Mycobacterium bovis Bacille-Calmette-Guérin Infection Aggravates Atherosclerosis

Tuberculosis has been associated with increased risk of atherosclerotic cardiovascular disease. To examine whether mycobacterial infection exacerbates atherosclerosis development in experimental conditions, we infected low-density lipoprotein receptor knockout (Ldlr -/-) mice with Mycobacterium bovis Bacille-Calmette-Guérin (BCG), an attenuated strain of the Mycobacterium tuberculosis complex. Twelve-week old male Ldlr -/- mice were infected with BCG (0.3–3.0x106 colony-forming units) via the intranasal route. Mice were subsequently fed a western-type diet containing 21% fat and 0.2% cholesterol for up to 16 weeks. Age-matched uninfected Ldlr -/- mice fed with an identical diet served as controls. Atherosclerotic lesions in aorta were examined using Oil Red O staining. Changes induced by BCG infection on the immunophenotyping profile of circulating T lymphocytes and monocytes were assessed using flow cytometry. BCG infection increased atherosclerotic lesions in en face aorta after 8 weeks (plaque ratio; 0.021±0.01 vs. 0.013±0.01; p = 0.011) and 16 weeks (plaque ratio, 0.15±0.13 vs. 0.06±0.02; p = 0.003). No significant differences in plasma cholesterol or triglyceride levels were observed between infected and uninfected mice. Compared to uninfected mice, BCG infection increased systemic CD4/CD8 T cell ratio and the proportion of Ly6Clow non-classical monocytes at weeks 8 and 16. Aortic plaque ratios correlated with CD4/CD8 T cell ratios (Spearman’s rho = 0.498; p = 0.001) and the proportion of Ly6Clow non-classical monocytes (Spearman’s rho = 0.629; p < 0.001) at week 16. In conclusion, BCG infection expanded the proportion of CD4+ T cell and Ly6Clow monocytes, and aggravated atherosclerosis formation in the aortas of hyperlipidemic Ldlr -/- mice. Our results indicate that mycobacterial infection is capable of enhancing atherosclerosis development.


INTRODUCTION
It is estimated that a quarter of the world population has latent tuberculosis infection, and about 10 million people develop active tuberculosis each year globally (1,2). Patients with a history of active tuberculosis have an increased risk of myocardial infarction, ischemic stroke, and peripheral arterial disease, suggesting that mycobacterial infection has a role in atherosclerotic cardiovascular disease (3)(4)(5)(6). Although these studies accounted for common traditional cardiovascular risk factors, there is a possibility of residual confounding effects from measured and unmeasured characteristics that may not be fully controlled for in human population-based studies (7). Therefore, there is a need to explore the relationship between mycobacterial infection and atherogenesis in experimental animal models. Furthermore, perturbations in circulating T cell and monocyte subsets have been described in tuberculosis (8,9). These immune cells play an important role in atherosclerosis development (10), but their correlation with atherosclerotic plaque in the setting of mycobacterial infection is not well characterized.
The pro-atherogenic effects of infectious agents were first described more than 4 decades ago, when Fabricant et al reported experimental induction of atherosclerosis by Marek's virus in chickens (11). Bacterial pathogens including Chlamydia pneumoniae (12- 14), Helicobacter pylori (15), and periodontal organisms such as Porphyromonas ginvivalis (16) have been associated with increased atherosclerosis formation in hyperlipidemic animal models. Several potential mechanisms linking infection and atherosclerosis have been described, including direct pathogen invasion of vascular tissue and indirect effects via inflammatory and immune mechanisms (17)(18)(19). Previous animal studies using Mycobacterium bovis Bacille-Calmette-Gueŕin (BCG), an attenuated strain of the Mycobacterium tuberculosis complex, have shown different modulating effects of infection in atherosclerosis development. In hypercholesterolemic rabbits, subcutaneous M. bovis BCG injections enhanced atherogenesis (20). However, a recent study in ApoE*3 Leiden.CETP mice showed reduced atherosclerotic plaque formation after intravenous inoculation of M. bovis BCG (21). Notably, the latter study was confounded by a significant reduction of plasma cholesterol levels in the mice infected with M. bovis BCG, which likely mediated the final atherosclerosis outcome in this model (22).
To further assess the effects of mycobacteria in atherogenesis, we infected low-density lipoprotein receptor knockout (Ldlr -/-) mice with M. bovis BCG via the intranasal route. We used this route of infection to mimic the natural respiratory route of acquisition of mycobacterial infections (23). We used Ldlr -/mice, as M. bovis BCG infection was not expected to induce significant plasma cholesterol changes in this atherosclerosis model (24). We aimed at assessing whether M. bovis BCG infection exacerbates atherosclerosis development and induces changes on the immunophenotyping profile of circulating T cell and monocytes. We report that M. bovis BCG infection expanded the proportion of circulating CD4 + T cell and Ly6C low monocytes, and aggravated atherosclerosis formation in murine aorta.

Mice, Diet, and Study Setting
Twelve-week old male C57BL/6J Ldlr -/mice were purchased from the Jackson Laboratory and housed at the Laboratory Animal Medical Services (LAMS) within the University of Cincinnati. Mice were anesthetized and inoculated with M. bovis BCG [0.3-3.0x10 6 colony-forming units (CFUs)] via the intranasal route. Mice were subsequently fed a western-type diet (WD) containing 21% fat and 0.2% cholesterol (Envigo TD.88137 diet) for up to 16 weeks. Age-matched uninfected Ldlr -/mice fed with an identical WD served as controls. Mice were weighed every 2 weeks after initiation of WD to assess for differences in total body weight between groups. The protocols for animal experiments were conducted as per the University of Cincinnati Institutional Animal Care and Use Committee (IACUC) and National Institutes of Health (NIH) guidelines.

Atherosclerotic Lesion Assessment
Mice were euthanized at 8 and 16 weeks of WD to examine atherosclerotic lesions in aortic root sections and en face aorta using Oil Red O staining. Twelve frozen sections per sample were examined throughout the aortic root. Plaque ratios were determined based on plaque area per total area at the aortic root. For en face assessments of total aorta, the extent of plaque was measured using the plaque size per aorta area ratio using ImageJ (NIH, Bethesda, MD). Plaque composition was assessed in aortic root sections. We used CD68 antibody (Abcam ab955) at 1:200 dilution and SMC alpha actin (Abcam ab5694) at 1:100 dilution for immunofluorescence staining to assess macrophage and smooth muscle content, respectively. Images were captured using immunofluorescence microscopy (Olympus BX61) and the areas of positive fluorescence per total area of plaque ratios were estimated. Fibrosis was detected by Sirius red staining.

Circulating Lipids Assessment
Blood was collected via intra-cardiac puncture immediately after euthanasia. Plasma was separated for analysis of circulating lipids. Total plasma cholesterol and triglyceride levels were measured using enzymatic assays (Infinity TM reagents). Lipoprotein distribution was assessed using fast protein liquid chromatography (FPLC).

Immunophenotyping of Circulating T Cells and Monocytes
T cell and monocyte subsets were assessed using flow cytometry. After two rounds of red blood cell lysis, FcgII/III receptors were blocked using anti-mouse CD32/CD16 for 15 min at 4°C (Leinco clone YT1.24). Cells were stained using antibodies for 30 min at 4°C at 1:100 dilution in FACS buffer (buffered salt solution with 0.5% bovine serum albumin; Leinco). The T cell panel included the following antibodies and conjugated fluorochromes: CD45. Cells stained with the T cell panel antibodies were washed and fixed with Fix/Perm buffer and then stained for intracellular FoxP3 (eBioscience clone NRRF-30/PE). Cells stained with the monocyte panel antibodies were washed and fixed with Fix/Perm buffer and then stained for intracellular NOD2 (Novus Biologicals clone 2D9/Alexa Fluor 488). Flow cytometric data were acquired using a BD TM LSR II. Data were analyzed using FlowJo v10 software. We used Fluorescence minus one (FMO) controls to set our gates.

Statistical Analyses
We used unpaired Student's t-test for group comparisons of numeric variables and flow cytometry data. To assess the correlation between immune parameters and plaque ratio, we used the Spearman's correlation test. Analyses were carried out in Stata v12 (College Station, TX); p values <0.05 were considered statistically significant. All p values were 2-tailed.

M. bovis BCG Infection Induces an Expansion of CD4 + T Cells and Monocytes
In a prior study of rabbits inoculated with two subcutaneous injections of M. bovis BCG and fed with cholesterolsupplemented diet, infected rabbits displayed increased atherosclerotic lesions in thoracic aorta compared to uninfected controls with similar plasma cholesterol levels (20). However, a recent study in APOE*3-Leiden.CETP mice showed that intravenous inoculation of M. bovis BCG was associated with decreased atherosclerosis formation in the aortic root after 6 weeks of infection (21). In the latter model, infection induced lower plasma cholesterol levels compared to uninfected controls, which may have affected the atherosclerosis outcome of the experiments. Our data shows that M. bovis BCG infection is capable of increasing atherosclerosis formation in aorta, under similar hypercholesterolemia conditions. Furthermore, we show for the first time that inoculation with M. bovis BCG via the respiratory route (which is the most common route of acquisition of mycobacterial infection in humans) exacerbated atherosclerosis and thus supports a pathogenic role of mycobacterial infection in plaque formation.
Population-based studies have indicated an increased risk of atherosclerotic cardiovascular disease in persons with a history of tuberculosis disease (3)(4)(5)(6). In addition, recent studies have shown that latent tuberculosis infection is associated with higher rates of coronary artery stenosis and spontaneous acute myocardial infarction (27,28). These data in humans suggest that the interplay between mycobacteria and cardiovascular disease can occur at different stages of mycobacterial infection. We were able Aorta plaque ratio in mice grouped as below or above CD4/CD8 ratio mean. (C) Spearman correlation of the proportion of Ly6C low monocytes and aorta plaque ratio. (D) Aorta plaque ratio in mice grouped as below or above mean proportion of Ly6C low monocytes in blood. n = 40; data are pooled from 2 independent experiments. Significance was determined by Spearman's correlation test for (A, C). Significance was determined using Student's t-test for (B, D). ***p < 0.001.
to recover viable mycobacteria from lung and spleen tissues after 16 weeks of M. bovis BCG inoculation, suggesting that our model may be more representative of conditions where there is mycobacterial persistence. This is true for active tuberculosis disease, but may also occur within the spectrum of subclinical tuberculosis and latent tuberculosis infection, as the "latent" state encompasses a wide range of host-pathogen interactions (29), some of which may result in residual bacterial replication and/or enhanced systemic immune activation (30)(31)(32).
We found that M. bovis BCG infection induced monocyte and CD4 + T cell driven systemic immune activation, which is in line with results from prior studies of host immune responses to mycobacteria (20,33). The contribution of these immune cells in atherosclerosis development has been well characterized (34), and likely provides a mechanistic link between infection and atherogenesis. Both the CD4/CD8 T cell ratio and the proportion of Ly6C low monocyte were associated with plaque burden in our study. However, a limitation of our study is that we did not conduct in-depth mechanistic experiments to assess the role of specific immune or mycobacterial parameters in atherogenesis. Future studies detailing specific tissue dissemination-including interactions of M. bovis BCG, inflammatory cells, and other stromal cells within atherosclerotic plaque-are needed. Whether targeted immune mechanisms mediate the effects of mycobacterial infection in atherosclerosis can be assessed in future studies. Of note, an increased systemic CD4/CD8 T cell ratio was recently found to be strongly and independently associated with coronary artery disease in elderly individuals (35). In human atherosclerotic plaque, there is a progressive expansion of the CD4 + T cell compartment as atherosclerotic lesions evolve (36). Although a decreased systemic CD4/CD8 T cell ratio has also been associated with increased human atherosclerosis, this phenomenon has been observed primarily in persons living with HIV/AIDS (37).
M. bovis BCG infection increased the number of circulating monocytes in our model. Monocytosis has been observed in human and murine myocardial infarction (38). Furthermore, a recent study demonstrated that monocyte recruitment from the circulation into aortic plaque is required for atherosclerosis progression (39). When monocyte subsets were analyzed, M. bovis BCG-infected mice showed a higher proportion of nonclassical Ly6C low circulating monocytes, which may be in response to increased endothelial injury, and could drive increased fibrosis in vascular tissue (40). Alternatively, our results may indicate a pro-atherogenic contribution of non-classical Ly6C low monocytes in atherogenesis. Despite Ly6C low monocytes being known to promote endothelial repair (41), a recent study indicated that these cells are involved in early plaque development (42). Furthermore, experimental inhibition of CCR5, a chemokine receptor preferentially involved in Ly6C low monocyte recruitment to atherosclerotic plaque (43), has resulted in decreased atherosclerosis formation (42,44). Of note, triggering of the NOD2 receptor has been reported to promote conversion of Ly6C high into Ly6C low monocytes with patrolling properties (45).
We did not see significant differences in NOD2 receptor expression between BCG-infected and uninfected mice; however, these results do not exclude the possibility of differential downstream NOD2 signaling. Monocytes exposed to M. bovis BCG develop a prolonged pro-inflammatory phenotype via epigenetic changes in histone methylation at the level of bone marrow progenitors, a phenomenon coined as trained immunity (46). In addition to central trained immunity, peripheral trained immunity of blood monocytes and tissue macrophages has also been described (47,48). Trained monocyte-derived macrophages have an augmented production of pro-atherogenic cytokines including IL-1b, IL-6, and tumor necrosis factor-a, and are more prone to foam cell formation upon exposure to a second non-specific stimulus (49,50). Thus, trained immunity is gaining recognition as a plausible mechanistic link between infection and atherosclerosis development (22,51). Furthermore, classical pro-atherogenic stimuli such as oxidized-LDL also induce a trained immune phenotype in monocytes (50), suggesting that both infectious and non-infectious triggers might contribute to atherosclerosis through shared disease pathways. Future studies assessing trained immunity and epigenetic reprogramming of Ly6C low and Ly6C h i gh monocytes and their effects in atherosclerotic plaque may provide insights into mechanisms of atherosclerosis in the setting of mycobacterial infection.
In conclusion, M. bovis BCG infection increased the extent of atherosclerosis development in the aortas of WD-fed hyperlipidemic Ldlr -/mice. Our results indicate that mycobacterial infection is capable of enhancing atherosclerosis development, and provide experimental evidence for previously reported links between tuberculosis and atherosclerotic cardiovascular disease in humans.

DATA AVAILABILITY STATEMENT
The primary data supporting the conclusions of this article will be made available by the authors, upon reasonable request.

ETHICS STATEMENT
The animal study was reviewed and approved by University of Cincinnati Institutional Animal Care and Use Committee (IACUC).

AUTHOR CONTRIBUTIONS
MH, JQ, CF, GD, and DH contributed to the conception and design of the study. MH, JQ, SJ, SS, AM, DK, EK, and RK conducted the investigation and experiments. MH, SJ, DK, and EK performed statistical analyses. All authors interpreted data. MH wrote the first draft of the report. All authors contributed to the article and approved the submitted version.

FUNDING
This work was supported by the National Center for Advancing Translational Science (grant number KL2 TR001426 to MH), the National Institute of Allergy and Infectious Diseases (grant number R01 AI116668 to JQ), and the National Institute of Diabetes and Digestive and Kidney Diseases (grant number R01 DK074932 to DH) at the National Institutes of Health. MH also received support from the Department of Internal Medicine at the University of Cincinnati College of Medicine. The contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health or the institutions with which the authors are affiliated. The funding source had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.

ACKNOWLEDGMENTS
All flow cytometric data were acquired using equipment maintained by the Research Flow Cytometry Core in the Division of Rheumatology at Cincinnati Children's Hospital Medical Center.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2020. 607957/full#supplementary-material SUPPLEMENTARY FIGURE 1 | Detection of mycobacteria in atherosclerotic lesions. Frozen sections of aortic roots isolated from M. bovis BCG-infected and uninfected Ldlr -/mice were histologically stained with acid fast reagents by placing in Carbol-fuchsin solution (Sigma-Aldrich, catalog No. HT801) for 1 min followed by 2 min of counterstain with Malachite Green solution (Sigma-Aldrich, catalog No. HT802). The frozen sections were also subjected to immunofluorescence staining with the anti-Mycobacterium tuberculosis antibodies Ag85B (Abcam, catalog No. ab43019) with DAPI counterstain to identify the nucleus. The scale bars represent 100 µm. Note that no mycobacteria were detectable in the atherosclerotic lesions from either M. bovis BCG infected or uninfected mice.