Augmenting muscle MHC expression enhances systemic pathogen control at the expense of T cell exhaustion

Abstract

Myocytes express low levels of MHC I, perhaps influencing the ability of CD8⁺ T cells to efficiently detect and destroy pathogens that invade muscle. Trypanosoma cruzi infects many cell types but preferentially persists in muscle and we asked if this tissue-dependent persistence was linked to MHC expression. Inducible enhancement of skeletal muscle MHC I in mice during the first 20 days of T. cruzi infection resulted in enhanced CD8-dependent reduction of parasite load. However, continued overexpression of MHC I beyond 30 days ultimately led to a collapse of systemic parasite control associated with immune exhaustion, which was reversible in part by blocking PD-1:PD-L1 interactions. These studies demonstrate a surprisingly strong and systemically dominant effect of skeletal muscle MHC expression on maintaining T cell function and pathogen control and argue that the normally low MHC I expression in skeletal muscle is host-protective by allowing for pathogen control while preventing immune exhaustion.

Introduction

The protozoan Trypanosoma cruzi is the highest impact parasitic disease agent in the Americas (1, 2) and the cause of Chagas Disease, a syndrome typified by muscle damage of the heart, colon, and esophagus. Although, T. cruzi demonstrates impressive promiscuity in cell range for invasion and can replicate within nearly all nucleated cells, parasites in chronically infected hosts are most frequently detected in muscle tissue (3–5). This study seeks to determine how T. cruzi is largely eliminated from other tissues but avoids elimination from muscle for the lifetime of most infected hosts. MHC class I expression on the surface of mammalian cells is essential for efficient immune control of many intracellular infections (6–8). Under basal conditions MHCI is nearly undetectable in muscle (9, 10). And although T. cruzi proteins are available for processing and presentation on host surface MHC I-peptide complexes (11–13) and are targeted by a robust CD8⁺ T cell response that is essential for host survival (14–16), T. cruzi nonetheless manages to evade elimination from nearly all hosts, and the persistent infection ultimately compromises muscle integrity (reviewed in 17).

Here we describe a model for modulating class I MHC expression specifically in skeletal muscle and document an initially productive but ultimately highly damaging impact of this modulation during the course of T. cruzi infection.

Materials and Methods

Mice and pathogens

Mice capable of inducible MHC I overexpression in skeletal muscle myocytes were generated by the crossing of HSA-rtTA/TRE-Cre (The Jackson Laboratory)(18) and mice carrying the tetracycline-response element (TRE-H2K^b) transgene, a kind gift from Dr. Kanneboyina Nagaraju (Children’s National Medicine Center, Washington, D.C.) (19). Mice positive for both transgene constructs (H2K^b-rtTA) were bred and maintained under specific pathogen-free conditions at the Coverdell Vivarium (University of Georgia, Athens, GA) per the animal welfare guidelines outlined by the University of Georgia Institutional Animal Care and Use Committee (UGA IACUC).

For T. cruzi infections, mice were infected via intraperitoneal (i.p.) injection with 10³ Brazil wild-type or Brazil ova expressing (T. cruzi-ova), or 5×10⁴ Colombiana strain luciferase expressing trypomastigotes (20). For Listeria monocytogenes infections, mice were infected by oral gavage with 10⁴ Listeria-ova bacteria.

Lymphocyte isolation and T cell analysis

Lymphocytes were isolated from muscle as previously described (21) and CD8⁺ T cell specificity was determined by staining with MHC I tetramer TSKB20 (ANYKFTLV/K^b) synthesized at the Tetramer Core Facility (Emory University, Atlanta, GA). Lymphocytes were stimulated with 1 μM T. cruzi peptide TSKB20 (ANYKFTLV) (GenScript) or 1.5 μg of plate-bound anti-mouse CD3ε (eBioscience) for 5h at 37°C in the presence of 1 μg/ml Golgi Plug (BD Pharmingen). To evaluate in situ cytokine production, a direct intracellular staining (dICS) protocol was applied (22).

Antibody treatment

200 μg of anti-CD8a (Clone: YTS 169.4) was given via i.p. injection every third day for 10 days and anti-PD-L1 (Clone: MIH5 and 10F.9G2) was administered every third day for 20 days.

Quantitation of parasite burden

Parasite equivalents in tissue were determined by qPCR as previously described (23) and systemic parasite levels were quantified by detection of luminescent signal following injection of 0.2 mg of D-luciferin (Gold Bio)(20).

Determination of MHC class I expression

Skeletal muscle was enzymatically digested prior to surface staining of tissue homogenate with H2K^b antibody (BD Biosciences)(24). For western blot analysis, muscle was dissociated using an electric tissue homogenizer in a 50 mM Tris/10 mM EDTA solution and then mixed with a 0.125M Tris/4% SDS/20% glycerol/10% MCE treatment buffer. Specific antibody staining for α-tubulin (Sigma-Aldrich) and H2K^b was measured using a GE Typhoon 7000 and band intensity calculated using Image J software.

Statistical analysis

We calculated statistical significance with a Student’s two-tailed t-test or by one-way ANOVA with Tukey post-test analysis. * indicates values (mean+ SEM) that are significantly different between specified groups (* P≤ 0.05, ** P≤ 0.01, ***P≤ .001).

Results

Inducible upregulation of MHC class I on skeletal muscle results in improved infection control

We generated a transgenic “tet-on” mouse model where MHC I expression on skeletal muscle myocytes is transiently increased by the administration of the tetracycline-analog doxycycline (Supplemental Fig. 1A) (18, 19). Administration of doxycycline results in a nearly two-fold increase in H2K^b expression (Supplemental Fig. 1B,C). Administration of doxycycline beginning on the day of T. cruzi infection resulted in >5-fold reduction in the muscle parasite load by day 20 of infection (qPCR) (Fig. 1A) and this impact was also evident systemically (Fig. 1B) and by histological examination (Fig. 1C,D,E). Depletion of CD8⁺ T cells reversed the impact of MHC overexpression on parasite control (Fig. 1F). MHC I overexpression also resulted in a selective increase in the number and activation level of T. cruzi-specific CD8⁺ T cells (Fig. 1G). These data indicate that increased MHC expression facilitates improved CD8⁺ T cell surveillance of and response to T. cruzi-infected cells, leading to enhanced parasite control.

Fig. 1. Increased antigen availability during early acute infection results in improved control and influences CD8⁺ T cell in vivo IFNγ production in muscle.

(A) Dox-induction of skeletal muscle MHC expression during 19 days of infection results in a significant decrease in parasite load (qPCR) in skeletal muscle, (B) systemically (whole-body luciferase imaging) and (C) histologically by the number of infected cells per 100 fields of skeletal muscle sections. (D, E) Histological sections showed a decrease in the number of heavily infected cells with MHC I overexpression at approximately 20 days post-infection. (F) Transgene induction was initiated upon infection and continued throughout the course of the experiment, and anti-CD8 monoclonal antibodies were administered every 3d between days 12 to 21 post infection. Depletion of CD8⁺ T cells reverses improvements in the parasite control observed during MHC I overexpression and results in significant increases in parasite number via qPCR (G). Mice were infected with 10⁴ Listeria monocytogenes bacteria expressing OVA to generate a SIINFEKL-specific CD8⁺ T cell population 17 days prior to infection with T. cruzi and the start of doxycycline administration for MHC overexpression. The number of T. cruzi-specific (TSKB20⁺) CD8⁺ T cells present and producing IFNγ in situ in skeletal muscle increased significantly in mice with transgene induction, while the frequency of OVA-specific population remained unchanged at 20 days post T. cruzi infection. Data are representative of at least 4 experiments with n=3–5 with the exception of panels B and F&G, where data are representative of one experiment with n=4–6, and G is representative of two experiments with n=3–4. Data are mean + SEM. * indicates percentage levels that are significantly different (* P ≤ 0.05) between specified groups.

Prolonged MHC overexpression results in compromised parasite control

Continued induction of class I MHC overexpression was expected to yield a further reduction of parasite numbers. However, mice induced for 30 days or more showed a massively elevated parasite load (Fig. 2A,B) and a dramatic increase in inflammation in muscle, ultimately requiring euthanasia (Fig. 2C). Higher parasite levels were also observed in the heart (Fig. 2D) and systemically (Fig. 2E). This lost control was also associated with a decrease in the frequency of IFNγ producing CD8⁺ and TSKB20-specific T cells in muscle tissue (Fig. 2F) and reduced potential of T cells systemically to respond to ex vivo stimulation (Fig. 2G,H).

Fig. 2. Prolonged overexpression results in compromised CD8⁺ T cell function and parasite control.

At 30 days post-infection and the start of transgene induction, the impact of increased antigen availability was assessed. (A) Parasite burden in the skeletal muscle was significantly increased in mice with ≥30 days of MHC I overexpression when assessed by qPCR and analysis of histology slides (B). (C) Prolonged (≥30 days) MHC overexpression results in severe inflammation and increased parasite levels in skeletal muscle. Black arrows point to T. cruzi infected cells. ≥30 days of skeletal muscle MHC I overexpression compromises T. cruzi control in the hearts (D) and systemically (E). Parasite-specific CD8⁺ T cell IFNγ production in situ in muscle is compromised following prolonged MHC I overexpression (F). (G) Response to ex vivo αCD3 stimulation is compromised in the CD8⁺ T cells from skeletal muscle of mice with prolonged MHC overexpression. (H) Response to ex vivo parasite specific-peptide stimulation (TSKB20) is compromised in splenic CD8⁺ T cells in mice with prolonged antigen exposure. Data are representative of 2–3 experiments with n=3–5, with the exception of E and bars are mean + SEM. * indicates averages that are significantly different (* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001) between specified groups.

Immune regulation via PD-1 contributes to compromised parasite control

The rapid collapse in parasite control associated with reduced T cell function suggested that increased antigen presentation due to prolonged upregulation of MHC class I in skeletal muscle might be exhausting T. cruzi–specific T cells. PD-1 and LAG3 are key regulators of T cell function, particularly in systems characterized by high antigen load (42–44) and CD8⁺ T cells expressing these inhibitory receptors were evident in the blood with continued MHC overexpression post-infection/induction (Fig. 3A) as well as in skeletal muscle (Fig. 3B,C). Blocking PD-1:PD-L1 interactions using anti-PD-L1 during days 15–35 of MHC overexpression (Fig. 3D) resulted in a restoration of T. cruzi-specific CD8 T cells frequency (Fig. 3E) although not full effector function (Fig. 3F). More importantly, PD-L1 blockade reestablished control of tissue parasite numbers and substantially reduced tissue inflammation in the muscle (Fig. 3G,H).

Fig. 3. Regulation via PD-1 contributes to lost parasite control during prolonged MHC I overexpression.

MHC I overexpression was initiated at the time of infection and was continued throughout the course of the experiment. (A) An increased frequency of CD8⁺ T cells in peripheral blood express the inhibitory receptors PD-1 (left) and LAG-3 (right) during extended MHC I overexpression. (B) PD-1 expression by muscle-derived CD8⁺ T cells is elevated after 27 days of MHC overexpression. (C)The frequency of TSKB20+ cells expressing LAG3 in skeletal muscle was also elevated at day 27 of infection with MHC induction. (D) Beginning on the 15^th day of T. cruzi infection and MHC I overexpression (in appropriate groups), αPD-L1 antibody was administered every 72 hours for the next 20 days. (E) PD-L1 blockade during MHC overexpression increased the frequency of T. cruzi-specific CD8⁺ T cells in spleen and skeletal muscle at 35 days post infection. (F) Cytokine production (IFNγ and TNFα) by muscle-derived CD8⁺ T cells following TSKB20 peptide stimulation is not restored by αPD-L1 treatment. (G) The presence of muscle inflammation and infection intensity is significantly reduced following αPD-L1 treatment in H2Kb-rtTA mice with MHC I overexpression at 35 days post infection. (H) Control of muscle T. cruzi infection (qPCR) is restored upon PD-L1 blockade in mice with increased MHC I expression on day 35 post infection. Data are representative of 2 experiments with n≥4 and bars are mean + SEM. * indicates averages that are significantly different (* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001) between specified groups.

Attempts to enhance pathogen control without exhausting the T. cruzi-specific T cell compartment, by limiting the degree of MHC I overexpression or combining doxycycline-induced MHC upregulation with anti-PD-L1 blockade during chronic infection, failed to generate overall better control of the infection (Supplemental Fig. 2). Lower doxycycline doses resulted in the expected reduced levels of PD-1-expressing, T. cruzi –specific CD8⁺ T cells and mice treated at 0.5 mg/ml doxycycline controlled parasite load early in the infection and maintained that control after 26 days of treatment—despite above normal inhibitory receptor expression (Supplemental Fig. 2A,B,C). However, this and lower levels of MHC induction still resulted in much reduced survival compared to uninduced mice and comparable tissue parasite load (Supplemental Fig. 2D,E). Increased MHC expression was also not productive when delayed until day 15 or >150 days post-infection (Supplemental Fig. 2F). Combining low-level MHC induction with anti-PD-L1 blockade reduced the degree of T cell reactivation in chronically infected mice but these animals succumbed to infection earlier than mice not receiving anti-PD-L1 antibodies.

Discussion

Pathogens have evolved many mechanisms to resist elimination from their hosts, with the most successful pathogens, like T. cruzi, able to persist for decades. T. cruzi predominantly persists in tissues with highly regulated MHC expression (e.g. muscle, nervous system; (3, 5, 25), despite the presence of highly functional CD8⁺ T cells in those same tissues (21). Additionally, we have failed to detect the involvement of other immune modulating mechanism, including the regulatory cytokines IL-10 and TGF-β (26), T regulatory cells (Tregs) (27), and most recently T cell exhaustion in general, in limiting CD8 function and impacting parasite levels during T. cruzi infection (21).

Here we show that increasing muscle MHC I expression substantially boosts control of T. cruzi in a dose-dependent manner and in association with increased CD8⁺ T cell effector function. However, this improved parasite control is short-lived as continued MHC I over-expression rapidly exhausts the parasite-specific CD8⁺ T cells, resulting in a massive increase in parasite load systemically. These results confirm the hypothesis that the normally low MHC expression in muscle contributes to the propensity of T. cruzi to persist selectively in tissues with low MHC I expression as well as the suggestion that the relative absence of T cell exhaustion in T. cruzi infection is related to low antigen load/presentation (17). These results also reveal the potential negative impact that checkpoint inhibitors and perhaps other immune enhancers have if employed as therapeutics in T. cruzi infection.

This study demonstrates the tight linkage between antigen presentation and maintenance of infection control and the loss of that control. The preponderance of parasites within skeletal muscle (versus in other tissues) as well as the relative abundance of skeletal muscle in the body – humans are approximately 40% muscle – may account for the systemic nature of and rapid change in effector T cell functionality during the period of MHC I upregulation.

The ability of high antigen levels to compromise CD8⁺ T cell function has been substantiated in a wide range of infections, but best studied via viral infections with strains that achieve different levels of viral load or producing epitopes that vary in abundance (28–30). To our knowledge this is the first demonstration that enhancing epitope presentation via modulation of MHC I expression can drive T cell exhaustion. In this system the induction of T cell regulation is rapid, tunable by using different levels of doxycycline, reversible through the administration of a classical immune checkpoint molecule (PD-1), and impacts pathogen control beyond the tissue with increased class I expression and thus may be useful to explore further the general impact of MHC class I expression on CD8⁺ T cell function in non-lymphoid tissues in general.

Key Points.

MHC class I can be induced in skeletal muscle in a doxycycline dose-dependent manner

Elevated muscle MHC I results in initial enhanced immune control of Trypanosoma cruzi

Sustained elevation in muscle MHC I ultimately exhausts pathogen-specific responses