Key Roles of CD4⁺ T Cells and IFN-γ in the Development of Abdominal Aortic Aneurysms in a Murine Model¹

Abdominal aortic aneurysms (AAA)³ represent a common and lethal disorder. Earlier concepts of aneurysm expansion envisaged a simple degenerative process but immunohistochemical, cellular, and molecular biological studies of human tissues and animal models of AAA have consistently shown large numbers of inflammatory cells, elevated levels of cytokines and matrix metalloproteinases (MMPs), and destruction of the elastic media (1, 2). Although many aspects of AAA development are undefined, these observations have led to a paradigm shift in which AAA is seen as a complex remodeling rather than a simple degeneration process. The inflammatory infiltrate, which is temporally and spatially associated with disruption of the orderly lamellar structure of the aortic media, appears to play an etiologic role in AAA development and progression directly through its ability to secrete elevated levels of MMPs (3), or indirectly, by secreting cytokines, including IFN-γ and TNF-α, which induce resident mesenchymal cell MMP productions (4, 5, 6). Increased proteolytic activity of MMPs results in weakening of the structural matrix of the abdominal aortic wall. The MMPs that have been implicated directly in AAA are MMP-9, MMP-2, MMP-1, and MMP-12 (7, 8, 9, 10, 11, 12).

MMPs are a family of Ca²⁺-activated, Zn²⁺-dependent endopeptidases that are able to degrade components of extracellular matrix (ECM) by their concerted actions (13). MMP-9 is one of the most abundant elastolytic proteinases secreted by human AAA tissues. It is primarily produced by aneurysm-infiltrating macrophages at the sites of tissue damage and its expression appears to correlate with increasing aneurysm diameter (14, 15). MMP-2 expression is elevated in human AAA (9, 16, 17). It is primarily produced by resident mesenchymal cells in AAA (9).

Analysis of the inflamed aneurysm wall has revealed the presence of a large number of activated T lymphocytes and macrophages (18, 19), implicating these cells as possible mediators in the disease processes. Monocytes are recruited to sites of tissue injury or chronic inflammation by cell-derived cytokines and chemotactic factors. Recent studies using a murine aneurysm model have revealed that macrophage-derived MMP-9 and smooth muscle cell (SMC)-secreted MMP-2 work in concert to produce aneurysms (8). However, the role of CD4⁺ T cells, the most abundant infiltrates in aneurysmal aortic tissue, remains elusive. IFN-γ, a major inflammatory product of T cells, is a potent activator of macrophage and a cytokine known to regulate MMPs (20). IFN-γ levels are elevated in AAA (21). Therefore, it is reasonable to speculate that the IFN-γ may stimulate MMP production from macrophages and SMC, orchestrating progressive destruction of the normal orderly lamellar architecture in AAA.

The studies on aortic tissue samples from human aneurysms inevitably deal with a late phase of disease and do not necessarily reflect the conditions that initiated aortic dilatation. A reliable animal model is essential to better understand the mechanisms that both initiate and lead to progression of aortic dilatation. Currently, there are three common methods of aortic aneurysm induction including: transient intraluminal elastase perfusion; periaortic application of CaCl₂; and angiotensin II-infusion in aopE-deficient mice. To study the mechanisms of aneurysm formation and development, we have been using and characterized the CaCl₂-induced aneurysm murine model, which emphasizes the role of the inflammatory infiltrate and metalloproteinases in the aneurysmal degeneration process that is remarkably similar to that found in human AAA (8).

In an attempt to understand the inflammatory reaction and subsequent matrix degradation mediated by CD4⁺ T lymphocytes, we have investigated the role of the CD4⁺ T lymphocytes in AAA pathogenesis using the murine model. These studies demonstrated that CD4-deficient mice were resistant to aneurysm formation. Because IFN-γ levels are increased in human AAA and IFN-γ is a product of T cells, we went on to define the specific role of IFN-γ. We found that deficiency of IFN-γ in mice prevent aneurysm formation. Importantly, aneurysms can be largely reconstituted in the CD4^−/− mice by IFN-γ injection or infusion of competent splenocytes into IFN-γ^−/− mice. CD4^−/− mice and IFN-γ^−/− mice demonstrate decreased MMP production in the aorta. Conversely, in vitro treatment of macrophages and SMC by IFN-γ increases MMP-9 and MMP-2 production, respectively. The CD4⁺ T cells and its product, IFN-γ, are important factors in the pathogenesis of AAA by virtue of their ability to induce MMP expression. Thus, these studies delineate the key role invading T lymphocytes play in regulation of the extracellular matrix.

Recombinant murine IFN-γ (rmIFN-γ) and human IFN-γ (hIFN-γ) were purchased from BD PharMingen (San Diego, CA). Tissue culture medium and FBS were purchased from Life Technologies (Grand Island, NY). The collagen preparation used was Vitrogen 100 (Cohesion Technology, Palo Alto, CA), which is ∼95% type I collagen and ∼5% type III collagen.

The homozygous CD4 gene knockout (CD4^−/−) mice and IFN-γ gene knockout (IFN-γ^−/−) mice bred on a C57BL/6 background were obtained from The Jackson Laboratory (Bar Harbor, ME). Normal C57BL/6 mice (The Jackson Laboratory) were used as controls for CD4^−/− and IFN-γ^−/− mice. Both male and female knockout mice were used in a random fashion.

Mice, at age 8 wk, underwent surgery as described previously (8). Briefly, the mice were anesthetized and underwent laparotomy. The abdominal aorta between the renal arteries and bifurcation of the iliac arteries was isolated from the surrounding retroperitoneal structures. The diameter of the aorta was measured in triplicate midway between the renal artery origin and iliac artery bifurcation. After baseline measurements, 0.25 M CaCl₂ was applied to the external surface of the aorta for 15 min. The aorta was rinsed with 0.9% sterile saline and the incision was closed. NaCl (0.9%) was substituted for CaCl₂ in sham control mice. Eight weeks later, the mice underwent laparotomy and dissection. Measurements were repeated at the same location in the mid-infrarenal aorta. Typically, there was diffuse, homogeneous dilatation of the infrarenal aorta. The aorta was collected for zymographic analysis of MMP proteins. For histological studies, the aorta was perfusion-fixed with 10% neutral buffered formalin.

Masson’s trichrome staining: After perfusion-fixation with 10% neutral buffered formalin, abdominal aortic tissues were embedded in paraffin and cut into 4-μm sections. The slides were stained with hematoxylin, crocein scarlet, acid fuchsin, and aniline blue (Sigma-Aldrich, St. Louis, MO). Each staining cycle alternated between fixing and washing procedures. The slides were examined and photographed using light microscopy (Kodak, Tokyo, Japan) (×20).

Mice underwent AAA induction according to the method described above. Four mice in each group were sacrificed at 8 wk for T cell staining performed on paraffin-embedded 4-μm aortic sections. The sections were incubated with a polyclonal rabbit anti-CD3α Ab (BD PharMingen) and diluted 1/20 for 30 min at 37°C. The sections were then briefly washed in citrate solution and subsequently incubated with the secondary Ab which is a biotin-conjugated goat anti-rabbit IgG. T cell staining was examined using light microscopy (×100). Positive controls and nonimmune negative controls were performed. CD3-positive cells were graded in the aortic media and adventitia by a pathologist unaware of the genotype or treatment. Four separate sections from each aorta were stained and evaluated, and the mean grade was reported.

The day after NaCl or CaCl₂ treatment, groups of eight CD4^−/− mice were injected i.p. with 50,000 U of rmIFN-γ, twice weekly for a period of 3 wk. For injection control, after CaCl₂ aneurysm induction, CD4^−/− mice (n = 5) were injected i.p. with the same volume of vehicle, PBS. Eight weeks later, the mice underwent a second laparotomy and the aortas were exposed and measured before being removed for zymography and histology.

C57BL/6 and CD4^−/− mice were sacrificed. Spleens were aseptically removed and teased apart between two sterile slides. Cells were isolated and resuspended in 1 ml RBC lysis buffer (Tris and NH₄CL, pH 7.2). After 1 min, cells were washed with RPMI 1640 medium twice and pelleted to remove cellular debris. Cells were plated in RPMI 1640 supplemented with 10% heat inactivated FBS and incubated at 37°C, 5% CO₂ for 1 h. The suspending cells were collected, pelleted, and resuspended in PBS. Groups of IFN-γ^−/− mice were injected with 5 × 10⁷ splenocytes from wild-type (WT) or CD4-deficient mice via the tail vein 1 day before AAA induction. A second splenocyte infusion was repeated a week later to booster cell number.

Mouse macrophages were isolated from peritoneal fluid. C57BL/6 mice were sacrificed. Peritoneal macrophages were collected, washed, and resuspended in RPMI 1640 medium. Cells were then plated at 1 × 10⁵ cells/well on 12-well plates, and plates coated with fibrillar collagen (1.5 mg/ml, prepared as suggested by the manufacturer; 0.15 ml/cm²). Cells were incubated in collagen-coated plates containing RPMI 1640 with 10% FBS medium at 37°C for 2 h, followed by rinses to remove nonadherent cells. The cells then were incubated in RPMI 1640 and treated with or without rmIFN-γ at a concentration of 0.1 and 1.0 ng/ml for 24 h. The conditioned medium was collected and spun at 3000 × g at 4°C for 30 min to remove cell debris. The effects of IFN-γ on MMP production were compared with untreated cells. The specific activity of mrIFN-γ was >2 × 10⁷ U/mg.

Infrarenal aortic tissues were obtained at organ procurement for transplantation. Informed consent was obtained for tissue collection in accordance with a protocol approved by the Institutional Review Board and Research Committee of the University of Nebraska Medical Center (Omaha, NE). Isolation and culture of aortic SMC was established by using previously described techniques (9, 22). SMC were plated on collagen-coated 12-well plates at 1 × 10⁵ cells/well. Cell cultures were treated with or without hIFN-γ at a concentration of 0.1 and 1.0 ng/ml in 1 ml/well of M-199 medium for 24 h. The medium was collected and spun at 3000 × g at 4°C for 30 min to remove cell debris.

Aortic proteins were extracted as previously described (9). The protein concentration for aortic proteins and cultured media from peritoneal macrophages and aortic SMC treated with or without IFN-γ was standardized with Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). Gelatin zymography was performed as described previously by Longo et al. (8), with 0.8% gelatin in a 10% SDS-polyacrylamide gel. The molecular sizes of gelatinolytic activities were determined using protein standards (Bio-Rad).

Measurements of aortic diameter are expressed as mean ± SEM. A paired Student’s t test was used to compare original and final diameter. Statistical significance was accepted at a p < 0.05.

A role for CD4⁺ T lymphocytes in aneurysm development is implied, as CD4⁺ T lymphocytes are the major component of the cellular infiltrates present in AAA. To study the role of the CD4⁺ T cells in the development of AAA, we used CD4-deficient (CD4^−/−) mice to test their response to CaCl₂ aneurysm induction. Eight weeks after periaortic application of CaCl₂, there was no significant change in aortic diameter or histology in CD4^−/− mice (Table I) which were similar to the NaCl-treated group. However, the control C57BL/6 WT mice showed a 45.8% increase (p < 0.01) in aortic diameter after CaCl₂ treatment (Table I). Connective tissue staining of aortic sections from these mice showed disruption and fragmentation of medial elastic fibers (Fig. 1,b), while NaCl-treated controls show intact medial elastic lamellae (Fig. 1,a). These data demonstrate that CD4⁺ T lymphocytes have a central role in aneurysm development. We also examined the aortic tissues for the presence of T cells (CD3⁺). CaCl₂-induced WT mice show T cell infiltration (Fig. 1,e and Table II), while CD4^−/− mouse aortic tissue showed no T cell infiltration (Table II).

Based on results obtained with CD4^−/− mice, we surmised the effects of the CD4⁺ cells was to secrete cytokines that promoted the local inflammatory response. IFN-γ can up-regulate MMPs and is increased in AAA, implicating it in AAA pathogenesis (4, 21, 23, 24). To examine the role of IFN-γ in AAA development, IFN-γ-deficient (IFN-γ^−/−) mice underwent CaCl₂ aneurysm induction. Aneurysm development was suppressed in IFN-γ-deficient mice (Fig. 2,a). Histologic analysis revealed minimal damage to the medial elastic lamellae in those mice (Fig. 2,c), which is similar to the NaCl-treated control (Fig. 2,b). T cell infiltration in CaCl₂-treated IFN-γ^−/− mice is significantly decreased compared with CaCl₂-induced WT mice (Table II). The expression levels of MMP-9 and -2 from aortic tissues were examined in WT, CD4^−/−, and IFN-γ^−/− mice. Both CD4 and IFN-γ gene-targeted mice exhibited significantly lower production of MMP-2 and MMP-9 in the aorta compared with WT mice after CaCl₂ induction (Fig. 3, lanes 1–6). These observations support previous observations that MMP-2 and MMP-9 are functionally significant in the connective tissue degradation in AAA and also indicate a mechanism by which CD4⁺ T cells and IFN-γ regulate matrix metabolism through production of MMPs.

Because CD4⁺ cells produce an array of cytokines, we next attempted to determine the specific contribution of IFN-γ. Recombinant murine IFN-γ was administrated i.p. to CD4^−/− mice on the day after CaCl₂ or NaCl treatment. CD4^−/− mice received 50,000 U of rmIFN-γ twice a week for 3 consecutive weeks. Table III showed that administration of rmIFN-γ alone partially reconstituted the CaCl₂-induced aneurysm in CD4-deficient mice. Connective tissue staining of aortic sections for those mice showed flattening of medial elastic fibers (Fig. 4,b) compared with control (Fig. 4,a). MMP-2 and -9 levels in aortic tissues of AAA were determined by zymography. The decreased production of MMP-2 and -9 in CD4^−/− mice were restored by infusion of IFN-γ (Fig. 3, lanes 7 and 8). These results demonstrate that T cell cytokines, and IFN-γ in particular, are critically involved in AAA formation, in part, through their ability to regulate tissue-degrading enzymes (MMP-2 and -9) produced in the aortas. To confirm that IFN-γ from T cells alone was adequate to reconstitute aneurysm in IFN-γ^−/− mice, we injected IFN-γ^−/− mice with splenocytes from their corresponding WT background mice. This was done via a tail vein 1 day before calcium chloride aneurysm induction. A second splenocyte infusion was done after a week to booster cell number. IFN-γ^−/− mice infused with WT splenocytes developed aneurysms that did not differ in size from competent WT mice treated with CaCl₂ (Table III). Histological analysis showed the typical elastin destruction following splenocyte infusion (Fig. 4,d). Furthermore, T lymphocyte infiltration was similar to CaCl₂-induced aneurysms in WT mice (Fig. 4,e). To insure that the aneurysm formation did not represent a nonspecific response to splenocyte infusion, splenocytes harvested from CD4^−/− mice were infused into IFN-γ^−/− mice 1 day before and 1 wk after CaCl₂ aneurysm induction. There was no increase in aortic diameter or significant elastin destruction in mice infused with CD4^−/− splenocytes (Table III and Fig. 4 c). The partial reconstitution of AAA with IFN-γ infusion indicates that the products by CD4⁺ cells are important in AAA formation. The partial aneurysm reconstitution by recombinant IFN-γ and full reconstitution by WT splenocytes infusion suggest that other cytokines from lymphocytes may also contribute to AAA formation or that i.p. IFN-γ dosing did not adequately replace intrinsic IFN-γ production. These results provide evidence that IFN-γ from T cells is critical for the development of calcium chloride-induced aneurysms.

To fully understand the role of IFN-γ in the regulation of monocyte/macrophage expression of MMP-9, peritoneal macrophages from C57BL/6 mice were isolated and cultured on polymerized collagen which mimics the physiological milieu of the arterial wall. Macrophages were treated with 0, 0.1, and 1.0 ng/ml of IFN-γ for 24 h. MMP-9 and MMP-2 secreted by macrophages was examined by gelatin zymography (Fig. 5,a). We observed that IFN-γ stimulated the synthesis of MMP-9 and MMP-2. Only low levels of MMP-2 were produced by macrophages. To test the effect of IFN-γ on MMP-2 and MMP-9 expression in SMC, human aortic SMC were cultured and treated with 0, 0.1, and 1.0 ng/ml of IFN-γ for 24 h. Conditioned medium were analyzed by gelatin zymography (Fig. 5 b). Production of MMP-2 was induced by IFN-γ treatment. No MMP-9 production in SMC was detected. These data suggest that IFN-γ induces production of MMP-9 from infiltrating macrophages and MMP-2 from local SMC. These observations combined with the findings of decreased MMP-2 and MMP-9 in the aorta of IFN-γ^−/− and CD4^−/− mice, support the hypothesis that IFN-γ from CD4⁺ cells induces AAA by local up-regulation of two critical MMPs, MMP-2, and MMP-9.

AAA involve disruption and attenuation of the elastic media and excessive production of matrix-degrading proteinases. The prominent inflammatory infiltrate has been implicated in the pathogenesis of AAA (16, 18). While studies of human tissues demonstrate the association between matrix destruction and inflammatory cells, animal data is even more compelling in demonstrating a causal role for local inflammation in AAA pathogenesis (8, 25, 26).

The inflammatory cell infiltrate in AAA is predominately composed of T lymphocytes and macrophages (18, 19). Elucidating specific contributions of subsets of inflammatory cells and their cytokines found in AAA will be one key to understanding AAA and other chronic inflammatory disease processes associated with destruction of the extracellular matrix. CD4⁺ T cells are the predominate immune subset in tissues from human AAA (18), suggesting a potentially important role in AAA formation. T cell infiltration is also one of the major characteristics in our murine aneurysm model (Fig. 1 e). To understand the role of CD4⁺ T cells in the pathogenesis of AAA, we used CD4-deficient mice, testing their response to CaCl₂ aneurysm induction. The hypothesis that tissue-infiltrating CD4⁺ T cells play a central role in AAA development was confirmed by our finding that the absence of CD4⁺ T cells prevents aneurysm development in a murine model. This aneurysm resistance is associated with decreased T cell infiltration. This is the first study to demonstrate that CD4⁺ T cells are essential for aneurysmal degeneration. It is well-known that CD4⁺ T cells are one of the major sources of IFN-γ (20). Furthermore, IFN-γ is a potent activator of macrophages and inducer of MMP-12 and cysteine proteinases (27). Importantly, circulating levels of IFN-γ are elevated in patients with AAA (21). In addition, Schönbeck et al. (28) have recently shown increased tissue levels of IFN-γ. Based on these observations, we hypothesized that IFN-γ was a pivotal mediator of AAA. To test this hypothesis, we injected CD4^−/− mice with IFN-γ after aneurysmal induction. IFN-γ was able to reconstitute aneurysms in those mice. Our observations are consistent with those of Tillides et al. (29) who reported that IFN-γ can mediate atherosclerotic changes when T lymphocytes were absent. The observation that IFN-γ^−/− mice are resistant to aneurysm formation lends further support to the concept that IFN-γ has a causal role in aneurysm formation. This effect was not a nonspecific because IL-6^−/− mice developed aneurysms comparable to control mice. Despite this compelling evidence that IFN-γ is a pivotal mediator of AAA, we further demonstrated that competent splenocyte infusion, but not CD4^−/− splenocyte, into IFN-γ^−/− mice can reconstitute aneurysm formation in those mice. These results demonstrate that IFN-γ produced by CD4⁺ T cells is critical for aneurysm degeneration.

MMP-9, one of the most abundant elastolytic proteases secreted by human AAA tissues, is produced by aneurysm-infiltrating macrophage at the sites of tissue damage (14). Studies from our laboratory and other investigators (7) have shown that macrophage-derived MMP-9 is responsible for the local matrix destruction seen in experimental AAA. We have gone on to show the mesenchymal cell MMP-2 is also essential for aneurysm development in experimental animal models. To gain insight into the mechanisms of IFN-γ-mediated aneurysm, we determined whether IFN-γ altered the levels of expression of aortic MMPs. These studies demonstrate, for the first time, that elastic lamellar preservation in CD4^−/− and IFN-γ^−/− mice correlates with decreased local MMP-2 and MMP-9 expression. Aneurysm reconstitution in CD4^−/− mice by IFN-γ infusion corresponds to restoration of MMP-2 and MMP-9 expression. This is consistent with the observation by Wang et al. (27), that IFN-γ stimulates the release of MMP-9 in vivo. Although several investigations have reported that IFN-γ down-regulates MMP-9 after it is maximally stimulated by Con A or mycobacterium (30) (31), we see evidence that IFN-γ regulates MMP-9 and MMP-2 both in vivo and in vitro. Macrophages and human SMC cultured on a collagen matrix designed to mimic the cells in the arterial wall exhibit increased MMP expression in response to IFN-γ. Tamai et al. (32) also observed that IFN-γ significantly induced the mRNA levels of interstitial collagenase and stromelysins in keratinocytes. Taken together, these data suggest that IFN-γ is a potent stimulator of MMP-2 and MMP-9. The findings demonstrate that the role of T lymphocytes in AAA is mediated largely by IFN-γ. CD4⁺ lymphocytes promote degradation of the extracellular matrix of aortic wall through IFN-γ induced MMP expression. This study highlights a pivotal cytokine in the pathway of AAA formation that involves the IFN-γ signaling. It will provide a novel mechanism for understanding and treating AAA and other chronic inflammatory diseases.