Introduction

Schistosomiasis affects more than 250 million people worldwide. This disease is caused by infection with trematode parasites of the Schistosoma species. Considerable hepatic morbidity in terms of fibrosis, variceal bleeding and host mortality are caused by the inflammatory granulomatous response to tissue entrapped Schistosoma mansoni eggs [36]. In addition, 85% of S. mansoni-infected individuals suffer from gastrointestinal manifestations caused by parasite egg-induced granulomas in the gastrointestinal tract, with symptoms such as bloody diarrhoea (frequently seen in children) [33], abdominal pain, nausea, dysentery, malnutrition, iron deficiency and anaemia [27, 31].

Extensive epidemiological studies on human populations in endemic countries have shown that adults generally have lower intensities of schistosome infection than children do [2]. Observation of the effects of host age, per se, on the susceptibility to human schistosomiasis is usually confounded by differences between adults and children with regard to: (1) water contact patterns, (2) development of acquired immunity due to previous infections and (3) age-related, neurohormonal influences. Much discussion has been prompted on this controversial issue [12, 13, 14, 15]. How this differential susceptibility pattern defines the egg-induced inflammatory response, how the dynamics of granulomogenesis differ with age and how this antigenic profile provides the prerequisite condition for the development of adult resistance to schistosomiasis are difficult to identify and quantify in human studies.

It is reported [9] that the peak of re-infection is coincident with puberty (about 12–14 years of age), suggesting that, after puberty, a mechanism may mediate the development of resistance to infection in adults. The corollary of this is that “juveniles” (i.e. those aged less than 12 years) may be innately more susceptible to schistosome infection than adults in terms of the inflammatory granulomatous response to parasite eggs and/or are unable to develop the acquired immunological responses that provide protection in adults.

In this study, we have used rhesus monkeys (Macaca mulatta), a widely used primate model for studies on primary infection with schistosomes [30], to investigate the morphology of the hepatic and intestinal granulomatous response and the nature of age-related differences in susceptibility to a primary infection with S. mansoni.

Materials and methods

Experimental animals and infection

Ten rhesus monkeys were used in this study. They were matched for sex (all females) and housed in the Biomedical Primate Research Center (Rijswijk, The Netherlands). The five adult monkeys were aged 13–19 years (mean age±SD, 16.1±3.1 years) and weighed 5.5±1.1 kg. Five juvenile monkeys were selected by age to be at the pre-pubertal to mid-pubertal stages of development (range, 1–3 years; mean±SD, 2.4±0.6 years) and weighed 3.0±0.9 kg. Before the start of the study, a veterinarian examined the animals. Levels of serum dehydroepiandrosterone (DHEA) were measured to ensure the correct categorisation of animals as juvenile or adult. Measurement of serum DHEA sulfate (DHEAS) was performed using standard methods in the clinical chemistry department of the district laboratory (SSDZ, Renier de Graaf Hospital, Delft, The Netherlands) [22]. DHEA levels confirmed pre-pubescent (DHEAS 0.9±1.7 μmol/l) versus adult (DHEAS <0.8 μmol/l) status of the animals [20, 21].

A Puerto Rican strain of S. mansoni, maintained at the Department of Parasitology (University of Leiden, The Netherlands) was used for infections. Animals were percutaneously exposed to 1000 S. mansoni cercariae on the shaved abdomen. After 30 min, the cercarial suspension was recovered, and the number of non-penetrating cercariae were counted. Both groups were sacrificed under anaesthesia after an infection period of 8 weeks. The animals were perfused with citrated saline, and worms were collected from the perfusion fluid and counted.

Morphological experiments

Histology

From each animal, a specimen of the liver, proximal and distal small bowel, proximal and distal colon and lung were investigated. A part of each organ was fixed in 4% buffered formalin, and another part was snap frozen in liquid nitrogen. The tissues fixed in buffered formalin were paraffin embedded, and 4-µm-thick sections were made. These sections were routinely stained using haematoxylin-eosin (HE), sirius-red-haematoxylin and Dominici-stain (for the detection of eosinophilic granulocytes) [3]. Cryostat sections were stained with toluidine-blue (for the detection of mast cells).

Cryostat sections were also used for immunohistochemistry, according to a routine procedure (indirect immunoperoxidase technique) using mouse anti-human antibodies against CD68 (clone KP-1, 1:80; Dako, Glostrup, Denmark), CD3 (clone F7.2.38, 1:50; Dako) and CD20 (clone L26, 1:80; Dako), respectively, for the detection of macrophages, T-lymphocytes and B-lymphocytes. These antibodies cross-react with rhesus monkey epitopes. The Peyer’s patches showed strong immunoreactivity for CD20 and the lamina propria T-lymphocytes for CD3, serving as an internal control.

Morphometry

The number of granulomas was counted on the HE sections using a Zeiss light microscope (objective ×10/0.32; Zeiss Oberkochen, Germany), and the number was corrected for the amount of tissue investigated. At least ten sections per intestinal region were investigated with a total mean length of 4.5 cm tissue sampled per region. For the intestine, the number was expressed as mean number of granulomas per centimetre tissue±standard error of mean.

Statistical analysis

For basis statistical analysis, a Graph-Pad Prism Program was used. Continuous data were tested for normality. Normally distributed continuous data were compared by a Student’s t-test for unpaired values (STT) and two-way analysis of variance followed by a Bonferroni post-hoc test where indicated. P values of less than 0.05 were considered significant.

Results

Macroscopy

In both age groups, the liver surface showed miliary patterns with 1–2 mm large, white nodules but with no fibrosis. The colon showed a diffuse punctate, mucosal haemorrhage. The hyperaemic mucosa was visible along the total length of the large bowel. No polyps or ulcerations were detected. Two animals (one adult, one pre-pubescent) had severe diarrhoea at the time of sacrifice.

Histopathology

Two distinct morphological stages in the granuloma formation were observed. One type showed a centrally located egg surrounded by lymphocytes, histiocytes (macrophages) and mainly eosinophilic granulocytes staining brightly red in the Dominici-stained sections. Among the eosinophilic granulocytes, rare neutrophilic granulocytes that stained faintly blue with Dominici stain were seen. This type of granuloma was termed the exudative-stage granuloma (ESG) (Fig. 1a).

Fig. 1
figure 1

Micrograph of the different types of intestinal granulomas detected in the S. mansoni-infected rhesus monkey. The exudative-stage granuloma (ESG; a) shows a centrally located egg surrounded by eosinophilic granulocytes. The outer rim is formed by lymphocytes and histiocytes. The productive-stage granuloma (PSG; b) containing an egg (centre) surrounded by histiocytes and foreign-body giant cells. The outer rim of the granuloma is mainly formed by lymphocytes. The eggshell is surrounded by eosinophilic rays of the Splendore–Hoeppli phenomenon as the result of an antigen–antibody reaction. Both (a and b) were treated with Dominici-stain. c and d ESG and PSG, respectively, using CD68 immunohistochemistry. The epithelioid histiocytes and the foreign-body giant cell show strong immunoreactivity. e The very small number of CD3-positive T lymphocytes, located at the boundary of the ESG. This is in contrast to S. mansoni-induced granulomas in rodents, which are mainly a T-cell-dependent reaction. There is no immunoreactivity for CD20 (B-cell marker) in the ESG. The location of B- and T-lymphocytes in PSG showed similar results (not shown). Final magnification ×200 (a, b); ×75 (c, d, e, f)

The other granuloma type consisted of parts of eggshell surrounded by numerous foreign body-type giant cells, often partly engulfing the eggshell or eggshell remnants. Lymphocytes, eosinophilic granulocytes and acidophilic, eosinophilic rays of the Splendore–Hoeppli phenomenon were noted. This stage was termed the productive-stage granuloma (PSG), (Fig. 1b).

Both types of granulomas were found in the ileum, colon and liver of pre-pubescent and adult animals. In the intestinal wall, the granulomas were almost exclusively detected in the submucosa.

Both stages of granulomas were also present in the lungs of the animals. On an average, one pulmonary granuloma was noted per five ×10-magnification fields. The Splendore–Hoeppli phenomenon was frequently observed in the lung specimens.

The fibrotic reaction in the liver and intestinal granulomas was minimal at 8 weeks post-infection in both age groups. In the small and large intestine, no mast cells were detected in the toluidine blue-stained cryostat sections, whereas the lamina propria showed a normal lymphoplasmocytic infiltrate.

Morphometry

Total granuloma number

The number of colonic granulomas was significantly higher in the pre-pubescent animals (Table 1, P<0.05, STT) than in the adult animals. When the granuloma numbers were compared between the ileum and colon, higher numbers were found in the colon, this difference being statistically significant (Fig. 2). There was no significant difference between the number of granulomas in the proximal and the distal colon (Fig. 2), nor between the numbers of granulomas in the liver of adult versus pre-pubescent monkeys (data not shown).

Table 1 Total granuloma number in the ileum and the colon. The pre-pubescent animals showed significantly more colonic granulomas than the adult animals. The number of granulomas in the ileum was very small and showed considerable variation. Results are presented as number of granulomas per centimetre of intestinal tissue and expressed as mean±standard error of mean (n=5 for both groups)
Fig. 2
figure 2

Bar graph representing the number of granulomas in the different intestinal regions. The difference in the number of granulomas between the ileum and the colon, both in pre-pubescent as well as in adult animals, is statistically significant (two-way analysis of variance). Results are presented as number of granulomas per centimetre of intestinal tissue and are expressed as mean±standard error of mean (n=5 for both groups)

When the worms were recovered from the experimental animals by perfusion at the end of the study, there were 38% fewer female worms, as a measure of egg-producing worm pairs, in the adult than in pre-pubescent animals. However, this difference in the mean number of recovered female worms was not statistically significant. The number of granulomas per centimetre colonic tissue per worm couple was calculated and showed no significant difference between the two groups (0.17±0.009 and 0.25±0.03 granulomas per centimetre colonic tissue per female worm for pre-pubescent versus adult animals respectively; P=0.44, STT).

Stage of granuloma

When the dataset for the large intestine was divided on the basis of the two granuloma stages (ESG and PSG), there was a significant difference between the number of ESG and the number of PSG in the pre-pubescent animals (Fig. 3; P<0.001, STT); there was, however, no significant difference in the adult age group (Fig. 3). When the number of PSG was compared, pre-pubescent animals showed significantly more colonic granulomas (Fig. 3; P<0.01, STT) than the adult animals. In the liver, no significant difference was seen (data not shown).

Fig. 3
figure 3

Bar graph showing the number of granulomas in the colon of S. mansoni-infected rhesus monkeys divided by subtype. The difference of exudative-stage granuloma (ESG) versus productive-stage granuloma (PSG) in pre-pubescent animals is highly significant [P<0.001, student’s t-test for unpaired values (STT)]. In the adult age group, there is no difference. The number of PSG between the pre-pubescent and the adult animals is also statistically significant (P<0.01, STT). Results are presented as number of granulomas per centimetre of intestinal tissue and expressed as mean±standard error of mean (n=5 for both groups)

Immunohistochemistry

The two granuloma stages detected in the morphological investigation could also be separated on immunohistochemical grounds. The ESG showed a centre of non-immunostained eosinophils surrounded by a rim of a small number of CD3-positive T-lymphocytes (Fig. 1e). CD20-positive B-lymphocytes were extremely rare (Fig. 1f). The outer rim was formed by CD68-positive macrophages (Fig. 1c).

The PSG showed a strong central immunoreactivity for CD68-positive cells, including the multinucleated giant cells (Fig. 1d). CD3-positive T-lymphocytes were detected mainly on the boundary and rarely mixed with the CD68-positive macrophages. The number of B-lymphocytes in or in the vicinity of the granulomas was very small. The T- and B-lymphocytes showed no quantitative difference between the two age groups.

Discussion

Rhesus monkeys (Macaca mulatta) have been extensively used in the past to study experimental S. mansoni infection [28]. Keeping in mind the implications of age-dependent susceptibility on schistosoma vaccine development and the implementation of control measures, our study delineated whether naive, pre-pubescent animals were innately more susceptible to infection with S.mansoni and whether age-related changes in enteric granulomogenesis were evident in these animals.

Previous experiments in small laboratory animals (mainly mice) have shown specific organ-dependent granulomatous patterns towards S. mansoni egg antigens [34, 35]. Colonic granulomas were comparable with the liver granulomas, whereas granulomas in the ileum were smaller and showed no down-modulation during the course of the infection [18]. A different cellular composition of the granulomas in the various organs has also been described [18, 34]. Granulomas were abundantly present in the murine ileum and were accompanied by a diffuse mucosal inflammation, a diffuse mast cell infiltrate—mainly in the muscularis propria—and specific ganglionitis [1].

In the present study, granulomogenesis induced by an acute, primary S. mansoni infection in adult versus pre-pubescent rhesus monkeys was almost exclusively located in the colon, in contrast with the results of Sadun et al. [28], who found entrapped eggs mainly in the small intestine of primates. Our observations are, however, confirmed by results from a large autopsy series on 197 human patients reported by Cheever [4]. In his study, Cheever found most of the granulomas in the sigmoid colon of a population of not heavily infected patients. When infected patients developed Symmer’s fibrosis, etiologically linked with heavy infection and a worm load of at least 160 worm pairs [4], the eggs shifted from the sigmoid colon towards the small intestine. The predilection of the sigmoid colon, however, could not be corroborated in the present study, where there were no significant differences in the number of granulomas in the proximal versus the distal part of the colon. Investigating the difference of infection intensities, Cheever et al. [6] confirmed the shift of worms and eggs from the colonic towards the small intestinal venules in heavily infected rhesus monkeys (600 cercariae) but only after an infection period of 12–27 weeks. Although the infective dose in the present study was 1000 cercariae per animal, the animals were investigated after an infection period of 8 weeks. It, therefore, seems that the preferential habitat of worms shifts from the colonic towards the small intestinal circulation, both in man as well as in rhesus monkey, in cases of heavy infection and/or after an infection period of at least 10 weeks.

In the present study, two distinct stages in the development of granulomas were detected. The first stage (ESG) comprises a central, mainly eosinophilic reaction surrounded by a rim of lymphocytes and some macrophages. This granuloma resembles the experimentally induced S. mansoni granulomas seen in mice and smaller laboratory animals (e.g. hamsters).

The second stage (PSG) resembles foreign body-type granulomas as well as the granulomas often seen in S. japonicum infection in man and smaller animals [17]. PSG S. mansoni granulomas appear to be encountered more frequently in primate hosts [11]. In both granuloma types, the main polymorphonuclear cell is the eosinophilic granulocyte, a clear difference with S. japonicum granulomas, in which the granulomas mainly contain neutrophilic granulocytes [17].

The development of S. mansoni-induced granulomas is well studied and described by Hsu et al., who recognised five stages in the hepatic granuloma development in the rhesus monkey [16]. One of the earlier stages is the exudative stage, which is followed by the productive stage, via an intermediate, mixed stage, before complete healing without fibrosis. Indeed, both in our study as well as in the study by Hsu et al. and other studies [28], no clear fibrosis was detected [5]. Because the animals were all infected at the same time and sacrificed after the same infection period, the higher number of intestinal granulomas in the productive, more advanced stage in pre-pubescent animals suggests that the intestinal granulomogenic process in pre-pubescent animals is faster than in adult animals.

The two stages in the intestinal granulomatous inflammation presented in the present study are similar to the evolution of human colonic schistosomiasis [24], stressing the importance of data acquired in the rhesus model for human pathology. Smaller colonic lesions consisted mainly of eosinophilic granulocytes, while larger granulomas often contained multinucleated giant cells. Immunotyping of these cells showed the transition of a hypersensitivity granuloma to a foreign-body granuloma [10].

In the present work, the number of granulomas in the large intestine was significantly higher in the pre-pubescent group than in the adult group. Interestingly, the number of female worms also showed an increased trend, without reaching significance. These findings might be explained both by a higher susceptibility to infection in the younger age group as well as by a higher fecundity of the female worms. Although the total number of eggs produced by the worms could not be calculated from the present data, the number of granulomas per centimetre colonic tissue per worm showed no significant difference in the two age groups, suggesting a similar fecundity. These findings support the hypothesis of a higher degree of susceptibility in younger animals.

Ageing plays an important role in the efficacy of the gastrointestinal immune system. Taylor et al. [32] demonstrated a decrease of luminally secreted immunoglobulins with age, due to alterations in the process of maturation and homing of specific B-lymphocytes, while the total number of CD4+ T lymphocytes showed a senescence-associated decline [19]. These data, mainly on the mouse model, are in accordance with the present study, showing a more intense response in younger animals. The total number of CD3+ T-lymphocytes, however, showed no difference. The mast cell response, which we detected in the mouse small intestine and the specific ganglionitis of the myenteric plexus, was not seen in the tissues investigated from the rhesus monkey [1].

On the same animals studied here, Fallon et al. [8] delineated the immune responses after primary S. mansoni infection. They showed that, in contrast with adults, juvenile (pre-pubescent) rhesus monkeys failed to develop type-2 cytokine responses [normally associated with high interleukin (IL)-4, IL-5 and low interferon (IFN)-gamma levels] and showed markedly reduced parasite-antigen-specific antibody responses and significantly limited overall IgE production. In their study, no differences were observed between adult and juveniles in the number of non-penetrating cercariae. They speculate, however, that a number of other physiological non-immunological responses could account for the differences in susceptibility between juvenile and adult, including age-associated changes in skin thickness, body fat or even size of the animal.

It has been hypothesised that the increased susceptibility to re-infection in young children is caused by an inability to induce appropriate—that is, protective—immune responses [26]. Our data indicate that juvenile monkeys elicit a higher colonic inflammatory response towards an increased number of entrapped S. mansoni eggs. It has yet to be elucidated whether the impaired immune responses in juveniles is a result of the differential granulomogenesis we observe, or whether senescence-related immunomodulation is involved. Yet other innate differences in susceptibility cannot be excluded as causes, including circulating putative schistosomicidal factors and age-associated hormones [7, 25] and neurohormones. In utero sensitisation to schistosomes could also be a significant factor in the subsequent susceptibility of children to schistosome infection [23].

There is ongoing debate whether the increased resistance of adult humans to schistosome infection is due to the age of the host, per se, or to previous experience of infection(s) [12, 37]. Recent evidence from a new focus of infection in Senegal showed that children were still at a higher risk of schistosoma infection with higher intensities of infection than adults [29]. Such studies stress that the changes associated with the age of the host determine susceptibility patterns to primary infection. Future research is required to give appropriate answers regarding the physiological differences between juveniles and adults with respect to susceptibility to schistosome infection and generation of anti-parasitic immune responses.