Blockade of Tim-1 and Tim-4 Enhances Atherosclerosis in LDL Receptor-Deficient Mice

Abstract

Objective

T cell immunoglobulin and mucin domain (Tim) proteins are expressed by numerous immune cells, recognize phosphatidylserine (PS) on apoptotic cells and function as costimulators or coinhibitors. Tim-1 is expressed by activated T cells but is also found on dendritic cells and B cells. Tim-4, present on macrophages and dendritic cells, plays a critical role in apoptotic cell clearance, regulates the number of PS-expressing activated T cells and is genetically associated with low LDL and triglyceride levels. Since these functions of Tim-1 and Tim-4 could affect atherosclerosis, their modulation has potential therapeutic value in cardiovascular disease.

Approach and Results

ldlr^−/− mice were fed a high-fat diet for 4 weeks while being treated with control (rat IgG1) or anti-Tim-1 (3D10) or -Tim-4 (21H12) mAbs that block PS recognition and phagocytosis. Both anti-Tim-1 and anti-Tim-4 treatments enhance atherosclerosis by 45% compared with controls by impairment of efferocytosis and increasing aortic CD4⁺T cells. Consistently, anti-Tim-4-treated mice show increased percentages of activated T cells and 'late' apoptotic cells in the circulation. Moreover, in vitro blockade of Tim-4 inhibited efferocytosis of oxLDL-induced apoptotic macrophages. Whereas anti-Tim-4 treatment increased Th1 and Th2 responses, anti-Tim-1 induced Th2 responses but dramatically reduced the percentage of Tregs. Finally, combined blockade of Tim-1 and Tim-4 increased atherosclerotic lesion size by 59%.

Conclusion

Blockade of Tim-4 aggravates atherosclerosis likely by prevention of phagocytosis of PS-expressing apoptotic cells and activated T cells by Tim-4-expressing cells, whereas Tim-1-associated effects on atherosclerosis are related to changes in Th1/Th2 balance and reduced circulating Tregs.

Introduction

Cardiovascular disease is the leading cause of death worldwide and is mainly caused by atherosclerosis and its thrombotic complications. Atherosclerosis is characterized by the combination of lipid accumulation and inflammatory immune processes in the arterial wall.¹ Current treatment options, primarily based on lipid-lowering, are inadequate to halt progression of cardiovascular disease with respect to lesion size or to reverse existing lesions, emphasizing an urgent need for new therapeutic strategies to inhibit atherosclerosis.

Apoptotic cell death is an important feature for maintenance of immune homeostasis in the atherosclerotic lesion.^{2, 3} Early in the process of atherosclerotic lesion development, clearance of apoptotic cells, such as oxLDL-loaded macrophages (foam cells), in the intima induces an anti-inflammatory response that limits lesion growth.⁴ However, in later stages of atherosclerosis, the clearance of apoptotic cells, or efferocytosis, is defective, causing secondary cellular necrosis that leads to the formation of necrotic cores and induces inflammatory responses that contribute to atherosclerosis progression.^{5, 6}

Apoptotic cells can be recognized by different ‘eat me’ signals, such as surface exposure of plasma membrane phosphatidylserine (PS), which facilitates uptake by phagocytes.⁷ T cell immunoglobulin and mucin domain (Tim) proteins are type 1 transmembrane proteins expressed on various immune cells that can recognize PS-exposing cells and can also mediate coinhibitory and costimulatory signals.⁸ Three TIM genes have been identified in humans (TIM-1, −3 and −4) and are associated with enhanced susceptibility to allergy and several autoimmune diseases, such as EAE and diabetes.^{8, 9} It has been shown that targeting of Tim-3 aggravates early and advanced atherosclerosis in ldlr^−/− mice¹⁰, however, the role of Tim-1 and Tim-4 in atherosclerosis is not known.

Tim-1 is expressed on activated T cells, and regulatory B cells, whereas Tim-4 expression is restricted to antigen-presenting cells. Tim-4 has particularly high expression on peritoneal macrophages¹¹ and splenic marginal zone macrophages¹² that are involved in maintaining immune homeostasis. Mice deficient for Tim-1 or Tim-4 have increased susceptibility for autoimmunity, as shown by increased Th2 responses in tim-1^−/− mice¹³ and hyperactive T and B cell responses in tim-4^−/− mice.¹⁴ Recently, Lind et al. found an association between Tim-1 and plaque occurrence in carotid arteries in a human population-based study using proteomic arrays.¹⁵ Interestingly, a SNP in the TIM-4 encoding gene TIMD4 is associated with lowered LDL, triglycerides and cardiovascular disease.^{16, 17} However, it is not known if this SNP affects the expression or function of Tim-4. Another study showed that tim-4 mRNA negatively correlated with LDL levels in mice suffering from type 2 diabetes.¹⁸

Given their immunosuppressive potency by regulating clearance of PS-expressing cells, affecting adaptive immune responses and their possible association with LDL and triglycerides, modulation of Tim-1 and Tim-4 may represent a novel therapeutic target to treat cardiovascular disease. In the present study, we therefore investigated the role of Tim-1 and Tim-4 in atherosclerosis using blocking antibodies against Tim-1 and Tim-4.

Materials and Methods

Materials and Methods are available in the online-only Data Supplement.

Results

Increased percentages of Tim-1⁺ T cells and decreased percentages of Tim-4⁺ macrophages in atherosclerotic mice

Whereas Tim-1 is mainly expressed on T cells, Tim-4 expression is restricted to antigen-presenting cells and is highly expressed particularly on splenic marginal zone macrophages. To determine the percentage of Tim-1⁺ T cells in the spleen of atherosclerotic mice, we fed ldlr^−/− mice a HFD for 0 (n=7), 4 (n=9) or 10 weeks (n=6). As shown in Figure 1A, CD3⁺CD4⁺Tim-1⁺ cells but not CD3⁺CD8⁺Tim-1⁺ cells were increased after 10 weeks of HFD. Moreover, Tim-1 was also expressed on splenic B cells and CD19⁺Tim-1⁺ B cells were decreased after 10 weeks of HFD. Tim-4⁺ macrophages defined as F4/80⁺CD11c⁻Tim-4⁺ cells were decreased in spleens of mice fed a HFD for 10 weeks (Figure 1B). Since macrophages are abundantly present in atherosclerotic lesions, we analyzed Tim-4 expression on macrophages from aortic digests (Supplemental Figure I). Aortas were isolated from ldlr^−/− mice fed a control or HFD for 10 weeks and as shown in Figure 1C, the percentage Tim-4⁺ macrophages in the high fat diet group, which will include largely lesional macrophages was 57.55%, slightly lower than the 66% Tim-4⁺ macrophages found in control diet group, which are mainly adventitial macrophages. Moreover, we measured Tim-4 expression on CD11c⁺MHCII⁺ dendritic cells (DCs) and CD11b-B220⁺ B cells (Supplemental Figure I). As shown in Figure 1D, Tim-4 is also expressed on DCs isolated from the aorta, however, the expression is not affected by HFD. We did not detect Tim-4 on B cells found in the aorta of either control or atherosclerotic mice (Figure 1E).

Figure 1. Tim-1 and Tim-4 expression on T cells and macrophages in atherosclerotic mice.

ldlr^−/− mice were fed a HFD for 0, 4 or 10 weeks (n=9 per time-point). At sacrifice, splenocytes were isolated and percentages of CD4⁺Tim-1⁺ or CD8⁺Tim-1⁺ cells within CD3⁺ cells were determined with flow cytometry. CD19⁺Tim-1⁺ B cells in spleen were determined at 0 and 10 weeks of HFD (n=6 per group) (A). Tim-4⁺ macrophages were determined by staining for Tim-4 and F4/80 (B). Tim-4⁺ macrophages in aortic digests of ldlr^−/− mice fed a HFD for 0 or 10 weeks (n=4-5; each dot is a pool of 3 mice) were determined (C). Tim-4⁺ dendritic cells (D) and Tim-4⁺ B cells (E) are shown in aortic digests of ldlr^−/− mice fed a HFD for 10 weeks. *P<0.05, ***P<0.001

Blockade of Tim-4 inhibits uptake of oxLDL-induced apoptotic macrophages and induces a pro-inflammatory phenotype

Tim-4⁺ macrophages are known to play an important anti-inflammatory role by engulfing PS-expressing apoptotic cells.^{11, 19} In early atherosclerotic lesions, apoptotic oxLDL-loaded macrophage foam cells are efficiently cleared by neighboring phagocytes, but interference with this process can lead to necrotic death of foam cells, stimulating inflammatory responses by other macrophages.² To investigate the importance of Tim-4 in this process, we isolated peritoneal macrophages and loaded them with an amount of oxLDL which causes macrophage apoptosis (100μg/ml) for 48 hours (Figure 2A). Subsequently, we added these apoptotic macrophages to freshly isolated peritoneal macrophages that were pre-incubated for 30 minutes with 20μg/ml rat IgG or anti-Tim-4 (21H12). As shown in Figure 2B, Tim-4 blockade significantly inhibits uptake of oxLDL-induced apoptotic macrophages by 69%. Furthermore, 24 hours after removing apoptotic macrophages from the coculture, we measured cytokine production in the supernatant of the phagocytes. Blockade of Tim-4 increased macrophage secretion of the pro-inflammatory cytokines IL-6 and MCP-1 (Figure 2C), suggesting that defective clearance of oxLDL-induced apoptotic macrophages mediated by Tim-4 blockade could have enhanced macrophage activation through the accumulation of necrotic cells or through oxidized phospholipids on the surface of the apoptotic cells.

Figure 2. Blockade of Tim-4 inhibits uptake of oxLDL-induced apoptotic macrophages and induces a pro-inflammatory phenotype.

Macrophages isolated from the peritoneum of B6 mice (n=3) were exposed to 5, 50 or 100 μg/ml of oxLDL for 48 hours. The percentage of apoptotic cells (AnnexinV⁺PI⁺ cells) was determined with flow cytometry (A). To assess whether Tim-4 blockade could interfere with uptake of oxLDL-induced apoptotic macrophages by macrophages, we first pre-incubated freshly isolated peritoneal macrophages with rat IgG or anti-Tim-4 (21H12, 20 μg/ml) for 30 minutes in triplo. Subsequently, CFSE-labeled oxLDL-induced apoptotic macrophages were added in a 1:1 ratio and after 3 hours, cells were stained for F4/80 and uptake of apoptotic cells was determined by immunohistochemistry (B). Cytokines (IL-6 and MCP-1) were measured in the supernatant of peritoneal macrophages preincubated with rat IgG (n=3) or anti-Tim-4 (21H12, n=3), 24 hours after a 3 hour incubation with oxLDL-induced apoptotic peritoneal macrophages (C). *P<0.05, ***P<0.001

Blockade of Tim-1 and Tim-4 aggravates the development of atherosclerosis

To assess the role of Tim-1 and Tim-4 in atherosclerosis, we blocked these molecules with rat anti-mouse mAbs in ldlr^−/− mice while they were fed a HFD for 4 weeks (Figure 3A). The 4 week endpoint was chosen because after this point mouse anti-rat Ig responses would impair the effectiveness of the mAbs and lead to immune complex formation that could affect the disease process. At sacrifice we did not observe any difference in body weight, and serum cholesterol and triglyceride levels were unaffected by blockade of Tim-1 or Tim-4 (Figure 3B and Supplemental Figure IIA). As shown in Figure 3C, lesion area as a percentage of the lumen was significantly higher (45% increase) in mice treated with anti-Tim-1 (7.76±0.78%) or anti-Tim-4 (7.76±0.67%) in comparison with control mice (5.37±0.37%, P<0.05). No significant difference in macrophage content (% lesion area) was observed between control mice (83.12±0.69%), anti-Tim-1-treated mice (81.02±2.38%) and anti-Tim-4-treated mice (80.89±1.63%, Figure 3D).

Figure 3. Blockade of Tim-1 and Tim-4 aggravates atherosclerosis.

ldlr^−/− mice were treated twice a week i.p. with anti-Tim-1 (clone 3D10, n=14), anti-Tim-4 (clone 21H12, n=20) or an isotype control (rat IgG1, n=19) for 4 weeks while being fed a HFD (A). At sacrifice, mice were weighed and serum cholesterol and triglyceride levels were determined (B). Representative cross-sections of lesion formation in the aortic valve area stained with Oil-Red-O and hematoxylin are shown and lesion size was determined (C). Sections of the aortic root were stained for macrophages using MOMA-2 staining and the percentage of macrophages relative to the lesion size was determined (D). Bars=100μm. *P<0.05

Impaired efferocytosis in anti-Tim-4-treated mice

Since Tim-1 and Tim-4 are involved in the recognition and removal of apoptotic cells, we first performed TUNEL staining to detect apoptotic cells in atherosclerotic lesions. We observed a 3.2-fold increase in TUNEL⁺ cells in lesions of anti-Tim-1-treated mice (97.0±33.2 cells/mm²) and a significant 4.3-fold increase in TUNEL⁺ cells in lesions of anti-Tim-4-treated mice (129.1±22.6 cells/mm²) compared with control mice (30.0±18.5 cells/mm², Figure 4A). Subsequently, we investigated the effect of Tim-1 and Tim-4 blockade on efferocytosis locally within the atherosclerotic lesion, as measured by the percentage of free TUNEL⁺ cells versus macrophage-associated TUNEL⁺ cells. As shown in Figure 4B efferocytosis is strongly impaired by Tim-4 blockade and although not significant, is also impaired in anti-Tim-1-treated mice. Moreover, we observed a defect in clearance of apoptotic cells in the circulation (Figure 4C) and more specifically of CD4⁺ T cells in anti-Tim-4-treated mice (Figure 4D).

Figure 4. Impaired efferocytosis and increased aortic T cells in mice treated with anti-Tim 1 or anti-Tim 4.

TUNEL⁺ apoptotic cells (white arrows) were stained in the atherosclerotic lesions (n=6-7 per group) (A) and efferocytosis was determined by calculating the ratio of free versus macrophage-associated TUNEL⁺ cells (B). At sacrifice, blood was isolated and the percentage of apoptotic cells (C) or apoptotic CD4⁺ T cells (D) in the blood was determined with an Annexin-V/PI staining and analyzed with flow cytometry (n=10 per group). The percentage of activated T cells (CD4⁺CD69⁺ cells) in rat IgG (n=19), anti-Tim-1 (n=14) and anti-Tim-4 mice (n=20) was determined with flow cytometry (E). Representative cross-sections of lesions in the aortic valve area of rat IgG (n=19), anti-Tim-1 (n=14) and anti-Tim-4 mice (n=20) were stained for CD4⁺ T cells and the number of T cells in the intima and perivascular tissue were manually analyzed (F). Bars=100μm. *P<0.05, **P<0.01

Increased T cell numbers in the aortic root of anti-Tim-1 and anti-Tim-4-treated mice

Previously, it has been described that Tim-4 expressing macrophages can engulf activated T cells that express PS at their surface shortly after activation.¹² In line with this finding, we observed increased circulating CD4⁺CD69⁺ T cells in anti-Tim-4-treated mice (6.47±0.54%) compared to control mice (4.83±0.34%, P<0.05, Figure 4E). Since both Tim-1 and Tim-4 have been implicated in T cell function and survival^{12, 13}, we investigated whether blockade of these molecules affects T cell numbers in atherosclerotic lesions of ldlr^−/− mice. Representative cross-sections of the aortic root were stained for CD4 and CD8 (Figure 4F, red cells). Treatment with anti-Tim-1 significantly enhanced CD4⁺ T cell numbers within the lesions (10.25±1.06 cells, P<0.05) in comparison with control mice (6.61±1.11 cells), whereas numbers of CD4⁺ T cells in the surrounding perivascular tissue were elevated in both anti-Tim-1 (13.53±1.59 cells, P<0.05) and anti-Tim-4-treated mice (15.68±1.75 cells, P<0.01) in comparison with control mice (7.55±1.29 cells). In contrast, we did not observe significant differences between groups in the number of CD8⁺ T cells in atherosclerotic lesions or in the perivascular tissue (Supplemental Figure IIB).

Blockade of Tim-1 and Tim-4 affect different subsets of effector T cells

Tim-4 expression on marginal zone macrophages in the spleen is essential for the maintenance of immune homeostasis by clearance of apoptotic cells and regulation of the number of activated T cells. Blockade of Tim-4 inhibits clearance of these cells and thereby promotes inflammation. Consistent with this, we found slightly enlarged spleens and an increased percentage of proliferating CD4⁺Ki-67⁺ cells in anti-Tim-4-treated mice in comparison with control mice (Figure 5A). However, the percentage of CD4⁺ and CD8⁺ T cells in the spleen did not differ (Supplemental Figure IIC).

Figure 5. T cell subset changes in mice treated with anti-Tim-1 or anti-Tim-4.

At sacrifice, spleens were weighed and the percentage of proliferating CD4⁺ T cells (CD4⁺Ki-67⁺ cells) was determined by flow cytometry (n=5 per group) (A). Splenic T cell subsets were determined by staining for CD4 and the hallmark cytokines IFNγ (Th1 cells), IL-4 (Th2 cells) or the transcription factor Foxp3 (Tregs) and analyzed with flow cytometry (n=5-6 per group) (B). Cytokines (IFNγ, IL-4, IL-13, IL-6 and IL-17) present in supernatant of anti-CD3/CD28-stimulated splenocytes (n= 5 per group) were determined with a Luminex bead-based multiplex assay (C). Total IgG2c, IgG1, oxLDL-specific IgG1 and MDA-LDL-specific IgG1 were detected in serum of control (n=19), anti-Tim-1-treated (n=14) and anti-Tim-4-treated mice (n=20) (D). *P<0.05, **P<0.01, ***P<0.001

To evaluate whether blockade of Tim-1 and Tim-4 affects specific effector T cell subsets, at sacrifice, splenocytes were stimulated with PMA/ionomycin/Brefeldin A for 4 hours and stained for IFNγ-secreting CD4⁺ cells (Th1 cells), IL-4-secreting CD4⁺ cells (Th2 cells) and IL-17-secreting CD4⁺ cells (Th17 cells) (Figure 5B). The majority of the pathogenic CD4⁺ T cells in atherosclerosis are Th1 cells. Mice treated with anti-Tim-4 showed elevated Th1 cells (5.37±0.51%) in comparison with control mice (2.32±0.37%, P<0.01). Whereas blockade of Tim-1 did not affect the percentage of Th1 cells (2.84±0.12%), Th2 cells were strongly increased (0.71±0.06%) in comparison with control mice (0.31±0.11%, P<0.05). Although not significant, the percentage of Th2 cells in anti-Tim-4-treated mice (0.59±0.09%) were also increased compared with control mice. We did not detect any Th17 cells by flow cytometry. To determine if blockade of Tim-1 and Tim-4 affects Tregs, we stained blood for CD4, CD25 and the transcription factor Foxp3. We detected a significant 76% reduction of circulating Tregs in Tim-1-treated mice (0.46±0.21%) compared with control mice (1.90±0.22%, P<0.001). In line with our flow cytometry data, Th2 cytokines were increased in the supernatants of anti-CD3/CD28-stimulated splenocytes of anti-Tim-1-treated mice (IL-4, IL-13), whereas supernatant of splenocytes from anti-Tim-4-treated mice contained high levels of both Th1 and Th2 cytokines (IFNγ, IL-4, IL-13) (Figure 5C). Moreover, IL-6 and IL-17 levels were significantly increased in cultures from anti-Tim-4-treated mice in comparison with control mice.

Humoral responses in anti-Tim-1 and anti-Tim-4-treated mice

Blockade of Tim-1 or Tim-4 during atherosclerosis strongly induced Th2 and Th1/Th2 responses respectively. Since isotype switching of B cells is dependent on cytokine secretion by T cells, we investigated whether B cell antibody production was affected in control, anti-Tim-1 and anti-Tim4-treated mice after 4 weeks of treatment and HFD. IL-4 secretion can induce an IgG1 isotype switch, whereas IFNγ can induce an IgG2c switch. Accordingly, Figure 5D shows a significant increase in total IgG2c and IgG1 in serum of anti-Tim-4-treated mice in comparison with control mice. Both anti-Tim-1 and anti-Tim-4 treatments induced IgG1 antibodies against oxLDL and MDA-LDL in comparison with control treatment. Notably, IgM antibodies against oxLDL and MDA-LDL, as well as the IgM NAb E06, were also elevated after Tim-1 and Tim-4 blockade (Supplemental Figure IIC).

Tim-1 and Tim-4 act nonredundantly in aggravating atherosclerosis

Since both Tim-1 and Tim-4 are important regulators of immune homeostasis and have been described to interact with each other, but also have distinct functions, we were interested to determine if blockade of both Tim-1 and Tim-4 at the same time would have a greater effect than either alone on atherosclerosis. To assess the role of combined blockade of Tim-1 and Tim-4 in atherosclerosis, we treated ldlr^−/− mice twice weekly i.p. with 200μg of anti-Tim-1 and 200μg anti-Tim-4 or 400μg of the isotype control (rat IgG1), while the mice were fed a HFD for 4 weeks (Figure 6A). At sacrifice we did not observe any difference in body weight and serum cholesterol levels (Figure 6B and Supplemental Figure IIIA). Interestingly, triglyceride levels were decreased in anti-Tim-1+4-treated mice (200.30±15.34mg/dl) in comparison with control mice (279.40±26.64mg/dl, P<0.05). Lesion area was 59% greater in mice with combined blockade of Tim-1 and Tim-4 (9.12±1.12%) in comparison with control mice (5.75±1.02%, P<0.05, Figure 6C). No significant difference in lesional macrophage content was observed (control: 77.06±2.32% vs. anti-Tim-1+4: 78.05±2.52%, Figure 6D).

Figure 6. Combined blockade of Tim-1 and Tim-4 aggravates atherosclerosis.

ldlr^−/− mice were treated twice a week i.p. with anti-Tim-1+anti-Tim-4 (n=10) or an isotype control (rat IgG1, n=10) for 4 weeks while being fed a HFD (A). At sacrifice, mice were weighed and serum cholesterol levels were determined (B). Representative cross-sections of lesion formation in the aortic valve area stained with Oil-Red-O and hematoxylin are shown and lesion size was determined (C). Sections of the aortic root were stained for macrophages using MOMA-2 staining and the percentage of macrophages relative to the lesion size was determined (D). Furthermore, lesions were stained for CD4⁺ T cells and the number of T cells in the intima was manually analyzed (E). TUNEL⁺ apoptotic cells were stained in the atherosclerotic lesions and efferocytosis was determined by calculating the ratio of free versus macrophage-associated TUNEL⁺ cells (n=5 per group) (F). At sacrifice, Tregs in blood were determined by staining for CD4, CD25 and the transcription factor Foxp3 and analyzed with flow cytometry (G). Cytokines (IFNγ, IL-4) present in supernatant of anti-CD3/CD28-stimulated splenocytes were determined with a Luminex bead-based multiplex assay (n=10 per group) (H). oxLDL-specific IgG2c and IgG1 were detected in serum (I). Bars=100μm. *P<0.05, **P<0.01, ***P<0.001

Moreover, treatment with anti-Tim-1+4 significantly enhanced CD4⁺ T cell numbers within the lesions (11.00±1.75 cells, P<0.05) and in perivascular tissue (9.56±1.28 cells, P<0.05) in comparison with control mice (5.80±1.16 cells and 5.30±1.28 cells, respectively, Figure 6E and Supplemental Figure IIIB), whereas CD8⁺ T cells were unaffected (Supplemental Figure IIIC). Simultaneous blockade of Tim-1 and Tim-4 results in a 7.8-fold increase in TUNEL⁺ cells (154.3±27.7 cells/mm² vs. 19.7±8.5 cells/mm² in control mice, Figure 6F). This increase in CD4⁺ T cells in anti-Tim-1+4-treated mice corresponds to impaired clearance of apoptotic T cells (Supplemental Figure IIID) and is consistent with enlarged spleens, increased spleen cellularity and enhanced proliferation in these mice (Supplemental Figure IIIE). Combined blockade of Tim-1 and Tim-4 also increased inflammatory CD11b⁺Ly6G⁻Ly6C^hi monocytes in the spleen (Supplemental Figure IIIF), whereas there is no difference in blood (data not shown). Furthermore, circulating Tregs were decreased in anti-Tim-1+4-treated mice (Figure 6G).

Surprisingly, both Th1 and Th2 cytokine secretion (IFNγ and IL-4) was reduced after treatment with anti-Tim-1+4 (Figure 6H), whereas, similar to our atherosclerosis experiments with individual Tim-1 and Tim-4 treatments, total IgG1 and IgG1 specific for oxLDL and MDA-LDL were increased in anti-Tim-1+4-treated mice (Figure 6 I and Supplemental Figure IV). Additionally, oxLDL-specific IgG2c and total IgM was increased after combined blockade of Tim-1 and Tim-4. We also did not observe a difference in the total percentage of CD19⁺ B cells or B cell subsets as defined by CD19⁺CD3⁻IgM^hiIgD^low and CD19⁺CD3⁻IgM^low/intIgD^hi cells (Supplemental Figure IVE).

Discussion

Tim-1and Tim-4 have diverse, and partly overlapping functions related to regulation of inflammatory and immune responses, including T cell costimulation, and clearance of apoptotic cells. Some of the functions of Tim-1 and Tim-4 suggest that they may be useful targets for therapy of immunological diseases. Given the central role of innate and adaptive immunity, and clearance of apoptotic cells in atherosclerotic lesion development and phenotype, it is likely that Tim-1 and Tim-4 influence this disease process. Nonetheless, the net effect of blockade of either Tim-1 or Tim-4 on atherosclerosis is not predictable from prior studies of the functions of these molecules in other disease contexts. We therefore determined how atherosclerosis is affected in mice treated with blocking antibodies, modeling the hypothetical treatment of humans with similar reagents.

In this study, we demonstrate that treatment with anti-Tim-1 (3D10) or anti-Tim-4 (21H12) aggravates atherosclerosis, independent of cholesterol and triglyceride levels. Two important processes contributing to atherosclerosis development were affected by Tim-1 or Tim-4 blockade: efferocytosis and adaptive immune responses.

Previously, it has been shown that peritoneal macrophages and B1 cells from TIM-4^−/− mice do not clear apoptotic bodies in vivo.¹⁴ In addition, studies using anti-Tim-4 (21H12) have shown that blockade of Tim-4 reduces phagocytosis of apoptotic cells by peritoneal or splenic marginal zone macrophages.^{11, 12} In line with these findings, we observed that Tim-4 blockade with 21H12 potently inhibits uptake of oxLDL-induced apoptotic macrophages by peritoneal macrophages in vitro and induces secretion of pro-inflammatory cytokines. Consistently, anti-Tim-4-treated mice have increased percentages of circulating late apoptotic cells, indicative of impaired efferocytosis that contributes to lesion growth. Interestingly, a recent paper describes a subset of tissue-resident Tim-4⁺CD169⁺ subset that are immunoregulatory by inducing Tregs, reducing T cell proliferation and by promoting a higher rate of death of activated T cells.²⁰ These Tim-4⁺CD169⁺ cells are highly susceptible for apoptosis and if Tim-4⁺CD169⁺ cells would not undergo apoptosis when Tim-4 is blocked, we predict we would have observed reduced apoptosis in our experiment. However, it has been shown that depletion of CD169⁺ macrophages increases necrotic core size and apoptotic cell content of atherosclerotic lesions²¹ and also in our study, we observed a significant increase in apoptotic TUNEL⁺ cells in lesions of anti-Tim-4-treated mice and show that Tim-4 blockade impairs efferocytosis. This indicates that the macrophages in the absence of Tim-4 are not able to recognize PS-expressing apoptotic cell, resulting in accumulation of apoptotic cells, which undergo secondary necrosis and trigger inflammation. Additionally, we also observe a trend towards increased apoptotic cells in atherosclerotic lesions of anti-Tim-1-treated mice. Although Tim-1 is not expressed on macrophages, it has been shown that Tim-1 can be present on CD11c⁺ dendritic cells and on B cells²², which both have the ability to phagocytize apoptotic cells and apoptotic bodies. Xiao et al. have shown previously that B cells of Tim-1 deficient mice showed a defect in binding and uptake of apoptotic cells.²³ It is therefore likely that the increase in numbers of TUNEL⁺ cells in atherosclerotic lesions of anti-Tim-1-treated mice is a consequence of impaired uptake of apoptotic cells by Tim-1⁺ B cells. Moreover, there is a certain subset of CD11c⁺ cells that co-express F4/80, which may explain the observed impaired efferocytosis capacity of F4/80⁺ cells in anti-Tim-1-treated mice.

Previous studies have shown that splenocyte cultures from mice immunized with apoptotic cells spontaneously release high levels of Th1 and Th2 cytokines.²⁴ Moreover, tim-1^−/− mice and tim-4^−/− mice have hyperactive T and B cells, as shown by enhanced proliferation, increased IFNγ and IL-17 secretion and elevated circulating immunoglobulins^{13, 14}, whereas Tim-4 Tg mice have reduced memory T cell responses, and their T cells do not produce IL-4 or IFNγ.¹² In our study, we show that blockade of either Tim-1 or Tim-4 increased the number of lesional or perivascular CD4⁺ T cells. Whereas anti-Tim-4 treatment increased splenic IFNγ-secreting Th1 cells and IL-4-secreting Th2 cells, anti-Tim-1 treatment only induced Th2 cells. IL-17 secretion was also increased by anti-CD3/CD28 stimulated splenocytes from anti-Tim-1 or anti-Tim-4-treated mice. In atherosclerosis, it has been well established that Th1 cells are pro-atherogenic^{25, 26}, and some studies show that IL-4-deficient ldlr^−/− and apoe^−/− mice have reduced atherosclerosis.^{27, 28}

The role of IL-17 in atherosclerosis remains controversial but exogenous IL-17 administration has been shown to promote the formation of atherosclerotic lesions.²⁹ In addition to increased Th2 responses, anti-Tim-1-treated mice had strongly reduced circulating Tregs which exert a protective role in atherosclerosis.³⁰ Consistent with our findings that Tim-1 and Tim-4 are atheroprotective by their effects on T cell responses, Xiao et al. show that Tim-1-deficient B cells promote Th1 and Th17 cells, inhibit Tregs and enhance the severity of experimental autoimmune encephalomyelitis.²³ Blockade of Tim-4 with RMT4-53 exacerbates the induction phase of collagen-induced arthritis by enhanced T cell proliferation and increased IFNγ and IL-17 secretion.³¹ Furthermore, Albacker et al. showed that Tim-4 blockade (21H12) reversed the induction of intranasal tolerance, resulting in increased proliferative and cytokine responses (IL-4, IFNγ) of T cells.³² In addition, reduced CD4 T cell responses in Tim-4 Tg mice lead to reduced airway hyper-responsiveness in a murine model of asthma.¹²

Previously it has been shown that T cells also express PS shortly after activation and anti-Tim-4 (21H12) treatment was shown to specifically reduce phagocytosis of antigen-specific T cells.³² Indeed, anti-Tim-4-treated ldlr^−/− mice had elevated levels of circulating activated T cells and their T cells showed enhanced proliferative capacity, which may have contributed to enhanced aortic T cells.

Moreover, immunoglobulin levels associated with Th1 and Th2 responses, IgG2c and IgG1 respectively, were elevated in anti-Tim-1 or anti-Tim-4-treated mice. Interestingly, compared to control mice, we also observed increased IgM levels against oxLDL and MDA-LDL in the anti-Tim-1 or anti-Tim-4 mice, as well as an increase in the E06 NAb against phosphocholine of oxPL, as found on oxLDL. It is notable that apoptotic cells express increased epitopes of both MDA and oxPL.²⁴ Both enhanced atherosclerosis and enhanced content of apoptotic cells secondary to anti-Tim-1 and anti-Tim-4 could have provided the antigen exposure to increase titers of these IgM antibodies. It has previously been shown that these specific IgMs have anti-atherosclerotic potential, due to their ability to enhance clearance of apoptotic cells and their inhibitory effect on foam cell formation.³³ Despite this atheroprotective effect, we observe enhanced atherosclerosis in anti-Tim-1 or anti-Tim-4-treated mice. Previously it has been shown that mice immunized with apoptotic cells develop high IgM and IgG titers to oxidation-specific epitopes of oxLDL which contributes to their clearance.²⁴ Thus, while the titers of these IgM antibodies were increased, presumably they were insufficient to fully compensate for the increased apoptotic cell number and enhanced macrophage foam cell generation. Moreover the skewing toward proatherogenic T cell responses could have dominated over an atheroprotective effect of increased IgM.

Moreover, we show that treatment of ldlr^−/− mice with anti-Tim-1 and anti-Tim-4 does not affect the total percentage of CD19⁺ B cells and does not affect B cell subsets as defined by CD19⁺CD3⁻IgM^hiIgD^low and CD19⁺CD3⁻IgM^low/intIgD^hi cells. It has been described that Tim-1 is expressed by a subset of IL-10-producing B cells; Bregs.³⁴ However, recently it has been shown that IL-10 producing B cells, or Bregs, are dispensable for atherosclerosis development in mice.³⁵ It is however possibly that reduced Bregs affect the Treg population and thereby indirectly affect atherosclerosis. Further research will clarify the importance of Tim-1⁺ Bregs in atherosclerosis.

Other than binding to PS and blocking apoptotic cell clearance, Tim-1 and Tim-4 have many nonoverlapping functions and are expressed on different cell types. Therefore, we reasoned that the proatherogenic effects of blocking both molecules would be greater than blocking only one. Indeed, we found that combined blockade had a greater effect on lesion development than each single molecule blockade and this correlated with more robust effects on apoptotic cell clearance, lesional T cell numbers, and serum anti-oxLDL antibodies. In addition, combined blockade of Tim-1 and Tim-4 increased inflammatory CD11b⁺Ly6G⁻Ly6C^hi monocytes in the spleen. Previously, Robbins et al. have shown that Ly6C^hi monocytes that originate from the spleen accumulate in the arterial wall and enhance atherosclerotic lesion formation.³⁶ Since we did not detect Tim-1 and Tim-4 expression on monocytes (unpublished data), the increase in monocytes upon Tim-1 and Tim-4 blockade is likely caused by enhanced inflammation following impaired clearance of apoptotic cells and changes in Th1/Th2 cell ratios.

In summary, treatment with a blocking Tim-4 antibody aggravates atherosclerosis by prevention of phagocytosis of PS-expressing apoptotic cells and activated T cells by Tim-4-expressing cells whereas Tim-1-associated effects on atherosclerosis are related to changes in Th1/Th2 balance and reduced circulating Tregs. Future experiments could focus on stimulating Tim-1 or Tim-4 signaling, which may lead to new therapeutic insights in the prevention of cardiovascular diseases.

Significance.

Myocardial infarction and stroke are the leading cause of mortality worldwide and are generally triggered by rupture of an atherosclerotic plaque. Current treatment consisting of statins and lifestyle advice is inadequate to halt progression or induce regression of atherosclerotic lesions, indicating an urgent need for new therapeutic strategies to inhibit atherosclerosis. Transmembrane T cell immunoglobulin and mucin domain (Tim) proteins are expressed on various immune cells and exert functions related to regulation of inflammatory and immune responses, including T cell costimulation, and clearance of apoptotic cells, which are processes of known importance in atherosclerosis. Therefore, we determined how atherosclerosis is affected in mice treated with blocking antibodies, modeling the hypothetical treatment of humans with similar reagents. We show that blockade of Tim-1 and Tim-4 aggravates atherosclerosis, suggesting that stimulating Tim signaling might be protective and could be of therapeutic value in cardiovascular disease.

Abbreviations

FACS

fluorescence activated cell sorting

HFD

high-fat diet

IFN

interferon

Ig

immunoglobulin

IL

interleukin

ldlr

low density lipoprotein receptor

MCP

monocyte chemoattractant protein

MDA-LDL

malondialdehyde-modified low density lipoprotein

oxLDL

oxidized low density lipoprotein

PMA

phorbol 12-myristate 13-acetate

PS

phosphatidylserine

SNP

single nuclear polymorphism

Tim

T cell immunoglobulin and mucin domain

Th

T helper cell

TNF

tumor necrosis factor

Treg

regulatory T cell