Contributions of diverse models of the female reproductive tract to the study of Chlamydia trachomatis-host interactions

Abstract

Chlamydia trachomatis is a common cause of sexually transmitted infections in humans with devastating sequelae. Understanding of disease on all scales, from molecular details to the immunology underlying pathology, is essential for identifying new ways of preventing and treating chlamydia. Infection models of various complexity are essential to understand all aspects of chlamydia pathogenesis. Cell culture systems allow for research into molecular details of infection, including characterization of the unique biphasic Chlamydia developmental cycle and the role of type III secreted effectors in modifying the host environment to allow for infection. Multi-cell type and organoid culture provide means to investigate how cells other than the infected cells contribute to the control of infection. Emerging comprehensive three-dimensional biomimetic systems may fill an important gap in current models to provide information on complex phenotypes that cannot be modeled in simpler in vitro models.

Introduction

Chlamydia trachomatis is a human-specific, obligate intracellular bacterium that causes the most frequently reported bacterial sexually transmitted infection [1]. The female reproductive tract is composed of a series of connected organs that each have distinct cell composition and architecture (Figure 1a). In women, C. trachomatis infections are initiated within the lower reproductive tract (vagina and cervix) and can ascend to organs of the upper reproductive tract (uterus and fallopian tubes). In all cases, the bacteria replicates within epithelial cells. If the bacteria are not cleared by the immune system, infection induces substantial inflammation within reproductive tissue and can lead to sequelae including pelvic inflammatory disease and infertility [2], as well as a potential increased risk of developing cervical cancer [3]. Despite the availability of effective antibiotic treatments, C. trachomatis infections still represent a substantial public health problem due to the sequelae, the prevalence of asymptomatic infections, and the lack of a vaccine in humans [4].

Figure 1.

Current models of the female genital tract to study Chlamydia infections. (a) The human female reproductive tract. Organs relevant to Chlamydia infection are labeled, with overall cellular composition of the cervix, uterus, and fallopian tube lining shown at right. The cervix contains squamous (light orange, at left) and columnar (orange, at right) epithelial cells, a layer of mucus (green), the cervicovaginal microbiota (red), and a layer of fibroblasts (brown) containing blood vessels lined with endothelial cells (red) and containing various immune cells (purple). The uterus includes columnar epithelial cells (orange), endometrial stromal cells (brown), muscle cells of the myometrium (dark purple), and spiral arteries (red). The fallopian tube lining contains secretory (orange) and ciliated (yellow) epithelial cells as well as fibroblasts (brown). Chlamydia inclusions containing infectious (green) and replicative (red) bacteria are depicted in some of the epithelial cells. (b) Top panel: The most common in vitro immortalized cell culture model of the female reproductive tract, consisting of a monolayer of immortalized epithelial cells (orange) on the bottom of a plastic dish (gray) covered in cell culture media (pink). Such systems are used to gain insights into molecular details of Chlamydia infection (bottom panel): at left, the developmental cycle of Chlamydia elementary bodies (green) and reticulate bodies (red) within the lumen of the inclusion and, at right, the secretion of various type 3 secretion system (black) effectors (teal). (c) Top left panel: Common in vitro co-culture models include immortalized epithelial cells (orange), sometimes on a porous transwell membrane (black dashed line), with fibroblasts (brown) at the bottom of the plastic dish (gray), all within cell culture media (pink). Top right panel: Other co-culture models include organoid systems, containing spheroids of stem cell-derived epithelial cells (orange) within a gel matrix (pink) in a plastic dish (gray), with the potential addition of other components such as immune cells (purple). Bottom panel: Co-culture systems are useful to investigate proposed cell-cell interactions during infection, such as the role of stromal cells (brown) as intermediates in relaying signals, such as (left) hormones (grey spheres) to modulate Chlamydia infection in epithelial cells (orange) or (right) investigating how Chlamydia impedes neutrophil (purple) recruitment and activation to the site of infection. (d) Top panel: An in vitro explant model of fallopian tube, consisting of excised pieces of fallopian tube in a dish (gray) within cell culture media (pink). Bottom panel: Explant models have allowed for proposed insights into complex phenotypes, such as tissue damage, caused by infection.

There has been substantial research into chlamydia pathogenesis to understand how to prevent and treat infections and their sequelae. As with other infectious diseases, a variety of models are used to investigate Chlamydia-host interactions, ranging from simple monoculture systems in vitro to in vivo work in mice and humans. Here, we describe how different models, particularly in vitro systems, have aided in dissecting the molecular basis of chlamydia pathogenesis. We additionally provide a perspective on emerging and future models that will be valuable for human-relevant investigations of Chlamydia-host interactions, focusing on systems made possible by recent advances in three dimensional (3D) complex cell culture models.

Insights into Chlamydia-host interactions gained from immortalized cell culture systems

Insights from commonly used monoculture systems

Cancer-derived (e.g. HeLa) and artificially immortalized (e.g. COS-7) cell lines are used to provide insights into host-Chlamydia interactions at the cellular and molecular level (Figure 1b). Immortalization provides the cells with the ability to expand for generations, allowing for many replicable experiments. Typical monoculture systems are often readily genetically manipulable, which has been especially useful in Chlamydia research as Chlamydia spp. were genetically intractable until recently. Such monoculture systems are also amenable to microscopy, a cornerstone of Chlamydia research for decades, that has allowed for key insights into Chlamydia cell biology. Monoculture systems have long been, and will remain to be, attractive models for dissecting fundamental molecular aspects of Chlamydia-host interactions.

Chlamydia spp. undergo a unique biphasic developmental cycle wherein the bacterium enters the host cell as an infectious elementary body (EB), forming a membrane bound compartment termed the inclusion within which the EB transitions to the replicative, but non-infectious reticulate body (RB) (Figure 1b) [5]. After replicating, RBs transition back to EBs, which are released from the infected cell via extrusion or lysis. Transcriptomic and proteomic based studies provided a wealth of information about the distinct properties of RBs and EBs [6-8] and were instrumental in generating fluorescent reporters of Chlamydia developmental forms to aid in quantifying the kinetics of the development cycle [9-11]. While little is conclusively known about the factors that regulate transition from one developmental form to another, several models have recently been proposed. Investigations utilizing fluorescent reporters suggest a model by which EBs are the result of asymmetric division of RBs [11]. Another recently proposed model, based on 3D electron microscopy of infected HeLa cells, predicts size-dependent regulation of RB to EB transition [12]. Additional data indicate that the metabolic state of host cells may be one signal that promotes the initial EB to RB transition after infection [13]. Specific Chlamydia genes, including clpX [14,15], incS [16], and euo [17], were also recently identified as regulating development through incompletely understood mechanisms. Regulation of the unique developmental cycle is of key interest to ongoing research, with monoculture systems providing the ideal system for detailed molecular dissection of pathways which could be targeted to limit the spread of the bacteria during infection.

Many host-Chlamydia interactions are dependent on the bacterial type III secretion system (T3SS), which translocates effector proteins from within the bacteria into the inclusion membrane and the host cell [18] (Figure 1b). To date, up to 100 effectors, including 50 predicted (~ 40 confirmed) inclusion membrane-associated (Inc) proteins, have been identified [19]. The recent development of genetic tools including expression plasmids and inducible expression systems [20-22], random mutagenesis [23-25], and targeted knockout and knockdowns [26-29] for Chlamydia have been instrumental for characterizing the roles of Chlamydia effectors.

Beside promoting uptake of the bacteria [30-33], many Chlamydia effectors interface with host proteins to promote interactions, directly or indirectly, between the inclusion and host organelles to favor the intracellular Chlamydia lifestyle [19,34] (Figure 1b). Some of the latest studies show that several T3SS substrates including the inclusion membrane proteins IncD [35,36], IncV [37,38], and IncS [39] mediate interactions between the inclusion and the endoplasmic reticulum (ER), establishing ER-inclusion membrane contact sites, putatively promoting processes such as lipid acquisition [40]. Multiple effectors target the Golgi apparatus, with Cdu1 inducing Golgi fragmentation [41,42] and InaC coordinating with host ARF GTPases to control positioning of Golgi ministacks around the inclusion [23,43]. At least two effectors localize to the mitochondria, which may contribute to altered mitochondrial protein composition during infection [44]. Host cell division is substantially altered, such as Chlamydia effector CteG inducing abnormal centrosome amplification [45] and IncM inducing multinucleation and inhibiting cytokinesis [46].

Perhaps not surprisingly for an obligate intracellular pathogen with a reduced genome, several secreted effectors have multiple roles during infection, both at distinct times and sites. IncM impacts a plethora of aspects of host cell biology, altering interactions with the Golgi apparatus in addition to its roles in modulating cell division [46]. IncS promotes the EB to RB transition early in infection and maintains the integrity of the inclusion late in infection [16,39]. CTL0390, subsequently renamed GarD, inhibits both ubiquitylation and subsequent clearance of the inclusion [47], as well as interacting with the STING pathway to promote host cell lysis [48]. CpoS also interacts with the STING pathway and suppresses host interferon signaling [49], while additionally regulating formation of microdomains on the inclusion surface [50]. Moonlighting roles of effector proteins may be more common than currently appreciated, providing a valuable vein of further investigation into molecular host-Chlamydia interactions.

The emergence of more physiologically relevant cell monoculture systems

Cancer-derived cell lines exhibit a variety of defects compared to cells in vivo, including altered metabolism, chromosomal abnormalities, lack of polarization, and altered immune responses to pathogens, including Chlamydia [51]. These significant limitations have driven the identification and development of more physiologically relevant cell lines to enhance the human relevance of in vitro studies.

A2EN and End1 cells are endocervical epithelial cells immortalized by expression of the human papillomavirus genes E6 and E7 [52,53]. They closely mimic epithelial cells in terms of polarization, metabolism, expression of key cell markers and cytokines, and hormone responses [52-55], thereby providing attractive alternatives to assess phenotypes observed in less physiologically relevant cell lines, or to reveal phenotypes that differ or would not be detectable in typical, non-polarized cell lines [45,49,55,56]. For example, the Chlamydia effector protein TepP is crucial for replication in A2EN, but not in HeLa cells [56]. Tight junctions present in A2EN, but not HeLa, monolayers, were destabilized by TepP, identifying a putative role in promoting the spread of infection in vivo [57]. Although only utilized in a limited number of studies so far, newer, more relevant cell lines that provide much of the ease of use of cancer-derived cell lines, without the altered physiology, will be key as the field moves forward in elucidating basic concepts of Chlamydia-host interactions.

Insights into Chlamydia-host interactions gained from co-culture systems

In vivo, infected epithelial cells are not found in isolation. Rather, the outcome of infection depends on interactions between infected cells and the surrounding environment, including stromal cells, immune cells, and the microbiota. Whole-organism models, such as mice or humans, overcome this limitation, but also come with significant disadvantages (Box 1 and Box 2).

Box 1. Murine models of Chlamydia infection of the female reproductive tract.

Various small animal models have been used for Chlamydia research, with mouse models being the most common. Murine models have been instrumental for dissecting the contribution of various immune factors in restricting Chlamydia infections. Amongst the many insights, it has become clear that a few key host factors, including T cells and interferon gamma (IFN-γ) are especially important in controlling Chlamydia [89-91]. Additionally, with the recent advances in genetic tools for Chlamydia and the expansion of mutants, mouse models have been used to confirm that genes important for infection in vitro also matter in vivo [16,45,57,92-95].

Despite their utility, murine models of Chlamydia remain limited in their human relevance. One major limitation is the fact that intravaginal infection of a mouse with C. trachomatis does not resemble human infection, as the bacteria are rapidly cleared and do not efficiently cause pathology [91,96,97]. Rapid clearance has been linked to an increased susceptibility of C. trachomatis to murine innate immune responses, such as an inability to counteract the murine IFN-γ response [96,98]. Sustained murine infections require the utilization of a distinct Chlamydia species, C. muridarum, for which the mouse is the natural host. Like C. trachomatis, C. muridarum is species-restricted by its sensitivity to human, but not murine, IFN-γ responses [99]. Differential susceptibility to infection by the two species can be partially ameliorated by utilizing a transcervical infection route, introducing C. trachomatis directly into the mouse uterine horns and resulting in substantial pathology [91,100,101].

Box 2. Insights into Chlamydia gained from human studies.

Observational studies utilizing human subjects can overcome Chlamydia species specificity (Box 1). Such studies have identified various host genetic factors involved in controlling susceptibility to infection and pathology [102-106]. One recent study identified host genes associated with infertility, allowing more focused investigation of sequelae of Chlamydia with clinical relevance [105].

Human studies have also demonstrated key roles for the cervicovaginal microbiota in modulating susceptibility to infection [107,108]. These studies demonstrated that communities of bacteria dominated by Lactobacillus spp. within the lower reproductive tract are associated with reduced risk of Chlamydia infection and increased ability to clear infection. Lactobacillus-dominated cervicovaginal communities (community state type (CST) I, II, III, and V) are generally considered healthy, in contrast to the diverse communities of predominantly anaerobic bacteria, termed CST IV, which are associated with the dysbiotic condition bacterial vaginosis and an increased risk of Chlamydia infection [109,110]. The protective role for Lactobacillus spp. has been investigated using cell culture models [111], identifying D (−) lactic acid as a key mediator of anti-Chlamydia activity [61,112].

Further human studies identifying host and bacterial factors that may be involved in disease will likely continue to be key sources of hypotheses to be tested in controlled laboratory environments.

Integrating multiple cell types into a single cell culture dish is a convenient middle-ground to study inter-cell type interactions at the molecular level. Co-culture systems described so far typically consist of epithelial cells grown on the permeable membrane of a transwell insert with stromal cells either as a feeder layer at the bottom of the well [58-60], or on the other side of the membrane [61,62] (Figure 1c). Alternatively, stromal cells can be embedded within a collagen matrix with the epithelial cells seeded on top [63]. Such systems allowed for investigating the role of female sex hormones and inter-cell signaling on Chlamydia infection, revealing the stromal cell-dependent pro-Chlamydia effects of estrogen upon co-culture of HEC-1-B or Ishikawa endometrial cells along SHT-290 endometrial fibroblasts (Figure 1c) [58]. With increasing cellular complexity, co-culture systems will be instrumental in teasing apart how inter-cell interactions influence Chlamydia infection, in relatively simple and tractable setups.

Towards more physiologically relevant cells and structure: primary cells, organoids, and explants

Many current systems remain limited in their resemblance to in vivo conditions, due to lack of physiological cell behavior, architecture, or both. Additionally, cell lines in common use are predominantly derived from the cervix and uterus preventing modeling of infection within portions of the upper reproductive tract, such as the fallopian tube, that are relevant for clinical pathology of Chlamydia infections.

Primary cells:

Primary cells are defined as cells isolated from biopsy material or whole organs that are subsequently propagated in vitro without immortalization. Models utilizing primary cells derived from relevant sites of the reproductive tract are key to increasing the physiological relevance of in vitro models of Chlamydia infection. Commercial sources are a good alternative to the potentially challenging acquisition of patient-derived cells, but do not alleviate the limited replicative capacity and culturing requirements of primary cells. Regardless, primary cells from throughout the female reproductive tract, including cervix and fallopian tube, are available and susceptible to Chlamydia infections [47,64,65]. The benefit of using primary cells over immortalized cells was recently highlighted by the replication defect of the cteG mutant displayed in primary cervical cells and in vivo, but not in HeLa or A2EN cells [45]. These results reinforce that, in addition to validating key phenotypes in primary cells to increase confidence in human relevance, investigators should also consider primary cells in their overall experimental design.

Organoids:

Organoids are 3-dimensional simulacra of in vivo tissue generated from stem cells cultured in vitro. The added benefits of organoids, compared to primary cells, are increased self-renewal and the possible inclusion of multiple cell types. Murine organoids, despite their limitations (Box 1), can offset the major hurdle of sourcing human reproductive tract tissues. Organoids derived from several female reproductive organs including human and murine cervix [66,67], human and murine fallopian tubes [13,68,69], and murine uterine endometrium [57,70,71] have been used to successfully model Chlamydia intracellular replication, and to confirm functions of host genes, e.g. SLC1A5 [13], or Chlamydia genes, e.g. tepP [57]. Organoids may also provide a model to investigate the link between Chlamydia infections and gynecological cancers, as highlighted by the increased stemness of epithelial cells upon infection of human fallopian tube organoids [68]. Immune cells can also be incorporated into organoid culture models, which has already revealed the role of the chlamydial effector TepP in driving neutrophil infiltration [71] (Figure 1c).

Explant cultures:

Explant cultures consist of the in vitro culture of small intact portions of surgically resected tissue or organs collected from an organism, allowing for the faithful preservation of tissue architecture (Figure 1d). However, key factors such as immune cells may be lacking. Sourcing, culture conditions, and longevity are also significant limitations. Fallopian tube explant cultures are most commonly used in Chlamydia research, allowing for investigation of an organ not modeled by cell culture lines [72-76]. Explant cultures are attractive models to investigate tissue disruption during infection and were instrumental in revealing roles for Wnt and IL-1 signaling in mediating tissue damage [75,76].

Biomimetics as potential future models for Chlamydia trachomatis infection

Advantages and limitations of biomimetics:

Organ-on-a-chip or biomimetic models (Figure 2) have emerged in recent years as attractive options for modeling human diseases in vitro [77]. The integration of diverse cell types into an in-house made or commercially available single microfluidic device mimics the in vivo tissue structure allowing for in-depth, human-relevant studies in a reductionist system.

Figure 2.

Key features of future biomimetic models to study Chlamydia infections. (a) A representative biomimetic chip containing two stacked channels (red and blue, respectively) separated by a porous membrane (not shown). Inlets and outlets allow for the introduction and/or removal of media to/from the channels. (b) Longitudinal side view through the center of the biomimetic chip shown in (a). An upper layer of columnar epithelial cells (orange) with mucus (green) are separated from a lower layer of fibroblasts (brown) via a porous membrane (black dashed line). Each channel can be supplied with a distinct media type (pink and blue), with additional biological or chemical alterations as needed. Here, the upper channel contains bacterial members of the microbiota (red), while the lower channel contains an assortment of immune cells (purple).

Bioengineering considerations are instrumental in creating a 3D simulacrum of tissue structure within a biomimetic device. Physical factors, such as stretch or flow, which are key modulators of tissue physiology, can be integrated to improve similarity to in vivo tissue [78]. The addition or removal of specific biological components including immune cells [79,80] and the microbiota [79,81,82], can reveal their roles in disease processes. Like any model, biomimetics are only as good as the cells that constitute them and carry the inherent limitations of each kind of cell culture as discussed above. Immortalized cancer-derived cell lines [83], primary cells [84], or organoid-derived cells [85] are used within biomimetics and should be chosen based on the goals of the study.

While biomimetic models are getting traction in various biomedical fields, the high cost of commercially available devices and the lack of technical expertise is perceived as a significant limitation, especially among scientists who do not use such platforms [86]. Current users also identify a number of technical limitations such as lack of long-term stability and insufficient complexity as prominent drawbacks of current systems [86]. Collaborations between groups with overlapping interests, and complementary expertise in bioengineering and biology, will be instrumental in overcoming perceived and technical obstacles to promote the widespread adoption of biomimetic devices.

Biomimetics and infectious diseases research:

Biomimetics of organs such as the lung, gut, and liver are making their way into infectious diseases research to dissect cellular details of host-pathogen interactions [87]. Additionally, the development of biomimetics of various organs of the female reproductive tract [88] is promising for Chlamydia research, providing potential new models of infection. Advancements such as including the microbiota and immune cells, as recently described in other culture systems for Chlamydia [62,71], will be instrumental in utilizing biomimetic technology to probe aspects of Chlamydia infections that are often not possible with current models but are key for infection in vivo. It opens the possibility of modulating immune cell populations in ways that are not readily feasible in vivo or in explant cultures, or of investigating the role of stromal cells during infection when structure more closely mimics in vivo organization than current co-culture models.

Conclusions

All models for studying a specific disease, including chlamydia, possess inherent biological and technical limitations that may decrease relevance of a model to human disease and that impede the use of these models. However, most models are still useful for studying certain aspects of a disease depending on the level of complexity required for the study at hand. Care should be given to ensure that a chosen model is both biologically and technically appropriate for the question being asked.

Monoculture of immortalized cells is best suited for interrogating molecular interactions and performing large-scale experiments, such as screens, that would be unwieldy in other models, but may require validation in primary cells or organoids. Co-culture systems allow for investigation of interactions between cell types, while explant cultures and organoid systems provide the ability to interrogate phenotypes such as tissue damage that are not found in simpler culture models. We further propose that biomimetic models are an attractive middle ground, overcoming the limitations of current mono- and co-culture models and offering the tractability and the relevance needed to further explore the molecular mechanisms underlying chlamydia pathogenesis.

Highlights.

• Chlamydia infections have damaging consequences on female reproductive health

• Diverse human model systems are required for a comprehensive understanding of Chlamydia-host interactions

• Simple cell culture systems permit studies at the molecular level

• Multi-cell systems allow insights into cell-cell interactions during infection

• Biomimetic systems provide tractable models for complex host-pathogen interactions

Data Availability

No data were used for the research described in the article.