VSIG4, a B7 family–related protein, is a negative regulator of T cell activation

T cell activation by APCs is positively and negatively regulated by members of the B7 family. We have identified a previously unknown function for B7 family–related protein V-set and Ig domain–containing 4 (VSIG4). In vitro experiments using VSIG4-Ig fusion molecules showed that VSIG4 is a strong negative regulator of murine and human T cell proliferation and IL-2 production. Administration to mice of soluble VSIG4-Ig fusion molecules reduced the induction of T cell responses in vivo and inhibited the production of Th cell–dependent IgG responses. Unlike that of B7 family members, surface expression of VSIG4 was restricted to resting tissue macrophages and absent upon activation by LPS or in autoimmune inflammatory foci. The specific expression of VSIG4 on resting macrophages in tissue suggests that this inhibitory ligand may be important for the maintenance of T cell unresponsiveness in healthy tissues.

T cell responses are regulated by a complex network of activating and inhibitory signals. Recognition of peptides presented by MHC molecules is usually not sufficient for full T cell activation, but additional signals from costimulatory molecules are required (1–4). The most prominent costimulatory molecule expressed on T cells is CD28, interacting with the B7 family members CD80 and CD86 (5, 6). Engagement of CD28 facilitates T cell activation by enhancing TCR-mediated signaling and reducing the number of TCRs that need to be engaged for activation (7, 8). CTLA-4, a close homolog of CD28, also engages CD80 and CD86 (5, 6). Yet it serves a completely different function, since it reduces rather than enhances T cell responses.

Novel members of the CD28/B7 families have been identified recently. ICOS, engaging ICOSL (9, 10), has a function homologous to that of CD28 and generally enhances T cell responses; under some conditions, ICOS stimulation appears to selectively favor induction of Th2 cells (11, 12). Moreover, ICOS has been shown to mediate CD28-independent antiviral responses (13, 14) and to enhance antibody responses and germinal center formation (15, 16). Another new member of the family is the inhibitory receptor programmed death 1 (PD-1), which interacts with PD-ligand 1 (PD-L1) (B7-H1) and PD-L2 (B7-DC) (17–21). PD-1 has a function similar to that of CTLA-4 and downmodulates T cell responses (18, 19). The same is true for BTLA, a CD28 homolog interacting with herpesvirus entry mediator on APCs (22, 23). There are 2 more B7 homologs with unknown receptors on T cells, called B7-H3 (24) and B7-H4 (B7x, B7S1) (22, 25, 26). Their function is less well established. B7-H3 is upregulated upon inflammation and has been suggested to function as both a positive and negative regulator of T cell responses (27, 28). B7-H4 is also expressed on DCs upon activation and is thought to function as a negative regulator (25).

Here we report the identification of a novel function of V-set and Ig domain–containing 4 (VSIG4). In vitro experiments showed that VSIG4 is at least as potent at inhibiting T cell responses as PD-L1. Furthermore, VSIG4 inhibited proliferation of mouse as well as human T cells. In vivo, administration of VSIG4-Ig fusion molecules was able to inhibit the induction of CTL responses as well as the development of Th cell–dependent IgG responses. Hence, VSIG4 is a potent negative regulator of T cell responses.

VSIG4, a B7 family–related protein. Considering the importance of B7 family members as regulators of immune responses, we set out to screen for members of this protein family. To do so, a search using HMMR software was performed in silico on a translated expressed sequence tag (EST) database using a hidden Markov model (HMM) profile of the ectodomain of all known B7 family members. Obtained hits were further narrowed using different filters as outlined in Methods. Two of the remaining hits turned out to be the mouse clones BC025105 and NM_177789, which were almost identical and obviously derived from the same mRNA encoding the protein VSIG4. The protein sequence encoded by these cDNAs displayed about 20% identity and shared conserved amino acids with known B7 family members (Figure 1). Based on this homology, we concluded that VSIG4 was a B7 family–related protein 3. In contrast to the B7 family members, which contain 2 IgG domains, VSIG4 contains 1 complete IgV-type domain and a truncated IgC-type domain.

Figure 1

Sequence and homology of VSIG4. (A) Amino acid sequence alignment of murine VSIG4 with the putative human ortholog Z39Ig. The N-terminal signal sequences determined according to von Heijne (53) are underlined. The 2 Ig domains are in italics, and identical amino acids are indicated with dots. The gaps are indicated with dashes. Bold letters correspond to the predicted transmembrane domain. Overall, the 2 proteins show 44% identity, and within the common extracellular domain (aa 1–139), 78% identity was found between VSIG4 and Z39Ig. (B) Percentage of identity between the extracellular domains of known B7 superfamily members.

Further screening with the mouse sequence led to the identification of the putative human ortholog named Ig superfamily protein 39 (Z39Ig; GeneBank accession number NM_007268). The amino acid sequence of Z39Ig shows 44% identity with mouse VSIG4 (Figure 1A). Although identities in a similar range are observed between human and mouse sequences for B7 family members (Figure 1B), this degree of identity is relatively low. However, the moderate homology may be explained by the different length of the mouse (280 aa) and the human (399 aa) open reading frames. The putative human ortholog contains 2 complete extracellular IgG domains, like the members of the B7 family. Detailed analysis of the common sequence within the extracellular domain (aa 1–139) reveals 78% identity between the mouse and the human protein (Figure 1A). Moreover, Vsig4 and Z39Ig are found in corresponding gene clusters on the mouse and human X chromosome, respectively (data not shown). Although Z39Ig has been described previously by others, little about its biological function is currently known (29–31). Recently, monocytes isolated from PBMCs and mature DCs derived from umbilical cord hematopoietic progenitor cells were shown to express Z39Ig, and Z39Ig has been proposed to be involved in the regulation of immune responses mediated by phagocytosis and/or antigen presentation (32).

In summary, VSIG4 contains typical features of B7 family members, suggesting that it is a B7 family–related protein. The degree of sequence identity and the chromosomal location identify Z39Ig as the human ortholog of VSIG4.

VSIG4 is specifically expressed on resting tissue macrophages. To characterize the expression pattern of Vsig4, the expression of Vsig4 in various tissues and cell types was evaluated by quantitative PCR. Vsig4 was found to be expressed at high levels in the liver, DCs, neutrophils, and macrophages. Low expression was found in various other tissues, including lung, heart, spleen, and lymph nodes, while no Vsig4 expression was found in T and B cells (Figure 2A). Hence, Vsig4 expression appears to be present in macrophages, neutrophils, and DCs and in a variety of peripheral tissues.

Figure 2

VSIG4 is expressed on resting peritoneal macrophages and downregulated upon activation. (A) VSIG4 copies normalized to GAPDH in the respective tissues or cell types are shown. Sm muscle, smooth muscle. (B) Characterization of anti-VSIG4 rabbit serum. Left: EL4 (filled histogram) or EL4-VSIG4 cells (black lines) were stained with anti-VSIG4. Right: Western blot analysis of lysates from 293 and 293-VSIG4 cells with anti-VSIG4. (C) Peritoneal macrophages (Mϕ), monocytes, and neutrophils from blood and spleen and activated and naive DCs were stained with anti-VSIG4 (black lines) and preimmune serum (filled histograms). Gating markers are indicated above the histograms. Half of the CD11b⁺/Gr1^– cells from the peritoneum (resident macrophages) showed VSIG4 expression. (D) Resting peritoneal macrophages and macrophages activated in vivo by thioglycolate (Thio) or LPS for 3 days were isolated, stained for the expression of VSIG4, and analyzed by FACS as described for C. Histograms of the gated CD11b⁺/Gr1^– cell population are shown. (E) Resting peritoneal macrophages and macrophages activated in vivo by thioglycolate and PBMCs from the same animals were isolated and analyzed for VSIG4 expression by Western blotting using polyclonal anti-VSIG4 antibodies. PerC, peritoneal cells. (F) Dot plots of cells isolated from the peritoneum and stained with the indicated cell-surface marker and VSIG4 rabbit serum. (G) Freshly isolated peritoneal macrophages were either stained directly or cultivated in the presence or absence of 10 μg/ml LPS for 3 days and stained with anti-VSIG4 (black lines) or preimmune serum (filled histogram). Histograms of the gated CD11b⁺/Gr1^low cell population are shown. Representative stainings of at least 2 independent experiments are shown.

To determine cell-surface expression of VSIG4, we generated a polyclonal rabbit serum against the extracellular domain of VSIG4. We first tested the specificity of the rabbit serum in cell lines ectopically expressing VSIG4. To this end, EL4 cells transduced with a recombinant retrovirus expressing VSIG4 were generated (EL4-VSIG4). As shown in Figure 2B, VSIG4 surface expression was detected with anti-VSIG4 rabbit serum on EL4-VSIG4 cells, whereas no staining was observed on EL4 cells transduced with empty expression vector. No staining was observed with preimmune serum on either cell population (data not shown). To test whether the serum also recognizes VSIG4 in cell lysates, 293 cells were transiently transfected with a VSIG4 expression vector, and cells were lysed and analyzed by Western blot. A distinct band of 46 kDa was detected in cell lysates of 293 cells transfected with the VSIG4 expression vector, and no signal was observed in nontransfected 293 cells (Figure 2B). Taken together, these data demonstrate that the anti-VSIG4 rabbit serum specifically recognizes surface VSIG4 as well as VSIG4 protein in cell lysates.

Having established that anti-VSIG4 serum was specific, we set out to analyze VSIG4 surface expression in various cell types. Several bone marrow–derived cell types were analyzed for VSIG4 cell-surface expression. In accordance with the expression data shown in Figure 2A, naive and activated T and B cells did not show VSIG4 surface expression (data not shown). Moreover, naive and activated DCs, blood and splenic neutrophils, and granulocytes did not show VSIG4 surface expression (Figure 2C), although some of these cell types were positive for Vsig4 mRNA expression. In contrast, a subpopulation of resting peritoneal macrophages expressed high levels of VSIG4 (Figure 2C). In the next step, we analyzed VSIG4 expression on resting and activated macrophage populations. To this end, VSIG4 expression in resting peritoneal macrophages and peritoneal macrophages obtained after in vivo activation by LPS or thioglycolate for 3 days was assessed. In contrast to resting macrophages, inflammatory macrophages activated in vivo did not show VSIG4 expression (Figure 2D). These results were confirmed by Western blotting, since VSIG4 protein could only be detected in cell lysates of naive peritoneal macrophages, whereas no VSIG4 could be detected in cell lysates of thioglycolate-treated macrophages. PBMCs from control and thioglycolate-injected mice also did not express VSIG4 (Figure 2E). The slightly faster migration of VSIG4 observed in lysates of peritoneal macrophages compared with the one observed in 293 cells (Figure 2B) could be explained by different glycosylation of the protein in the different cell types. Taken together, these results suggest that VSIG4 expression is restricted to resting macrophages and that the expression is lost upon activation. Since VSIG4 was only observed in a subpopulation of resting peritoneal macrophages, we wondered whether the VSIG4-positive population differed from the VSIG4-negative cells in the expression of different surface markers. To this end, cells from the peritoneum were isolated and stained with anti-VSIG4 and anti-CD11b, -GR1, -F480, and –MHC class II antibodies and analyzed by FACS. As shown in Figure 2F, the VSIG4-positive population was CD11b^high, GR1^low, and F480^high, demonstrating that only resident macrophages of the peritoneal lavage are positive for VSIG4 expression. However the VSIG4-positive population did not differ in the expression of either CD86 or MHC class II from the VSIG4-negative population of resident peritoneal macrophages (Figure 2F).

Taken together, these results suggest that VSIG4 expression is restricted to resting macrophages and is lost upon activation. To test this further, resting peritoneal macrophages were isolated from naive mice and either stained directly or cultivated for 3 days in the presence or absence 10 μg/ml LPS. As shown in Figure 2G, VSIG4 expression was reduced upon activation with LPS. Hence, VSIG4 is downregulated in vitro upon activation of the cells. The slightly reduced expression of VSIG4 observed in the cells that had not been activated compared with freshly isolated peritoneal macrophages may be explained by partial activation during cultivation of the cells in tissue culture plates. In summary, these results demonstrate that VSIG4 is expressed on naive resting macrophages and that the expression is lost upon activation of these cells. Moreover, infiltrating activated macrophages did not display detectable VSIG4 expression. The more restricted expression of VSIG4 assessed by antibody staining compared with that assessed by RT-PCR is not without precedent, since mRNA for B7-H1 is found in many tissues, while protein expression is largely restricted to professional APCs (17–21). In addition, most tissues, in particular the liver, contain resting macrophages as a potential source of VSIG4 expression. Indeed, histological analysis showed expression of VSIG4 macrophages in many tissues, including liver, thymus medulla, adipose tissue, and the heart (Figure 3; see also Figure 4 for costaining with CD68 and CD11b). In myocardium, a slightly uneven, at times patchy distribution of VSIG4-positive macrophages was seen. VSIG4 was not detected in intestines, kidney, skeletal muscle, salivary gland, lymph node, splenic white pulp, lung, and brain (data not shown).

Figure 3

VSIG4 is expressed on resting tissue macrophages. Organs of untreated mice were assessed for VSIG4 expression by histology. (A and E) Kupffer cells lining the sinusoids of the liver were evenly positive for VSIG4. (B and F) Occasional macrophages of the red pulp of the spleen were positive, while macrophages of the white pulp were negative for VSIG4 (B). Within the red pulp (F), iron-laden macrophages (a weak granular signal was derived from the iron) were negative (small arrow), while other macrophages were weakly positive for VSIG4 (large arrowheads). (C and G) The myocardium showed an uneven distribution of VSIG4-positive macrophages. VSIG4 was also detected in tissue-resident macrophages of adipose tissue (D). VSIG4 was absent in thymic cortex and detected in rare macrophages of the thymic medulla (H). Representative stainings of at least 2 independent experiments are shown. Original magnification, ×60 (A–D); ×150 (E–H).

Figure 4

Macrophages in autoimmune infiltrates of the heart do not express VSIG4. Myocarditis was induced by immunization with α–myosin heavy chain–derived peptide in CFA (CFA/Myhca), and as a control, untreated mice were used. Heart sections from inflamed hearts and control hearts were stained with VSIG4 (green) together with CD68, MHC class II (MHC II), or CD11b (red) as indicated. In untreated mice, a fraction of CD68- or CD11b-positive tissue-resident macrophages coexpressed VSIG4. (Note that all VSIG4-positive cells coexpressed Cd11b; in some cases, Cd11b expression was low, precluding visualization of the expression in an overlay image). In contrast, MHC class II–positive DCs showed no coexpression of VSIG4. In CFA/Myhca-immunized mice, inflammatory foci included numerous CD68-positive macrophages that showed no coexpression of VSIG4. Occasional VSIG4-positive macrophages were detected in the myocardium surrounding the inflammatory focus. On a consecutive histological section, abundant MHC class II expression was detected within the inflammatory infiltrate, probably mostly on activated macrophages. Representative stainings of at least 2 independent experiments are shown. Original magnification, ×150.

VSIG4 is downregulated in autoimmune tissue. Since VSIG4 was downregulated on macrophages activated by LPS or thioglycolate, we wanted to assess whether VSIG4 may also be absent on macrophages in autoimmune tissue. Thus, female BALB/c mice were immunized and boosted with myosin-derived peptides, leading to autoimmune inflammation of the heart (33). Three weeks later, VSIG4 expression was assessed on control heart and diseased heart. In the healthy tissue, VSIG4 expression was found on a fraction of CD68⁺ and CD11b⁺ macrophages but not on MHC class II^high DCs (Figure 4). In contrast, VSIG4 expression was not found in the confluent inflammatory infiltrates of hearts with myocarditis (Figure 4). Very rarely, a single VSIG4-positive macrophage could be detected within an inflammatory focus, and macrophages outside inflammatory foci expressed VSIG4 in normal numbers and unremarkable distribution (data not shown). This indicates that absence of VSIG4 expression was restricted to actual sites of inflammation. These histological data further support the notion that VSIG4 is downregulated upon the activation of the cells.

VSIG4 is a negative regulator of T cell proliferation. To study the function of the putative member of the B7 family, the extracellular domain of VSIG4 (aa 1–186) was fused to the human IgG1 constant region and expressed in 293-EBNA cells. The soluble fusion molecule was named VSIG4-Ig. Bearing in mind the importance of costimulatory molecules in positive and negative regulation of T cell activation, we first tested the role of VSIG4 in vitro in T cell proliferation assays. Briefly, increasing concentrations of anti-CD3 antibody (0, 0.25, 5, and 1 μg/ml) were coated in the presence of 5 μg/ml VSIG4-Ig or control IgG1. Purified CD4⁺ T cells were added, and proliferation was assessed after 48 hours by [³H]thymidine incorporation. As shown in Figure 5A, dose-dependent proliferation of T cells was observed with increasing concentrations of anti-CD3 antibody. Inclusion of immobilized VSIG4-Ig led to a strong reduction in T cell proliferation as compared with IgG1 at all 3 anti-CD3 concentrations used. IL-2 levels were also strongly reduced in culture supernatants of CD4⁺ T cells incubated in the presence of VSIG4 (Figure 5B). The inhibitory activity of VSIG4 on CD8⁺ T cells was analyzed next (Figure 5D). Proliferation assays as described above were performed with a fixed concentration of anti-CD3 (0.5 μg/ml) in the presence of 5 or 10 μg/ml VSIG4-Ig or control IgG1. As shown in Figure 5B, the inclusion of immobilized VSIG4-Ig fusion protein strongly inhibited proliferation of CD8⁺ T cells at both concentrations of VSIG4-Ig.

Figure 5

VSIG4 inhibits T cell proliferation and IL-2 production in vitro. (A) Plates were coated with anti-CD3 at the indicated concentrations in the presence of 5 μg/ml VSIG4-Ig or control IgG1. Proliferation of purified CD4⁺ T cells was monitored after 2 days by [³H]thymidine incorporation. (B) Plates were coated with anti-CD3 at 0.5 μg/ml in the presence of 5 or 10 μg/ml VSIG4-Ig or IgG1 at the indicated concentrations. CD8⁺ T cell proliferation was monitored after 2 days by [³H]thymidine incorporation. (C) Costimulation assays were performed with purified CD4⁺ T cells with a fixed concentration of anti-CD3 (0.5 μg/ml) in the presence or absence of anti-CD28 (2 μg/ml) with either 5 μg/ml VSIG4-Ig or control IgG1. (D) IL-2 production of CD4⁺ T cells stimulated with 0.5 μg/ml anti-CD3 in the presence of 5 μg/ml VSIG4-Ig or control IgG1. (E) Proliferation assays were performed in the presence of 0.5 μg/ml anti-CD3 and VSIG4-Ig, PD-L1–Ig, PD-L2–Ig, or control IgG1 at a concentration of 10 μg/ml. (F) Proliferation assays were performed in the presence of 0.5 μg/ml anti-CD3 and VSIG4-Ig or control IgG1 at a concentration of 5 or 10 μg/ml in the absence or presence of 10 ng/ml IL-2. Error bars represent standard deviations for data from experiments performed in triplicate. Representative data from least 2 independent experiments are shown. The difference between VSIG4-Ig– and IgG1-treated cells was significant in all panels shown (P < 0.05). In F, the difference between VSIG4-Ig– and IgG1-treated cells in the presence of IL-2 was not significant (P > 0.05).

Having established that VSIG4-Ig is a strong inhibitor of T cell proliferation induced by anti-CD3, we wondered whether VSIG4-Ig could also inhibit T cell proliferation induced by anti-CD3 and anti-CD28. For this purpose, costimulation assays were performed with a fixed concentration of anti-CD3 (0.5 μg/ml) in the presence or absence of anti-CD28 (2 μg/ml) with either 5 μg/ml VSIG4-Ig or control IgG1. Purified CD4⁺ T cells were added, and proliferation was assessed by [³H]thymidine incorporation. VSIG4-Ig strongly inhibited proliferation of anti-CD3 as well as anti-CD3/anti-CD28–stimulated T cells (Figure 5C).

To benchmark the potency of VSIG4 as a negative regulator of T cell activation, we compared its inhibitory properties with those of PD-L1 and PD-L2, two well-known inhibitors of T cell activation. To this end, costimulation assays were performed with a fixed concentration of anti-CD3 (1.0 μg/ml) in the presence of 10 μg/ml VSIG4-Ig, PD-L1–Ig, or PD-L2–Ig. VSIG4 inhibited T cell proliferation at least as potently as PD-L1 and PD-L2 (Figure 5E).

Since incubation of T cells with VSIG4-Ig led to the inhibition of T cell proliferation with a concomitant reduction in IL-2 secretion, we wondered whether T cell proliferation could be rescued by the addition of IL-2. To this end, costimulation assays were performed with a fixed concentration of anti-CD3 with either 5 or 10 μg/ml VSIG4-Ig or, as a control, human IgG1 in the presence or absence of 10 ng/ml IL-2. Purified T cells were added, and proliferation was assessed by [³H]thymidine incorporation. The inhibitory activity of VSIG4-Ig could be abolished by the addition of IL-2 (Figure 5F). These results confirm the specific inhibitory activity of VSIG4-Ig and rule out unspecific toxic effects of the protein preparation.

As outlined earlier, based on the high degree of conservation as well as the genomic locus, we postulated that Z39Ig is the human ortholog of VSIG4. To test whether Z39Ig also negatively regulates T cell proliferation, the extracellular domain of Z39Ig (aa 1–280) was fused to the human IgG1 constant region and expressed in 293-EBNA cells. Costimulation assays as described above were performed with a fixed concentration of anti-CD3 (0.5 μg/ml) in the presence or absence of anti-CD28 (2 μg/ml) with either 5 μg/ml Z39Ig-Ig or control IgG1. Purified murine CD4⁺ T cells were added, and proliferation was assessed by [³H]thymidine incorporation. As shown in Figure 6A, Z39Ig-Ig strongly inhibited proliferation of anti-CD3– as well as anti-CD3/anti-CD28–stimulated murine T cells, documenting that the inhibitory function of VSIG4 is conserved across the species barrier. Furthermore, simulation of human CD4⁺ T cells with anti-CD3 resulted in reduced proliferation in the presence of plastic-bound Z39Ig-Ig (Figure 6B). These results demonstrate that VSIG4 functions as a negative regulator of mouse as well as human T cells.

Figure 6

Z39Ig, the human homolog of VSIG4, inhibits T cell proliferation. (A) Proliferation assays were performed with purified mouse CD4⁺ T cells with 0.5 μg/ml anti-CD3 in the presence or absence of 2 μg/ml anti-CD28 with Z39Ig-Ig (human VSIG4) or control IgG1 at a concentration of 5 μg/ml. (B) Proliferation assays were performed with CD4⁺ T cells purified from human PBMCs in the presence of the indicated amounts of anti-CD3 together with Z39Ig-Ig (human VSIG4) or control IgG1 at a concentration of 20 μg/ml. Error bars represent standard deviations for data from experiments performed in triplicate. Representative data from at least 2 independent similar experiments are shown. The differences between Z39Ig-Ig– treated cells compared with control IgG1–treated cells were significant (P < 0.05) in both A and B.

In conclusion, these data demonstrate that VSIG4 is a strong negative regulator of T cell activation, providing further evidence that VSIG4 is indeed a B7 family–related protein.

VSIG4 negatively regulates CD8⁺ T cell responses and Th cell–dependent IgG responses in vivo. Having established the function of VSIG4 in vitro, we further explored the function of VSIG4 in vivo. We have previously shown that immunization of mice with peptide p33 derived from lymphocytic CMV glycoprotein coupled to virus-like particles (VLPs) derived from the bacteriophage Qβ induce strong primary CTL responses in mice (34, 35). To evaluate the in vivo effect of VSIG4 on the induction of CTLs, mice were treated with soluble VSIG4 (VSIG4-Ig) or a control protein (human IgG1 antibody) during vaccination with VLPs (Figure 7A). Accordingly, mice were injected i.p. on days –1, 1, and 3 with either mVSIG4-Ig or control IgG1. On day 0 mice were immunized with 50 μg of p33-Qβ mixed with CpGs as an adjuvant. To analyze specific cellular immune responses, p33-specific CD8⁺ T cells were measured in the 2 experimental groups by p33-specific tetramer staining 7 days after immunization. As shown in Figure 7B, immunized mice treated with VSIG4-Ig had a roughly 2.3-fold reduction in the number of p33-specific T cells in the blood. To further assess cellular T cell responses, intracellular IFN-γ production by splenic CD8⁺ T cells was measured by intracellular staining followed by FACS. As shown in Figure 7C, a 3.5-fold reduction in frequencies of IFN-γ–producing cells was observed in mice treated with VSIG4-Ig. Hence, similar to what was observed in vitro, VSIG4 inhibits CD8⁺ T cell activation in vivo.

Figure 7

Soluble VSIG4 inhibits induction of CTL responses and Th cell–dependent humoral immune responses in vivo. (A) Experimental diagram. Mice were injected with VSIG4-Ig or control IgG i.p. on days –1, 1, and 3 and immunized with Qβ-p33 VLPs with CpGs on day 0. Qβ-specific IgM and IgG responses were measured after 4 and 10 days. p33-specific CD8⁺ T cells were determined in the blood by tetramer staining on day 7. Mice were sacrificed on day 10, and levels of Qβ-specific B cells, AFCs, and p33-specific IFN-γ–expressing CD8⁺ T cells were determined from spleens. (B–G) Averages of 2 independent experiments ± SEM are shown (n = 8). Differences between VSIG4-Ig– and IgG1-treated mice were significant in B, C, and the right panel of D, as well as F and G. P values obtained by Student’s t test are shown. (B) Percentage p33-specific CD8⁺ T cells. (C) Percentage p33-specific CD8⁺ T cells after in vitro stimulation with p33-pulsed DCs 10 days after immunization. (D) ELISA titers of the Qβ-specific IgM and IgG responses. (E) Staining and gating strategy for the detection of isotype-switched B cells by FACS. Activated, isotype-switched B cells were found in the indicated gate. A representative example of Qβ-binding B cells in the indicated gate is shown in the right panel. (F) Percentage of Qβ-specific B cells within the indicated gate in E in mice 10 days after immunization. (G) Number of Qβ-specific AFCs per 10⁶ splenocytes as determined by ELISPOT are shown.

Humoral immune responses in the same mice were studied next. We have previously shown that Qβ VLPs induce an early, Th cell–independent IgM response followed by a Th cell–dependent IgG response (34, 36). To evaluate the in vivo effect of VSIG4 on the induction of humoral responses, IgG antibody titers, the number of antibody-forming cells (AFCs), and the number of Qβ-specific B cells were determined 10 days after immunization. Moreover, Qβ-specific IgM antibody titers were measured on day 4. The Th cell–independent IgM response was only marginally affected by treatment with VSIG4-Ig. In contrast, the Th cell–dependent IgG response against Qβ was reduced about 2-fold in mice treated with VSIG4-Ig (Figure 7D).

To further characterize B cell responses, Qβ-specific B cells were enumerated by FACS (36). Activated and isotype-switched B cells downregulate surface expression of IgM and IgD. Thus, in order to specifically stain Qβ-specific, isotype-switched B cells, a population of CD19⁺ B cells expressing low levels of IgM and IgD was assessed for its ability to bind Qβ (Figure 7E). As shown in Figure 7F, the percentage of Qβ-specific, isotype-switched B cells was reduced by a factor of about 2 in the VSIG4-treated animals. Numbers of Qβ-specific AFCs were assessed by ELISPOT and found to be decreased by a factor of about 3 (Figure 7G).

In summary, treatment of immunized mice with VSIG4-Ig inhibited Th cell–dependent IgG responses, as documented by reduced specific IgG titers and reduced frequencies of Qβ-specific B cells and AFCs. In contrast, the Th cell–independent Qβ-specific IgM response was only marginally affected by VSIG4. Thus, VSIG4-Ig appears to inhibit B cell responses by blocking T help.

The induction of specific T cell responses is a tightly regulated process. While research initially focused on costimulatory molecules enhancing T cell activation, the list of inhibiting receptor-ligand pairs is growing (23, 37, 38).The present study describes the identification of VSIG4 as potent inhibitor of T cell activation.

All B7 homologs described so far share a characteristic expression pattern: they are expressed at relatively low levels on resting cells and are upregulated in response to inflammatory signals. Some of the B7 family members, such as B7-1 and B7-2 as well as B7-H4, are largely restricted to professional APCs, such as DCs and macrophages, while others, e.g., B7-H1, are expressed rather in nonimmunological tissues, especially upon stimulation with inflammatory cytokines. It is thought that most of the B7 family members have the ability to engage both activating and inhibitory receptors on T cells; whether or not T cells become activated appears to depend on the timing of the interaction, since inhibitory receptors are preferentially expressed on activated cells (6). Interestingly, VSIG4 has a different expression pattern; it is highly expressed on resting, peripheral tissue macrophages and is downregulated or lost upon activation. In addition, little expression is found on macrophages in lymphoid organs. Hence, the main function of VSIG4 may not be the downregulation of an ongoing T cell response but rather the prevention of T cell activation in the intact tissue. Only upon inflammatory insult is VSIG4 downregulated or superseded by de novo infiltrating VSIG4-negative macrophages, which subsequently allow for proper T cell responses in peripheral tissues. It is tempting to speculate that VSIG4 may therefore be involved in the maintenance of peripheral T cell tolerance and/or nonresponsiveness. Indeed, VSIG4 was not present in the inflammatory infiltrates in tissue affected by autoimmune disease. This concept is supported by the finding that resting DCs induce T cell tolerance rather than T cell activation (39, 40). Additionally, and not mutually exclusively, it is possible that VSIG4 induces the differentiation of regulatory T cells. However, in preliminary in vitro experiments, we did not observe an increased production of antiinflammatory cytokines by T cells upon stimulation with VSIG4 (data not shown).

The mechanism of VSIG4-mediated inhibition of T cell activation remains to be established. Although our data show that VSIG4 is able to directly inhibit T cell proliferation in vitro, it remains possible that VSIG4 may influence T cell activation indirectly in a manner similar to that described for CTLA-4. CTLA-4 is able to inhibit T cell activation through recruitment of the phosphatase shp-2, which reduces TCR-dependent signal transduction by dephosphorylation of signaling molecules (41, 42). Yet T cell responses mounted by CTLA-4–deficient, TCR-transgenic T cells are remarkably normal (43, 44). The more important function of CTLA-4 may therefore be to maintain peripheral tolerance. Indirect and direct mechanism(s) may account for this function, since CTLA-4 has been shown to be involved in the generation of regulatory T cells as well as inducing T cell anergy (45–49).

While we were revising the current manuscript, a study was published describing VSIG4 as a novel complement-binding protein, CRIg (binding C3b and iC3b), specifically expressed by tissue macrophages, in particular Kupffer cells. This is an intriguing finding demonstrating that the complement system has not only receptors for fine-tuning of B cell responses (CD21) but also links the innate immune system with the regulation of T cell activation. Interestingly, binding of C3b or iC3b to VSIG4/CRIg triggers internalization of the receptor (50), suggesting that activation of the complement cascade removes the inhibitory properties of tissue macrophages, promoting inflammation. Hence, VSIG4/CRIg may be a new pathway of adaptive immune regulation by the innate immune system.

In conclusion, VSIG4 was shown to be a strong inhibitor of T cell proliferation in vitro and in vivo. The restricted expression of VSIG4 on resting tissue macrophages suggests that VSIG4 plays an important role in the maintenance of T cell unresponsiveness in healthy tissue.

We would like to thank Annette Oxenius for helpful discussions.

Address correspondence to: Martin F. Bachmann, Cytos Biotechnology AG, Wagistrasse 25, 8952 Zürich-Schlieren, Switzerland. Phone: 41-1-733-47-07; Fax: 41-1-733-47-57; E-mail: martin.bachmann@cytos.com.

Nonstandard abbreviations used: AFC, antibody-forming cell; HMM, hidden Markov model; PD-1, inhibitory receptor programmed death 1; PD-L1, PD-ligand 1; VLP, virus-like particle; VSIG4, V-set and Ig domain–containing 4; Z39Ig, Ig superfamily protein 39.

Conflict of interest: L. Vogt, N. Schmitz, M. Bauer, H.I. Hilton, D. Gatto, P. Sebbel, R.R. Beerli, P. Saudan, and M.F. Bachmann are employed by Cytos Biotechnology AG and own stock or stock options in the company.

Reference information: J. Clin. Invest.116:2817–2826 (2006). doi:10.1172/JCI25673

See the related Commentary beginning on page 2590.

1. Langman, R.E., Cohn, M. 1993. Two signal models of lymphocyte activation? Immunol. Today. 14:235-237.
   View this article via: PubMed Google Scholar
2. Lenschow, D.J., Walunas, T.L., Bluestone, J.A. 1996. CD28/B7 system of T cell costimulation. Annu. Rev. Immunol. 14:233-258.
   View this article via: PubMed CrossRef Google Scholar
3. Chambers, C.A., Allison, J.P. 1999. Costimulatory regulation of T cell function. Curr. Opin. Cell Biol. 11:203-210.
   View this article via: PubMed CrossRef Google Scholar
4. Schwartz, R.H. 2003. T cell anergy. Annu. Rev. Immunol. 21:305-334.
   View this article via: PubMed CrossRef Google Scholar
5. Sharpe, A.H., Freeman, G.J. 2002. The B7-CD28 superfamily. Nat. Rev. Immunol. 2:116-126.
   View this article via: PubMed CrossRef Google Scholar
6. Coyle, A.J., Gutierrez-Ramos, J.C. 2001. The expanding B7 superfamily: increasing complexity in costimulatory signals regulating T cell function. Nat. Immunol. 2:203-209.
   View this article via: PubMed CrossRef Google Scholar
7. Viola, A., Lanzavecchia, A. 1996. T cell activation determined by T cell receptor number and tunable thresholds. Science. 273:104-106.
   View this article via: PubMed CrossRef Google Scholar
8. Bachmann, M.F., et al. 1997. Distinct roles for LFA-1 and CD28 during activation of naive T cells: adhesion versus costimulation. Immunity. 7:549-557.
   View this article via: PubMed CrossRef Google Scholar
9. Hutloff, A., et al. 1999. ICOS is an inducible T-cell co-stimulator structurally and functionally related to CD28. Nature. 397:263-266.
   View this article via: PubMed CrossRef Google Scholar
10. Yoshinaga, S.K., et al. 1999. T-cell co-stimulation through B7RP-1 and ICOS. Nature. 402:827-832.
    View this article via: PubMed CrossRef Google Scholar
11. Coyle, A.J., et al. 2000. The CD28-related molecule ICOS is required for effective T cell-dependent immune responses. Immunity. 13:95-105.
    View this article via: PubMed CrossRef Google Scholar
12. Tafuri, A., et al. 2001. ICOS is essential for effective T-helper-cell responses. Nature. 409:105-109.
    View this article via: PubMed CrossRef Google Scholar
13. Kopf, M., et al. 2000. Inducible costimulator protein (ICOS) controls T helper cell subset polarization after virus and parasite infection. J. Exp. Med. 192:53-61.
    View this article via: PubMed CrossRef Google Scholar
14. Bertram, E.M., et al. 2002. Role of ICOS versus CD28 in antiviral immunity. Eur. J. Immunol. 32:3376-3385.
    View this article via: PubMed Google Scholar
15. Suh, W.K., et al. 2004. The inducible costimulator plays the major costimulatory role in humoral immune responses in the absence of CD28. J. Immunol. 172:5917-5923.
    View this article via: PubMed Google Scholar
16. Mak, T.W., et al. 2003. Costimulation through the inducible costimulator ligand is essential for both T helper and B cell functions in T cell-dependent B cell responses. Nat. Immunol. 4:765-772.
    View this article via: PubMed CrossRef Google Scholar
17. Ishida, Y., Agata, Y., Shibahara, K., Honjo, T. 1992. Induced expression of PD-1, a novel member of the immunoglobulin gene superfamily, upon programmed cell death. Embo J. 11:3887-3895.
    View this article via: PubMed Google Scholar
18. Latchman, Y., et al. 2001. PD-L2 is a second ligand for PD-1 and inhibits T cell activation. Nat. Immunol. 2:261-268.
    View this article via: PubMed CrossRef Google Scholar
19. Freeman, G.J., et al. 2000. Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads to negative regulation of lymphocyte activation. J. Exp. Med. 192:1027-1034.
    View this article via: PubMed CrossRef Google Scholar
20. Dong, H., Zhu, G., Tamada, K., Chen, L. 1999. B7-H1, a third member of the B7 family, co-stimulates T-cell proliferation and interleukin-10 secretion. Nat. Med. 5:1365-1369.
    View this article via: PubMed CrossRef Google Scholar
21. Tseng, S.Y., et al. 2001. B7-DC, a new dendritic cell molecule with potent costimulatory properties for T cells. J. Exp. Med. 193:839-846.
    View this article via: PubMed CrossRef Google Scholar
22. Watanabe, N., et al. 2003. BTLA is a lymphocyte inhibitory receptor with similarities to CTLA-4 and PD-1. Nat. Immunol. 4:670-679.
    View this article via: PubMed CrossRef Google Scholar
23. Sedy, J.R., et al. 2005. B and T lymphocyte attenuator regulates T cell activation through interaction with herpesvirus entry mediator. Nat. Immunol. 6:90-98.
    View this article via: PubMed CrossRef Google Scholar
24. Chapoval, A.I., et al. 2001. B7-H3: a costimulatory molecule for T cell activation and IFN-gamma production. Nat. Immunol. 2:269-274.
    View this article via: PubMed CrossRef Google Scholar
25. Sica, G.L., et al. 2003. B7-H4, a molecule of the B7 family, negatively regulates T cell immunity. Immunity. 18:849-861.
    View this article via: PubMed CrossRef Google Scholar
26. Prasad, D.V., Richards, S., Mai, X.M., Dong, C. 2003. B7S1, a novel B7 family member that negatively regulates T cell activation. Immunity. 18:863-873.
    View this article via: PubMed CrossRef Google Scholar
27. Sun, M., et al. 2002. Characterization of mouse and human B7-H3 genes. J. Immunol. 168:6294-6297.
    View this article via: PubMed Google Scholar
28. Suh, W.K., et al. 2003. The B7 family member B7-H3 preferentially down-regulates T helper type 1-mediated immune responses. Nat. Immunol. 4:899-906.
    View this article via: PubMed CrossRef Google Scholar
29. Langnaese, K., Colleaux, L., Kloos, D.U., Fontes, M., Wieacker, P. 2000. Cloning of Z39Ig, a novel gene with immunoglobulin-like domains located on human chromosome X. Biochim. Biophys. Acta. 1492:522-525.
    View this article via: PubMed Google Scholar
30. Ahn, J.H., et al. 2002. Identification of the genes differentially expressed in human dendritic cell subsets by cDNA subtraction and microarray analysis. Blood. 100:1742-1754.
    View this article via: PubMed Google Scholar
31. Walker, M.G. 2002. Z39Ig is co-expressed with activated macrophage genes. Biochim. Biophys. Acta. 1574:387-390.
    View this article via: PubMed Google Scholar
32. Kim, J.K., et al. 2005. Characterization of monoclonal antibody specific to the Z39Ig protein, a member of immunoglobulin superfamily. Immunol. Lett. 99:153-161.
    View this article via: PubMed CrossRef Google Scholar
33. Eriksson, U., et al. 2003. Activation of dendritic cells through the interleukin 1 receptor 1 is critical for the induction of autoimmune myocarditis. J. Exp. Med. 197:323-331.
    View this article via: PubMed CrossRef Google Scholar
34. Jegerlehner, A., et al. 2002. Regulation of IgG antibody responses by epitope density and CD21-mediated costimulation. Eur. J. Immunol. 32:3305-3314.
    View this article via: PubMed CrossRef Google Scholar
35. Storni, T., et al. 2004. Nonmethylated CG motifs packaged into virus-like particles induce protective cytotoxic T cell responses in the absence of systemic side effects. J. Immunol. 172:1777-1785.
    View this article via: PubMed Google Scholar
36. Gatto, D., Ruedl, C., Odermatt, B., Bachmann, M.F. 2004. Rapid response of marginal zone B cells to viral particles. J. Immunol. 173:4308-4316.
    View this article via: PubMed Google Scholar
37. Khoury, S.J., Sayegh, M.H. 2004. The roles of the new negative T cell costimulatory pathways in regulating autoimmunity. Immunity. 20:529-538.
    View this article via: PubMed CrossRef Google Scholar
38. Chen, L. 2004. Co-inhibitory molecules of the B7-CD28 family in the control of T-cell immunity. Nat. Rev. Immunol. 4:336-347.
    View this article via: PubMed CrossRef Google Scholar
39. Bonifaz, L., et al. 2002. Efficient targeting of protein antigen to the dendritic cell receptor DEC-205 in the steady state leads to antigen presentation on major histocompatibility complex class I products and peripheral CD8+ T cell tolerance. J. Exp. Med. 196:1627-1638.
    View this article via: PubMed CrossRef Google Scholar
40. Probst, H.C., Lagnel, J., Kollias, G., van den Broek, M. 2003. Inducible transgenic mice reveal resting dendritic cells as potent inducers of CD8+ T cell tolerance. Immunity. 18:713-720.
    View this article via: PubMed CrossRef Google Scholar
41. Marengere, L.E., et al. 1996. Regulation of T cell receptor signaling by tyrosine phosphatase SYP association with CTLA-4. Science. 272:1170-1173.
    View this article via: PubMed CrossRef Google Scholar
42. Lee, K.M., et al. 1998. Molecular basis of T cell inactivation by CTLA-4. Science. 282:2263-2266.
    View this article via: PubMed CrossRef Google Scholar
43. Bachmann, M.F., et al. 1998. Normal responsiveness of CTLA-4-deficient anti-viral cytotoxic T cells. J. Immunol. 160:95-100.
    View this article via: PubMed Google Scholar
44. Chambers, C.A., Sullivan, T.J., Truong, T., Allison, J.P. 1998. Secondary but not primary T cell responses are enhanced in CTLA-4-deficient CD8+ T cells. Eur. J. Immunol. 28:3137-3143.
    View this article via: PubMed CrossRef Google Scholar
45. Bluestone, J.A. 1997. Is CTLA-4 a master switch for peripheral T cell tolerance? J. Immunol. 158:1989-1993.
    View this article via: PubMed Google Scholar
46. Bachmann, M.F., Kohler, G., Ecabert, B., Mak, T.W., Kopf, M. 1999. Cutting edge: lymphoproliferative disease in the absence of CTLA-4 is not T cell autonomous. J. Immunol. 163:1128-1131.
    View this article via: PubMed Google Scholar
47. Bachmann, M.F., et al. 2001. Normal pathogen-specific immune responses mounted by CTLA-4-deficient T cells: a paradigm reconsidered. Eur. J. Immunol. 31:450-458.
    View this article via: PubMed CrossRef Google Scholar
48. Tivol, E.A., Gorski, J. 2002. Re-establishing peripheral tolerance in the absence of CTLA-4: complementation by wild-type T cells points to an indirect role for CTLA-4. J. Immunol. 169:1852-1858.
    View this article via: PubMed Google Scholar
49. Greenwald, R.J., Boussiotis, V.A., Lorsbach, R.B., Abbas, A.K., Sharpe, A.H. 2001. CTLA-4 regulates induction of anergy in vivo. Immunity. 14:145-155.
    View this article via: PubMed CrossRef Google Scholar
50. Helmy, K.Y., et al. 2006. CRIg: a macrophage complement receptor required for phagocytosis of circulating pathogens. Cell. 124:915-927.
    View this article via: PubMed CrossRef Google Scholar
51. Durbin, R., Eddy, S., Kroegh, A., and Mitchison, G. 1998. Biological sequence analysis: probabilistic models of proteins and nucleic acids. Cambridge University Press. New York, New York, USA. 356 pp.
    View this article via: PubMed Google Scholar
52. Letunic, I., et al. 2002. Recent improvements to the SMART domain-based sequence annotation resource. Nucleic Acid Res. 30:242-244.
    View this article via: PubMed CrossRef Google Scholar
53. Nielsen, H., Engelbrecht, J., Brunak, S., von Heijne, G. 1997. Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Eng. 10:1-6..
    View this article via: PubMed CrossRef Google Scholar