B2 Cells Suppress Experimental Abdominal Aortic Aneurysms

Abstract

Recent reports of rupture in patients with abdominal aortic aneurysm (AAA) receiving B-cell depletion therapy highlight the importance of understanding the role of B cells (B1 and B2 subsets) in the development of AAA. We hypothesized that B2 cells aggravate experimental aneurysm formation. The IHC staining revealed infiltration of B cells in the aorta of wild-type (C57BL/6) mice at day 7 after elastase perfusion and persisted through day 21. Quantification of immune cell types using flow cytometry at day 14 showed significantly greater infiltration of mononuclear cells, including B cells (B2: 93% of total B cells) and T cells in elastase-perfused aortas compared with saline-perfused or normal aortas. muMT (mature B-cell deficient) mice were prone to AAA formation similar to wild-type mice in two different experimental AAA models. Contradicting our hypothesis, adoptive transfer of B2 cells suppressed AAA formation (102.0% ± 7.3% versus 75.2% ± 5.5%; P < 0.05) with concomitant increase in the splenic regulatory T cell (0.24% ± 0.03% versus 0.92% ± 0.23%; P < 0.05) and decrease in aortic infiltration of mononuclear cells. Our data suggest that B2 cells constitute the largest population of B cells in experimental AAA. Furthermore, B2 cells, in the absence of other B-cell subsets, increase splenic regulatory T-cell population and suppress AAA formation.

Abdominal aortic aneurysms (AAAs) remain a life-threatening disease. Rupture of AAAs causes acute hemorrhage leading to shock and death. Moreover, recent reports indicate AAA rupture in patients receiving rituximab (a commercially available anti-CD20 antibody) for the treatment of various autoimmune diseases and B-cell lymphomas.^1–3 With no available nonsurgical therapies to treat AAAs, there is a pressing need to understand the mechanisms of AAA growth and rupture. During AAA formation, aortic medial and adventitial layers undergo major pathological changes, such as degradation of extracellular matrix and marked infiltration of inflammatory cells, such as T cells, B cells, macrophages, mast cells, neutrophils, natural killer cells, and dendritic cells.^4–10 Therefore, understanding the role of the inflammatory cells is crucial to develop a therapy to suppress aneurysm growth.

B cells have long been known to be present in human AAA. By immunohistochemistry (IHC), an estimated 60% to 80% of the inflammatory infiltrate in the adventitia is classified as B cells (CD20⁺). By using flow cytometry, nearly 20% to 41.1% of total mononuclear hematopoietic cells (CD45⁺) are considered B cells (CD19⁺).^4,8 Flow cytometry analysis of lymphocytes⁴ in human AAA wall identified activated memory B cells with homing properties and expressing activation markers. The presence of B-cell–bound immunoglobulins, such as IgM and IgG, in human AAA tissue has also been well documented.^4,7,8,11 B cells are known to form lymphoid follicles called vascular-associated lymphoid tissue^12,13 with T cells, which are prominent in AAA. B cells are also detected in AAA of apolipoprotein E–deficient mice after 14 days of angiotensin II infusion.¹⁴

Similar to AAA, atherosclerosis is characterized by a chronic inflammatory cell infiltrate, including B and T cells.¹⁵ Genetically B-cell–deficient mice, muMT, develop atherosclerosis.¹⁶ Doran et al¹⁷ have reported adoptive transfer of total B-cell population to muMT mice protected mice against atherosclerosis with significant decrease in macrophage content in aortas. Specific roles for B-cell subsets B1a, B1b, and B2 in the pathogenesis and progression of atherosclerosis have also been identified.^16–26 Specifically, B1a cells appear to be protective in atherosclerosis via production of natural antibodies IgM, whereas B2 cells are proatherogenic via activation/proliferation of T cells. As far as role of T cells is concerned, CD4⁺ T cells are atherogenic,²⁷ whereas T-regulatory cells (Tregs; CD4⁺CD25⁺Foxp3⁺) are atheroprotective.^28,29

Although the role of T cells in AAA formation has been widely investigated,^30–35 the role of B cells or B-cell subsets in AAA remains elusive. We hypothesized that B cells were necessary for AAA formation, and the largest constituent of B cells in AAA would be B2 cells. We further hypothesized that adoptive transfer of B2 cells to a B-cell–deficient mouse would exacerbate AAA formation. We have previously detected neutrophils, macrophages, and T cells in the murine aorta at day 3 using a well-established model of aortic elastase perfusion.³⁶ Because B cells have never been fully characterized in experimental AAA, we first examined if B cells are present in murine AAA and, subsequently, developed a flow cytometry method to methodically phenotype and quantify B-cell subsets in AAA. To understand the role of B2 cells, we adoptively transferred isolated B2 cells to muMT mice and examined AAA formation. Our study addresses a putative protective role of B2 cells in experimental AAA.

Materials and Methods

Tissue Collection from Humans

Tissue collection from humans was approved by the University of Virginia (Charlottesville, VA) Institutional Review Board for Health Sciences Research (Institutional Review Board number 13178). AAA tissues from humans were collected from patients undergoing open AAA repair. The tissues collected were immediately rinsed with saline and stored at 10% formalin. The patient and tissue characteristics are given in Supplemental Table S1.

Mice

C57BL/6 (stock number 000664) and muMT mice on C57BL/6 background (stock number 002288) were obtained from Jackson Laboratory (Bar Harbor, ME). Only male mice were used for induction of AAA or cell isolation. All protocols were approved by the University of Virginia Animal Care and Use Committee. All mice were given water and standard chow diet (catalog number 7012; Teklad Diets, Madison, WI) ad libitum.

Elastase Perfusion and Topical Model of Mouse AAA

Elastase perfusion model of mouse AAA was performed using saline (control) or elastase, as previously described,^37,38 and aortas were harvested at day 0, 3, 7, 14, or 21. A topical elastase model of mouse AAA was performed as described by Bhamidipati et al,³⁹ and aortas were harvested at day 14. Video micrometry measurements were performed before perfusion, after perfusion, and at the time of harvest. Changes in aortic diameter were calculated as percentage dilation over baseline at day 14.

Phenotyping and Quantification of Immune Cells in AAA Samples

Dilated aortas were carefully collected from mice and digested with an enzymatic cocktail of 1000 U/mL collagenase type I, 400 U/mL collagenase type XI, 125 U/mL hyaluronidase type I-s, and 60 U/mL DNase.¹⁷ Subsequently, the samples were stained with fluorescent dye–conjugated antibodies and run on the 16-color flow cytometer machine LSRFortessa (equipped with laser lines 488, 405, 561, and 640 nm; Becton Dickinson, Franklin Lakes, NJ) located in the Flow Cytometry Core Facility at University of Virginia. The flow antibodies used were obtained from BD Pharmingen (San Jose, CA) [phosphatidylethanolamine (PE) rat anti-mouse Foxp3, MF23] and Biolegend (San Diego, CA) (purified anti-mouse CD16/32, clone 93; PE/Cy7 anti-mouse CD3ε, clone 145-2C11; allophycocyanin anti-mouse CD4, GK1.5; Alexa Fluor 647 anti-mouse CD5, 53-7.3; PerCP/Cy5.5 anti-mouse CD19, 6D5; PE anti-mouse CD23, B3B4; allophycocyanin/Cy7 anti-mouse CD45, 30-F11; Alexa Fluor 488 anti-mouse/human CD45R/B220, RA3-6B2; and PE anti-mouse IgM, RMM-1). Dead cells were eliminated from analysis by staining with DAPI (Sigma-Aldrich, St. Louis, MO) or LIVE/DEAD Fixable Dead Cell Stain Kit from Molecular Probes, Invitrogen (Eugene, OR). For Foxp3 staining, surface-stained cells were fixed with 4% paraformaldehyde and permeabilized with 0.5% saponin. The flow data were analyzed using FlowJo software. To count the number of cells, CountBright Absolute Counting Beads (Molecular Probes, Grand Island, NY) were added to the cell suspension before flow cytometry. Fluorescent minus one controls were used for each antibody in the experiment.

To quantify intracellular cytokines, isolated B cells were first stained with live/dead cell stain, then surface stained with PE/Cy7 anti-mouse CD19 (6D5; Biolegend), fixed, and permeabilized with saponin. Intracellular cytokines were stained with antibodies from eBiosciences (San Diego, CA) (anti-mouse IL-6 PE, MP5-20F3) and Biolegend [Alexa Fluor 700 anti-mouse IL-2, JES6-5H4; Alexa Flour 647 anti-mouse IL-4, 1B11; PerCP/Cy5.5 anti-mouse IL-10, JES5-16E3; and Alexa Fluor 488 anti-mouse interferon (IFN)-γ, XMG1.2], and counting beads were added before running on LSRFortessa.

IHC Data

Standard immunoperoxidase or immunofluorescence staining methods were performed on paraffin sections (5 μm thick) of human or mouse AAA. In human sections, B cells were detected using anti-CD79a antibody (HM47/A9; Abcam, Cambridge, MA) and T cells using anti-CD3 antibody (SP7; Abcam). In mouse aortic sections, B cells were detected using rat anti-mouse CD45R (RA3-6B2; BD Pharmingen, San Jose, CA), T cells using anti-CD3 antibody (sc-1127; Santa Cruz Biotechnology, Inc., Dallas, TX), macrophages using anti–Mac-2 antibody (Cedarlane, Burlington, NC), and neutrophils using rat anti-mouse Ly-6B.2 antibody (AbD Serotec, Kidlington, UK). Immunoperoxidase and epifluorescent images were acquired on a Nikon Eclipse Ti–U inverted microscope (Melville, NY) and NIS-Elements Br microscope imaging software. Confocal images were acquired on Zeiss LSM 700 (Peabody, MA) using 405-, 488-, and 633-nm lasers and ZEN 2012 lite software (Zeiss).

Isolation of B2 Cell

B2 cells were prepared from spleen of 8- to 10-week-old wild-type (WT) mice. Briefly, mice were euthanized by CO₂ inhalation, followed by cervical dislocation. Spleens were collected, and a single-cell suspension was prepared by passing through a 70-μm mesh. Red blood cells were lysed and B2 cells were isolated using CD43 (ly-48) microbeads (130-049-801) from Miltenyi Biotec (Auburn, CA) using negative selection strategy as the recommended protocol. Soon after isolation, B2 cells were suspended in phosphate-buffered saline (PBS) and transferred to muMT mice.

Isolation and Culture of B Cells from Aorta

AAA was induced in WT mice using an elastase perfusion method. For controls, aortas were perfused with saline. Day 14 after perfusion, aortas were harvested as described before. Two elastase- or saline-perfused aortas were pulled together and digested with the aorta-digesting enzymes. Thereafter, the cell suspension was washed with PBS and untouched pan B cells were isolated using pan B cell isolation kit (130-095-813; Miltenyi Biotec, San Diego, CA) using the recommended protocol. Immediately after purification, the cells were washed and suspended in B-cell medium [Iscove's modified Dulbecco's medium supplemented with 1 mmol/L sodium pyruvate, 2 mmol/L l-glutamine, 55 μmol/L β-mercaptoethanol, 0.1 mg/mL bovine serum albumin, and 1% antibiotic-antimycotic (Gibco, Grand Island, NY)] and incubated at 37°C. Two hours after incubation, the aortic B cells were cultured for a further 18 hours with 5 ng/mL phorbol 12-myristate 13-acetate (PMA; Sigma-Aldrich), calcium ionophore A23187, or 1.25 μmol/L ionomycin (Sigma-Aldrich) and 10 μg/mL brefeldin A (Sigma-Aldrich) and intracellular cytokines, which were quantified on a flow cytometer, as described earlier in Materials and Methods.

B-Cell Transfer and Induction of AAA

Seven-week-old muMT male littermates were transferred with 25 million of isolated B2 cells or PBS via tail vein, and AAA was induced using elastase perfusion model after 7 days. Fourteen days after elastase perfusion, mice were harvested.

Statistical Analysis

Means and SEMs were calculated using GraphPad Prism version 5 (La Jolla, CA). An unpaired nonparametric t-test was used to determine differences between two groups, whereas multiple groups were compared using a simple one-way analysis of variance. P < 0.05 was considered significant.

Results

B and T Cells Colocalize in Human AAA

We performed double-immunofluorescence staining to examine the abundance and localization of both B (anti-CD79α) and T (anti-CD3ε) cells in human AAA wall (Figure 1A). Both B and T cells were abundant and contacted each other, forming lymphoid-like structures in the aortic media of AAA. Verhoeff–Van Gieson (VVG) and α-smooth muscle actin staining revealed absence of thrombus in AAA wall (Supplemental Figure S1, A and B).

Figure 1.

Presence of B and T cells in the aortic wall of human and mouse AAA. A:Left: Human AAA section stained with B cells (CD79α, brown). Right: A consecutive section stained for both B cells (CD79α, green) and T cells (CD3ε, red); image was acquired on a confocal microscope. B: Mouse aortic sections from day 0, 3, 7, 14, and 21 after elastase perfusion stained for B cells (B220, brown) (n = 3). C: Mouse aortic section from 14 days after elastase perfusion stained for B cells (B220, green) and T cells (CD3ε, red); image was acquired on an epifluorescent microscope. The asterisk indicates lumen. Scale bars: 500 μm (A, left); 10 μm (A, right); 50 μm (B and C).

B Cells in Experimental AAA

To determine whether B cells are present in experimental models of mouse AAA, we induced AAA by elastase perfusion in WT mice and harvested the aortas at days 0, 3, 7, 14, and 21. Staining for CD45R/B220, a marker for B cells, demonstrated appearance of B cells at day 7, which persisted at day 21 in the adventitial layer (Figure 1B and Supplemental Figure S1C). Similar to the human AAA samples, we observed B and T cells are present together at day 14 (Figure 1C). To avoid model-specific effects, we used an alternate AAA model, which we recently described (using full-strength elastase placed adventitially on WT mice³⁹), and observed similar accumulation of B cells at day 14 (data not shown). Altogether, our results demonstrate prevalence of B cells in experimental models of mouse AAA.

Characterization of B-Cell Subsets in Mouse AAA

Next, we developed a unique method to perform flow cytometry on individual mouse AAAs to quantify B-cell subsets. Our optimized protocol for digestion allowed us to prepare a cell suspension from an approximately 5-mm segment of abdominal aorta (Figure 2A) from mouse. Unexpectedly, we observed that surface expression of CD23, a well-studied marker for B-cell phenotyping, was abolished (Supplemental Figure S2A) in our optimized protocol and in the protocol described by Butcher et al.⁴⁰ Therefore, we followed the gating strategy, as described by Thomas et al,⁴¹ which uses the markers CD19 and B220 to determine the B1 and B2 cell populations. In our gating strategy (Figure 2B), lymphocytes were gated first, followed by live cells, singlets, CD45⁺CD3⁻ mononuclear hematopoietic cells, and B cells (CD19⁺B220⁺). CD19^hiB220^lo cells were considered B1 cells, whereas CD19^loB220^hi cells were considered B2 cells. Furthermore, B1-gated cells were phenotyped as B1a (CD19^hiCD5^hi) and B1b (CD19^hiCD5^lo) cells. All B cells were found to express IgM (data not shown). We further observed that our surgical procedure, which involved laparotomy of mouse to perform saline or elastase perfusion of abdominal aorta, led to a decrease in B1 cell population (from 46% to 13%) and an increase in B2 cell population (from 21% to 49%) of total hematopoietic cells in peritoneal fluid (Supplemental Figure S2B). However, the populations of B-cell subsets in spleen and blood were unaffected after surgery (data not shown).

Figure 2.

A: Schematic shows a segment of dilated aorta harvested for phenotypic analysis of B cells. B: Representative flow plots and gating strategy to phenotype B cells from day 14 elastase-perfused aorta. C: Representative flow plots (already gated for lymphocytes, live and singlets) from day 14 saline-perfused aorta. FSC, forward scatter; IVC, inferior vena cava; SSC, side scatter.

B2 Cells Are the Largest Constituent of B Cells in Mouse AAA

We used counting beads in flow cytometry to determine the number of immune cells in unperfused normal abdominal aorta, and saline- and elastase-perfused aortas of WT mice. Elastase perfusion led to AAA formation in WT mice (Figure 3). Few immune cells were detected in the normal aorta, whereas the numbers of mononuclear hematopoietic cells, T cells, B cells, and B-cell subsets were significantly higher in elastase-perfused aortas compared with the saline-perfused aortas (Figure 3). In elastase-perfused aortas, 23% of mononuclear infiltrates were T cells, whereas 4.5% were B cells. B-cell numbers were 3631 ± 578 and 1476 ± 524 in elastase- and saline-perfused aortas, respectively. B2 cells were the largest constituent of B cells, both in elastase (B1 versus B2, 7.2% versus 92.8% of total B cell) and saline (B1 versus B2, 3.7% versus 96.3% of total B cell) perfused aortas (Figures 2, B and C, and 3). Among the B1 cell subsets, B1b cell number was higher than B1a cell. Correlation analysis between the immune cell number aorta diameters (irrespective of elastase or saline perfusion) revealed that total hematopoietic cell, T cell, B1, B1a, and B1b cell counts strongly correlated with increasing aorta diameter (Supplemental Figure S3). However, total B and B2 cell counts poorly correlated with aortic diameter. Because B2 cells were the largest constituent of B cells, B2 cell number affected correlation of total B cell with aorta diameter. Altogether, our data showed that elastase perfusion in WT mice led to increased aortic infiltration of B cells compared with saline-perfused or normal aortas, and many B cells are B2 cells; however, it is undefined whether B2 cells are protective or harmful.

Figure 3.

Increased infiltration of immune cells after saline or elastase perfusion of abdominal aorta of WT mice. Increase in aorta diameter and quantification of the number of immune cells in day 14 saline (n = 6) and elastase (n = 11) perfused aortas, and equivalent length of unperfused abdominal aorta (n = 6) from WT mice. Values are expressed as means ± SEM. ^∗P < 0.05, ^∗∗P < 0.01, and ^∗∗∗P < 0.001, respectively.

B Cells from Saline- and Elastase-Perfused Aortas Have a Similar Cytokine Secretion Profile

Cytokine-producing effector B cells (Be) have been shown to regulate many inflammatory diseases, which are independent of antibodies produced by B cells.⁴² Harris et al⁴³ divided Be cells into two subsets: Be1 cells that synthesize tumor necrosis factor-α, IFN-γ, and IL-12, and Be2 cells that synthesize IL-2, IL-4, IL-6, and IL-10. To determine the Be cell types present in saline- and elastase-perfused aortas, we purified total B cells using magnetic-activated cell sorting (MACS) columns (Figure 4A) stimulated with a mixture of PMA and ionomycin, blocked cytokine secretion using brefeldin A, and examined intracellular synthesis of IFN-γ, IL-2, IL-4, IL-6, and IL-10 using flow cytometry. We found a large population of B cells synthesized IL-6 and IL-10, indicating predominance of Be2 cell type; however, no difference was observed between these populations of B cells isolated from saline- and elastase-perfused mice (Figure 4B). This suggests that the B cells infiltrated to the aorta have similar Be populations irrespective of saline or elastase perfusion. However, more B cells in elastase-perfused aorta compared with saline-perfused aorta suggests a higher level of Be cytokines in the aneurysms.

Figure 4.

Cytokines synthesized by B cells. A: Elastase- or saline-perfused aortas were collected at day 14 and digested, and untouched total B cells were isolated using Miltenyi column. B: The isolated aortic B cells were cultured for 18 hours with 5 ng/mL PMA, 1.25 μmol/L ionomycin, and 10 μg/mL brefeldin A and then stained with live/dead dye and B cell surface marker (CD19). Subsequently, the cells were fixed, permeabilized, and stained for intracellular cytokines, which were quantified on a flow cytometer. B cells isolated from two aortas were pooled to make n = 4. Values are expressed as means ± SEM.

B-Cell Deficiency Cannot Protect Mice from Experimental AAA

muMT mice are deficient in mature B1 and B2 cells. To understand the role of B cell in AAA formation, we induced AAA in muMT mice using the elastase perfusion model. To our surprise, we observed muMT mice developed AAA similar to the WT mice at day 14 (Figure 5A). Histochemical staining on aortic cross sections revealed increased collagen degradation (VVG) and abundance of T cells, macrophages, and neutrophils (Figure 5A), which were also found in day 14 WT mice (Supplemental Figure S4). Our findings contradict a study by Zhou et al,⁴⁴ who reported that muMT mice were protected from AAA by elastase perfusion. Therefore, to further confirm our findings, we studied the effect of B-cell deficiency using a second experimental AAA model. Specifically, we observed that both WT and muMT mice developed AAA at day 14 using a novel topical elastase model that we reported previously³⁹ (Figure 5B). The IHC staining of aortic sections from muMT mice further demonstrated abundance of T cells, macrophages, and neutrophils (Figure 5B), similar to day 14 WT mice (Supplemental Figure S4). These results suggest that in the absence of B cells, other immune cells, such as neutrophils, macrophages, and T cells, potentially regulate aneurysm formation in muMT mice.

Figure 5.

muMT mice develop AAA with increased elastin degradation and increased infiltration of immune cells to aorta. muMT mice develop AAA similar to WT mice in elastase perfusion (A; n = 8 to 10) and topical elastase (B; n = 6 to 8) model. AAA sections from muMT mice were stained for VVG (elastic fibers, black), T cells (CD3ε, brown), macrophages (Mac2, brown) and neutrophils (Gr1, brown). Values are expressed as means ± SEM. There is no significant difference in AAA growth between WT/elastase and muMT/elastase animals in elastase perfusion or topical elastase model. ^∗∗∗P < 0.001. Scale bar = 50 μm (A and B).

Adoptive Transfer of B2 Cells Suppresses AAA Growth and Increases Peripheral Treg Population in muMT Mice

Because the muMT mice developed an aneurysm similar to WT mice, we speculated that B1 and B2 cells have opposing roles in AAA formation, as they do in atherosclerosis. Adoptive transfer of total spleen cells or splenic B cells has been shown to protect mice from development of atherosclerosis.²⁶ Because B2 cells worsen experimental atherosclerosis, we hypothesized that adoptive transfer of B2 cells would worsen aneurysm formation in muMT mice. Therefore, we adoptively transferred 25 million MACS column purified (purity >98%) (Supplemental Figure S5) splenic B2 cells in PBS to muMT mice. As described in experimental atherosclerosis studies,¹⁹ control muMT mice were injected with an equal volume of PBS. Transfer of isolated B2 cells reconstituted 8.8 ± 1.6 × 10⁵ B2 cells in the spleen of muMT mice (Figure 6, A and B). This led to a 2.2-fold increase in T cell and a twofold increase in the number of mononuclear hematopoietic cells in spleen (Figure 6, C and D). Contrary to our hypothesis, the aneurysm size was significantly smaller in B2 cell–transferred mice compared with PBS controls (75.2% ± 5.5% in B2 cell–transferred muMT versus 102.0% ± 7.3% in control muMT, P = 0.01) (Figure 6E). In accordance with decreased AAA size, the number of mononuclear hematopoietic cells and T cells trended lower in the elastase-perfused aortas of B2 cell–transferred mice (Supplemental Figure S6), although it did not reach to a level of statistical significance. This suggests adoptive transfer of B2 cell–induced generation or differentiation of a protective cell type.

Figure 6.

Adoptive transfer of B2 cells suppressed AAA formation and increased peripheral Treg population in muMT mice. A: Representative flow plots show reconstitution of B2 cells in spleen of muMT mice transferred with 25 million WT B2 cells. Increase in absolute numbers of B2 (B), T (C), and mononuclear hematopoietic (D) cells in spleen of muMT mice (n = 5) transferred with B2 cells. E: Increase in aorta diameter in PBS or B2 cell–transferred mice (n = 7 to 9). F: Representative flow plots show increased Treg (CD4⁺Foxp3⁺) population in B2 cell–transferred mice. G: Increase in Treg population in spleen of muMT mice (n = 5) transferred with B2 cells. Values are expressed as means ± SEM. ^∗P < 0.05, ^∗∗P < 0.01, and ^∗∗∗P < 0.001, respectively.

B2 cells have been shown to promote differentiation of naïve CD4 T cells to Tregs. Therefore, we determined Treg population (CD45⁺CD3ε⁺CD4⁺Foxp3⁺) in the spleen, draining lymph nodes, peritoneal fluid, and aneurysm tissues of these mice using flow cytometry. The B2 cell–transferred muMT mice showed an increased percentage (% of total CD3ε⁺ T cells), 0.24% ± 0.03% versus 0.92% ± 0.23% (P = 0.02), as well as total number (P = 0.01) of Tregs in spleen (Figure 6, F and G). The percentage of Tregs in infiltrated T cells trended to be increased in B2 cell–transferred aorta (Supplemental Figure S6); however, we did not observe significant differences in Treg population in lymph nodes or peritoneal fluids. The observed increase in Treg population in the muMT mice might be originated from the contaminating Treg in B2 cell preparation. Therefore, we examined our B2 cell preparation for the presence of Tregs. We did not find contaminating Tregs in isolated B2 cell preparations that were being transferred to the muMT mice (Supplemental Figure S5). These results altogether showed that adoptive transfer of isolated B2 cells to muMT mice suppressed aneurysm formation with a concomitant increase in peripheral Tregs and a decrease in aortic infiltration of immune cells.

Discussion

The role of B cells in AAA formation has been poorly understood. By using a mouse elastase perfusion-induced AAA, we first determined the location and phenotype of the B cells. We observed that B and T cells colocalized in human and mouse AAA and that the B2 cell population was significantly higher than B1 cells. By using two models of experimental AAA, we observed that the muMT mice are not protected from AAA and appear similar to the WT mice. Unexpectedly, adoptive transfer of B2 cells into muMT mice suppressed AAA and was associated with an increase in peripheral Treg cells.

B cells derived from spleen regulate experimental atherosclerosis.¹⁹ In atherosclerosis, B2 cells are considered harmful via activation/proliferation of T cells. However, in vitro, B2 cells promote Treg generation from naïve CD4⁺ T cells in the presence of transforming growth factor (TGF)-β and IL-2.⁴⁵ Moreover, resting B cells expand the Treg population via producing TGF-β3.⁴⁶ Others have also shown that splenic B cells proliferate splenic Tregs in the presence of soluble anti-CD3 and irradiated splenic antigen-presenting cells.⁴⁷ Furthermore, B-cell depletion by anti-CD20 antibody led to encephalomyelitis with concomitant significant reduction in peripheral Tregs.⁴⁷ In these studies, the B-cell–depleted mice have a normal thymic Treg content,^45–47 indicating a dominant role of B-cell–mediated generation of peripheral Tregs in immunosuppression. Consistent with a decreased Treg population in B-cell depletion studies, muMT mice were found to have significantly fewer peripheral Tregs compared with WT mice. A B-cell–mediated increase in Treg population was found to be independent of B-cell production of IL-10. Altogether, there are several lines of evidence that splenic B cells can increase Treg population either by differentiating naïve CD4 T cells to Tregs, expanding the existing population of Tregs, or recruiting Tregs to lymphoid organs. Moreover, these findings support our observation that adoptive transfer of B2 cells increases the splenic Treg population in muMT mice. Furthermore, TGF-β activity, which is required for Treg homeostasis, has been shown to protect WT mice from AAA, and neutralization of TGF-β using anti–TGF-β antibody led to development of aortic aneurysm.⁴⁸

Both atherosclerosis and AAA are aortic diseases, and may share common risk factors, such as smoking, hypertension, and obesity.^49,50 It is not clear whether atherosclerosis leads to AAA formation or vice versa. In the Tromsø (Norway) study, lack of consistent dose-response relationship between atherosclerosis and AAA suggests atherosclerosis may develop concomitantly with AAA or secondary to AAA growth.⁵⁰ Moreover, diabetes is positively associated with atherosclerosis, but negatively associated with AAA, which contradicts the consensus that AAA is a manifestation of atherosclerosis.⁵¹

Tregs have been shown to play an important role in suppression of atherogenesis in mice.²⁹ It has also been shown that depletion of Foxp3⁺ Tregs promotes atherosclerosis in mice.²⁸ As far as human AAA is concerned, Yin et al⁵² found that the population of CD4⁺CD25⁺Foxp3⁺ Tregs was significantly lower in the peripheral blood of patients with AAA compared with patients with abdominal aortic atherosclerotic occlusive disease or healthy controls.

Ait-Oufella et al⁵³ reported that both genetic and anti-CD25 antibody-mediated depletion of Tregs augmented angiotensin II–induced aneurysm formation, and supplementing genetically Treg-deficient mice with WT Tregs suppressed aneurysm formation. Although anti-CD25 antibody treatment causes a small decrease in peripheral Treg population (not the natural Treg content of thymus), it significantly affects host immune response.⁵⁴ Because B2 cells differentiate naïve CD4⁺ T cells to Tregs to a greater frequency than dendritic cells,⁴⁵ a potential mechanism for B2-induced suppression of AAA in muMT mice would be Tregs differentiated or expanded in lymphoid organs, such as spleen, suppress inflammation systemically, leading to inhibition of AAA growth. Alternatively, B2 cell–induced newly differentiated Tregs are recruited to the site of AAA and prevent aneurysm growth. In fact, our studies demonstrate increased population of Tregs in the spleen of muMT mice after B2 cell transfer.

We found the total B cell isolated from saline- or elastase-perfused aortas produced both proinflammatory (IFN-γ and IL-6) and anti-inflammatory (IL-2, IL-4, and IL-10) cytokines. B1 cells are well known to produce natural IgM and protect mice from atherogenesis. In vitro, in contrast to the effect of B2 cells, B1 cells promote the differentiation of proinflammatory type 1 helper T cell (Th1) and Th17 from naïve CD4⁺ T cells with a greater frequency than dendritic cells.⁴⁵ Although B1 cells are fewer than B2 cells, they have greater potency of differentiating Th17 than B2 cells differentiating Tregs. As far as regulation of T-cell phenotype is concerned, the effects of B1 and B2 may counteract each other via Th17 and Treg in WT mice.

There are several limitations to our study. In most of the studies involving AAA patients, B cells constitute a large percentage of mononuclear cell infiltrate. However, in the elastase perfusion model, we found the percentage of hematopoietic cells that were B cells (CD45⁺CD3⁻CD19⁺B220⁺) was only 4.5%. A plausible explanation for this is most of the human AAA samples were in late stages of aneurysm growth, which might have taken years to develop. On the other hand, elastase perfusion is a rapid model and it takes approximately 14 days for aneurysm development. Nevertheless, this model has been documented to recapitulate human AAA development^55,56 and widely used for experimental AAA studies. Furthermore, B-cell populations are identified by different markers and differ greatly between mice and humans.^57,58

To understand the role of B cells in aneurysm formation, we performed elastase perfusion–induced AAA in muMT mice and found that the muMT mice developed AAA similar to the WT mice. In accordance with this finding, a 62-year-old patient with a history of non-Hodgkin lymphoma receiving rituximab had ruptured AAA,² whereas patients with immune cytopenia or spondyloarthropathies died after receiving rituximab treatment.^1,3 In contrast to our findings, Zhau et al⁴⁴ have reported that muMT mice were protected from elastase perfusion–induced AAA and the absence of humoral immune response (IgG or IgM antibody) was attributed to the protection. Supplementing the muMT mice with total IgG antibodies or natural IgM antibodies induced AAA formation. To confirm our observation that muMT mice are susceptible to experimental AAA, we used a second model of AAA (topical elastase model developed in our laboratory)³⁹ and still found no protection in AAA in muMT mice. Overall, formation of AAA in muMT mice emphasizes dominance of other immune cells, such as neutrophils, macrophages, and T cells in AAA in absence of B cells.

Increasing B2 cells in the spleen of muMT mice via adoptive transfer can increase Treg population by differentiating CD4⁺ T cell, expanding the population of existing Tregs, or both. Further experimentation needs to be done to delineate the mechanism of B2 cell–mediated increase in Treg population. We observed that adoptive transfer of B2 cells not only increased splenic B2 and Treg populations, but it also increased total hematopoietic and T-cell population. Among the T-cell population, Tregs comprised only 0.9%. Therefore, the presence of other splenic T-cell subpopulations, such as Th1, Th2, and Th17, in spleen of B2 cell–transferred muMT mice needs to be determined. Furthermore, it is unknown how splenic Treg content regulates AAA formation. It is possible that adoptively transferred B2 cells may be synthesizing protective antibodies.

Although the potent immunosuppressive effects of Tregs can be used in the treatment of multiple inflammatory diseases, the scarce availability and unspecific immune response generate major obstacles in using autologous Treg cells as a means of treatment of AAA. Thus, various methods of Treg generation have been tested, including dendritic cell–induced Treg differentiation.⁵⁹ Because B2 cells modulate Treg differentiation from naïve CD4⁺ T cells at a greater frequency than dendritic cells, increasing B2 cell number may represent an alternative and effective approach to increase Treg number and prevent AAA growth.

Altogether, we have shown that both B1 and B2 cells are present in mouse experimental AAA, with significantly more B2 than B1 cells. Absence of total B cells does not affect aneurysm formation; however, adoptive transfer of isolated B2 cells can suppress AAA progression, with an increase in peripheral Treg population. Our study has highlighted a possible protective role of B2 cells in vascular diseases.

Supplemental Data

Supplemental Figure S1

A: Human AAA section stained with VVG (elastic fibers, black; nuclei, blue to black; collagen, red). B: Human aortic sections stained with α-smooth muscle actin (red), and H&E from AAA (left panel) and healthy aorta (right panel). C: Mouse aortic sections from day 7, 14, and 21 after elastase perfusion stained with VVG. Insets are enlarged. Scale bars: 250 μm (A and B); 50 μm (C and insets). The asterisk indicates lumen.

Supplemental Figure S2

A: Loss of surface expression of CD23 after treatment of mouse splenocytes with aorta enzyme digestion cocktail, without (left panel) and with (right panel) enzymes. B: Representative flow cytometry plots show change in percentage of B-cell population (expressed as percentage of total mononuclear hematopoietic cells) in peritoneal fluid of mice after surgery: unoperated on (left panel) and operated on (right panel) mice.

Supplemental Figure S3

Correlation of immune cell types with increase in aorta diameter at day 14 of saline or elastase perfusion in WT mice. The r² and P values are shown.

Supplemental Figure S4

AAA induced by elastase perfusion or topical elastase in WT mice was stained for VVG (elastic fibers, black), T cells (CD3ε, brown), macrophages (Mac2, brown) and neutrophils (Gr1, brown). Scale bar = 50 μm.

Supplemental Figure S5

Study design for investigating the role of B2 cells in AAA formation. B2 cells were isolated from mouse spleen using CD43 (ly-48) microbeads and MACS column. A part of the isolated B2 cells was examined for purity using flow cytometry. Representative flow plots show purity of B2 cells in percentage of total hematopoietic cells (left panel) and absence of contaminating Tregs (CD4⁺Foxp3⁺) in isolated B2 cell population (right panel). Isolated B2 cells (25 × 10⁶) in PBS or PBS alone were injected to muMT mice 7 days before elastase perfusion to abdominal aorta. Fourteen days after elastase perfusion, AAA size was determined, and aorta, spleen, peritoneal fluid, and lymph nodes were harvested.

Supplemental Figure S6

Number of mononuclear hematopoietic cells (A), T cells (B), and Treg cells (C, as % of T cells) in elastase-perfused segment of aorta of mice that received PBS or WT B2 cells. Values are expressed as means ± SEM (n = 4 to 5); no statistically significant differences are found in cell populations in A, B, or C.