft induced by rejection on Day 6 as detected by H&E staining and immunofluorescence staining for neutrophils (7/4) and T cells (CD3). Importantly, with immunohistochemistry, this lack of leukocyte infiltration in the cardiac parenchyma of IgM treated recipients was also associated with decreased CXCL-1 production and with no or minimal fragmentation of capillaries as identified by the endothelial cell marker, CD31 (Figure 5b).

Tregs in IgMko mice are competent but do not effectively suppress TH-17

Since Tregs are the predominant suppressors of inflammatory processes mediated by adaptive immune responses, it became necessary to determine if the enhanced inflammatory response in B6/S4-IgMko results from inadequate Treg levels or function. Such a possibility seemed unlikely as B6/S4-IgMko had similar levels of CD4+CD25⁺ splenic Tregs when compared to their WT counterparts (see Figure 6a) and B6/S4-IgMko Tregs possessed more suppressive activity when compared to their WT counterparts as evidenced by the suppressive ability of isolated CD4⁺CD25⁺ Tregs in inhibiting TH-1 differentiation of splenic leukocytes activated with LPS and anti-CD3 (Figure 6b). In figure 6b, one also observes more TH-17⁺ cell differentiation in setting of adding more Tregs, especially with IgMko Tregs. A potential explanation is that IL17⁺ cells could arise from the added Tregs that differentiate into TH-17 cells (see figure 2d), especially in setting of decreased IFN production induced by Tregs. IFN is known to regulate TH-17 differentiation (19). These in-vitro findings may explain our observations in-vivo, where we observed more severe rejection induced inflammation with TH-17 cells in IgMko recipients (Figure 5A and 6C) despite similar levels of Tregs in cardiac allografts from both IgMko and WT-B6/54 recipients.

Figure 6.

More severe rejection in B6/S4-IgMko is not due to deficient Treg function. Figure 6a compares the percent of CD4+CD25⁺ Tregs present in splenic leukocytes from B6/S4-IgMko and their WT counterparts (WT-B6/S4). In Figure 6b, different doses of magnetic bead purified Tregs from spleens of either WT-B6/S4 mice or IgMko mice were co-cultured with 250,000 WT-B6 splenic leukocytes in the presence of LPS and soluble anti-CD3 and cultured for 4 days prior to quantitating intracytoplasmic IFN-γ and IL-17 with flow cytometry. Data are representative examples from 3 to 4 experiments. Figure 6c – Severe cardiac allograft rejection in B6/S4-IgMko is associated with abundant TH-17 cells, as well as with abundant CD4⁺Foxp3⁺ T cells. B6-bm12 cardiac allografts from mice in Figure 5a were examined by immunohistochemical staining for CD4⁺Foxp3⁺ cells, CD4⁺ cells and TH-17 cells (stained brown). With immunofluoresence microscopy, 12 fields (x400 magnification) were evaluated on each slide to quantitate Foxp3⁺ cells and IL-17⁺ cells. The average number of Foxp3⁺ cells in each field for WT-B6 and IgMko was 17.5 and 18.6 respectively. Average number of IL-17⁺ cells in each field for WT-B6/S4 and IgMko was 3.1 vs 34.6. Representative examples from one of the 3 mice in each group are depicted.

IgM inhibits chemokine production by preventing TLR4 induced NF-kB translocation into the nucleus

Since the protective effect of IgM, in both renal IRI and cardiac allograft rejection, is associated with minimal leukocyte interstitial infiltration, it became important to determine if IgM inhibited early innate events that lead to cell recruitment to injured sites. The initial ischemic injury in both the kidney and donor heart, releases endogenous ligands, which bind to toll-like receptors, in particular TLR-2 and TLR-4 (20,21). Activation and up-regulation of TLR’s present on leukocytes, endothelial and epithelial cells, induces the expression of adhesion molecules (e.g. ICAM-1), chemokines and chemokine receptors that regulate cell migration to the sites of inflammation. Two approaches were used to determine if IgM inhibited activation of TLR’s and chemokine production. In the in-vivo approach, WT-B6 mice, subjected to severe renal IRI (32 minutes), were sacrificed 3 hours after ischemia reperfusion. At 3 hours, the ischemic reperfused kidney had increased TLR4 expression in both renal tubular cells and endothelial cells and with increased production of MCP-1 in renal tubular cells and CXCL-1 in endothelial cells but with minimal or noleukocyte infiltration (See Figure 7b). However, pre-treatment of mice with IgM, 24 hours before ischemia, partially inhibited up-regulation of TLR-4 expression on both endothelial cells and renal tubular cells but significantly reduced production of MC P-1 and CXCL1. Furthermore, pretreatment with IgM prevented the ischemia induced increase in mRNA levels of TLR-4 (Figure 7a).

Figure 7.

IgM protects against ischemia induced leukocyte infiltration in renal IRI by preventing upregulation of TLR-4 and suppressing chemokine production. In these studies, WT-B6 mice were pretreated with IgM or bovine albumin as in Figure 3b, kidneys were subjected to severe ischemia (32 minutes) and after 3 hours of reperfusion, kidneys were examined. Figure 7a depicts mRNA levels of TLR4 in kidney tissue after 3 hours reperfusion. **denotes p<0.01 using two-way ANOVA with 4 mice in each group. Figure 7b are representative examples from the 3 hour reperfusion kidneys. In Figure 7b (upper panel), note that CXCL1 (red) co-localizes with CD31 (white), a marker of endothelial cells. In Figure 7b (lower panel) note that TLR4 (red) is expressed by renal tubular cells and also in endothelial cells (CD31⁺). Co-localization of TLR4 (red) and CD31 (green) in endothelial cells creates a yellow fluorescence. Middle panel in Figure 7b depicts MCP-1 production by tubular cells. Note that albumin pretreated control mice have significantly more CXCL-1, TLR4 and MCP-1 when compared to IgM pretreated mice. Data in Figure 7b are representative examples from one of three mice.

In the second approach, the effect of IgM was tested on cultured renal glomerular endothelial cells, where TLR-4 was activated with the bacterial lipo-polysaccharide ligand (LPS) to induce chemokine production. As can be seen from Figure 8a, IgM (but not leukocyte adsorbed IgM deficient) significantly inhibited TLR4 activation induced translocation of the transcription factor NF-kB into the nucleus. As a result, NF-kB induced CXCL-1 and MCP-1 chemokine production was also inhibited by IgM (Figure 8b and 8c). Data in Figure 7, would indicate that IgM inhibits TLR4 induced NF-kB translocation by preventing upregulation of TLR4 in response to LPS. By flow cytometry, we could not show that IgM inhibited the binding of FITC labeled LPS to murine glomerular endothelial cells or to the murine macrophage cell line J774 (data not shown).

Figure 8.

IgM inhibits TLR4 induced NF-kB translocation into the nucleus of cultured endothelial cells (Figure 8a) with resulting decrease in MCP-1 (Figure 8b) and CXCL1 (Figure 8c) production. Cultured glomerular endothelial cells were activated with LPS in the presence or absence of 50μg IgM which was added 1 hour prior to adding LPS. One hour after LPS activation, cells were fixed to determine if NF-kB had translocated into the nucleus. In Figure 8a, intracytoplasmic NF-kB stains brownish-red and intra-nuclear NF-kB stains purple. Other endothelial cells were fixed at 12 hours to determine the presence of intracytoplasmic MCP-1 by immunofluorsence staining (Figure 8b, red) or to quantitate mRNA levels of CXCL-1 (Figure 8c). **denotes p-value <0.01 using two-way ANOVA. Data are representative examples from one of three separate experiments.

DISCUSSION

In this study we provide evidence to show that polyclonal IgM inhibits inflammatory processes that occur after renal ischemia reperfusion and cardiac allograft rejection indicating therefore that IgM inhibits inflammatory processes mediated by both innate and adaptive immune mechanisms. Mice lacking only secreted IgM (i.e. IgMko) had a more severe inflammatory response to both renal IRI and to cardiac allografts and replenishing IgM in these IgMko mice made them less sensitive to renal injury. Importantly, increasing IgM in WT-B6 made these mice more resistant to renal IRI and to cardiac allograft rejection. Our studies also indicate that the anti-inflammatory effect is mediated, in part, by a subset of polyclonal IgM that binds to leukocyte receptors (IgM-ALA).

The anti-inflammatory protection afforded by IgM was associated with a significant lack of infiltrating leukocytes in the affected organs suggesting therefore that IgM mediates the protective effect by inhibiting processes involved in leukocyte activation and migration. Our in-vitro and in-vivo data would indicate that IgM-ALA inhibits activation of TLR-4 receptors on antigen presenting cells (APC) and endothelial cells. IgM inhibited LPS induced translocation of NF-kB into the endothelial cell nucleus indicating therefore that IgM inhibits processes involved in TLR-4 activation. The transcription factor NF-kB plays a central role in the generation of an inflammatory response as it regulates production of (i) pro-inflammatory cytokines such as IL-1β, TNF-α and IFN-γ; (ii) chemokine production such as CXCL-1(KC), MCP-1 and RANTES; and (iii) expression of adhesion molecules, e.g. LFA-1 and ICAM-1. Leukocyte migration through capillaries is dependent on chemokines and upregulation of adhesion molecules on endothelial cells whereas activation of effectors that mediate the innate immune response (e.g. NK cells) or the adaptive immune response (e.g. TH cells) is dependent on cytokine production and antigen presentation by activated APC and endothelial cells. Data in this study would therefore indicate that IgM inhibits the inflammatory process, through inhibiting up-regulation of TLR4 (Figure 7) and preventing TLR-4 induced activation of APC and endothelial cells (Figure 8). In support of such a premise are other studies clearly showing that absence of TLR-4 or inhibition of TLR-4 activation in mice, will protect mice from renal ischemia reperfusion injury as well as from allograft rejections (20,21,22). IgM-ALA could also inhibit the inflammatory process by other mechanisms. One potential mechanism is that IgM could directly inhibit cell differentiation towards effectors that participate in the adaptive immune response and such a mechanism would explain the protective effect of IgM on cardiac rejection (Figure 5) even though IgM was administered 24 hours post-transplant, i.e., after APC and T cell activation has occurred. Our in-vitro data (Figure 2) would support this possibility as IgM inhibited proliferation and differentiation of cells into TH-1 and TH-17 effectors even when IgM was added 48 hours after T cells were activated in an MLR. Another mechanism is that IgM could inhibit leukocyte chemotaxis, by decreasing chemokine production (see Figure 4, 6, 7) or by inhibiting binding of chemokines to their receptors. We have previously demonstrated that IgM binds to chemokine receptors and inhibits chemokine induced chemotaxis (1).

Natural IgM can inhibit inflammatory processes, including TH-17 mediated inflammation, via several mechanisms. In our in-vitro and in-vivo inflammatory models, we show that IgM-ALA can directly inhibit activation, proliferation, differentiation and migration of inflammatory cells. Others have shown that natural IgM can inhibit altered self-antigen induced inflammation by another mechanism which involves clearing of self-antigens e.g. ds DNA, apoptotic cells and oxidized lipids (reviewed in 23,24,25). In these murine models of inflammation which include SLE, arthritis and atherogenesis, poly reactive monoclonal IgM with specificity for the self-antigen was used to show that natural IgM can ameliorate the inflammatory process. Additionally, in a murine model of complement mediated glomerular inflammation, natural IgM has been shown to ameliorate the inflammatory process by scavenging C3 and C4 (26). It is important to note that many of these anti-inflammatory mechanisms of natural IgM can operate simultaneously in an inflammatory disorder. For example, in SLE, high levels of IgM-ALA have been detected (27). Additionally, in our in-vivo murine models of renal IRI and rejection, we cannot exclude the anti-complement contribution of natural IgM.

Based on our findings, one would expect RAG-1^-/- mice, that lack both IgM and Tregs, and B cell deficient mice (μMT) that lack IgM and other immunoglobins, to be more sensitive towards renal ischemia when compared to their WT counterparts. However, this was not observed (13,28,29). The absence of NKT cell effectors in RAG-1^-/- mice could explain the lack of increased sensitivity to renal IRI in these mice (9). However in μMT mice, the concomitant lack of B cells or their secreted products could have altered the immune responsiveness of T cell effector cells that mediate the inflammatory process. There is evidence to support such an explanation in other inflammatory models involving μMT mice where effector T cell function was found to be significantly reduced (30,31,32). Such findings may explain why Burne-Taney et al., failed to observe renal IRI in their μMT mice, despite the presence in the renal parenchyma of infiltrating leukocytes and T cells which were quantitatively similar to those in WT controls with renal IRI (29). The IgMko mice that we used differs from the μMT mice in that the IgMko mice have intact B cells expressing IgM on their cell membrane but lack the ability to secrete IgM but not other immunoglobulins. Additionally, function of T cells and other effector cells in IgMko is not reduced as evidenced by the heightened inflammatory response in these mice after IRI or an allograft transplant.

Data in our studies differ from that of Thurman’s laboratory, where B1 cells producing natural IgM were depleted by lysing intraperitoneal B1 cells with distilled water (33). In their studies, reduction of peritoneal B1 cells did not worsen the ischemia induced acute renal injury. It is possible that their technique did not deplete the subset of B1 cells that produce IgM-ALA. In their studies, mice were not evaluated for IgM-ALA.

Data in our studies also differ from studies involving ischemia reperfusion injury of bowel, skeletal muscle or cardiac muscle in rodents (34,35,36). In these ischemic rodent models, IgM was found to be pathogenic and worsened the inflammatory process leading to more reperfusion injury. A likely explanation is that the mechanism of ischemia reperfusion injury (IRI) for these organsis not the same as that in the kidney, especially since these investigators convincingly showed that ischemia induced a complement mediated inflammatory response which occurs after a specific clone of natural IgM Ab binds to a neo-antigen which is exposed by ischemic injury. Furthermore, they show that this neo-antigen is common to bowel, skeletal muscle and cardiac muscle. IgM binding and C4/C3 (classical pathway) are readily detectable in these tissues (34,37). RAG-1^-/- mice, which lack B cells and IgM, are protected from IRI of bowel and cardiac muscle. In murine renal IRI, IgM is not pathogenic and one does not detect IgM binding to the tubular basement membrane or to other tubular structures. Additionally, RAG-1^-/- mice are not protected from renal IRI (13,28). Furthermore, peritubular C3 deposition in murine renal IRI does not involve natural IgM, but results from ischaemia induced tubular injury that activates complement via the alternate complement pathway independently of antibody (38,39).

Further studies are clearly necessary to better define the anti-inflammatory role of IgM-ALA in the context of Tregs, especially with suppression of excess inflammation mediated by innate and adaptive immune mechanisms (reviewed in 40,41). Prior observations would indicate that the lack of IgM in B6/S4-IgMko mice does not predispose these mice to T cell-mediated autoimmune diseases as seen with Treg deficiency in both humans and mice, implying therefore that Tregs in IgMko, do not need IgM to control the type of T cells that are autoagressive (8,42). However, our studies indicate that regulation of excess inflammation involving TH-17 effectors, as for example in allograft rejection, requires both Tregs and high levels of IgM-ALA. In both humans and rodents, the rejection induced inflammatory response involves TH-17 effectors primarily because IL-6 and IL-21 in the inflammatory environment enhances TH-17 differentiation and hence, in this environment, polyclonal natural Tregs and TGFβ cannot effectively suppress TH-17 differentiation (reviewed in 43,44,45). Furthermore, natural Tregs can convert into TH-17 cells and become pro-inflammatory with the amount of IL-6 present in the inflammatory environment (19,46,47). This latter observation has led to some concern regarding the use of cultured Tregs for therapy. Data in Figure 2d and 2e would suggest that natural IgM can directly inhibit TH-17 differentiation and prevent Foxp3⁺Tregs from becoming pro-inflammatory under TH-17 inflammatory conditions.

Our studies in mice (Figure 5) and observations in human renal transplants (reviewed in ref#1) would indicate that high IgM-ALA levels can, in the presence of Tregs, attenuate severe inflammatory processes, e.g., in rejection, by inhibiting TH-1 and TH-17 proliferation and differentiation. Allograft rejection of several different organs in humans also involves TH-17 cells thus making our observations with IgM even more relevant (48,49). High IgM-ALA levels can be naturally induced as we observed in a subset of end-stage renal dialysis patients or achieved by administering IgM as we did to protect WT-B6 mice, with competent Tregs, from developing an aggressive rejection with fully incompatible BALB/c cardiac allografts where rejection in this model also involves IL-17 (50).

These studies, prompted by observations in human renal and cardiac transplants, highlight the importance of naturally occurring IgM-ALA in providing another mechanism to regulate excess inflammation involving IL-17 producing cells and mediated by both innate and adaptive immune mechanisms and could explain why these antibodies (i) increase in various infective and other inflammatory conditions and (ii) are designed to be non-cytolytic to leukocytes at 37°C despite the presence of complement. Such antibodies could pre-emptively be used either alone or together with Tregs to prevent renal IRI and to treat severe inflammatory disorders involving TH-17 cells.