The Importance of Persistence and Dormancy in Trypanosoma cruzi infection and Chagas disease

Abstract

Trypanosoma cruzi typically establishes a life-long infection in its mammalian hosts, causing the destruction of muscle tissues and ultimately resulting in potentially fatal Chagas disease. In this review, we consider the array of avoidance mechanisms that allows for T. cruzi persistence, many of which are unconventional among protozoan pathogens but which collectively are highly effective in the face of otherwise potent host immune responses. We also reflect on the phenomenon of dormancy in T. cruzi amastigotes, which is likely not involved in the long-term persistence of infection. Lastly, we consider how these phenomena of persistence and dormancy complicate the effectiveness of potential therapeutic interventions to prevent Chagas disease.

Introduction

T. cruzi is a kinetoplastid protozoan and causative agent of Chagas disease. Despite the restriction of the insect-vectored T. cruzi transmission cycle solely to the Americas, Chagas disease is among the top causes of infection-induced cardiomyopathy globally. T. cruzi naturally infects a very wide range of mostly mammalian species and importantly, infections in these various natural hosts involve the same parasite populations and mimic the infection control and disease process in humans [1]. Thus, infections in these natural host species are not simply models of human infection, they faithfully represent the human infection and disease.

T. cruzi infection results in primarily muscle-focused pathology because 1) T. cruzi maintains a continuous cycle of infection and multiplication primarily in muscle tissue, and 2) the infection is generally life-long. The biological basis of these attributes – i.e. why the infection is controlled but not eliminated and why it is relatively restricted to muscle, are not fully understood. However, they are likely connected; T. cruzi is able to persist despite the generation of highly effective immune responses in part because of its ability to invade and replicate in muscle tissues. Here we refer to this ability to maintain a protracted, albeit frequently very low-level, but constantly active infection cycle in most hosts as persistence.

Below we discuss the biology of the T. cruzi: host interface that allows for persistence. But first it is important to discuss the factors that do not contribute to T. cruzi persistence. Multiple pathogens actively suppress immune responses and multiple mechanisms for immune regulation in T. cruzi infection have been proposed. However, there is little direct in vivo evidence for attenuated immune responses actively preventing the control of T. cruzi infection and abundant evidence of robust and highly effective anti-T. cruzi adaptive immune responses as discussed below. Furthermore, blocking potential immune regulatory pathways fails to further enhance parasite control [2,3]. Although immune exhaustion is evident in chronic T. cruzi infection, this appears to be the result of the long-standing infection, rather than its cause [4,5]. Thus, the persistence of T. cruzi is not caused by an active inhibition of host immunity.

Persistence of T. cruzi is also not the result of a parasite life-cycle option, such as the development of an arrested or encysted stage as in some other parasites (reviewed in [6]). Indeed the bulk of parasites in a persistently infected host continuously cycle in and out of host cells and are thus always targets for immune clearance. However occasionally an amastigote in host cells assumes a metabolically inactive state, a process we have termed dormancy [7]. This dormancy state was only recently identified and in part due to its infrequency, relatively little is understood about the causes or mechanisms. Unlike dormant states described in other pathogens, dormancy in T. cruzi is neither a programmed stage in the lifecycle nor a response to stressors, such as drugs. Most importantly for the discussion here, dormancy is not involved in T. cruzi persistence and so we prefer the designation of “dormants” to describe these temporarily inactive amastigotes, to the term “persisters” to avoid any confusion. Dormant amastigotes appear to become relevant to parasite persistence only when drug treatment is involved.

Here we review the data supporting the role of parasite biology in persistence of T. cruzi in highly immune competent hosts and the challenges that this biology may present with respect to attempts to tilt this host:parasite interaction toward complete parasite clearance – with the ultimate goal of preventing disease development. We then discuss the limited data on dormancy and possible future directions on this topic.

Immune control of T. cruzi infection

In vertebrate hosts, T. cruzi cycles between non-replicating extracellular trypomastigotes and intracellular amastigotes. Although T. cruzi amastigotes within host cells are sometimes referred to as being part of a “pseudocyst”, they are in fact actively replicating, expanding to ~500 progeny over 4-6 days before destroying the host cell upon their exit. Immune control of T. cruzi infection is primarily mediated by anti-parasite antibodies, that efficiently kill the extracellular trypomastigotes, and both CD4⁺ and CD8⁺ T cells that likely contribute to parasite control at multiple levels (reviewed in [8]). Mice deficient in any one of these adaptive effector mechanisms/cell types rapidly succumb to infection. However infection in the absence of either class I or class II MHC-restricted T cell population results in a much more rapid rise in parasite numbers and host death relative to the absence of antibody responses, emphasizing the importance of immune control at the level of the infected host cell. The absence of IFN-γ yields a similar outcome as total T cell depletion [9], indicating the dominant role of this cytokine in infection control, although the specific mechanism of this IFN-γ-dependent control is not clear.

The location of replicating amastigotes of T. cruzi in the host cell cytoplasm makes any released parasite proteins readily accessible to presentation by class I MHC (expressed to some degree on all nucleated host cells) and thus would explain why CD8⁺ T cells are crucial to parasite control. Although IFN-γ -producing CD4⁺ T cells have been shown to play an important role in control within the initial infection site (where the majority of infected cells are monocyte/macrophages [10]), CD8⁺ T cells appear responsible for the ongoing maintenance of parasite control within sites of persistence including skeletal muscle, fat, and gut tissues [11,12].

Established immunity in chronically infected hosts is highly efficacious, as attested to by the difficulty of detecting parasites in blood or tissues, even using the most sensitive methods. Our recent studies in non-human primates with naturally acquired T. cruzi infections indicate that detection of parasite DNA in blood is relatively stable over more than 1 year but differs by over 5 orders of magnitude between individual animals [13]. The ability of most hosts to restrict parasite load to very low levels is also demonstrated in studies in humans showing that repeat sampling and analysis by PCR [14] or xenodiagnosis [15], is required to demonstrate active infection in many individuals. And the rapid rise in parasite numbers in hosts in which immunity becomes impaired by immunosuppression (reviewed in [16]) further supports the highly active nature of the infection and the role of robust host immunity in its control. Thus, in order to sustain a continuous infection cycle within the host, T. cruzi must have equally effective immune evasion strategies. Here, we propose a set of mechanisms that T. cruzi utilizes to persist despite highly effective host immune responses.

Mechanisms of parasite persistence

-Limited activation of innate immune sensing

Host cell invasion by T. cruzi parasites is a remarkably silent process involving very weak triggering of innate immune recognition. Although several T. cruzi-derived pathogen-associated molecular patterns (PAMPs) have been described (reviewed in [8]), the infective trypomastigotes appear to lack expression of strong PAMPs as reflected in the very modest changes in host gene expression detected during the invasion process [17,18]. The bulk of the gene expression changes appear to be in response to low-level type I IFN production, only detectable at >24 hrs following host cell invasion [18], a time at which amastigotes are replicating in the host cell cytoplasm. This type I IFN response is dependent on the cytosolic DNA sensing cGAS/STING pathway, although the specific stimulus has not been determined and activation of this pathway has no effect on infection outcome [19–22]. However the potential enhancement of infection detection through the overexpression of bacterial PAMPs in T. cruzi increased parasite clearance, further suggesting that the absence of strong PAMPs in T. cruzi promotes parasite persistence [23]. Thus, the lack of strong innate activation is a clear advantage to T. cruzi, allowing the successful establishment of an infectious focus without prompting a host inflammatory response.

-Delayed generation and effector functions of T cell immunity

A major consequence of the feeble host cell response to parasite invasion is a delay in both the initial generation of and the acquisition of effector functions by parasite-specific T cells. Potential antigen-presenting cells (APCs) in the infection site remain unactivated for ~4 days post-infection as amastigotes replicate in host cells [24,25]. The exit of parasites from the (now dead) host cells at day 4-6 appears to be the first substantial stimulus for immune activation [24,26], likely driven by the release of host damage-associated molecular patterns (DAMPs) and possibly aided by parasite PAMPs. Priming of anti-T. cruzi T cells is evident in the draining lymph nodes by 8-9 dpi [24,26] followed by T cell migration to the infection site and the first hints of limits on parasite expansion there. However, by this point, the infecting parasites have had 8-10 days -long enough for two rounds of ~500-fold expansion each – and the opportunity to disseminate throughout the host.

The relatively stealth host cell invasion process by T. cruzi allows not only establishment and dissemination from the initial infection site, but also has a major role in infection persistence in hosts with established immunity. Previously primed T. cruzi-specific effector T cells within chronically infected mice also exhibit a sluggish response to even high numbers of parasites at a re-infection site, arriving at ~5 dpi, coincident with completion of a round of amplification and parasite release [10]. Thus, the lack of vigorous sensing of parasite invasion by host cells delays T cell infiltration and thus parasite control at each new infection site (in uninflamed regions of tissues), providing the means for maintaining a persistent infection by continuously establishing new infection sites within the host (Figure 1).

Figure 1. T. cruzi persistence and tissue damage in muscle.

T. cruzi invades, replicates and releases parasites in a non-inflamed site (1). The released parasites infecting other host cells in the same region are subject to T cells and other effectors recruited by the released DAMPs and PAMPs (2) and that inflammatory response eventually result in parasite clearance from this site (3). However, parasites that traffic to other non-inflamed tissue sites are able to repeat the process of relatively unrestrained initial replication at the new site (bottom panel). Created with BioRender.com

-Immunodominant CD8⁺ T cell targeting of highly variant and late-stage expressed epitopes

CD8⁺ T cells are important effectors in control of T. cruzi yet are dominantly focused on suboptimal targets. Antigens either secreted or released by T. cruzi parasites have been shown to be targeted by CD8⁺ T cell immunity, including surface proteins encoded by large, highly variable and constantly changing families of genes [27,28]. Although not the sole targets of CD8⁺ T cells, trans-sialidase (TS) gene family proteins appear to be the major targets in mice and are also well-recognized in humans (reviewed in [29]) [24,30]. Perhaps this is not surprising, given that the TS proteins are among the most abundantly expressed proteins in T. cruzi [31]. During acute infection in C57BL/6 mice, up to 50% of the entire CD8⁺ T cell population is focused on just three immunodominant epitopes: TSKb18 and TSKb20 from TS proteins and MUCKb25 from mucin family proteins [24,28]. All three CD8⁺ T cell responses, however, fail to coordinate parasite clearance, suggesting that the host targeting of this set of epitopes is an ineffective strategy.

One potential reason why the large gene family epitopes are suboptimal T cell targets may be the vast potential for epitope variation. TS and mucin proteins represent a complex set of antigens presented to the immune system – both families are comprised of ~500 to >1000 members, and unlike antigen variants in related pathogens such as the African trypanosomes, many of these variants are expressed simultaneously, thus potentially inundating the MHC-I presentation pathway with epitope variants competing for presentation [24,31,32] (Figure 2). There is also evidence within the TS genes repertoire of high levels of recombination leading to the generation of new TS variants, further adding to the diversity of potential T cell epitopes presented to the immune system [32,33]. Although the rate of new variant production is not known, it is remarkable that the carefully curated genomes of two parasite strains found no identical TS proteins between them [33]. Thus, the immunodominant T cells in T. cruzi likely encounter a multitude of epitope variants during infection, including some that could abrogate T cell recognition.

Figure 2. Summary of immune evasion mechanisms preventing efficient detection of T. cruzi-infected host cells.

The limited host cell sensing of T. cruzi invasion delays the recruitment of effector T cells to new sites of infection, providing a “free-pass” for parasites to replicate and further disperse (1 and 2). Once recruited to infection sites, T cells encounter a time-limited presentation of a complex array of parasite epitopes (3 and 4), or in the case of muscle, a limited number of MHC-presented peptides (5). Created with BioRender.com

In addition to the high level of epitope variation, these variant protein family epitopes appear to have limited and delayed presentation by class I MHC in infected host cells. TSKb20-specific T cells have been shown to recognize in vitro infected cells both early (16 hpi) and late (96 hpi) in the infection cycle, but surprisingly not during the amastigote replication phase (24-72 hpi) [28,34]. This could suggest that the many genes encoding theTSKb20 epitope may only be released by trypomastigotes either at the invasion step or just prior to their release following amastigote replication. The MUCKb25 epitope, however, does not appear to be directly presented by infected cells at all, as MUCKb25-specific T cells were shown to encounter antigen only after parasites were released from host cells (96 hpi) [28]. In this case, nearby APCs were activated to cross-present the MUCKb25 epitope – a process that may be similarly required for TSKb20 epitope presentation. Regardless, the lack of MUCKb25 epitope presentation by infected cells would render the MUCKb25 T cell response completely useless for the recognition of infected cells and the presentation of the TSKb20 epitope at limited times during the infection cycle would provide a very narrow window for detection by T cells. Thus, such ineffective targeting by the T. cruzi-specific CD8⁺ T cell response may allow infected cells to elude recognition by CD8⁺ T cells.

-Persistence in muscle

T. cruzi can infect a wide range of cell and tissue types but most frequently persists within muscle [35,36]. In addition to being by mass the most abundant tissue type in the body, muscle has an extremely low basal MHC-I expression [37,38] and presents class I MHC-associated peptides in levels similar to that of cells in immune privileged sites such as brain, testis, ovary [39] (Figure 2). Thus, even highly functional CD8⁺ T cells which are required to maintain infection control [11], may be particularly challenged to directly detect T. cruzi-infected myocytes. Increasing MHC-I expression selectively within skeletal muscle was shown to significantly enhance CD8⁺ T cell-mediated parasite control within the site, confirming that low tissue MHC-I is limiting T cell recognition and optimal effector activity in muscle [12]. Unfortunately, long term over-expression of MHC-I in muscle eventually led to T cell exhaustion and a consequential uncontrollable parasite load [12], establishing that low MHC expression in muscle is host-protective but also contributes to the persistence of T. cruzi.

T. cruzi dormancy

Some microorganisms possess the remarkable ability to enter a quiescent/dormant state, defined as the capacity to restrict metabolic activity and cease replication. In many cases, dormancy allows select individuals within the population to survive and is particularly advantageous when environmental conditions become more stressful. While most extensively studied in bacteria [40], versions of restricted metabolism and slowed or absent replication is important in the life history of some protozoans as well, such as Plasmodium hypnozoites, Toxoplasma gondii bradyzoites, and Leishmania amastigotes [6]. It was recently revealed that upon invading host cells, a very low number of amastigotes of T. cruzi can enter a transient quiescent state [7]. Following a variable period of dormancy, some of these amastigotes reactivate, differentiate, and exit the host cell to propagate the infection.

The transition into and out of dormancy in T. cruzi is a spontaneous and stochastic process not requiring triggering by external stimuli such as drug exposure [41], nutrient deprivation [42] or immune pressure. This is evidenced by the presence of dormant forms in vitro under non-stress conditions and alongside actively replicating parasites within the same cell [7]. Analysis using the methionine analogue L-Homopropargylglycine (HPG) reveals greatly reduced synthesis of new proteins in dormant amastigotes, both in vitro (Figure 3) and in mouse tissues. These findings align with the negligible fluorescence signal of dormant parasites expressing the tdTomato reporter protein under the otherwise potent ribosomal promoter. Most importantly, the dormant state of T. cruzi amastigotes allows resistance of this minor population to multiple trypanocidal drugs that are highly effective against actively replicating forms. [7,43].

Figure 3. Dormant and metabolically active T. cruzi amastigotes.

Detection of a single dormant amastigote (arrow) among metabolically active amastigotes in an infected Vero cell. Note the apparently normal nucleus and kinetoplast structure but the essential absence of new protein production (detected by incorporation of the methionine analogue L-Homopropargylglycine (HPG; orange) in the dormant amastigote). Cell monolayers were expanded using Ultrastructural Expansion Microscopy (U-ExM), stained with SYTOX Deep Red (DNA; cyan) and imaged using Airyscan microscopy.

The mechanisms that drive dormancy are not fully understood. A compelling hypothesis proposes that spontaneous DNA damage activates DNA repair mechanisms, leading to the assembly of repair enzymes within the nucleus and eventual cell cycle arrest until DNA repair is complete [44–47]. This model is consistent with the flexibility required of the T. cruzi genome to the continuous generation of new variants within the recombination-prone large gene families such as the TS, as described above. However, solid evidence for DNA damage causing dormancy is lacking. The study of dormant T. cruzi forms presents numerous challenges, including its rarity and the lack of cellular markers of dormancy. Staining infecting trypomastigotes with CellTrace or other proliferation dyes is a simple way to mark dormant cells, as replicating amastigotes progressively dilute the dye until it becomes non-detectable with only dormant forms retaining [7,47–49]. However, the dye dilution approach is only useful for tracking parasites that enter dormancy early after initiation of infection and does not allow for the detection of newly dormant amastigotes occurring at any point during in vitro or in vivo infection. An ideal dormancy marker might be an enzymatic probe detectable in live cells, so that time-lapse imaging techniques could be used to track transitions into and out of dormancy.

Conclusions

There is still much we do not understand about both persistence and dormancy in T. cruzi. These are priority research areas since both phenomena potentially impact how one approaches development and delivery of therapeutics for preventing or curing T. cruzi infection. Our understanding of the mechanisms of persistence suggest that it will be extremely difficult to develop vaccines that will prevent infection, or other therapeutics that will effectively boost immunity so as to achieve sterility. The fact that mice subjected to three rounds of long-term infection followed by drug-induced cure are still able to be infected a 4^th time with the homologous parasite strain is emblematic of this concern [50]. Alternatively, individuals differ significantly in the efficiency with which they control T. cruzi infection, and some do achieve parasitological cure, so immune-mediated elimination of T. cruzi is not impossible.

As noted, dormant stages do not appear to be required for infection persistence, but they are highly relevant to achieving consistent and rapid cure using typanocidal drugs. A better understanding of the molecular mechanisms mediating both the entry and exit of quiescence could facilitate the design of effective anti-quiescence compounds. Lacking this, we may have to accept that effective treatment will require a longer treatment period than is desired, effectively maintaining drug pressure until all dormant parasites have either died or have exited dormancy and become susceptible to killing by trypanocidal drugs.

Highlights.

• T. cruzi persists in muscle despite highly effective immune responses

• Chagas disease pathology is the result of this long-term persistence in muscle

• T. cruzi invades host cells without robust stimulation of innate immune sensing

• Persistence in muscle is abetted by the limited presentation of antigen variants

• Dormant amastigotes are irrelevant to persistence but problematic for drug treatment