Introduction
Cryptosporidium apicomplexan protists known to be causative agents of diarrhea in animals, have emerged as major causes of diarrhea in humans.1 Cryptosporidium species are able to cause self-limiting diarrhea in immunocompetent adults or life-threatening diarrhea in immune suppressed persons, particularly in acquired immunodeficiency syndrome (AIDS) patients.2 Within the genus Cryptosporidium, many morphologically similar species have been reported, but with varying degrees of disease severity according to host susceptibility to infection and to diversity in isolate pathogenicity.3 Interestingly, Cryptosporidium species have been related to carcinogenesis. The association of cryptosporidiosis and colonic adenocarcinoma was speculated in the case of a Spanish patient carrying both, who died rapidly after the onset of symptoms.4 More recently, an epidemiologic study in Poland reported a high frequency of cryptosporidiosis in patients with colorectal cancer.5 However, in these reports it was unclear if Cryptosporidium sp. behaved as a carcinogenesis factor or simply as an opportunistic agent whose development was enhanced by host immunosuppression. In this work, we described the potential role of Cryptosporidium parvum in the development of colon neoplasia in experimentally infected severe combined immunodeficiency (SCID) mice treated with dexamethasone.6 Herein, the ability of C. parvum to induce neoplastic changes could be established experimentally. In addition, we showed that deep immunosuppression alone did not entail neoplasia in uninfected hosts. Taking into account that C. parvum is able to modulate apoptosis of intestinal cells during its replication,7 a mechanism known to be involved in the carcinogenesis processes,8 and that SCID mice developed a chronic C. parvum infection with dose-dependent pathologic effects,9 we further studied C. parvum-induced gastrointestinal neoplasia using variable doses of C. parvum inocula and the cell proliferation marker Ki-67. The Ki-67 nuclear antigen has been shown to be expressed in proliferating cells during G1, S, G2, and mitosis stages of the cell cycle. In the normal gastrointestinal epithelium, Ki-67 is expressed in the nuclei of replicating cells around the base of the crypts; in dysplastic epithelia and advanced neoplasms, Ki-67 expresses a markedly abnormal pattern of proliferation in the middle and upper zones of the crypts.10
Materials and Methods
Overall study design and procedures.
The study targeted the ability of different inoculum sizes of C. parvum oocysts to induce gastrointestinal neoplastic changes in dexamethasone-treated or untreated SCID mice. Follow-up included evaluation of infection intensity (according to oocyst shedding counts and semi-quantification of parasites in the tissues, see below), histology, immunohistochemistry, and semi-quantitative as-sessing of the severity of lesions. The SCID mice are susceptible to Cryptosporidium infection and develop chronic disease caused by their defect in T and B lymphocytes,9 but dexamethasone treatment can also inhibit the priming of the innate immune response and interferon-γ (IFN-γ)-regulated gene expression.11
Animals and experimental design.
Cryptosporidium parvum IOWA oocysts were purchased from Waterborne TM, Inc. (New Orleans, LA). The suspension of oocysts was stored in a solution containing phosphate buffered saline (PBS), penicillin, streptomycin, gentamicin, amphotericin B, 0.001% Tween 20. Infective doses were prepared and were inoculated by oral–gastric gavages using 18–20 gauge feeding tubes to 7-week-old female CB17-SCID mice obtained from a colony bred at Pasteur Institute of Lille (France). Oocyst viability before inoculation was determined by a trypsin-taurocholate excystation test12 and absence of bacteria or fungi was assured by testing the oocyst suspensions on plate count agar (37°C, 1 week) and on Sabouraud plates (37°C, 1 week). When needed, animals were administered with Dexamethasone sodium phosphate (4 mg/L drinking water) (Dex) (Merck, Lyon, France). Dex administration started 2 weeks before inoculation and was maintained during the whole experimentation as previously described.6 Thirty-two SCID mice were randomly divided into eight groups of four in capped cages according to the dose of C. parvum oocysts inoculated and to the administration or not of Dex. Four groups of mice without Dex were inoculated by oral route with 105, 106, 107, or 108 oocysts, respectively, in 200 μL of PBS. The other four groups were inoculated similarly but they received oral Dex as explained previously (Table 1). Eight SCID mice constituted two control groups: four mice received oral Dex and were inoculated with PBS (CDex group), and four mice received Dex and an inoculum from which oocysts (equivalent to 106) were removed by filtration with a Nanosep MF tube (Pall Corporation, Port Washington, NY) with a 0.45 µm pore-size membrane (CDV group). On Days 20, 35, 46, and 57 post-inoculation (PI), one mouse from each group (including CDex and CDV) was euthanatized by a sodium pentobarbital (Ceva, Libourne, France) intra-cardiac injection. The rationale for selecting dates for euthanasia was based on our previous results.6 The SCID mouse colony source of the mice used in this work is regularly evaluated for the absence of microbial or parasitologic infections, including Helicobacter. Experiments were performed according to the European guidelines (Council directives on the protection of animals for experimental and other scientific purposes. J. Off. Communautés Européennes, 86/609/EEC, 18 December 1986, L358).
Collecting mouse fecal specimens.
Specimens were collected from the first day post-infection (PI) until the end of the experiment. Every 1–3 days each group of animals (initially four per group) was transferred to a clean cage during 30 to 60 minutes. Freshly collected fecal pellets (between 5 and 12 pellets) at each time point from each cage were placed in 1.5 mL conical tubes, weighed, and suspended in 500 to 1,000 μL deionized (Milli-Q, Millipore Corp.) water.
Processing of mouse fecal specimens and quantification of the oocyst shedding.
The feces were homogenized by extensive vortexing. To concentrate the oocysts, immunomagnetic separation (IMS) was done using Dynabeads anti-Cryptosporidium kit (Invitrogen, Cergy-Pontoise, France) according to the supplier recommendations. Briefly, 300 μL collected from each pool of fecal specimens were incubated with 100 μL of the Dynabeads anti-Cryptosporidium. The samples were then placed in a magnetic particle concentrator (MPC-1) to separate the bead-oocyst complex from the contaminating debris. The beads were resuspended and transferred into a 1.5 mL tube, and separated by using a magnetic particle concentrator (MPC-2). To dissociate the bead-oocyst complex 50 μL of 0.1 N HCl was used. Afterward, 30 μL of the purified oocyst suspension were placed in each well of a 10-well slide (Thermo Scientific, Portsmouth, NH) and allowed to dry. Slides were incubated at 37°C for one hour, fixed with methanol, and processed for immunofluorescence (IF) using an FITC conjugate anti-Cryptosporidium monoclonal antibody (Cellabs Pty. Ldt., Croissy-Beaubourg, France). The whole surface of each well was examined, the average number of fluorescing oocysts identified in 10 randomly selected microscopic fields at a magnification of ×400 was recorded, and the number of oocysts per milligram of pooled feces was calculated. The oocyst number per milligram of feces from all the mice in each cage was obtained. The mouse number in each cage decreased from 4 to 3 to 2, and finally to 1; therefore, the number of oocysts shown in Table 1 did not necessarily represent the specific shedding of each animal. It was assumed that the IMS technique using the Dynabeads anti-Cryptosopridium kit captured 100% of the oocysts in the suspension of feces.
Histologic examination.
Stomach, liver, pancreas, duodenum, samples of proximal, medium and distal parts of jejunum, ileocaecal region, colon, and lungs were removed from each mouse, fixed in 10% neutral formalin and embedded in paraffin. Sections 5 μm thick were either stained with hematoxylin and eosin (H&E) stains or used in immunohistochemistry (see below). Several stained sections of each solid organ and many longitudinal and transversal sections of gastrointestinal organs were examined microscopically using a Leica DMRB microscope equipped with a Leica digital camera connected to an Imaging Research MCID analysis system (MCID software, Cambridge, UK). The presence of parasites in the tissues was scored as follows13: 0, no parasites; 1+, small number of parasites focally distributed; 2+, moderate number of parasites widely distributed; 3+, abundant presence of parasites widely distributed throughout the tissue. Pathologic changes were classified according to the Nomenclature for Histologic Assessment of Intestinal Tumors in the Rodent.14 The importance of lesions was scored at the following sites: stomach, liver, duodenum, ileocaecal region, and colon. Lesions were scored as follows: 0, no lesion; 1, inflammation; 2, low grade intraepithelial neoplasia; 3, high grade intraepithelial neoplasia or adenoma with high-grade dysplasia (AHGD). The degree of severity of histologic damage for each mouse was calculated by the sum of individual scores over the five organs. This rating score, which was developed specifically to this work as no neoplastic lesions were reported before in Cryptosporidium infection, was based in the rationale of scoring for intestinal inflammation used previously.15
Immunohistochemistry.
The expression of Ki-67 antigen in mouse intestinal sections was assessed using a monoclonal rat anti-mouse Ki-67 antibody (dilution 1:25) (M7249, Dako, Denmark), following the procedure recommended by the supplier. Briefly, tissue sections were deparaffinized, rehydrated, and treated with a pH 6 citrate buffer for 40 minutes at 99°C. After incubation with the primary antibody, and after the secondary biotinylated rabbit anti-rat antibody (E0468, Dako)/streptavidin steps, the visualization was carried out by using DAB+ (Dako) as chromogen. The Ki-67 index was scored by counting (400× magnification) the total number of epithelial cell nuclei and that of Ki-67 immunostained nuclei in 10 randomly selected representative areas.16 Above all, rate and topography of Ki-67 immunostaining were carefully compared with the intestinal sections of control mice (Dex-treated or untreated uninfected SCID mice). For the calculation of the Ki-67 index, the areas were chosen in the sections where dysplastic lesions were identified. The Ki-67 index was scored according to the number of cells with positive nuclear staining per 1,000 cells in each area: +1, less than 10% of cells with positive nuclei; +2, 10–50% with positive nuclei; and +3, more than 50% of cells with positive nuclei and/or presence of proliferation in the upper third of the crypts and in the surface epithelium.10,17
Statistical analysis.
An analysis of variance (ANOVA) was conducted to account for the effects of relevant factors (inocula, Dex administration, day PI) and their interactions on average daily oocyst excretion. Data analysis was performed with the statistical software S-PLUS 2000 (MathSoft, Seattle, WA). Logarithmic transformations of oocyst excretion values were used. An average amount of oocysts excreted per day and per group of mice was estimated from all the unitary values. As stated previously, the amount of oocysts/mg of feces was estimated per day and per group of mice, with the number of mice per group decreasing over time. Significance was defined as P ≤ 0.05.
Results
Influence of C. parvum inoculum size and of dexamethasone administration on the intensity of oocyst shedding.
The shedding of oocysts in the feces determined by IMS-IF was positive in all groups after the first day of collection PI until the end of the study. However, the median oocyst excretion during the experiments increased relative to inoculum size (in the range between 105 and 107 oocysts), and was substantially higher in Dex-treated mice than in the untreated mice (Figure 1). The highest oocyst excretion was observed in Dex-treated SCID mice inoculated with 107 oocysts. Unexpectedly, SCID mice inoculated with 108 oocysts (especially in Dex-treated animals), shed lower oocyst numbers compared with mice inoculated with lower oocyst doses (Figure 1). Additionally, ANOVA of the whole dataset showed that the day PI, the administration of Dex, and the interaction between Dex and the inoculum size, significantly influenced the oocyst excretion (P < 0.001, P = 0.005, P = 0.05, respectively).
Daily Cryptosporidium oocyst excretion (natural logarithm of oocyst/mg feces) in different groups of severe combined immunodeficiency (SCID) mice over the 57 days of the study. Each box represents the median 50% of data, white line is the median. Whiskers represent the extreme values within 1.5 times the box height. Extra lines above and below the whiskers are the outliers.
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Daily Cryptosporidium oocyst excretion (natural logarithm of oocyst/mg feces) in different groups of severe combined immunodeficiency (SCID) mice over the 57 days of the study. Each box represents the median 50% of data, white line is the median. Whiskers represent the extreme values within 1.5 times the box height. Extra lines above and below the whiskers are the outliers.
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Daily Cryptosporidium oocyst excretion (natural logarithm of oocyst/mg feces) in different groups of severe combined immunodeficiency (SCID) mice over the 57 days of the study. Each box represents the median 50% of data, white line is the median. Whiskers represent the extreme values within 1.5 times the box height. Extra lines above and below the whiskers are the outliers.
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Histologic alterations.
In Dex-untreated SCID mice, regardless of the C. parvum inoculum size, mild histologic lesions were mostly observed in the stomach, liver, and duodenum (Table 1). In 8/16 Dex-untreated SCID mice (50%), parasites were detected in the antro-pyloric region in association with a slightly modified mucosa including the presence of glandular tubular structures with cellular atypias. These mucosal changes were found to be typical of low-grade intraepithelial neoplasia, and appeared as early as Day 35 PI with the highest inocula (107 and 108) (Table 1). In the liver of 13/16 (81.3%) Dex-untreated SCID mice, mild periportal inflammation, associated in some cases with parasite invasion of intrahepatic bile ducts, were detected. Interestingly, neither parasites nor inflammatory nor other lesions were detected in the large intestine (including caecum) of Dex-untreated SCID mice. When severity of histologic lesions was scored, Dex-untreated mice exhibited scores ≤ 4 (Table 1).
With regard to Dex-treated SCID mice, on the whole, they showed more extended and severe lesions than the untreated ones (Table 1). In 6/16 animals (37.5%) (mice nos. 7, 8, 15, 16, 24, and 32), lesions were detected in more than one organ with a maximum score of histologic severity of 11 in mouse no. 16 (Table 1). Specifically, 3 mice (nos. 24, 30, and 32) had stomach lesions suggestive of low-grade intraepithelial neoplasia, and one (no 16) presented high-grade intraepithelial neoplasia characterized by an increasing architectural distortion associating glandular crowding, prominent cellular atypias, and pseudostratified nuclei, without stromal invasion (Figure 2 A and B). Mice nos. 8 and 32 (Table 1) had low- and high-grade intraepithelial neoplasia in the duodenum, respectively. In the ileocaecal regions, 6/16 Dex-treated SCID mice (37.5%) (nos. 7, 8, 15, 16, 23, and 24) (Table 1) presented adenomas with high-grade dysplasia after Day 46 PI (Figure 2 C and D, Figure 3), coinciding with the period of highest oocyst shedding. Three mice (18.8%) (nos. 7, 15, and 16) (Table 1) had also this kind of lesion but located in the colon. Lesions in the large intestine of Dex-treated SCID mice were characterized by adenomatous masses (Figure 3) that appeared closely packed, branching sometimes dilated tubular structures, separated by normal or inflammatory lamina propria. Focal cystic dilation was observed (Figure 2C). Some tubules were covered by a low- or high-grade dysplastic epithelium, which showed mucin depletion and nuclear stratification. In some areas, architectural distortion was associated with cellular atypias. Epithelial cells showed loss of normal polarity. As well, abnormal nuclear changes consisting of prominent nucleoli and irregularly scattered chromatin were recorded. Foci of merged glands typical of high-grade dysplasia invading into lamina propria were found (Figure 2).
Dex-treated severe combined immunodeficiency (SCID) mice infected with 106 oocysts of C. parvum (euthanasia Day 57 post-infection [PI]). A, Antro-pyloric region showing high-grade intraepithelial neoplasia characterized by an architectural distortion of glands (Bar = 80 μm). B, Antro-pyloric region showing cellular atypias, loss of normal polarity of epithelial cells, and presence of numerous parasites inside the glands (Bar = 15 μm). C, Ileocaecal region showing a projection of a polypoid structure with focal cystic dilation developing inside the intestinal lumen (Bar = 200 μm). D, Areas of high-grade dysplasia (arrow) inside a neoplastic polypoid structure with numerous parasites (arrowhead) (Bar = 15 μm), Hematoxylin & Eosin.
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Dex-treated severe combined immunodeficiency (SCID) mice infected with 106 oocysts of C. parvum (euthanasia Day 57 post-infection [PI]). A, Antro-pyloric region showing high-grade intraepithelial neoplasia characterized by an architectural distortion of glands (Bar = 80 μm). B, Antro-pyloric region showing cellular atypias, loss of normal polarity of epithelial cells, and presence of numerous parasites inside the glands (Bar = 15 μm). C, Ileocaecal region showing a projection of a polypoid structure with focal cystic dilation developing inside the intestinal lumen (Bar = 200 μm). D, Areas of high-grade dysplasia (arrow) inside a neoplastic polypoid structure with numerous parasites (arrowhead) (Bar = 15 μm), Hematoxylin & Eosin.
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Dex-treated severe combined immunodeficiency (SCID) mice infected with 106 oocysts of C. parvum (euthanasia Day 57 post-infection [PI]). A, Antro-pyloric region showing high-grade intraepithelial neoplasia characterized by an architectural distortion of glands (Bar = 80 μm). B, Antro-pyloric region showing cellular atypias, loss of normal polarity of epithelial cells, and presence of numerous parasites inside the glands (Bar = 15 μm). C, Ileocaecal region showing a projection of a polypoid structure with focal cystic dilation developing inside the intestinal lumen (Bar = 200 μm). D, Areas of high-grade dysplasia (arrow) inside a neoplastic polypoid structure with numerous parasites (arrowhead) (Bar = 15 μm), Hematoxylin & Eosin.
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Ileocaecal region of a severe combined immunodeficiency (SCID) mouse infected with 105 oocysts of C. parvum and euthanatized after Day 45 PI. Arrows show the presence of adenomatous masses in the intestinal lumen (Bar = 1,000 µm).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Ileocaecal region of a severe combined immunodeficiency (SCID) mouse infected with 105 oocysts of C. parvum and euthanatized after Day 45 PI. Arrows show the presence of adenomatous masses in the intestinal lumen (Bar = 1,000 µm).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Ileocaecal region of a severe combined immunodeficiency (SCID) mouse infected with 105 oocysts of C. parvum and euthanatized after Day 45 PI. Arrows show the presence of adenomatous masses in the intestinal lumen (Bar = 1,000 µm).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
On the whole, ileocaecal neoplastic lesions were found after Day 46 PI and only in Dex-treated mice (Table 1). However, gastric dysplastic lesions were observed as early as Day 35 PI (Table 1) in both Dex-untreated (nos. 18 and 26) and Dex-treated mice (no. 30). Hepatobiliary inflammation was more frequent in Dex-untreated mice (13/16 versus 3/16) and it was observed with parasite colonization in two Dex-untreated mice (nos. 3 and 17).
A significant correlation (Figure 4) was observed between the histologic lesion severity scores and oocyst excretion rate (r: 0.622, P < 0.001). Consistently, most severe lesions were systematically associated with high parasite scores (all life cycle stages) in the tissues (Table 1). Thus, for lesions with severity scores of 2 or 3, the presence of parasites in the tissues was always either moderate or abundant, respectively, and this correlation was highly significant (r = 0.83, P < 0.001).
Correlation between the severity of histologic lesions and the intensity of daily oocyst excretion. Each point represents one mouse. Circles: Dex-untreated mice, Triangles: Dex-treated mice. Note that most animals euthanatized between Days 20 and 35 post-infection [PI] have a score of severity < 2 (see Materials and Methods). All animals euthanatized after Day 35 PI had a severity score ≥ 2 (= neoplastic lesions). There was a significant correlation between severity score and oocyst excretion (r = 0,622, P < 0.001).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Correlation between the severity of histologic lesions and the intensity of daily oocyst excretion. Each point represents one mouse. Circles: Dex-untreated mice, Triangles: Dex-treated mice. Note that most animals euthanatized between Days 20 and 35 post-infection [PI] have a score of severity < 2 (see Materials and Methods). All animals euthanatized after Day 35 PI had a severity score ≥ 2 (= neoplastic lesions). There was a significant correlation between severity score and oocyst excretion (r = 0,622, P < 0.001).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Correlation between the severity of histologic lesions and the intensity of daily oocyst excretion. Each point represents one mouse. Circles: Dex-untreated mice, Triangles: Dex-treated mice. Note that most animals euthanatized between Days 20 and 35 post-infection [PI] have a score of severity < 2 (see Materials and Methods). All animals euthanatized after Day 35 PI had a severity score ≥ 2 (= neoplastic lesions). There was a significant correlation between severity score and oocyst excretion (r = 0,622, P < 0.001).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
At the histologic examination of the jejunum, a small number of parasites attached to the epithelial cells and villus atrophy and crypt hyperplasia were found without significant differences between groups of Dex-untreated or treated SCID mice. These infections were not associated with dysplastic changes. No parasites or morphologic changes were observed in other organs, such as lungs or pancreas. Mice of control groups did not develop Cryptosporidium infection or histologic gastrointestinal lesions.
Immunohistochemical assessment of Ki-67 antigen expression.
An apparent increase of the mitosis number was ob-served in H&E stained sections of gastric (Dex-treated or untreated mice) or ileocaecal (Dex-treated mice) epithelia between Days 20 and 35 PI, before histologic evidence of neoplasia. To confirm that the dysplastic compartment at the top of the crypts represented abnormally proliferating cells, we stained the sections with an anti-Ki-67-specific antibody. In control or low infected animals euthanatized at Day 20 PI, Ki-67 staining was observed at the lower third of gastric (data not shown) or ileocaecal epithelium crypts (Figure 5A–D). Among infected animals, the extension of Ki-67 staining increased progressively after 20 days PI, especially in mice with higher parasite loads, leading to proliferation in the upper third of the crypts and in the surface epithelium after 35 days PI (Figure 5E–H). Thus, immunohistochemical results using Ki-67 indicate that the alteration in cell proliferation occurs before the histologic diagnosis, which was currently established at Day 35 PI in the stomach, and at Day 46 PI in the ileocaecal region (Table 2).
Expression of Ki-67 in the epithelium of Dex-treated severe combined immunodeficiency (SCID) mice infected with C. parvum. A, Dex-treated uninfected mouse (control) euthanatized at Day 57 post-infection (PI): Ki-67 immunoreactive nuclei are observed at the lower third of the crypts (2–6 immunoreactive cells per glandular section) (Bar = 100 µm). B, Detail of A (delimited area) (Bar = 50 µm). C, Infected mouse euthanatized at Day 20 PI: Ki-67 immunoreactive cells confined to the lower third of the crypts (7–10 immunoreactive cells per glandular section) (Bar = 100 µm). D, Detail of C (delimited area) (Bar = 50 µm). E, Infected mouse euthanatized at Day 35 PI: increased extension of the staining (20–30 immunoreactive cells per glandular section) (Bar = 100 µm). F, Detail of E (delimited area). Proliferating cells are observed in the upper third of the crypts and in the surface epithelium (Bar = 50 µm). G, Infected mouse euthanatized at Day 57 PI: Ki-67 immunoreactive cells are present all over the crypt including the surface epithelium. The architecture of the mucosa is highly altered by the neoplastic process (Bar = 100 µm). H, Detail of G (delimited area) (Bar = 50 µm).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Expression of Ki-67 in the epithelium of Dex-treated severe combined immunodeficiency (SCID) mice infected with C. parvum. A, Dex-treated uninfected mouse (control) euthanatized at Day 57 post-infection (PI): Ki-67 immunoreactive nuclei are observed at the lower third of the crypts (2–6 immunoreactive cells per glandular section) (Bar = 100 µm). B, Detail of A (delimited area) (Bar = 50 µm). C, Infected mouse euthanatized at Day 20 PI: Ki-67 immunoreactive cells confined to the lower third of the crypts (7–10 immunoreactive cells per glandular section) (Bar = 100 µm). D, Detail of C (delimited area) (Bar = 50 µm). E, Infected mouse euthanatized at Day 35 PI: increased extension of the staining (20–30 immunoreactive cells per glandular section) (Bar = 100 µm). F, Detail of E (delimited area). Proliferating cells are observed in the upper third of the crypts and in the surface epithelium (Bar = 50 µm). G, Infected mouse euthanatized at Day 57 PI: Ki-67 immunoreactive cells are present all over the crypt including the surface epithelium. The architecture of the mucosa is highly altered by the neoplastic process (Bar = 100 µm). H, Detail of G (delimited area) (Bar = 50 µm).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Expression of Ki-67 in the epithelium of Dex-treated severe combined immunodeficiency (SCID) mice infected with C. parvum. A, Dex-treated uninfected mouse (control) euthanatized at Day 57 post-infection (PI): Ki-67 immunoreactive nuclei are observed at the lower third of the crypts (2–6 immunoreactive cells per glandular section) (Bar = 100 µm). B, Detail of A (delimited area) (Bar = 50 µm). C, Infected mouse euthanatized at Day 20 PI: Ki-67 immunoreactive cells confined to the lower third of the crypts (7–10 immunoreactive cells per glandular section) (Bar = 100 µm). D, Detail of C (delimited area) (Bar = 50 µm). E, Infected mouse euthanatized at Day 35 PI: increased extension of the staining (20–30 immunoreactive cells per glandular section) (Bar = 100 µm). F, Detail of E (delimited area). Proliferating cells are observed in the upper third of the crypts and in the surface epithelium (Bar = 50 µm). G, Infected mouse euthanatized at Day 57 PI: Ki-67 immunoreactive cells are present all over the crypt including the surface epithelium. The architecture of the mucosa is highly altered by the neoplastic process (Bar = 100 µm). H, Detail of G (delimited area) (Bar = 50 µm).
Citation: The American Society of Tropical Medicine and Hygiene 82, 2; 10.4269/ajtmh.2010.09-0309
Occurrence of gastrointestinal neoplasia.
On the whole, gastrointestinal neoplastic lesions were observed in mice euthanatized at Day 46 or after in both Dex-untreated and Dex-treated SCID mice (Table 1). Gastric intraepithelial neoplasia of low- or high-grade was more frequent in Dex-untreated SCID mice (6/8, 75%) than in Dex-treated animals (3/8, 37.5%). In contrast, only Dex-treated SCID mice developed duodenal (2/8, 25%) or colonic (3/8, 37.5%) intraepithelial neoplasia (of low or high grade).
With regard to the ileocaecal region (Table 1), only Dex-treated SCID mice developed neoplastic lesions described as high-grade intraepithelial neoplasia (or AHGD) in all the cases. Thus, 6/6 of Dex-treated SCID mice (100%) infected with 105–107 oocysts and euthanatized after Day 46 PI showed ileocaecal lesions with a severity score of 3 (Figure 2; Table 1). Neoplastic lesions in more than one organ were observed in 5/6 of these mice (nos. 7, 8, 15, 16, and 24) (Table 1).
Furthermore, efforts to identify biotic or abiotic factors in the oocyst inoculum responsible for the observed neoplastic lesions were unsuccessful. No tissue lesions and no evidence of cryptosporidial infection were observed in either CDex or CDV control mice inoculated with filtrates of the oocyst inocula.
Discussion
This work belongs to a series of experiments exploring the potential of C. parvum to induce neoplastic changes in the digestive epithelium of SCID mice treated or not with Dex. We chose this animal model based on the lack of functional B and T cells in SCID mice. Consequently, they are unable to have antibody or cell mediated immune responses, remaining infected by Cryptosporidium for long periods.9 However, it has been shown that some SCID mice can express detectable levels of immunoglobulin (leaky mice), especially when they get older.18 Treatment of animals with Dex was then based on previous reports showing that Dex administration reinforces immunosuppression, and strongly expands parasite rates in SCID mice infected with Pneumocystis, another opportunistic agent.19 Furthermore, a Dex-treated adult SCID mouse model supported good propagation of at least two species of Cryptosporidium.6 Indeed, SCID mice treated with Dex showed less signs of inflammation and had higher risk of developing severe C. parvum infections.6 Consistently, in this study, regardless of the inoculum size, a slightly increased inflammatory response was observed in Dex-untreated mice. This observation is in agreement with the capacity of corticosteroids to inhibit innate immunity decreasing IFN-γ gene expression,11 and with the ability of Cryptosporidium infection to induce increased interleukin-15 (IL-15) and IFN-γ expression to contribute to the elimination of parasites.20
The first part of the study examined the influence of C. parvum inoculum size on the intensity of parasite shedding with or without Dex treatment. It was found that oocyst excretion increased according to inoculum size ranging between 105 and 107 and was substantially higher in Dex-treated mice than in the untreated ones. Thus, this animal model (especially with inoculum < 107) could improve current models to perpetuate and amplify parasite isolates of C. parvum or Cryptosporidium muris. Moreover, another advantage of this model is that animals become chronically infected, and have a significant oocyst shedding during at least 120 days (Certad G and others, unpublished data).
Of interest, in our study the mean number of excreted oocysts tended to decrease in Dex-treated mice receiving the highest challenge inoculum (108). A likely similar phenomenon was evoked in C. parvum experimental infection of healthy humans.21 Furthermore, in studies of Eimeria infection of chickens and rats, it was especially noticeable that with the greatest infecting dose, the number of oocysts produced per oocyst fed was smaller.22 Sequestration of the parasites out of the intestine and into other sites could also explain a diminished detection of Cryptosporidium oocysts in the faeces.23 However, in our study, mice infected with 108 oocysts showed no more evidence of extra-intestinal parasites or more clinical signs than mice in other groups (Table 1).
Traditionally, in SCID mice C. parvum develops usually in enterocytes of small intestine and colon; gastric and duodenal locations are reported less often.9 In our study, the examination of tissue sections revealed marked differences between groups in relation with parasite distribution in the gastrointestinal tract. In Dex-untreated C. parvum-infected mice, parasites were localized predominantly in the stomach, as early as 35 days PI, and in less degree in the liver. Mead and others24 noted gastric colonization in C. parvum-infected SCID mice (without corticosteroid administration) since the eighth week, with an intensity of the infection that increased henceforth.24 Others authors have reported that liver involvement occurs later, in some cases 13–26 weeks PI.25 However, we found that C. parvum parasites under Dex develop predominantly in the ileocaecal region and in the colon of SCID mice (Reference 6 and present work) (Table 1).
As described in our previous work, we confirmed that C. parvum induced neoplastic changes. Particularly notable, the association between C. parvum infection and the generation of ileocaecal tumors was shown in most animals, as we reported before.6 Additionally, for the first time C. parvum-induced neoplastic changes were also found in the stomach (12/16 SCID mice) and in the duodenum (2/16). Neoplastic lesions in the stomach were more frequent in Dex-untreated (8/16) than in Dex-treated mice (4/16) (Table 1). Indeed, when we assessed the intensity of the infection according to both oocyst shedding and extension of parasite invasion in the digestive tract (Table 1), a highly significant correlation (Figure 4) was found between intensity of cryptosporidiosis and severity of neoplastic lesions in Dex-treated or untreated mice. This observation suggests a direct role of C. parvum in the development of neoplastic lesions. Although Mead and colleagues9 did not report neoplastic changes in SCID mice experimentally infected with C. parvum (without corticosteroid administration), their results also suggested a correlation between inoculum size and severity of disease/clinical signs. After a challenge with different inoculum sizes, they noted that the extension of the infection, including colonization of gallbladder and hepatobiliary duct epithelium, was pronounced in SCID mice receiving the highest oocyst dose (107).9
Our study showed that animals not treated with Dex had alterations in the liver and duodenum rather associated with inflammation than with neoplasia (13/16 and 8/16 Dex-untreated mice had inflammation in liver and duodenum, respectively). However, hepatic histopathologic changes were associated with parasite detection in only two Dex-untreated mice and no parasite was detected in duodenum (Table 1). Inflammatory changes in the same organs were absent in Dex-treated SCID mice (Table 1). This divergence could be explained by the anti-inflammatory effect of Dex. Differences in the severity score of lesions between the groups of Dex-untreated and Dex-treated animals would therefore be more marked if purely inflammatory changes were not taken into account.
The present experiments show also that C. parvum infection induced early the emergence of an abnormal pattern of enterocyte proliferation, even before the identification of histologic changes revealing neoplasia (Figure 5). Indeed, the actual presence of neoplasia either in the stomach or in the ileocaecal region was consistent with the extension of Ki-67 staining above the basal third of the crypt. It has been suggested that a Ki-67 abnormal staining reflects an altered proliferative epithelium rather connected with neoplasia pathogenesis than with hyperplastic lesions.17 According to the standards for histologic assessment of intestinal tumors in mice,14 mitosis in hyperplastic epithelium are typically located in the lower two-thirds of the mucosa, nuclei lack significant atypia, are basally located, ovoid to round, and are usually uniformly dark with occasionally visible nucleoli.14 In our study, the dysplastic epithelium at the top of the crypts exhibited a markedly abnormal pattern of proliferation, similar to that observed in advanced neoplasms.26 Additionally, early increases of Ki-67 scores preceded local detectable parasite proliferation in about half of mice.
According to the current classification of intestinal tumors in rodents, the C. parvum-induced changes found in Dex-treated SCID mice correspond to gastrointestinal neoplasia (= microadenoma, microcarcinoma, carcinoma in situ, focal areas of dysplasia or adenoma, i.e., circumscribed broad-based, sessile, or pedunculated tumors lined by dysplastic epithelium, with either low-grade or high-grade dysplasia).14 As stated by the Vienna classification of human gastrointestinal neoplasia most severe lesions we found correspond to subcategory 4.2: “non-invasive carcinoma (carcinoma in situ).”27,28 On this last point, the histologic changes described here can be considered as putative precursors to digestive neoplasia. Actually, human colorectal tumorigenesis is believed to involve a series of genetic changes leading to the progression from normal epithelium to carcinoma, via the intermediate steps of dysplasia and adenoma.29 Consistently, we have recently observed invasive adenocarcinoma in a larger series of Dex-treated SCID mice infected with C. parvum and euthanatized at Day 120 PI (Certad G and others, unpublished data).
Previous studies have associated cryptosporidiosis with the development of tumor lesions in vertebrates. One report described the association between Cryptosporidium sp. and aural-pharyngeal polyps in iguanas.30 Cystic hyperplasia of the colonic mucosa was also described in nude mice.31 None of these studies has described the presence of pre-malignant lesions associated to cryptosporidiosis. However, the presence of portal fibrosis, biliary sclerosis, and necrosis with dilation of ductlike structures lined by highly atypical biliary epithelial cells was found in IFN-γ knockout mice infected with Cryptosporidium.25 These changes, classified as low-grade dysplasia,25 are consistent with our results. Interestingly, similar histologic findings (especially Figure 3 in Mead and others23) were associated with chronic C. parvum infection in NIH-III nu/nu mice.23
Despite the scarcity of information about links between human cryptosporidiosis and digestive neoplasia, some previous data suggested a possible causal association. An epidemiologic study of 55 patients with colorectal cancer and before chemotherapy, reported 18% of cryptosporidiosis prevalence.5
No data about the mechanism of C. parvum-induced neoplasia are available. Nevertheless, it is well known that Theileria parva, another apicomplexan parasite, is responsible for a lymphoproliferative disorder of cattle.32 This organism infects and transforms bovine lymphocytes resulting in tumors with metastatic/invasive potential by a mechanism associated to an inhibition of apoptosis.32 Inhibition of apoptosis has also been reported in other apicomplexan protozoa including Cryptosporidium33 that is able to activate NF-κB pathway, preventing the induction of cell death early after infection. Apoptosis prevention probably benefits the parasite by stabilizing the host cell long enough to permit the completion of the life cycle.33 Interestingly, resistance to apoptosis could be an essential step in the progression to malignancy.8 Therefore, persistent infection with C. parvum could be a risk for gastrointestinal neoplasia as an adverse effect caused by the inhibition of intestinal cell apoptosis.
In summary, despite the relatively small number of mice we used, statistical correlations were found between the intensity of Cryptosporidium infection and the score of severity of dysplastic lesions, especially in Dex-treated SCID mice. Moreover, adenomas with low- or high-grade intraepithelial neoplasia associated with numerous C. parvum life stages were observed in different areas of the digestive tract, including stomach, duodenum, and ileocaecal region. The use of Ki-67 supported the neoplastic nature of the described Cryptosporidium-induced epithelial transformation, and showed that potential neoplastic alterations begin before histopathologic lesions can be detected with standard stains. Further studies should be done to characterize at molecular level the Cryptosporidium-induced gastrointestinal neoplasia, and to explore its potential occurrence in humans.
Acknowledgments:
We thank Stephanos Papadopulos and Pierre Gosset, pathologists, for their precious points of view, very helpful for this work, and El Moukhtar Aliouat for his valuable advice. We also want to acknowledge the reviewers for their helpful suggestions and highly constructive criticism.
- 1.↑
Sunnotel O, Lowery CJ, Moore JE, Dooley JS, Xiao L, Millar BC, Rooney PJ, Snelling WJ, 2006. Cryptosporidium. Lett Appl Microbiol 43: 7–16.
- 2.↑
Pozio E, Morales MA, 2005. The impact of HIV-protease inhibitors on opportunistic parasites. Trends Parasitol 21: 58–63.
- 3.↑
Okhuysen PC, Chappell CL, 2002. Cryptosporidium virulence determinants–are we there yet? Int J Parasitol 32: 517–525.
- 4.↑
Izquierdo J, Antunez I, Calderon MT, Perez Giraldo C, Munoz Sanz A, 1988. Diarrhea caused by Cryptosporidium and colonic neoplasia. Rev Clin Esp 182: 393–394.
- 5.↑
Sulzyc-Bielicka V, Kuzna-Grygiel W, Kolodziejczyk L, Bielicki D, Kladny J, Stepien-Korzonek M, Telatynska-Smieszek B, 2007. Cryptosporidiosis in patients with colorectal cancer. J Parasitol 93: 722–724.
- 6.↑
Certad G, Ngouanesavanh T, Guyot K, Gantois N, Chassat T, Mouray A, Fleurisse L, Pinon A, Cailliez JC, Dei-Cas E, Creusy C, 2007. Cryptosporidium parvum, a potential cause of colic adenocarcinoma. Infect Agent Cancer 2: 22.
- 7.↑
Mele R, Gomez Morales MA, Tosini F, Pozio E, 2004. Cryptosporidium parvum at different developmental stages modulates host cell apoptosis in vitro. Infect Immun 72: 6061–6067.
- 9.↑
Mead JR, Ilksoy N, You X, Belenkaya Y, Arrowood MJ, Fallon MT, Schinazi RF, 1994. Infection dynamics and clinical features of cryptosporidiosis in SCID mice. Infect Immun 62: 1691–1695.
- 10.↑
Kang M, Mitomi H, Sada M, Tokumitsu Y, Takahashi Y, Igarashi M, Katsumata T, Okayasu I, 1997. Ki-67, p53, and Bcl-2 expression of serrated adenomas of the colon. Am J Surg Pathol 21: 417–423.
- 11.↑
Stojadinovic O, Lee B, Vouthounis C, Vukelic S, Pastar I, Blumenberg M, Brem H, Tomic-Canic M, 2007. Novel genomic effects of glucocorticoids in epidermal keratinocytes: inhibition of apoptosis, interferon-gamma pathway, and wound healing along with promotion of terminal differentiation. J Biol Chem 282: 4021–4034.
- 12.↑
Guyot K, Gireaudot-Liepmann MF, Cabon A, Riveau-Ricard I, Lange M, Delattre JM, Dei-Cas E, 2000. Influence of US Epa 1622 method successive steps on the viability of Cryptosporidium oocysts. Water Sci Technol 41: 189–196.
- 13.↑
Rasmussen KR, Healey MC, 1992. Experimental Cryptosporidium parvum infections in immunosuppressed adult mice. Infect Immun 60: 1648–1652.
- 14.↑
Boivin GP, Washington K, Yang K, Ward JM, Pretlow TP, Russell R, Besselsen DG, Godfrey VL, Doetschman T, Dove WF, Pitot HC, Halberg RB, Itzkowitz SH, Groden J, Coffey RJ, 2003. Pathology of mouse models of intestinal cancer: consensus report and recommendations. Gastroenterology 124: 762–777.
- 15.↑
Dommels YE, Butts CA, Zhu S, Davy M, Martell S, Hedderley D, Barnett MP, McNabb WC, Roy NC, 2007. Characterization of intestinal inflammation and identification of related gene expression changes in mdr1a(−/−) mice. Genes Nutr 2: 209–223.
- 16.↑
Aparicio T, Kotelevets L, Tsocas A, Laigneau JP, Sobhani I, Chastre E, Lehy T, 2005. Leptin stimulates the proliferation of human colon cancer cells in vitro but does not promote the growth of colon cancer xenografts in nude mice or intestinal tumorigenesis in Apc(Min/+) mice. Gut 54: 1136–1145.
- 17.↑
Polyak K, Hamilton SR, Vogelstein B, Kinzler KW, 1996. Early alteration of cell-cycle-regulated gene expression in colorectal neoplasia. Am J Pathol 149: 381–387.
- 18.↑
Nonoyama S, Smith FO, Bernstein ID, Ochs HD, 1993. Strain-dependent leakiness of mice with severe combined immune deficiency. J Immunol 150: 3817–3824.
- 19.↑
Aliouat EM, Martinez A, Jimenez E, Dei-Cas E, Mullet C, Delcourt P, Gargallo-Viola D, 1997. Development of Pneumocystosis animal models: corticosteroid-treated Wistar rat; SCID mouse and nude rat. J Eukaryot Microbiol 44: 41S–42S.
- 20.↑
Dann SM, Wang HC, Gambarin KJ, Actor JK, Robinson P, Lewis DE, Caillat-Zucman S, White AC Jr, 2005. Interleukin-15 activates human natural killer cells to clear the intestinal protozoan Cryptosporidium. J Infect Dis 192: 1294–1302.
- 21.↑
Chappell CL, Okhuysen PC, Sterling CR, DuPont HL, 1996. Cryptosporidium parvum: intensity of infection and oocyst excretion patterns in healthy volunteers. J Infect Dis 173: 232–236.
- 22.↑
Levine ND, 1973. Protozoan Parasites of Domestic Animals and of Man. Minneapolis, MN: Burgess Publishing Company, 156–254.
- 23.↑
Mead JR, Arrowood MJ, Sidwell RW, Healey MC, 1991. Chronic Cryptosporidium parvum infections in congenitally immunodeficient SCID and nude mice. J Infect Dis 163: 1297–1304.
- 24.↑
Mead JR, Arrowood MJ, Healey MC, Sidwell RW, 1991. Cryptosporidial infections in SCID mice reconstituted with human or murine lymphocytes. J Protozool 38: 59S–61S.
- 25.↑
Stephens J, Cosyns M, Jones M, Hayward A, 1999. Liver and bile duct pathology following Cryptosporidium parvum infection of immunodeficient mice. Hepatology 30: 27–35.
- 26.↑
Shih IM, Wang TL, Traverso G, Romans K, Hamilton SR, Ben-Sasson S, Kinzler KW, Vogelstein B, 2001. Top-down morphogenesis of colorectal tumors. Proc Natl Acad Sci USA 98: 2640–2645.
- 27.↑
Scoazec JY, 2007. Dysplasia in glandular digestive tissues: new concepts, new classifications. Ann Pathol 27: 398–416.
- 28.↑
Stolte M, 2003. The new Vienna classification of epithelial neoplasia of the gastrointestinal tract: advantages and disadvantages. Virchows Arch 442: 99–106.
- 29.↑
Luperchio SA, Schauer DB, 2001. Molecular pathogenesis of Citrobacter rodentium and transmissible murine colonic hyperplasia. Microbes Infect 3: 333–340.
- 30.↑
Uhl EW, Jacobson E, Bartick TE, Micinilio J, Schimdt R, 2001. Aural-pharyngeal polyps associated with Cryptosporidium infection in three iguanas (Iguana iguana). Vet Pathol 38: 239–242.
- 31.↑
Heine J, Moon HW, Woodmansee DB, 1984. Persistent Cryptosporidium infection in congenitally athymic (nude) mice. Infect Immun 43: 856–859.
- 32.↑
Dobbelaere DA, Fernandez PC, Heussler VT, 2000. Theileria parva: taking control of host cell proliferation and survival mechanisms. Cell Microbiol 2: 91–99.
- 33.↑
Heussler VT, Kuenzi P, Rottenberg S, 2001. Inhibition of apoptosis by intracellular protozoan parasites. Int J Parasitol 31: 1166–1176.