Abstract
Cryptosporidium is a leading infectious cause of diarrhea around the world associated with waterborne outbreaks, community spread, or zoonotic transmission. The parasite has significant impact on early childhood mortality, and infection is both consequence and cause of malnutrition and stunting. There is currently no vaccine, and treatment options are very limited. Cryptosporidium is a member of the Apicomplexa, and as typical for this protist phylum relies on asexual and sexual reproduction. In contrast to other Apicomplexa, like malaria parasite Plasmodium, Cryptosporidium’s entire lifecycle unfolds in a single host in less than three days. Here we establish a model to image lifecycle progression in living cells, and observe, track, and compare nuclear division of asexual and sexual stage parasites. We establish the length and sequence of the cell cycles of all stages and map the developmental fate of parasites across multiple rounds of invasion and egress. We determine that the parasite executes an intrinsic program of three generations of asexual replication, followed by a single generation of sexual stages that is independent of environmental stimuli. We find no evidence for a morphologically distinct intermediate stage (the tetraploid type II meront) but demonstrate direct development of gametes from 8N type I meronts. The progeny of each meront is collectively committed to either asexual or sexual fate, but importantly, meronts committed to sexual fate give rise to both males and females. We define a Cryptosporidium lifecycle matching Tyzzer’s original description and inconsistent with the coccidian lifecycle now shown in many textbooks.
Introduction
Pathogen associated diarrheal disease is one of the leading causes of mortality in children under the age of 5 years [1]. While efforts to improve sanitation, hygiene and access to clean water have reduced the number of diarrheal deaths, nearly half a million children under the age of 5 died due to diarrheal diseases in 2015. The three most common causes of pathogen associated diarrhea in children under 5 are rotavirus, Cryptosporidium, and Shigella, which together account for more than half of all diarrheal deaths in this age group [1, 2]. Malnourished children are particularly susceptible to severe cryptosporidiosis [2–4], and in turn, infection with this pathogen can have long lasting consequences for the nutritional status and overall growth and development of children [5, 6]. Children develop non-sterile immunity to Cryptosporidium that protects from severe disease and malnutrition [7], however this immunity is slow to develop and currently no vaccines are available to prevent the infection [3]. Nitazoxanide has been approved by the US Food and Drug Administration for the treatment of cryptosporidiosis, but this drug is not effective in immunocompromised or malnourished individuals [8]. The last five years have seen a significant push towards better treatments for cryptosporidiosis (see [9] for a succinct review). Several of these efforts took advantage of recent advances in the development of antimalarials by using cherry picked compound collections initially derived in phenotypic screens against Plasmodium falciparum [10–15]. The value of targeting multiple lifecycle stages is a clear lesson that emerged from the malaria drug development effort [16]. A more comprehensive understanding of the biology of the Cryptosporidium lifecycle and the relative susceptibility of its different segments is required to discover and improve drugs to establish effective treatments for this disease [3].
Cryptosporidium is a single cell protist parasite and a member of the phylum Apicomplexa, organisms that undergo complex lifecycles of asexual and sexual reproduction. In the well-studied apicomplexans Plasmodium and Toxoplasma (the causative agents of malaria and toxoplasmosis), this lifecycle unfolds in different hosts: mice and cats or humans and mosquitoes, respectively. In contrast, Cryptosporidium replicates both asexually and sexually within a single host and transmission between hosts occurs through meiotic spores called oocysts. In humans and animals, Cryptosporidium infects the intestinal epithelium and oocyst shedding begins on the third day of infection, in the absence of cell mediated immunity infection is chronic and parasite growth continues unabated [17–19]. Here we study Cryptosporidium parvum, a parasite of cattle and humans that is experimentally tractable [20]. At any given time, roughly one third of the parasites within the small intestine of an infected mouse replicate asexually, one third appear to be sexual stages, and one third represent post-fertilization stages that are in the process of forming oocysts [21]. In cell culture systems, including the widely used human adeno carcinoma cell line HCT-8, Cryptosporidium is limited to approximately 3 days of growth. Inoculation of HCT- 8 cells with oocysts or sporozoites released from oocysts produces robust infection with parasites that reproduce asexually, however after two days the culture abruptly sexualizes and is dominated by male and female gametes and growth ceases. Post-fertilization stages are not observed in culture, likely due to a block in the fertilization step [21]. Interestingly, in organoid-based cultures longer-term growth has been observed as has fertilization and oocyst formation [22–24]. Sex thus appears to reset the lifecycle and initiate subsequent rounds of asexual growth and expansion.
Different apicomplexans have evolved diverging mechanisms to accommodate the progression of developmental stages to their respective host niches. Commitment to sexual development in Plasmodium occurs at varying frequencies depending on species and strain, suggesting an underlying inherited developmental threshold [25]. A small portion of each asexual generation commits to sexual development and initiates gametocyte development, these cells will mature into gametes once ingested by mosquitoes with a bloodmeal and then undergo fertilization [26]. Tissue and biochemical cues have been identified that impact on the likelihood of the developmental switch that results in exit from the asexual cell cycle and differentiation into the growth arrested gametocyte stage [27, 28]. In contrast, conversion of Hammondia from fast growing tachyzoites to slow growing bradyzoites appears to be governed by a molecular clock [29, 30]. Similarly, Eimeria executes a predetermined number of asexual cycles prior to the emergence of gametes [31, 32].
The Cryptosporidium lifecycle is remarkably short and much of it unfolds in tissue culture. Here we establish a long-term live-cell microscopy model to directly observe the lifecycle and to fate map developmental progression. We find no evidence of environmental induction of gametogenesis, but strict adherence to a timed developmental program. The intracellular development of all stages unfolds in roughly twelve-hour intervals, with three generations of asexual meronts followed by a single generation of gametes. Merozoites emerging from one parasite cell are collectively committed to either an asexual or sexual fate, but sexually committed meronts give rise to both males and females. We rigorously demonstrate that gametes develop directly from asexual stages that produce eight merozoites, known as type I meronts, and we refute a role for a morphologically distinct type II meront as an intermediate stage between the asexual and the sexual phase of Cryptosporidium development.
Results
Sexual differentiation of Cryptosporidium parvum follows a parasite intrinsic program
Cryptosporidium parvum differentiates from the asexual to sexual phase of its lifecycle 48 hours into culture and parasites cease to replicate. We wondered how this transition may be triggered and considered the presence of a parasite extrinsic stimulus (Fig 1A). This might include changes in the physicochemical properties of the environment [33], the depletion or accumulation of a metabolite [27, 34], or the activity of a dedicated density-dependent quorum sensing mechanism [35]. Alternatively, C. parvum may follow an intrinsic program that is independent of extracellular factors. To test for differentiation stimuli, we performed experiments with conditioned media (Fig 1B). Media were conditioned by growing HCT-8 cells with or without C. parvum infection for 48 hours, the time point when differentiation occurs. The media was filtered (0.45 μm) to remove extracellular parasites, and then transferred to fresh cultures. These cultures were infected with a C. parvum reporter line expressing nanoluciferase and we monitored parasite growth over 72 hours by luciferase assay [20]. Use of infection conditioned media did not result in an earlier arrest of parasite growth (Fig 1C).
We also conducted experiments using a C. parvum reporter strain [21] and recorded the developmental progression through different stages. We again used conditioned media, this time added to infected coverslip cultures, which were processed for immunofluorescence assays 24, 48, and 72 hours after infection. The number of male gamonts (blue), female gametes (pink) and asexual meronts (green) was scored at each time point by microscopy (n=3) and is displayed as a fraction of all parasites encountered (Fig 1D). Conditioning did not hasten lifecycle progression and the representation of different stages was indistinguishable between conditioned and unconditioned media, and similar to that previously reported [21].
We considered that a sex inducing factor might be unstable, poorly soluble, or remain cell associated, and thus is not transmitted well by media transfer. We used superinfection of the same culture to test this (Fig 1E). Host cells were first infected with unmarked wildtype C. parvum, 24 hours later they were infected again, this second time with a transgenic parasite strain expressing a fluorescent reporter. We then performed immunofluorescence assays to score stages and assessed life cycle progression of both infections separately. Despite the presence of a sexualized primary infection, the second infection again produced gametes only after 48 hours (Fig 1F), both waves showed similar kinetics but were offset by their 24 hours difference of time in culture. Taken together, we did not find evidence for an external induction mechanism, and thus propose that C. parvum is following an intrinsic program of lifecycle progression.
Imaging the intracellular development of asexual and sexual stages of C. parvum
Next, we wanted to observe the C. parvum lifecycle program in real time by live-cell microscopy. We engineered parasites to express a mScarlet fluorescent protein in the cytoplasm in addition to a mNeon-tagged H2B histone labeling the nuclei in green (S1 Fig). These parasites were used to infect HCT-8 cells grown in 8-well chamber slides and imaged using an GE DeltaVision OMX Structured Illumination Microscope controlling temperature and atmosphere (see Materials and Methods for detail). In preliminary experimentation, we established that imaging every 30 minutes permitted continuous recording for up to 42 hours while maintaining parasite and host cell viability. We began imaging at 11 hours of culture which we had previously established as the end of the first sporozoite initiated merogony cycle [36] and also conducted experiments imaging from 29 or 40 hours of culture onwards. We used multiple point visiting and autofocus routines to allow us to observe cells in parallel and in significant numbers collecting a total of 6171 hours of time lapse data, of which 4542.5 hours were suitable for analysis. We were able to consistently distinguish the intracellular development of asexual meronts, from that of male gamonts and female gametes. Asexual parasites were observed prior to 40 hours post infection, sexual stages after that point.
Figure 2A shows selected still images from two representative movies of asexual development (see S1 Movie for multiple additional cells). We analyzed the intracellular development of a total of 380 meronts and found a mean time to egress of 12.57 hours, which is similar to the timing of sporozoite initiated C. parvum merogony [36]. C. parvum replicates by schizogony [37], resulting in a cell with eight nuclei. We found that the increase in the number of nuclei strictly followed geometric progression, indicating that in contrast to Plasmodium [38], in C. parvum nuclear divisions are highly coordinated and synchronous throughout. Using nuclear divisions as landmarks, we decerned three phases: a long initial establishment phase, a relatively brief mitotic phase, and a final budding phase. Meronts completed the first nuclear division after 7.92 hours, and then ran through two additional complete mitotic cycles, taking about 1 hour for each (Fig 2C, D and E). There was a lag phase of 2.77 hours between the last division and egress.
The development of male gamonts followed a pattern similar to that of meronts (Fig 2B and S2 Movie). After a 6.28-hour establishment phase, four rapid mitotic cycles produced 16 nuclei, and the average time to egress was 12.08 hours (Fig 2F, G and H). Up to the 8N stage, the nuclei of male gamonts were round; only the last division produced the highly condensed spindle shaped nuclei, that are characteristic for male gametes [39, 40]. As seen for meronts the time required increased slightly from division to division (Fig. 2F). While immature male gamonts appeared overall similar to meronts, they can be distinguished. The nuclei of gamonts clustered to the center, while nuclei of meronts showed greater dispersion (see S2 Fig). We note that half of the male parasites observed failed to egress (93 out of 172 times), this may be typical for male gamonts, associated with the culture model, or could reflect additive photodamage due to longer overall imaging times.
Female parasites did not undergo nuclear division and remained intracellular, allowing us to image them for 24 hours (the time the experiment ended). However, the size of their nuclei increased 3-4 times in area, and the overall size of the gamete grew 6-8 times in this time (Fig 2I, J and S3 Movie). We made two additional observations. First, while female gametes grew rapidly initially, growth plateaued after 12-15 hours. Female gametes are the transcriptionally and translationally most active of all lifecycle stages as they produce essentially all components of the oocyst [21, 41] and this was evident in their production of fluorescent protein. We observed a robust increase in cytosolic mScarlet fluorescence beginning at 8 hours (Fig 2K), this fluorescence reached peak intensity 12-14 hours after invasion (note that we did not correct for photobleaching and that fluorescence thus diminishes in the absence of new synthesis). We note that both these timeframes are match the time to egress for male gametes. Interestingly, all three parasite stages are morphologically indistinguishable for the first 6-8 hours after invasion, after which male and asexual parasites begin to divide their nuclei, while female parasites continue to increase in size (S3 Fig).
Long term imaging reveals three generations of meronts followed by gametes
Our imaging covered a total of 60 hours of parasite development in multiple overlapping experiments. We tracked hundreds of individual parasites and mapped them onto the overall lifecycle timeline. Figure 3A shows each cell as a line, with the start representing invasion, and the end showing the time of egress (egress is only shown for meronts (green) and male gametes (blue) as female parasites do not egress). We next mapped all observed nuclear divisions, as well as invasion and egress events as individual time points (Fig 3 B-E). These analyses revealed three waves of merogony followed by a single wave of gamete development (please note that we only observed the tail end of the first wave as we start imaging at 11 hours). Waves begin with a relatively sharp line followed by a trail of ‘late comers’. We note a consistent shift to gametes at 40 hours in line with previous studies using immunofluorescence assays [21, 42, 43].
Meronts release merozoites committed to either asexual or sexual fate
Next, we searched our image dataset for instances in which we could track successive generations of parasite development. Figure 4A shows selected frames over 27 hours of a time lapse movie (Movie S4). The fate of mature cells is indicated by a colored arrowhead, newly invaded next generation stages are highlighted by white arrowheads. We used this information to derive trees mapping the fate of each parasite cell (Fig 4B shows examples with the interpretation of the cells shown in Fig 4A in the middle). Note that we cannot unambiguously map the origin or fate of all parasites. Merozoites left the field of view, cells were lost due to egress and/or cell death, the simultaneous egress of multiple meronts obscured origin, or phototoxicity stalled the development of some parasites before we could assign fate. Nonetheless, we were able to track the fate of the progeny of 49 meronts. During earlier time points, all merozoites emerging from a meront again gave rise to asexual meronts in the next generation (Fig 4C). As the cultures shifted from their asexual to their sexual phase at 40 hours, we found all progeny of individual meronts to give rise to sexual stages. Importantly, when tracking the offspring of individual meronts, we never observed both asexual and sexual stages developing from the same parental meront. We conclude that merozoites are collectively committed to lifecycle transition which suggests that commitment is most likely to occur during the intracellular development of the meront.
Meronts committed to sexual development give rise to both male and female gametes
While merozoites from a single meront were strictly committed to either asexual or sexual fate, they were not collectively committed to a single specific sex. We observed both male and female offspring from the same meront (Fig 4C). Out of 26 individual meronts committed to sexual fate for which we were able to observe the development of at least 2 offspring, we observed both male and female offspring 19 times. We were never able to observe all 8 progeny from any meront, the maximum number of observed offspring was 6, in which 2 developed into males and 4 became female. To determine the male to female ratio of offspring from a single meront, we conducted a weighted confidence interval based on all 73 observed offspring from the 26 sexually committed meronts. Based on our observations, the 95% confidence interval suggested that out of the 8 progeny in each meront, between 1.92 and 3.11 will be male. This is consistent with the fact that we never observed more than 2 male progeny from a single meront.
Gametes develop directly from type I meronts, type II meronts are not apparent
Currently many lifecycles of Cryptosporidium depict a morphologically distinct tetraploid generation of asexual parasites called the type II meront, as an intermediate between asexual meronts and gametes (see e.g. the widely reproduced lifecycle from the Centers of Disease Control and Prevention at https://www.cdc.gov/parasites/crypto/pathogen.html). Type II meronts are shown to give rise to four merozoites in contrast to asexual type I meronts which produce eight. This model predicts sexual differentiation to be preceded by a wave of tetraploid meronts (Fig 5A), surprisingly, we did not observe this in a previous study that used molecular markers to define stages [21]. Our live-cell imaging experiments used a nuclear marker that clearly distinguished 4 and 8 nuclei stages and thus provided the opportunity to test this rigorously using a large dataset. We analyzed the eventual fate of 1095 parasites that reached the 4N stage across the 60 hours observed. Parasites were binned by the time at which four nuclei were observed, and then categorized into one of three outcomes (Fig 5B): 1) Disappearance after the 4N stage consistent with egress predicted by type II merogony, 2) progression to 8N prior to egress (predicted by type I merogony), or 3) progression to 16N prior to egress (male gamogony). Parasites that remained 4N until the end of the imaging experiment were excluded from analysis. From 11-40 hours post infection, the vast majority of parasites that reach the 4N stage continued past that stage to the 8N stage prior to egress. During this time, we did not observe male (16N) parasites. After 40.5 hours in culture, the proportion of the population that egresses at the 8N stage decreased markedly over time, while the proportion of the population that develop into males increased at the same rate. Parasites with apparent egress at the 4N stage were rare and importantly, their frequency did not change over the culture time and lifecycle.
Loss of fluorescence, used here as a proxy for egress, is indistinguishable from host cell death or lysis due to the egress of another parasite in cells carrying multiple infections. We thus sought to independently evaluate whether 4N meronts do or do not produce merozoites and egress using molecular markers. Apicomplexan parasites assemble the organelles required for invasion at the very end of the cell cycle, and in those parasites replicating by schizogony, rhoptries and micronemes are only apparent in formed merozoite immediately preceding egress [37, 44]. Several rhoptry bulb proteins were recently described for C. parvum, and their expression was found to be similarly restricted to parasites harboring merozoites poised to egress [36]. We evaluated two time points (20 and 34 hours) to observe meronts committed to asexual or sexual fate in the next generation, respectively and scored the expression of ROP3-HA in 4N and 8N parasites by immunofluorescence. Labeling was exclusively found in 8N parasites regardless of the point in the progression of the lifecycle they were observed (Fig 5D and G).
Key transitions in the intracellular development of apicomplexan parasites are regulated by the activity of calcium dependent kinases (CDPKs) [45]. CDPK1, an important drug target in C. parvum [46], was recently shown to underly strict cell cycle regulation. The protein was only detectable in late stages prior to egress and in very young stages following invasion [47]. We scored the presence of CDPK1-HA in 4N and 8N parasites by immunofluorescence at 20 and 34 hours. CDPK1 staining was only found in 8N stages (Fig 5E), and quantification showed this to be highly reproducible (Fig 5H). We considered that there might be differences between parasite development in vitro and in vivo and thus infected mice with CDPK1-HA parasites. Mice were sacrificed at the peak of infection, small intestines were resected, fixed, frozen and sectioned (see Material and Methods for detail). Cryo-sections of the tissue were incubated with antibodies to HA, lactate dehydrogenase (a marker of all parasite stages), and Hoechst to label DNA. We identified a total of 159 CDPK1 positive cells in 26 fields of view for which we then scored the number of nuclei in the Hoechst channel. All positive cells have a single or eight nuclei and again we did not detect CDPK1 in 4N parasites (Fig. 5F). Taken together, our culture and animal studies find no evidence for a tetraploid type II meront stage, and we conclude that development to gametes occurs directly from meronts that produce eight merozoites. Importantly, we directly observed this transition (e.g. 4A and B) in our time lapse experiments 26 times.
Discussion
Apicomplexa undergo a cascade of developmental changes as they transition through their lifecycles. More than a century of investigation described a complex succession of morphological types that are specifically adapted to the tasks of invasion and intracellular replication in different hosts, organs, and tissues. This includes specialized transmission stages that carry the infection from one host to the next. As these are single celled organisms, differentiation is not terminal or rigidly inherited, but rather a continuous flow in which each generation elaborates a transient fate. Edward Tyzzer in his initial description of the Cryptosporidium muris lifecycle identified three intracellular stages: microgamonts that produced 16 microgametes, macrogamonts that produced single macrogametes, and asexual schizonts. He commented that “the number of merozoites produced in this process of schizogony is almost invariably eight” [48] he also described fertilization and oocyst formation resulting in parasite stages containing four sporozoites. The concept of the tetraploid type II meront as a developmental intermediate between the asexual and sexual reproduction was introduced by John Vetterling in 1971 [49, 50] in studies on C. wraii a parasite of guinea pigs. This might have been inspired by his extensive work on Eimeria in various animals where distinct meront types occur [51, 52]. At the time Cryptosporidium and Eimeria were seen as closely related members of the Coccidia (a phylogenetic view no longer held [53, 54]). Vetterling’s two meront model has been cited widely since [55] and has become the text book lifecycle for Cryptosporidium. The core of the argument between these authors was how to interpret the tetraploid intracellular parasites found in infected animals and cultures. Are they mature meronts that will yield four merozoites committed to sexual differentiation, or are these immature stages that will undergo further nuclear divisions or form oocysts? This was difficult to resolve using fixed samples, and we therefore chose to study living cells. We documented the fate of more than a thousand tetraploid parasites by time-lapse microcopy and our observations are entirely consistent with Tyzzer’s original assertion that all meronts produce eight merozoites – we find no evidence for a type II meront. Molecular markers that report on parasite cell cycle progression further support this in culture and in infected animals. We note that we have not tested C. wrai, the guineapig parasite Vetterling used in his original work, however, for C. parvum, the most widely studied species of this parasite genus we demonstrate a simple and direct lifecycle of only three morphologically distinct intracellular stages: meronts that yield eight merozoites, male gamonts, and female gametes (Fig. 6A).
The intracellular development of Cryptosporidium appears modular. For the initial eight hours all intracellular stages regardless of their eventual fate are morphologically indistinguishable. We discern a uniform phase establishing the intracellular reproductive niche that is associated with initial growth and biomass increase. The parasites inject host modulating factors that likely play a role in establishing their replicative niche during and following invasion [36, 56]. Establishment of a protein export system is required for the delivery of the parasite protein MEDLE2 to the host cytosol. This takes about five hours to become operational and is active in both asexual and sexual stages. Following the initial establishment phase, the cellular programs diverge markedly leading to asexual merozoites, or male and female gametes. This process is highly synchronous yielding a stage specific number of progeny, over a similar roughly 12 hour time frame (eight merozoites, 16 male gametes, and a single female gamete). This is fast when compared to other apicomplexans where intracellular cycles unfold over days, and may represent an adaptation to the intestinal epithelium with its high directional turnover with cells being constantly shed at the tip of each villus.
A striking feature of C. parvum development is the dramatic switch from asexual to sexual reproduction following three generations of meronts, with gametes discernable at 48 hours [21, 42, 43]. In our experiments we found no evidence for an environmental sex inducing factor, the parasite adhered to a rigid timetable of differentiation in different scenarios of media transfer or coinfection. This suggests an intrinsic developmental program that is reset by sex and is consistent with the link between sex and growth observed by multiple investigators [21–23]. This contrast with Plasmodium where transition to sex is sensitive to environmental and metabolic indicators and stressors [27, 28] which is critical to achieving balance between colonization and transmission [57]. To understand how Cryptosporidium might be able to forgo such regulation it is important to consider that, in contrast to Plasmodium, for this parasite sex is not solely linked to transmission but contributes to continued infection. This may shift the colonization/transmission balance to the oocyst and the likelihood that sporozoites will excyst immediately. The parasite may be able to integrate environmental cues into this step. Interestingly, some authors reported thin shelled and thick shell oocyst that may morphologically reflect the dichotomy of local infection and transmission [40] but experimental confirmation of such a mechanism is still missing.
It is technically difficult to establish whether the parasite adheres to its rigid pattern or timing in animals. However, we believe this to be likely, and note that oocyst shedding is detectable on the third day of infection of mice consistent with the time frame observed in culture [18]. The mechanism underlying this intrinsic lifecycle transition is unknown, the parasite may measure time, the number of intracellular cycles, or the accumulation or depletion of a particular molecule or epigenetic mark. A variety of such mechanisms have been explored in the context of the self-limiting expansion of stem cell populations [58] and this may stimulate future studies. Among Apicomplexa the transition from asexual to sexual stages has been most intensively studied in Plasmodium [59], where the transcription factor AP2-G was shown to be required for gametocyte production [60, 61]. AP2-G acts as a master regulator of sex-specific gene expression through a cascade of transcription factors and additional regulatory genes [61]. AP2-G itself is epigenetically silenced in a Heterochromatin protein1 (HP1) dependent manner during asexual growth [62] and HP1 silencing is removed by the protein gametocyte development 1 (GDV1) [63], which is expressed when Plasmodium parasites are grown under conditions that favor sexual stage development. While there are no obvious homologs of HP1 and GDV1 in Cryptosporidium, a similar epigenetic switch could nonetheless underlie the sexual commitment that we observed here. Very little is known about epigenetic gene expression regulation in Cryptosporidium, beyond that the C. parvum genome encodes some histone modification enzymes [64] and that the parasites are susceptible to pharmacological inhibition of these pathways [65, 66]. Future studies analyzing the level of histone modification across the genome and across the lifecycle could reveal such a mechanism.
While sex is an ancient phenomenon found in most eukaryotes, how the specific sex or mating type of individuals is determined is varied and evolutionary malleable [67]. Sex can be inherited, be determined by environmental factors like temperature, or be ‘negotiated’ between different members of a population by social behavior, biochemical clues, or cell-cell interaction and associated signaling events [68–73]. How sex is determined in Apicomplexa is unknown [31, 74] but sexual differentiation and sex-ratio play important roles in transmission and environmental adaptation [57]. In Plasmodium the offspring of individual schizonts is thought to be collectively committed to a male or female fate in a mutually exclusive fashion [75]. One of the most interesting observations of our study in Cryptosporidium, however, is that the sexually committed meront consistently gives rise to both males and females (Fig. 4). We propose that this makes it likely that the future sex of individual merozoites and thus the sex ratio is determined in the meront prior to egress (Fig. 6B). The mechanism by which this is achieved remains to be discovered. The fact that commitment occurs over the backdrop of cell division may inform the discussion and offers hypotheses. Unequal inheritance or asymmetrical segregation of fate determining organelles or molecules is well established in the differentiation of mammalian cells and embryos [76, 77] and could be at play here. We note an abundance of sex-specific non-coding RNAs in Cryptosporidium [78] as one set of potential candidates. There is also an emerging understanding of mechanisms that yield differential inheritance of epigenetic histone modification, and this can occur at different steps of DNA synthesis and mitotic segregation [79]. The mechanism could be stochastic instead of directive, it may not matter which specific merozoite adopts a certain fate as long as the desired sex ratio is achieved. Overall, the model predicts merozoite heterogeneity, and experiments that would reveal the molecular nature of such heterogeneity could provide clues as to how sex is determined. An important biological consequence of the Cryptosporidium commitment model is that male and female gametes by default will develop in close spatial proximity. Even at very low initial parasite burden this mechanism provides a safe route towards fertilization through selfing. This may be critical for a parasite that relies on a sexual lifecycle reset for continued growth [21–23]. Cryptosporidium provides the opportunity to analyze the fundamental molecular tenants of apicomplexan lifecycle progression, sex determination, and gamete interaction in a stripped-down single host lifecycle. In this study, we showed it to unfold in culture and over the course of only four infectious cycles in less than 72 hours, and to employ a minimal cast of three morphological types.
Materials and Methods
Generation of Transgenic Parasites
Transgenic parasites were generated using previously described methods [20, 80]. Briefly, 5 x 107 C. parvum oocysts (Iowa II strain obtained from Bunchgrass Farms) were bleached and washed prior to incubation in 0.8% sodium taurocholate for 1 hour at 37 C to induce excystation. Following excystation a CRISPR guide plasmid and repair template with 50 base pair homology arms were introduced via nucleofection using the AMAXA 4D Nucleofector (Lonza). Transfected sporozoites were then diluted in PBS and used to infect a cage of mice. Here we used the modified protocol for infection in which mice were given 100 μL of 8% sodium bicarbonate solution by oral gavage 10 minutes prior to oral gavage with the transfected parasites [80]. Stable transformants were selected with 16 mg/mL paromomycin (in drinking water at libitum) and parasite shedding was monitored via nanoluciferase activity in the feces of infected mice [20].
Transgenic parasites were isolated from feces by sucrose floatation followed by cesium chloride gradient [81] and then stored at 4 C in PBS until used for each experiment.
Immunofluorescence Assay
Human adeno carcinoma HCT-8 (ATTC CCL-244) cells were maintained by serial passage in RPMI-1640 containing 10% fetal bovine serum (FBS). Immediately prior to infection, the medium was exchanged to RPMI-1640 containing 1% FBS. HCT-8 cells were infected with oocysts that were bleached using 10% bleach for 10 minutes at 4 C, washed 3 times with cold PBS, excysted with 0.8% sodium taurocholate at 37 C for 10 minutes, and washed once with PBS before addition to the culture. Infected HCT-8s were fixed using 4% paraformaldehyde for 10 minutes, then washed and permeabilized with 0.25% triton x-100 for 10 minutes. Fixed and permeabilized cells were then blocked with 3% bovine serum albumin (BSA) for 1 hour at room temperature prior to incubation with primary antibodies for 1 hour at room temperature.
For immunohistology, ifng-/- mice (Jackson Laboratories Strain 002287 bred inhouse) were infected with 10,000 CDPK1-HA oocysts (a kind gift of Dr. Sumiti Vinayak, University of Illinois Urbana-Champaign) and their intestines were resected and ‘swiss-rolled’ prior to fixation overnight in formalin. Cryo-sectioning was performed by the PennVet Pathology core facility and immunofluorescence was performed as described [18].
Antibodies and dye used: anti-HA (Roche clone 3F10), VVL-FITC (Vector FL1231), anti- LDH (a kind gift of Dr. Guan Zhu Texas A&M University, now Jilin University [82]).
Following incubation with primary antibodies, coverslips were washed three times with PBS at room temperature, then incubated with appropriate secondary antibodies in 3% BSA for one hour at room temperature. Cells were counterstained with DAPI or Hoechst for 10 minutes, washed with PBS twice and then mounted on slides with Fluoromount or Vectashield. Coverslips were observed using a Leica DM6000B Upright Widefield Microscope using 63x or 100x objectives or a GE DeltaVision OMX Structured Resolution Microscope using a 60x objective. Both microscopes are maintained by the Penn Vet Imaging Core.
Acquisition and Processing of Live-Imaging Data
8-well chamber slides (Ibidi) were seeded with HCT-8 cells, which were then grown in RPMI-1640 supplemented with 10% FBS at 37 C, 5% CO2 for 24-48 hours prior to infection. Host cells were switch to pre-warmed RPMI-1640 supplemented with 1% FBS immediately prior to infection with bleached, washed and excysted sporozoites. Imaging was preformed beginning at various time points using a GE DeltaVision OMX Structured Resolution Microscope using the conventional light path. Growth conditions were maintained throughout each imaging experiment at 37 C, 5% CO2 and 40-60% humidity. Using the AquireSR Acquisition control software, Z-stacks were taken in both the 488 and 568 channels for multiple points of interest every 30 minutes for up to 42 hours. Images were processed using the softWoRx image reconstruction and analysis software. Briefly, images were deconvolved, the channels were then aligned, and the z-stacks were compressed to generate a 2-channel image for each time point. Manual drift correction was applied in ImageJ to generate movies and stills of individual developing parasites. Ten independent experiments were performed, we collected a total of 6171 hours of images and were able to analyze 4542.5 hours containing growth and replication information for 1365 individual intracellular parasites.
Analysis of Live-Imaging Data
Analysis was performed using 2-channel time series data obtained as described above. Parasites were manually tracked and recorded for time of invasion, any subsequent nuclear replication events, and apparent egress. Data was graphed and analyzed using GraphPad Prism and Microsoft Excel software.
Statistical Analysis
The weighted confidence interval was performed in Microsoft Excel using the observational data from 73 offspring of 26 sexually committed meronts. All other statistical analyses were performed using GraphPad Prism.
Supporting information
S1 Movie. Growth and division of asexual C. parvum. Five representative videos of individual asexual meronts from first appearance to apparent egress.
S2 Movie. Growth and division of male C. parvum. Five representative videos of individual male gamonts from first appearance to apparent egress.
S3 Movie. Growth and development of female C. parvum. Five representative videos of individual female gamonts from first appearance until maximum fluorescence intensity has passed.
S4 Movie. Multiple generations of asexual growth and sexual development of C. parvum. A representative image of two generations of asexual growth, followed by the development of male and female gamonts.
Acknowledgements
This work was supported in part by grant R01AI127798 from the National Institutes of Health to BS and a postdoctoral fellowship from EMBO to AG (ALTF 58-2018). We are grateful to Drs. Vinayak and Zhu for sharing reagents with us and to the PennVet pathology and imaging core for assistance. We would also like to thank members of our laboratory, particularly Dr. Katelyn Walzer for ongoing discussions and feedback.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.
- 12.
- 13.
- 14.
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.
- 70.
- 71.
- 72.
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵