Nontuberculous mycobacterium M. avium infection predisposes aged mice to cardiac abnormalities and inflammation

Abstract Biological aging dynamically alters normal immune and cardiac function, favoring the production of pro‐inflammatory cytokines (IL‐1β, IL‐6, and TNF‐α) and increased instances of cardiac distress. Cardiac failure is the primary reason for hospitalization of the elderly (65+ years). The elderly are also increasingly susceptible to developing chronic bacterial infections due to aging associated immune abnormalities. Since bacterial infections compound the rates of cardiac failure in the elderly, and this phenomenon is not entirely understood, the interplay between the immune system and cardiovascular function in the elderly is of great interest. Using Mycobacterium avium, an opportunistic pathogen, we investigated the effect of mycobacteria on cardiac function in aged mice. Young (2–3 months) and old (18–20 months) C57BL/6 mice were intranasally infected with M. avium strain 104, and we compared the bacterial burden, immune status, cardiac electrical activity, pathology, and function of infected mice against uninfected age‐matched controls. Herein, we show that biological aging may predispose old mice infected with M. avium to mycobacterial dissemination into the heart tissue and this leads to cardiac dysfunction. M. avium infected old mice had significant dysrhythmia, cardiac hypertrophy, increased recruitment of CD45+ leukocytes, cardiac fibrosis, and increased expression of inflammatory genes in isolated heart tissue. This is the first study to report the effect of mycobacteria on cardiac function in an aged model. Our findings are critical to understanding how nontuberculous mycobacterium (NTM) and other mycobacterial infections contribute to cardiac dysfunction in the elderly population.


| INTRODUCTION
Biological aging dynamically transforms the molecular, cellular, and physiological paradigms of homeostasis, often biasing organisms toward dysregulation (Li et al., 2015;Pattabiraman, Palasiewicz, Galvin, & Ucker, 2017;Pomatto & Davies, 2017). In regard to immune function, chronic exposure to noxious environmental stimuli, as well as the intrinsic physiological changes associated with aging, incurs a continual decrease in immune responses that is seemingly coupled with chronic increase in production and circulation of pro-inflammatory cytokines such as IL-1β, TNF-α, and IL-6. This maladaptation has been termed inflammaging (Franceschi & Campisi, 2014), and this unresolved inflammatory perturbation progressively contributes toward chronic multi-system dysfunctions (Parkinson disease, Alzheimer, cardiovascular diseases [CVD], arthritis, and diabetes) that impact normal neurological, pulmonary, vascular, immunological metabolic and cardiac status (Xia et al., 2016).
We have previously shown that pulmonary infections with the bacterium Franscisella tularensis subspecies novicida (Ft.n) induced cardiac damage. More specifically, intranasal infection of mice with Ft.n resulted in altered cardiac electrophysiology (increased heart rate, QRS duration, and PR intervals), significant formation of cardiac micro-lesions, cardiac fibrosis, as well as immune cell infiltration, myocarditis, and bacterial colonization of cardiac tissue (Makara et al., 2016). Studies by other groups investigating bacterial infections associated with sepsis and cardiac failure (Bergounioux et al., 2016;Brown et al., 2014) have paralleled our findings. Additionally, bacterial products can activate toll-like receptor-mediated inflammatory cytokines that contribute to cardiomyocyte contractile dysfunction and subsequent cardiac damage (Boyd, Mathur, Wang, Bateman, & Walley, 2006;Fillon et al., 2006;Rolli et al., 2010). In short, there is growing consensus that the dissemination of bacteria into cardiac tissue may be a pivotal step in the onset of cardiac failure in sepsis patients. Dissecting the underlying interactions among bacterial infections, their immuno-pathological consequences and cardiac function in old age are necessary to develop effective therapeutic strategies.
In this study, we examined cardiac function of young and old mice during nontuberculous mycobacteria (NTM) infection. NTM are usually opportunistic pathogens and are ubiquitously found in nature (Bermudez, Wagner, & Sosnowska, 2000;Whiley, Keegan, Giglio, & Bentham, 2012). Individuals with chronic lung infections such as patients with HIV, COPD, cystic fibrosis, and/or medications that dampen proper immune response have increased susceptibility to contract NTM infections. Nevertheless, age is also a major prognostic factor for contraction of NTM disease and recent clinical data reflect increased incidences in elderly immunocompetent individuals (Prevots & Marras, 2015;Stout, Koh, & Yew, 2016).
Once internalized by tissue resident macrophages, M. avium can inhibit phagosome acidification, is seemingly impervious to the antimicrobial oxidative burst (ROS) and nitric oxide (NO), can persist and multiply within the phagocytic vacuole of resting macrophages, and can induce the production of immuno-suppressive cytokines such as IL-10 and TGF-β (Appelberg, 1994(Appelberg, , 2006Appelberg et al., 1994;Bermudez et al., 2000).
Herein, we have uncovered that M. avium infection causes premature atrial contraction and cardiac dysrhythmia in old mice. Intranasally infected old mice suffered from mycobacterial dissemination in the superior pericardium, increased infiltration of immune cells and cardiac fibrosis in the heart. We additionally show an increased expression of inflammatory genes and genes related to cardiac fibrosis in the heart of M. avium infected old mice. The findings from this NTM infection model are critical to further our comprehension of cardiac dysfunction in the elderly.

| M. avium disseminates into the cardiac tissue of old mice
Intranasal infection with M. avium strain 104 (1.2 × 10 5 CFU) resulted in detectable bacterial loads as early as 10 days postinfection in the lung, spleen, and liver homogenates of young and old mice. Compared to young mice, the lung CFU count in old mice was significantly lower at 10 days postinfection (p.i.) and remained significantly lower in the lungs of old mice through 30 days p.i (Figure 1a).
The mycobacterial CFU load in the spleen and liver of old and young infected mice remained comparable at the measured time points . We also performed immunoflourescence microscopy on heart tissue sectioned from infected old and young mice 35 days p.i.
Notably, M. avium infected old mice had disseminated mycobacteria in the superior region of the heart, particularly in the pericardial sac ( Figure 1d), but the bacterial load in the heart was sublethal.

| M. avium dissemination causes premature atrial contractions and cardiac hypertrophy in old mice
Since M. avium disseminated into the heart of old mice, we next To determine whether there were multiple RR intervals among treatment groups, 10 min of ECG recordings was overlaid in 30-s increments and analyzed using LabChart Pro-8 software. M. avium infected old mice were found to have three distinct RR intervals (Supporting Information Figure S2C), while M. avium infected young mice and both young and old uninfected groups have similar RR interval (Supporting Information Figure S2A-C). We also determined the heart rate (HR) from recorded ECG data and compared the experimental groups. M. avium infected old mice showed significant reduction in HR as well as increased variability (236-400 bpm; Supporting Information Figure S2D). The HR in young mice was not affected by infection (Supporting Information Figure S2B-D). The heart function of sham treated and M. avium infected young and old mice was also assessed by echocardiography (Echo) (Figure 3a,b).
Our results revealed that the heart function was normal in sham treated young and old mice ( Figure S3D

| M. avium infection aggravates systemic inflammation in old mice
Since it is well established that aging is associated with chronic systemic inflammation, we questioned how infection with M. avium influenced the inflammatory state of old mice. We analyzed the serum of M. avium infected young and old mice 30 days p.i. for levels of pro-inflammatory cytokines IL-1β and TNF-α. Consistent F I G U R E 1 Mycobacterium avium bacterial burden in lungs of young and old mice. C57BLC57BL/66 mice intranasally infected with M. avium strain 104 (1.2 × 10 5 CFU) were euthanized at 10, 20, and 30 days postinfection, and lungs were homogenized in sterile saline buffer. Serial lung (a), spleen (b), and liver (c) tissue dilutions were plated onto OADC supplemented 7H11 plates and incubated for 21 days at 37°C. CFU were counted and expressed as Log 10 CFU; 4-5 mice per group per time point (mean ± SEM; *p < 0.05). The images shown in (d) are hearts of M. avium infected young and old mice, stained with anti-Mycobacterium tuberculosis polyclonal antibody followed by anti-rabbit Alexa flour 568 conjugated secondary antibody. The section was examined for M. avium by confocal microscopy. Images shown are representative of hearts from five young and old animals. (e) Randomly selected confocal images (5 images per heart, n = 5) from M. avium infected young and old mice hearts were analyzed by ImageJ, and the mean fluorescent intensities were plotted in the graph (mean ± SEM; *p < 0.05) with previous studies regarding chronic systemic inflammation in mice (Starr, Saito, Evers, & Saito, 2015), the serum of old uninfected mice contained higher levels of IL-1β and TNF-α compared to uninfected young mice (Supporting Information Figure S1A,B). M. avium infected old mice had significantly higher levels of IL-1β and TNF-α compared to M. avium infected young mice (Supporting Information Figure S1A,B). The levels of IL-1β and TNF-α in M. avium infected old mice were roughly 2.5-to 3-fold higher than in the age-matched uninfected controls. Collectively, these results suggest that pulmonary NTM infection further aggravates systemic inflammation in old mice.

| Enhanced immune cell infiltration in the heart of M. avium infected old mice
Mycobacteria disseminated into the heart of old mice, and these mice also showed abnormal ECGs and Echo. We questioned whether there was any leukocyte infiltration of into the cardiac muscle of infected mice. We stained heart sections (four chamber view) from

| Dissemination of M. avium triggers the inflammatory gene expression in the hearts of infected mice
To determine whether cardiac infiltration of immune cells causes car-

| Validation of mRNA expression patterns by qRT-PCR
To further validate the gene expression profiling results that we obtained from NanoString technology, we used qRT-PCR to assay expression of selected upregulated genes (TNF-α and IL-1β). The gene expression profiles of TNF-α and IL-1β obtained by NanoString immunoassay were in agreement with the gene profiles obtained by qRT-PCR (Figure 6c).

| M. avium dissemination in old mice upregulates canonical pathways related to immune cell recruitment, inflammation, fibrosis, and cell death
Using Ingenuity Pathway Analysis (IPA), we examined the relationship between these differentially expressed and highly significant genes (230 genes) to determine the most significant canonical pathways and biological networks involved in M. avium infected old mice hearts ( Table 1). The top enriched categories of canonical pathways (p < 0.05) were associated with granulocyte adhesion, macrophages, fibroblasts, and endothelial cell activation PRRs in recognition of bacteria and viruses, and cardiac hypertrophic signaling, apoptosis and death receptor signaling, and TGFβ signaling end of sentence.
F I G U R E 4 NTM infection induces cardiac fibrosis in old mice. Heart sections from sham and Mycobacterium avium infected young and old mice were stained with Masson's trichrome to identify fibrosis in cardiac tissue. Image shown here are representative of whole heart and interstitial fibrosis from sham-treated young (a) and old (c) mice that were infected with M. avium (b and d). The image shown in right panel is 40× magnification and representative of five animals/group. The trichrome staining (blue color) in the heart sections was isolated using Photoshop CC in color range selection mode. The total intensity of the stained area in the heart sections was further quantified via ImageJ using color intensity to multiply area with staining. Graph shown in (e) young mice sham-treated and M. avium infected, (f) old mice shamtreated M. avium infected, and (g) comparison of M. avium infected young and old mice. Data shown in graphs are cumulative data from five animals (mean ± SEM; ** p < 0.005; ***p < 0.0005; N = 5)

| Disease and function analysis and top upstream regulators
In addition to canonical pathways, the differentially expressed genes in M. avium infected old mice hearts were also categorized to related disease and functions, and upstream regulators (Table 2). We found that cellular movement, development, cellular growth and proliferation, cell-to-cell signaling and interaction and cellular function and maintenance were considerably activated with increased number of interacting molecules in M. avium infected old mice. Notably, young mice infected with M. avium showed only few molecules that were activated in the above listed functional pathways (Supporting Information Table S5). The IPA upstream functional analysis was used to predict the top upstream transcriptional regulators from differentially

| Analysis for cardiotoxicity
To further identify the key pathways that are involved in the cardiac dysfunction, we examined the cardiotoxicity functions of F I G U R E 5 Changes of immune-related gene expression in the hearts of Mycobacterium avium infected young and old mice. Hearts were collected from sham-treated or M. avium infected (30 days postinfection) young and old mice to isolate mRNA. The mRNA samples were analyzed using NanoString assay on the mouse pan-cancer immune panel, followed by data analysis in nSolver software. The heat map showing genes whose induction in heart tissue by M. avium infection was more than 1.5-fold different in young and old mice (a), p < 0.05. The vein diagrams shown in (b-e) is the number of genes that were upregulated, downregulated, or no change in expression in young and old mice that were infected or uninfected with M. avium. Two-tailed Student's t test is used to select differentially expressed genes with values p < 0.05 differentially expressed gene sets. IPA analysis data reveal that gene network pathways linking to cardiac toxicity such as cardiac infarction, cardiac necrosis/cell death, cardiac fibrosis, inflammation, dysfunction, and enlargement are activated in M. avium infected old mice. In contrast, young mice infected M. avium showed very poor linkage with pathways connected cardiotoxicity (Table 2).

| DISCUSSION
Bacterial infections and bacteremia significantly escalate the risk of cardiac failure in the elderly (Blecker et al., 2013;Court et al., 2002;Parrillo et al., 1990;Rudiger & Singer, 2007), and development of effective therapeutic interventions relies on comprehensive understanding of host-pathogen interactions. With the exception of epidemiological and case studies (Cordioli et al., 2015;Prevots & Marras, 2015), basic scientific information regarding cardiac dysfunction during mycobacterial infections is sparse. We are the first to show that the opportunistic pathogen M. avium, which is more closely associated with pulmonary and gastrointestinal colonization (Appelberg, 2006;Bermudez et al., 2000;Reddy, 1998), incurs significant cardiac damage and dysfunction in infected old mice.
The strain of M. avium (strain 104) used in this study was originally isolated from an AIDS patient, and its virulence and immunopathology have been previously characterized (Saunders, Dane, Briscoe, & Britton, 2002;Torrelles et al., 2002). As little as 100 CFU (via aerosol delivery) or a many as 5 × 10 7 CFU (intraperitoneal injections) have been previously used to model chronic infection and immune responses (Saunders et al., 2002;Torrelles et al., 2002). The mild dose (1.2 × 10 5 CFU) used in our experiments resulted in detectable CFU in surveyed organs (lung, liver, and spleen), with no apparent differences in pathogen-specific adaptive immune responses (CD4 + T cell count, levels of INFɣ in lung homogenate; data not shown) in M. avium infected young and old mice.
We also found that bacterial burden in the lung is decreased in old mice, while no difference in spleen and liver which may be due to the difference in circulating antimycobacterial antibodies in young and old mice. Since it has been shown earlier that containment of M. avium infection was due to increased levels of circulatory Cardiac hypertrophy is complex and influenced by genetic, physiological, and environmental factors involving particular transcriptional factors and contractile proteins (Hunter & Chien, 1999). Although cardiac hypertrophy is the heart's primary response to stress and an adaptive mechanism, prolonged stimulation of hypertrophy has adverse consequences that are linked to heart failure and sudden death (Braunwald & Bristow, 2000). Pleiotropic pro-inflammatory cytokines like IL-1β and TNF-α aggravate endothelial cells, cardiomyocytes, fibroblast, and leukocytes into activated states. The activated endothelium enhances recruitment of circulating leukocytes and impacts cardiac fibroblast activity favoring the production of collagen (Biernacka & Frangogiannis, 2011;Kong et al., 2014;Suthahar et al., 2017). Excess collagen impacts the ability of cardio myocytes to propagate the cardiac action potential generated by the sinoatrial node, ultimately leading to abnormal diastolic function, and cardiac failure (Nguyen, Kiriazis, Gao, & Du, 2017;Nguyen, Qu, & Weiss, 2014). Cardiac hypertrophy is typically associated with diastolic T A B L E 1 Top significantly enriched canonical pathways in Mycobacterium avium infected old animal hearts by IPA
In summary, we show that increased inflammation

| Echocardiography
To assess cardiac function in vivo, 2D-echocardiography (Vevo 2100, Visualsonics) was performed in M. avium infected young and old mice at 30-day postinfection. Mice were anesthetized in an induction chamber at 2% isoflurane in oxygen at a flow rate of 1.0 L/min.
Mice were then placed in supine position on a heated stage and hair was removed from the chest using depilatory lotion. Anesthesia was maintained at 1.5% isoflurane for the duration of the experiment.
Heart rate was monitored throughout to ensure proper anesthetic dosage. Using a MS-400 transducer, proper anatomical orientation was determined via imaging of the long axis of the heart. Once proper orientation was achieved, the transducer was turned 90 degrees to visualize the short axis of the left ventricle. M-mode images were recorded at the level of the papillary muscles. Images were analyzed to assess ejection fraction, fractional shortening, chamber diameters, and left ventricular wall thicknesses.

| Pathological methods and immunohistochemistry
At day 35 postinfection, the hearts of young and old M. avium infected mice and age-matched uninfected controls were isolated and washed in cold PBS and then fixed in 10% formalin overnight at 4°C.
The hearts were then switched to 20% sucrose solution. The fixed hearts were embedded in paraffin and sectioned for further analysis.
To examine leukocyte infiltration, the four chamber view heart sections from M. avium infected and uninfected controls were stained with αCD45 antibody (Abcam, clone EP322Y). The heart tissue sections were deparaffinized and antigen epitopes were retrieved via heat-induced antigen retrieval prior to incubation with αCD45 and biotinylated α-rat antibody for 60 and 30 min, respectively. The numbers of αCD45+ cells in the heart sections of M. avium infected and uninfected controls (5 heart sections per group) were quantified using Aperio Image Scope software and converted to number of cells per µm 2 . To examine bacterial dissemination in the heart, heart sections were incubated with anti-Mycobacterium tuberculosis polyclonal antibody (ab905; Abcam, Cambridge, MA, USA) followed by anti-rabbit Alexa fluor 568 secondary antibody. The nuclei were stained with DAPI. To examine tissue fibrosis, heart sections from M. avium infected young and old mice and uninfected controls were stained with Masson's trichrome stain (Schipke et al., 2017). Trichrome stained heart sections were imaged, and positive staining (indicated by blue color) was quantified using Photoshop CC color range selection mode and ImageJ software. Trichrome color intensity was multiplied by measured areas to determine the mean fluorescent intensity (MFI) of heart sections (5 heart sections per group).

| Cytokine assay
Blood samples from M. avium infected young and old mice and uninfected controls were collected immediately after euthanization. After a 1 hr incubation at 4°C, the blood samples were centrifuged for 10 min at 5,000 g.

| NanoString nCounter assay
To investigate the gene expression in the hearts of M. avium infected young and old mice and age-matched uninfected controls, we performed the NanoString nCounter assay (NanoString Technologies, Seattle, WA) using the mouse pan-cancer immunology panel. All procedures were performed according to the manufacturer's protocol (NanoString Technologies). Briefly, 200 ng of high-quality heart tissue RNA (3 mice per group) was hybridized to NanoString probes by HEADLEY ET AL. Calculations were performed using the R statistical computing environment. Bioinformatic analysis was performed to analyze differentially expressed genes in M. avium infected young and old mice.
Briefly, outcomes form nanoString data analysis were first uploaded into Qiagens' IPA system for core analysis and then with the global molecular network in the Ingenuity Pathway Knowledge Base (IPKB).
IPA was performed to identify canonical pathways, disease and functions, and gene network that is related to cardiovascular diseases.

| NanoString data validation
The 100 ng of total heart RNA was reverse transcribed to cDNA by reverse transcriptase enzyme (SuperScript III; Invitrogen), and qRT-PCR was performed using mouse IL-1β or TNF-α TaqMan gene expression kit (Applied Biosystems). IL-1β and TNF amplification was normalized to β actin as a housekeeping gene for relative gene expression. Triplicate samples were analyzed in duplicate wells in each experiment. (n = 3).

| Statistical analysis
Statistical analysis was carried our using student t test or 2-way ANOVA with the Tukey post hoc test, when appropriate. The results are shown as the Mean ± SEM values of p < 0.05 were considered to be significant.