T. gondii ingress to the intestinal stroma early in the infection

Studies have demonstrated that the jejunum is the primary intestinal niche of T. gondii growth in mice at day 6 post-oral infection [32]. To investigate the site of bradyzoite infection and eventual differentiation to tachyzoites, we analyzed the early time points of the infection in the jejunum of mice by following the natural route of infection through the consumption of brains containing cysts. To closely mimic natural conditions, we used the cystogenic M4 strain (type II strain), recently isolated from oocysts in our laboratory. This strain underwent minimal cell culture passaging, thereby more accurately representing a wild non-laboratory strain. The M4 strain exhibits a cystogenic phenotype in mice, providing a high yield of cysts per brain [36].

We infected a total of six mice across three independent experiments, each mouse fed two brains containing approximately 4 × 10² brain cysts of the M4 strain. To detect the parasites in the jejunal villi, we sacked the mice after 3 days post-feeding and scrapped the jejunal villi. We used polyclonal antibodies that recognize both stages of T. gondii due to the uncertain parasite stage at 3 days post-feeding, which precluded the use of stage-specific markers such as SAG1 or BAG1 antibodies. We employed actin staining to visualize the apical side of the enterocytes and villi structures (Figs 1A, 1B and S1). Although we collected numerous scraped villi from the jejunum of a total of six infected mice, detecting extracellular parasites or infected cells was challenging. We found single parasites in the jejunum, with diverse locations in the villi. We observed parasites on the luminal side of the enterocytes (Figs 1Bi and S1B), inside an enterocyte as single parasites close to the nucleus (S1C Fig), near cell-to-cell junctions -suggesting transmigration between adjacent cells- (S1D Fig), in the stroma (Figs 1Bii, S1E and S1F); and, within a cell, most likely an immune cell which is characterized by its rounded shape and multi-lobed nuclei (S1G Fig). We did not see evidence of parasite replication in any of the 3-day samples. We observed infrequent foci of infection after examining hundreds of scraped villi from each infected mouse (S1 Fig). We could not quantify the findings, given that the number of observed events were limited.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. T. gondii localization in mouse jejunal villi after ingesting cysts-containing brains. Jejunal localization of T. gondii after the natural route of infection.

(A) Schematic representation of (i) the villus and its structures and (ii) the experimental setup. Images represent single events captured in independent infected mice after 3 days post-ingestion. Single mouse ingested brains containing ~4 × 10² cysts. (B) Confocal images of infected villus with T. gondii after 3 days post-ingestion of cysts-containing brains. Each image represents single and isolated events captured in independent mice. T. gondii is localized in the intestinal lumen (i) or in the jejunal stroma (ii). In (B) the villi are stained for actin (red, rhodamine phalloidin), T. gondii (green, polyclonal antibodies), and nuclei (blue, DAPI). (Ci) Schematic representation of the experimental setup for detecting T. gondii infection by IVIS in the small intestine of mice fed cysts-containing brains. Single mouse ingested brains containing ~4 × 10²–~5 × 10² cysts. (Cii) Representative image of the distribution of T. gondii infection by IVIS in the small intestine of mice fed cysts-containing brains after 5 days post-ingestion. (D) Grouped scatter plot showing the number of infection spots detected in the small intestine of seven mice in two independent experiments. (E) Grouped scatter plot of luminescence flux measured in the small intestine of infected mice. Data in (D) and (E) represent the observation of the events registered in seven infected mice in a combination of two independent experiments, with values presented as means ± SD (*p ≤ 0.05). (F) Representative epifluorescent images of T. gondii infection in the jejunum of mice fed brain containing cysts at 3- and 5-days post-feeding. SAG1-positive parasites (tachyzoites) in the stroma. (G) Confocal 3D projection of tachyzoites replicating in the stroma. (H) Epifluorescent image showing the absence of BAG1 in a villus of the infected jejunum. Lines in yellow define the villus edges. (I) Grouped scatter plots showing the fluorescence intensity of SAG1 in infection spots (left plots) or the number of SAG1-positive PVs detected in either the stroma or enterocytes within the jejunum (right plots). (J) Quantification of the observed infected villi positive for SAG-1 parasites per section of intestine. (K) Representative of an epifluorescent image showing tachyzoites replicating within regions of excessive cell infiltration at 5 days post-feeding with brains. Intestines are stained for WGA (green), SAG1 (red), BAG1 (red), and nuclei (blue, DAPI). Data in (I) represents a combination of at least fifteen different intestinal structures, with values presented as means ± SD (***p ≤ 0.001, ****p ≤ 0.0001). Data in (J) represents a combination of at least 12 independent observations in different sections of the same or different infected mouse. Schematic representations in (A and C) were as created with BioRender.com.

https://doi.org/10.1371/journal.pntd.0012855.g001

Bioluminescence imaging reveals dynamic distribution of T. gondii infection in mice intestines

Challenges in detecting parasites and infected cells in the scraped intestines included loss of villi directionality, villi/crypt integrity, and possible stroma damage due to the mechanical force when scraping. These issues necessitated the development of alternative methods to study early infection. We therefore infected mice with PruΔHPT:luciferase strain to obtain brain cysts. After day 28 post-infection, we isolated the brains and fed mice with brain containing cysts (Fig 1Ci). Each mouse ingested an approximate dose of 4 × 10²–5 × 10² brain cysts. Then, we used the in vivo imaging system (IVIS) to actively detect parasite bioluminescence following injection of luciferin [37,38], identifying infection sites in the peritoneal cavity of the mice and in their isolated small intestines at 3- and 5-days post-feeding (Figs 1Cii and S2).

To localize regions of infection, we euthanized the mice and exposed their peritoneal cavities to access the intestine (Figs 1Cii and S2). We detected T. gondii infection in the intestines of mice at 5 days post-feeding (Fig 1Cii). The distribution of the infection manifested as evenly distributed bioluminescent spots in mice along the small intestine (Figs 1Cii and S2). We detected poor to no bioluminescence signal in small intestines at 3 days post-feeding, suggesting that the infection was still in the early stages with little to no replication (S2B and S2D Fig). At 5 days post-feeding, intestines showed infected areas predominantly in the jejunum (Fig 1D and 1E). After 5 days post-feeding, the infection levels in jejuna were higher than those observed in ilea (Fig 1D and 1E). No signal was detected in food and a piece of muscle isolated from the leg of infected mice, both used as a control (Figs 1Cii and S2). Upon isolating the mouse intestines, we observed enlarged lesions that were visible to the naked eye and colocalized with the bioluminescent infection foci detected by IVIS (S2B Fig). These lesions were consistently identifiable in all infected mice at 5 days post-infection and were absent in the intestines of uninfected controls.

Parasite distribution and replication dynamics in mouse intestinal tissues

To investigate parasite replication and stage conversion, we collected the bioluminescent areas in the jejunum and ileum at 3 and 5 days-post feeding for paraffin embedding (represented in Fig 1Ci). We stained the sections with antibodies against SAG1 and BAG1 to determine the parasite stages (Figs 1F, 1G, 1H, 1K, S3 and S4), or with polyclonal antibodies against T. gondii for a general overview of the infection (S5 and S4 Figs). Of note, we used wheat germ agglutinin to stain the intestinal membranes, as actin staining, which worked well for fresh or fixed villi, was ineffective for paraffin-embedded intestinal sections, possibly due to the long dehydration time during sample processing.

At 3 days post-feeding, we only observed SAG1-positive parasites in the jejunal stroma with the absence of BAG1-positive parasites (Figs 1F, 1G, 1H, S3A and S4A) and within the enterocytes (Fig 1F, 1G and 1H). Additionally, we detected single parasites or small parasitophorous vacuoles (PVs) near the muscular layer beneath the crypts (S4A Fig). Parasites located in the stroma were replicating, as evidenced by vacuoles containing more than one parasite (Figs 1G, S4B, S4C and S1 Movie). Scarce parasites were found replicating in the enterocytes (Fig 1I). Despite the ingestion of hundreds of cysts by the mice, infection foci were confined only to one or two villi per intestinal section in more than 95% of the sections analyzed (Fig 1J). We found evidence of parasites in the lamina propria and inside a rounded cell, which most likely are immune cells (S4A Fig). For the rest of the parasites, we cannot state whether they replicated in immune cells or in intestinal cells (S4A Fig).

At 5 days post-feeding, we noted a prominent area full of cells between the crypts and the muscular layer (Figs 1K and S5A) and detected SAG1-positive/BAG1-negative parasites replicating in the area (Figs 1K, S3B and S2 Movie). We confirmed the presence of immune cells in the bumped areas by staining against CD45 (S5B Fig). The enlarged areas were detected by the naked eye, colocalized with bioluminescent areas seen by IVIS and present in all the intestinal sections analyzed at 5 days post-ingestion. Hundred percent of the bumps were infected or surrounded by parasites. Our samples showed CD45-positive cells adjacent to infected host cells (S5B Fig); however, infected cells were CD45-negative which suggests that tachyzoite’s replication continues in intestinal cells even after 5 days-feeding (S5C Fig). CD45-positive cells were also found in uninfected intestines close to the crypts and in the villus (S5E Fig). We did not find any cell aggregate structures in uninfected intestines (S5A Fig), and we detected parasites close to endothelial tissue containing red blood cells (S5D Fig). An anti-CD31 antibody-stained erythrocytes (S5F Fig), but not the capillary network in the intestines (S2G and S5F Figs).

Optimization of culture conditions for bradyzoite differentiation

In vitro cultured bradyzoites have proven highly beneficial for studying bradyzoite biology, particularly due to the reduction in time and animals [39]. Our laboratory commonly grows tachyzoites at 33°C to slow parasite growth and reduce the number of parasite passages required for maintenance. We took advantage of slow replicating tachyzoites to induce bradyzoite formation using the widely used alkaline pH differentiation media [39]. The new differentiation condition includes culturing tachyzoites in media with a pH 8.1 at 33°C [39], in a thermostatic incubator under normoxic conditions with ambient CO₂ concentrations. We followed parasite replication in standard media (HFF media) and differentiation conditions using ME49 mCherry parasites at 33°C. ME49 mCherry parasites grown in HFF media lysed the host cells by day 8 (Figs 2A and S6A). In contrast, in the presence of differentiation media, parasites remained still intracellular after 8 days (Figs 2A and S6A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. In vitro bradyzoite differentiation of type I and II strains at 33°C.

(A) Plot represents the number of intact vacuoles of ME49 parasites in the presence of HFF media or differentiation media. Quantification was done in at least ten different fields after 6- and 8-days post infection. Images to the right show representative fluorescent imaging of infected HFF cells. (B) Plots to the left shows the quantification of positive vacuoles to SAG1, BAG1 or both in ME49 parasites after incubation for 6 days in differentiation media at 37°C or 33°C. Plots to the right show positive vacuoles to SAG1, BAG1 or both in type II and type I parasites after incubation for 12 days in differentiation media at 33°C. Images to the right show representative fluorescent imaging of infected HFF cells. (C) Plots to the left show the quantification of PVs positive to DBA after 7 days in differentiation media in ME49 and RH parasites. Images to the right show representative fluorescent images of DBA parasites (D) Plots to the left show the quantification of PVs positive to DBA after 8- and 12-days in differentiation media in ME49 and RH parasites. Images to the right show representative fluorescent images of DBA- or BAG1-positive type I parasites after 8- or 12-days in differentiation media. (E) Plots show the in vitro differentiation of EGS strain at 33°C and at 37°C. Plots show live fluorescent imaging of EGS parasites detecting mCherry-expressing tachyzoites or GFP-expressing bradyzoites. Quantifications were performed by quantifying the number of vacuoles (top plots) or by measuring the fluorescent area (bottom plots). (G) Survival curve of mice fed bread with a range of 1 × 10² – 1 × 10⁵ in vitro-generated bradyzoites. Three Swiss Webster mice were tested for each experimental condition.

https://doi.org/10.1371/journal.pntd.0012855.g002

Temperature played a large role in bradyzoite in vitro differentiation. ME49 parasites differentiated at 37°C only yielded ~30% of BAG1 positive parasites after 6 days in differentiation media; in contrast, almost a complete differentiation to bradyzoites (>95% BAG1 positive) were obtained by 33°C incubation in differentiation media (Figs 2B and S6B, BAG1/SAG1). Our new switching conditions yielded a stable high efficiency of bradyzoites differentiation (>97%) up to 12 days post-switching (Fig 2B, right graph). Dolichos biflorus agglutinin (DBA) decorated the cyst wall in 98% of ME49 parasites as soon as 7 days in differentiation media (Fig 2C) and remained still DBA-positive at 12 days in differentiation media (Fig 2D). Our conditions achieved 92–98% cysts formation in type II strains Pru and ME49 (Fig 2D). RH parasites rarely expressed BAG1 in pH 8.1 media at 37°C followed by rapid cell lysis (Fig 2B, left graph), but reduction of the temperature to 33°C generated ~90% of DBA positive vacuoles in RH parasites (Figs 2C and S6D).

To investigate real-time switching dynamics, we used the type I/III hybrid strain EGS LDH2p-GFP/SAG1p-mCherry parasites [40] and measured the expression of stage-specific fluorescent reporters to track the point of stage conversion (Fig 2E). While incubation of EGS at 37°C in differentiation media led to cell lysis after 3–4 days post-infection (Figs 2E and S7A), at 33°C complete bradyzoite differentiation occurred within 6 days in differentiation media, as evidenced by the presence of GFP/LDH2-positive cysts structures and the absence of SAG1/mCherry-positive parasites (Figs 2E and S7B). After 9 days in differentiation media, we detected free floating GFP-positive cyst structures, possibly due to host cells detaching from the substrate or cyst structures being expelled from the host cells (S7 Fig, 9 dpi). Given the consistent bradyzoite development across strains, we decided to harvest in vitro cysts between 7–9 days in differentiation media for all experimental procedures.

Evaluation of in vitro bradyzoites in host systems

In vitro bradyzoites differentiated at 33°C in switching media exhibited molecular and physiological features similar to brain cysts, such as: positive for DBA and BAG1, negative for SAG1, rounded shape structures, slow replication parasites, and resistance to pepsin digestion (addressed below). ME49 mCherry in vitro bradyzoites infected and replicated in both HFF and Caco-2 cells, indicating their viability after incubation at 33°C (S6F Fig). To assess the virulence of the in vitro cysts, we fed Swiss Webster mice with in vitro cysts on bread, mimicking the natural route of infection. One mouse fed ~1 × 10⁴ and three mice fed ~7 × 10⁴ ME49 mCherry in vitro-differentiated cysts developed chronic infection after 28 days, as evidenced by cyst isolated from the brain (S6G Fig). In contrast, mice fed ~1 × 10⁴ EGS in vitro cysts did not exhibit brain cysts formation or succumbed to the infection after 28 days post-feeding (Fig 2G). To evaluate the virulence of EGS in vitro cysts further, three mice per group were fed with EGS in vitro cysts ranging from ~1 × 10² to ~1 × 10⁵. Only mice fed with ~1 × 10⁵ in vitro cysts succumbed to the infection after 11–13 days (Fig 2G). Brain cysts were not detected in the remaining EGS-infected mice, likely due to the nature of this type I/III hybrid strain.

Bradyzoite activation depends on pepsin digestion

Our group has successfully used the microphysiological system (MPS) and Caco-2 cells to model T. gondii tachyzoite-specific interactions during the infection and study the innate immune response [35]. We turned to these devices to model early intestinal bradyzoite infection using Caco-2. MPS previously demonstrated confluent distribution throughout the lumen and along with the presence of mucus-producing goblet cells marker MUC2A, the tight junction protein ZO-1 (confirming the stability of the cells forming the lumen), and the microvilli marker villin [35]. We confirmed the homogeneous and close distribution of Caco-2 cells in the current study by actin staining (Fig 3Ai). Barrier functionality and luminal coverage were validated by the maintenance of FITC-dextran within the lumen (Fig 3Aii). We tested the viability of Caco-2 cells in the MPS by infecting them with RH mCherry tachyzoites and detecting active replication after 48 hours post-infection (Fig 3Aiii).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Bradyzoite activation for infection of human intestinal microphysiological systems to study bradyzoite interactions.

(A) Schematic representation of bradyzoite infection niche in human intestinal MPS. (i) Live staining of actin showing the coverage of the lumen by Caco-2 cells. (ii) FITC-dextran live imaging assessing the polarity of the Caco-2 cell line in the MPS. (iii) Parasite replication at 48 hours post-infection with 1 × 10³ RH tachyzoites as a control for host cell viability. (B) Representative fluorescent images of parasite replication in HFF cells infected with brain bradyzoites, either activated by pepsin or subjected to mechanical lysis, at 9 days post-infection. Graph shows parasite kinetics of growth in HFF for pepsin-activated versus inactivated bradyzoites. (C) Representative fluorescent images of replication in HFF, Caco-2, and HIEC-6 infected with activated bradyzoites at 7 days post-infection. Graph shows parasite kinetics of growth in the three different cell lines. Schematic representation in (A) was created with BioRender.com.

https://doi.org/10.1371/journal.pntd.0012855.g003

We infected Caco-2 lumens with in vitro-differentiated ME49 mCherry or EGS bradyzoites to study the early points of the infection. We simulated the oral infection by degrading the cyst structures with pepsin from porcine stomach for 2 min to exteriorize bradyzoites; however, we did not detect replicating parasites in any of our lumens after 5 or more days of infection (S8 and S9 Figs). This result was unexpected because bradyzoites from the same sample successfully infected and replicated in Caco-2 monolayers (S6B Fig). We confirmed that the in vitro bradyzoites were viable after 2 minutes of pepsin treatment through SYTOX green with 97% being viable (S8C Fig). We assessed Caco-2 cells viability and accessibility to T. gondii in lumens by infecting them with ME49 mCherry tachyzoites and observed replication after 3 days post-infection (S8D Fig), suggesting the defect in replication was specific to bradyzoites. We hypothesized that bradyzoites need to be activated by extending the treatment time with pepsin, as recently suggested [41].

We tested the role of pepsin in bradyzoite activation in 2D monolayers. We incubated ME49 mCherry brain cysts with pepsin from 2 min up to 2 h, and then monitored parasite replication over 9 days in HFF monolayers (Fig 3B). As a negative control of pepsin activation, we mechanically disrupted cysts by passing them through a 30-gauge needle. Bradyzoites stimulated with pepsin for 30 min displayed the highest replication, followed by those stimulated for 15 min and 1 h, which exhibited similar replication patterns (Fig 3B). Bradyzoites exposed to pepsin for 2 minutes and 2 hours showed lower levels of parasite replication (Fig 3B) as well as syringe-lysed cysts (Figs 3B and S10A). In vitro differentiated EGS bradyzoites were also effectively activated by incubating with pepsin for 30 minutes, compared to 2 minutes or syringe lysis (S10B Fig, top graph).

Optimization of human cell lines for use in microphysiological systems

Given that our MPS are seeded with Caco-2 cells, we ensured that bradyzoites infect Caco-2 cells under our new activation conditions. We compared 30 min pepsin-activated brain bradyzoite infection in Caco-2 vs HFF monolayers. HFF cells exhibited higher susceptibility to bradyzoite infection compared to a limited infection observed in Caco-2 cells (Figs 3C, S9A, S3 and S4 Movies). We then tested the human colon epithelial cell line (HIEC-6), which displayed high susceptibility to bradyzoite infection similar to HFF cells (S10A Fig and S5 Movie). However, the HIEC-6 cells both did not form uniform lumens in the MPS and migrated outside the lumen into the collagen I/fibronectin matrix (S6 Movie). As such, we used Caco-2 cells for generating human-intestine lumens, as they are more effective than HIEC-6 at accurately modeling luminal structures and maintaining epithelial integrity.

Modeling intestinal bradyzoite infection using human intestinal microphysiological systems

We examined bradyzoite-to-tachyzoite conversion in human intestinal MPS, by infecting Caco-2 lumens with brain-isolated ME49 mCherry bradyzoites (Fig 4A). At 3 days post-infection, parasites were fixed, mCherry fluorescence was quenched using methanol (S11 Fig), and then parasites were stained with SAG1 and BAG1 antibodies (Figs 4Ai-iii). Although most PVs were SAG1-positive, we also detected BAG1-positive PVs (Fig 4Ai). Notably, BAG1-positive PVs were often co-labeled with SAG1, indicating that they were converting to tachyzoites (Fig 4Ai-iii).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Bradyzoite differentiation to tachyzoite in human intestinal microphysiological systems.

(A) Fluorescent images showing replication of alive ME49 mCherry bradyzoites in the Caco-2 lumen. (i)(ii), and (iii) show immunofluorescent images of the parasite conversion of ME49 mCherry bradyzoites to tachyzoites, with markers for tachyzoites (SAG1, red) and bradyzoites (BAG1, green) at 3 days post-infection. (B) Fluorescent images of alive EGS LDH2p-GFP/SAG1p-mCherry parasites undergoing differentiation in MPS, with bradyzoites shown in green and tachyzoites in red. Punctuated lines represent the edges of the lumen and its boundary with the matrix. Lumens in (A) and (B) were infected with a total of 3 × 10⁴ bradyzoites ME49 mCherry or EGS LDH2p-GFP/SAG1p-mCherry parasites. Plots represent the quantification of mCherry positive PVs (expressing SAG1) from day 2- to 5-post infection as bradyzoites. Plots are graphed as the average of positive PVs per lumen. A total of seven plates were analyzed, each one with at least three infected lumens.

https://doi.org/10.1371/journal.pntd.0012855.g004

To monitor stage conversion in real-time, we infected Caco-2 lumens with in vitro bradyzoites of the EGS strain and quantified tachyzoite conversion by counting red PVs (mCherry/SAG1 expressing parasites). At 2 days post-infection, > 90% of PVs contained GFP-expressing bradyzoites (Fig 4B). Consistent with the ME49 strain, tachyzoite conversion was clearly detected by day 3 post-infection in 30% of vacuoles (Fig 4B). By day 4 post-infection, tachyzoites were present in 70% of PVs, and near complete tachyzoite conversion was evident by 5 days post-infection with only mCherry-expressing EGS parasites detected (Fig 4B). These findings mirror in vivo results where tachyzoites were detected in mouse intestinal sections at 3 days post-feeding (Fig 1). Our 2D models using HFF cells suggest that tachyzoite conversion from EGS bradyzoites occurs after approximately four replication cycles, as evidenced by the following the same PV for up to 4 days. Rapid tachyzoite replication was observed after 3 days post-infection, as shown in our Incucyte data obtained (S10C Fig).

Bradyzoites rapidly transmigrate the intestinal barrier without replicating in the enterocytes

During the early stages of bradyzoite infection in the MPS, we observed parasites localizing to the collagen I/fibronectin matrix within 20 hours post-infection. This finding suggests that bradyzoites may rapidly traverse the Caco-2 cell layer within the lumens (Fig 5A). To ensure the observed transmigration behavior reflects realistic biological processes in vivo and to confirm that these observations are not artifacts of the MPS system but reflect true bradyzoite behavior, we isolated mouse jejunal villi and infected them with in vitro-differentiated EGS bradyzoites. Our observational results showed that bradyzoites localize within the stroma in villi after 6 or 24 h post-infection (Fig 5Bi and 5Bii). We also observed bradyzoites migrating between enterocytes, indicating that the parasites may reach the stroma without necessarily infecting the enterocytes themselves (Fig 5Biii and 5Biv).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Bradyzoite transmigration of Caco-2 cells in human intestinal microphysiological systemsand intestinal jejunal villi explants.

(A) Representative fluorescent images showing bradyzoite transmigration through the Caco-2 cells after 20 h post-infection of lumens. Punctuated lines in white represent the edges of the lumen and its boundary with the matrix. (B) Confocal images of explants of villi infected for 6 h (i) or 24 h (ii-iv) and then fixed. (i) Shows EGS LDH2p-GFP/SAG1p-mCherry bradyzoites localizing in the jejunal stroma. (ii a) shows a 3D projection of a bradyzoite localizing in the stroma of villus. (ii b) represents a Z-section of (ii a). (iii) Shows a 3D projection of bradyzoites transmigrating through the enterocytes. (iv) Confocal image of a bradyzoite localizing in the stroma of villus. (C) Confocal images of explants of villi fixed and then infected for 6 h (i) or 24 h (ii-iv). (i) Shows EGS LDH2p-GFP/SAG1p-mCherry bradyzoites localizing underneath the enterocytes. (ii a) shows a 3D projection of a bradyzoite localizing in the stroma of the infected fixed villus. (ii b) represents a Z-section of (ii a). (iii) 3D projection of bradyzoites interacting with the enterocytes of the fixed villus. (iv) Confocal image of a bradyzoite localizing in the stroma of fixed villus.

https://doi.org/10.1371/journal.pntd.0012855.g005

To avoid the structural villi damage when scraping previously discussed (Figs 1 and S12), we opted to use fixed villi prior to bradyzoite infection. In fixed villi, we detected bradyzoites beneath the enterocytes after 6 h post-infection (Fig 5Ci) and an active bradyzoite internalization into the stroma (Fig 5Cii). As expected, we did not see bradyzoites within the cells, indicating that they cannot invade the fixed tissue. Consistent with observations in unfixed villi (Fig 5Biii), we found bradyzoites adjacent to areas of open enterocytes (Fig 5Ciii), suggesting a process of transient aperture of the enterocytes to reach the stroma (Fig 5iv).

Tachyzoites can migrate collagen and infect surrounding cells

In our intestinal sections, we observed both replicating and individual parasites within the stroma (Figs 1 and S1), but it was still unclear whether parasites migrated by entering immune cells or as extracellular parasites crossing biological barriers. The potential for migration in immune cells to reach the capillaries and organs appears highly feasible [42–45], but there is also evidence suggesting active parasite migration through the tissues [32]. To investigate whether parasites can disseminate without the aid of a host immune cell, we used an MPS with 2 lumens, each of which was seeded with Caco-2 cells with one lumen infected with in vitro EGS bradyzoites (Fig 6A) and the other left uninfected. We monitored the behavior of parasites exteriorized from Caco-2 cells throughout the experiment every 24 h. By 3 days post-infection, a mixture of tachyzoites and bradyzoites were detected in the lumen, with additional free bradyzoites observed in the collagen I/fibronectin matrix (Fig 6B). By 5 days post-infection, tachyzoites had migrated through the collagen I/fibronectin matrix and reached the adjacent uninfected Caco-2 lumen (Fig 6C). By 6 days post-infection, we observed replicating parasites in the uninfected lumen, indicating the ongoing expansion of the infection in vitro (Fig 6D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Active tachyzoite migration through 3D collagen I/fibronectin matrix and expansion of its infection in human intestinal MPS generated with Caco-2 double-lumens.

(A) Schematic representation of bradyzoite/tachyzoite infection niche in human intestinal MPS. (B) Fluorescent images showing the coexistence of EGS LDH2p-GFP/SAG1p-mCherry bradyzoites and tachyzoites in the infected intestinal lumen at 3 days post-infection, with insets highlighting the localization of bradyzoites in the 3D collagen I/fibronectin matrix. Fluorescent images tracking the migration in collagen and Caco-2 infection by EGS LDH2p-GFP/SAG1p-mCherry parasites after exteriorization from Caco-2 cells in the bottom Caco-2 lumen shown in (C) and (D). (C) Micrographs show tachyzoites localizing in the 3D collagen I/fibronectin matrix and actively migrating towards the top Caco-2 lumen by day 5 post-infection. (D) Right micrographs show tachyzoites expanding the infection to the adjacent top Caco-2 lumen. Punctuated circles show parasitic infection to the top Caco-2 lumen. Punctuated lines in white represent the edges of the lumen and its boundary with the matrix. Lumens were infected with 5 × 10⁴ bradyzoites of the EGS LDH2p-GFP/SAG1p-mCherry strain. Magnified photograph in (B) shows different parasites captured in different planes than those in unmagnified photograph. Prominent areas in (C) and (D) are the result from foci of infection. Schematic representation in (A) was created with BioRender.com.

https://doi.org/10.1371/journal.pntd.0012855.g006

Ethics statement

Animals were treated in compliance with the guidelines set by the Institutional Animal Care and Use Committee (IACUC) of the University of Wisconsin School of Medicine and Public Health (protocol #M005217), which adheres to the regulations and guidelines set by the National Research Council. The University of Wisconsin is accredited by the International Association for Assessment and Accreditation of Laboratory Animal Care.

Host cells lines cultivation

All cultures were maintained in a 37°C humidified 5% CO₂ incubator unless specified. Briefly, Human Foreskin Fibroblasts (HFF, ATCC SCRC-1041) and Human Colorectal Adenocarcinoma Cells (Caco-2, ATCC HTB-37) were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 1% penicillin-streptomycin. Human Intestinal Epithelial Cells (HIEC-6, ATCC CRL-3266) were cultured in OptiMEM 1 Reduced Serum Medium (Gibco) supplemented with 4% FBS, 10 mM GlutaMAX (Gibco), 20 mM HEPES (Gibco), and 10 ng/mL Epidermal Growth Factor (EGF). All cultures were periodically tested for mycoplasma and tested negative.

Experimental mice

Swiss Webster or C57BL/6 female and male mice, 8–12 weeks-old, were used for all the experiments. Infected or moribund mice were euthanized with CO₂.

Toxoplasma gondii tachyzoite cultivation

T. gondii tachyzoites were cultured in HFF cells and passed every two or three days. To avoid the appearance of any tissue culture specific characteristics, all experimental procedures were done using low passage parasites and used no longer than passage twenty. T. gondii strains: a) type I: RH ΔKU80ΔHPT, RH mCherry and RH GFP; b) type II: PruΔHPT:luciferase, ME49 ΔKU80DHPT (donated by Dr. S. Lourido), ME49 mCherry, and M4; and type I/III: EGS LDH2p-GFP/SAG1p-mCherry [40] (donated by Dr. L. Weiss).

Toxoplasma gondii in vitro bradyzoite cultivation

Briefly, HFF cells were infected with 5 × 10⁵ – 2 × 10⁶ tachyzoites for 2 h. Then, extracellular parasites were removed and differentiation media (RPMI 1640 without sodium bicarbonate (Sigma-Aldrich), 1% FBS, and 40 mM HEPES, at pH 8.1.) was added. Infected cells in differentiation media were incubated at 33°C for up to 12 days in a thermostatic incubator under normoxic conditions with ambient CO₂ concentrations. The flask cap was tightly closed to avoid gas exchange. Bradyzoite and cysts formation were confirmed by staining SAG1, BAG1, or DBA. In EGS parasites, bradyzoite differentiation was monitored in vivo by the expression of GFP and loss of mCherry by using EVOS FL Auto system (Life Technologies) scope or a live cell incubator (Incucyte, Sartorius). Quantification of in vitro encystment was performed by counting at least 100 positive or negative vacuoles. The statistics were performed using one-way ANOVA in the GraphPad Prism software.

Toxoplasma gondii in vivo encystment

To develop chronic infection and collect brain cysts, Swiss Webster or C57BL/6 mice were inoculated intraperitoneally (i.p.) with 1 × 10³ – 1 × 10⁴ of freshly egressed T. gondii tachyzoites of M4, PruΔHPT:luciferase, ME49 mCherry, or EGS LDH2p-GFP/SAG1p-mCherry parasite strains. Mice were euthanized 28 days post-infection, and the brains were removed to feed other mice or to isolate brain cysts.

Brain cysts isolation

Cysts were purified from infected mouse brains infected for at least 28 days. Brains were harvested and homogenized in ice-cold PBS. Brain material was passed through 18-, 20-, and 22-gauge needles three times to desegregate the brain. To purify cysts, brain material was centrifuged and resuspended in 30% of Percoll-90 (90% Percoll and 10% 1X PBS) in PBS. Cysts were pelleted by centrifugation at 1,000 × g for 15 min, washed with PBS, counted in a hemocytometer.

Survival curves of in vitro bradyzoite-infected mice

Three male Swiss Webster mice per dose were naturally infected with in vitro cysts of EGS strain: doses were 1 × 10², 1 × 10³, 1 × 10⁴, or 1 × 10⁵. For ME49 in vitro cysts, natural infection doses were 1 × 10⁴ (one mouse) or 7 × 10⁴ (three mice)”. Eight days in differentiation media for EGS and eleven days in differentiation media for ME49, monolayers containing in vitro cyst structures were scraped and lysed using a 27-gauge needle. In vitro cysts were quantified, adjusted to the corresponding burden, and centrifuged at 500 × g for 10 min. Pellets were resuspended in 100 µl of 2% sucrose and added into 0.5 cm³ fresh white bread cubes. Mice were fasted overnight [70] and consumed the bread containing in vitro cysts within a 1-hour window. ME49 mCherry in vitro cysts were used as a positive control of chronic infection. For both strains, mice health was monitored daily for health status. At 28 days post-infection, mice were euthanized, and their brains were collected and analyzed for the presence of cysts.

Villi isolation from infected mice

Swiss Webster mice were subcutaneously infected with oocysts of the M4 strain (type II, generously donated by Jeroen Saeji and David Arranz-Solís). Upon establishment of chronic infection, mice were euthanized, and their brains were isolated to release bradyzoites, following established protocols [55]. Bradyzoites were then propagated in HFF cells through one passage. Subsequently, 1 × 10⁴ tachyzoites of the M4 strain were infected in mice to develop chronic infection. At least four brains containing cysts were collected and served as food for two healthy mice, following the natural route of infection. Briefly, two mice were fasted overnight [70] and consumed the brains containing in vitro cysts within a 1-hour window. Each mouse was fed with two brains, each containing ~2 × 10² cysts per brain. The number of cysts were determined isolating brains from two infected mice from the same population, as above described. After feeding, mice were euthanized 3 days post-feeding, and their intestines were isolated. Sections of 6–8 cm of jejunum of infected mice was gently washed with ice-cold PBS, opened longitudinally, and mucus layer was removed by scraping superficially with a glass slide as previously described [33]. Villi were then detached by gentle scraping, collected in tubes, and fixed in 4% paraformaldehyde for 30 min. The fixed villi were centrifuged at 100 × g for 5 min, washed twice with PBS, and processed for immunofluorescence assay. Floating villi were placed in slides and imaged in a confocal microscope (ZeissLSM 800 Laser Scanning Microscope). This experiment was repeated three times to find foci of infection.

Bradyzoite excystation and pepsin activation

Bradyzoite exteriorization was performed by incubating purified cysts with digestion buffer containing NaCl and pepsin (5 mg/ml pepsin, 100 mg/ml NaCl, 0.14 N HCl, pH 2.1) as described previously [54,56]. Excystation was carried out for 2 min to 2 h. Digestion solution containing free bradyzoites was neutralized by addition of 1% Na₂CO₃ (v/v). Bradyzoites were counted using a hemocytometer, centrifuged at 2,000 × g for 10 min, and resuspended in the corresponding media.

Mice in vivo imaging system

To study T. gondii intestinal interactions followed by the natural route of infection, three males and four females Swiss Webster mice were infected by consuming brains containing cysts (refer to figure legends to see the mice number in each time point). Each mouse ingested two brains with a total approximated dose of 4 × 10^2–5 × 10² brain cysts. The number of brain cysts were determined by isolating brains from two infected mice from the same population, as above described. At 3 or 5 days post-infection, parasitemia in mice was detected by bioluminescence using the in vivo imaging system (IVIS, PerkinElmer) as previously reported [37]. Prior to imaging, mice were anesthetized with 4% isoflurane and then injected retroorbital and i.p. with 200 µL of D-Luciferin potassium salt (15.4 mg/mL in PBS). Immediately after the injection, mice were sacrificed, intestines were isolated and exposed to the IVIS detector, and images were collected. Background measurements were determined by injecting uninfected mice with D-Luciferin and collecting the images as previously described. Additional background controls included the imaging of food and pieces of muscle from infected mice. All conditions were repeated at least twice. For the analysis of infected areas, the small intestines was collected the duodenum, jejunum, and ileum which encompass 33 cm of length [71]. Duodenum was considered as the first 2 cm after the pyloric sphincter, jejunum constituted 15 cm after the end of the duodenum, and ileum constituted the last 15 cm before the cecum. Data quantification of the bioluminescent areas was performed by counting the number of events per intestinal region or by drawing a region of interest (ROI) box for each mouse and recording the total number of photons per second (Total Flux) [38].

Mice intestines immunohistochemistry

Areas of infection in intestines were detected by IVIS. Three to four infected sections of intestines of 5 mm long were cut and immediately fixed in 4% formalin in PBS overnight and dehydrated in 75% ethanol. Paraffin-embedding and sectioning of intestines were performed by request in the Translational Research Initiatives in Pathology (TRIPath lab, University of Wisconsin-Madison). Paraffin-embedded intestines were serial cut in at least 4 sections per slice to facilitate antibody comparisons. Infected sections of the intestine were processed at least three times for detecting parasites in different planes. Giving a total of at least 36 sections of foci of infection per condition. Intestinal sections were incubated at 65°C in a water bath for 1 h, washed with xylene 3 times for 10 min to remove the paraffin, twice for 1 min with 100% ethanol, twice with 95% ethanol in water, once with 75% ethanol in water, and finally in water. The epitopes were retrieved by exposure to the low-pressure setting in a pressure cooker for 6 min in urea buffer (1.25 M Tris Base, 1 M Urea, pH 9.5). Sections were blocked with 3% goat serum in 0.05 M TBS for 1 h. Primary antibody was incubated at 4°C overnight in 0.05 M TBS. Primary antibodies against- SAG1 (donated by John C Boothroyd,1:40, rabbit polyclonal antibody), BAG1 (donated by Louis Weiss, 1:40, rabbit polyclonal antibody), T. gondii (Invitrogen, 1:40, rabbit polyclonal antibody). CD31 (BD Pharmingen, 1:40, rat polyclonal antibody), and CD45 (BD Pharmingen, 1:40, rat polyclonal antibody). Dyes included: wheat germ agglutinin Alexa 488 (Invitrogen, 1:100) and DAPI (Millipore Sigma, 1:1000). Secondary antibodies against- rabbit or rat Alexa 594 (Thermo Fisher Scientific, 1:100). Washed 3 times with 0.05 M TBS and mounted in VECTASHIELD antifade mounting medium (VectorLabs). Each antibody was tested in at least twice biological replicates and by duplicate. Wheat germ agglutinin was preferred over phalloidin to counterstain the embedded tissue due to punctuated and unclear patterns detected in our samples when staining actin. Samples were imaged on a Zeiss Axioplan III (Imager.M2, Carl Zeiss) or a confocal microscope (ZeissLSM 800 Laser Scanning Microscope). 3D projections were reconstructed from z-stack sections using the Zen imaging software (Carl Zeiss).

Quantification of infected areas was conducted by measuring fluorescence intensity. Regions-of-Interest corresponding to the fluorescence signal of SAG1 in the stroma or enterocytes in infected intestines were identified. The integrated pixel intensity of the entire region-of-interest was measured, and background fluorescence was subtracted. Quantitative analysis of images was performed using ImageJ software as described previously [72]. In addition, a quantification of the location of SAG1-positive vacuoles in the stroma or enterocytes was quantified and graphed. The statistics were performed using one-way ANOVA in the GraphPad Prism software.

Ex-vivo infection of villi

Healthy C57BL/6 mice were euthanized, and their intestines were isolated on an ice-cold Petri dish with PBS. Intestines were gently washed with 1× Pen/Strep and 25 μg/ml gentamicin in sterile PBS. Intestines were opened longitudinally, and villi detached as above described. Villi were washed once in antibiotics/PBS solution, centrifuged at 100 × g for 5 min, and resuspended in Leibovitz L-15 media (Thermo Scientific 31415029) supplemented with 10% FBS and 1× GlutaMAX (Gibco). Villi suspension was placed in chamber slides (Thermo Scientific 177372PK) covered by 3 mg/ml rat tail collagen I (Col-I, Corning), 0.5 N sodium hydroxide (Thermo Fisher Scientific), 50 ng/ μl human fibronectin (Millipore), mixed with 7.5 pH 5× PBS, and complete DMEM medium. Villi were challenged with ~5 × 10⁵ of RH GFP tachyzoites or EGS LDH2p-GFP/SAG1p-mCherry in vitro bradyzoites for 6 h and 24 h. Samples were then fixed with 4% paraformaldehyde and processed for immunostaining. Viability assays were performed at the same time points by using SYTOX Green Nucleic Acid Stain (Thermo Scientific S7020). Samples were imaged on a Zeiss Axioplan III (Imager.M2, Carl Zeiss) or a confocal microscope (ZeissLSM 800 Laser Scanning Microscope). 3D projections were reconstructed from z-stack sections using the Zen imaging software (Carl Zeiss).

Parasite growth assays

To measure parasite growth, HFF, Caco-2, or HIEC-6 cells were seeded in 12- or 24- well plates. In vitro-differentiated or brain cysts-isolated bradyzoites of ME49 mCherry or in vitro-differentiated bradyzoites of EGS strain were isolated and excysted as above described using pepsin or by mechanical disruption using 30-gauge needle. Under-confluent cells were infected with 1 × 10⁴ – 1 × 10⁵ exteriorized bradyzoites. Plates were incubated in the Incucyte incubator for up to 9 days. For comparative assays, all cell types were seeded in the same plate, infected with the same parasites passage, and imaged and quantified under the same conditions as previously described [37,73]. Samples were quantified using the total mCherry area (µm²/image) as previously published [37,73]. The statistics were performed using one-way ANOVA in the GraphPad Prism software.

Generation of human intestinal microphysiological systems

MPS were fabricated using soft lithography techniques and filled with a collagen/fibronectin matrix to create a molded lumen that was seeded with a monolayer of Caco-2 cells to create an intestine mimic as previously described [35]. Briefly, the LumeNEXT device is created from two stacked PDMS layers, surrounding a removable PDMS rod, with separate inlet and outlet ports. After the main chamber is filled with extracellular matrix and polymerized, the PDMS rods are removed to form a molded lumen structure that can be seeded with cells. For specific features about MPS fabrication, please refer to Jimenez-Torres and collaborators [74].

Devices were UV-sterilized for 20 min. To promote matrix adhesion to PDMS, chambers were treated with 1% polyethylenimine (MilliporeSigma) in water solution for 10 min, followed by a treatment of 0.1% glutaraldehyde (MilliporeSigma) for 30 min, and then washed 5 times with water. ECM was prepared consisting of 3 mg/ml rat tail collagen I (Col-I, Corning), 0.5 N sodium hydroxide (Thermo Fisher Scientific), 50 µg/μl human fibronectin (Millipore), mixed with 7.5 pH 5× PBS, and complete DMEM medium. The pH of the ECM matrix was adjusted to pH 7.2 before loading the matrix into the central chamber of the device. ECM was polymerized at room temperature for 20 min and then moved to an incubator at 37°C for at least 1 hour. PDMS rods were then removed leaving molded lumen structures within the ECM gel that can be lined with cells.

To seed the lumens with epithelium, Caco-2 cells were resuspended at 20 × 10⁶ cells/ml of supplemented DMEM. 3 μl of cell suspension was introduced into the lumens through the inlet port. The MPS was incubated upside down for 30 min, then flipped and incubated for 1 h. Nonadherent cells were removed by washing with culture medium, and additional medium was added to the inlet/outlet ports. Devices were incubated for 24–48 hours to permit the formation of a confluent monolayer of cells that mimic human intestinal lumens. Intestinal lumens were infected with 1 × 10³ – 5 × 10⁴ exteriorized bradyzoites. For the quantification of tachyzoite conversion, seven plates were infected with 3 × 104 in vitro bradyzoites. EGS tachyzoites were switched to the conditions described above and maintained for at least 7 days. Bradyzoite conversion to tachyzoites was monitored in the device using the EVOS FL Auto system and assessed by detecting PVs positive to mCherry/SAG1. Results are shown as a percentage of the average of red vacuoles in each lumen per plate, 36 lumens total were quantified. An additional staining for cell coverture in the lumens was tracked by actin staining using fluorescein phalloidin (Invitrogen). Barrier stability of lumens was tested using 2 mg/ml of 4 kDa FITC-Dextran (Sigma) and monitoring dextran diffusion.

Immunofluorescence staining

For immunoassays, infected monolayers were processed following modified protocol based on previous works [47,75]. DBA, SAG1, or BAG1 positive vacuoles, infected monolayers were fixed with 4% paraformaldehyde for 30 min or ice-cold methanol for 20 min. Methanol was also used to quench mCherry or GFP fluorescence in parasites. Fixed monolayers were permeabilized with 0.5% Triton X-100 for 5 min and blocked with 3% BSA and 0.2% Triton X-100 for 30 minutes. For stating the cysts structures, we used Dolichos biflorus Agglutinin (DBA) which specifically binds the mucin domain of N-acetyl-galactosamine of the CST1 glycoprotein in the cyst wall [76]. DBA was used coupled to rhodamine (Vector laboratories) or fluorescein (Vector laboratories) was diluted in 3% BSA and 0.2% Triton X-100 and incubated for at least 1 h.

For immunostaining infected villi, fixed tissue was processed as previously described [37]. Fixed villi were permeabilized with 0.5% Triton X-100 for 5 min and blocked with 5% BSA for 30 min. Villi were incubated with polyclonal antibodies against- T. gondii (Invitrogen, 1:500, rabbit), overnight at 4˚C. Secondary antibody Alexa Fluor 488 (Thermo Scientific, 1:1,000, goat anti-rabbit) were incubated for 2 h at room temperature. For GFP or mCherry parasites no antibodies were used for detecting parasites. Villi were counterstained with rhodamine phalloidin (Cytoskeleton, Inc) or fluorescein phalloidin (Invitrogen), DAPI, and mounted using VECTASHIELD. Villi were imaged using a confocal microscope (ZeissLSM 800 Laser Scanning Microscope).

Immunostaining of parasites in MPS followed a modified protocol based on a previous report [35]. Briefly, cells were fixed with 4% paraformaldehyde for 20 min and permeabilized with 0.2% Triton X-100 (Fisher Scientific) for 10 min or ice-cold methanol for 20 min to quench mCherry fluorescence in parasites. Blocking was performed overnight at 4°C in a buffer solution (3% BSA and 0.1% Tween 20 (Thermo Fisher Scientific). Primary antibodies against- SAG1 (1:40, monoclonal mouse antibody DG52) or BAG1 (1:40, rabbit polyclonal antibodies) in buffer solution were added to the infected lumens and incubated at 4°C for 2 days. Secondary antibodies: rabbit Alexa Fluor 488 (Thermo Fisher Scientific, 1:100) and mouse Alexa Fluor 594 (Thermo Fisher Scientific, 1:100) were added to the buffer solution and incubated for 1 day at 4°C. Additionally, viability of cells in lumens were tracked by using CellMask Green Actin (Invitrogen), counterstained with DAPI, and imaged using a Nikon AXR Confocal Microscope.

Statistical analysis

Data were analyzed (Prism 9.0; GraphPad Software). Statistical significance was assessed using nonparametric Student’s t tests when comparing two conditions/groups, and when comparing more than two groups, significance was assessed using one-way analysis of variance (ANOVA).