The present study confirmed that T. spiralis is highly infective at very low doses in Wistar rats and can easily be detected by serology or pathogen detection using digestion. We describe the dynamics of T. spiralis infection in rats using doses ranging from very low (10 ML per rat) to very high (16 000 ML), with the aim of studying the usefulness of rats as sentinel animals in a wildlife monitoring program for Trichinella. Rats are sentinels for several human pathogens and have several advantages for epidemiological studies as mentioned by Psaroulaki et al. . This is the first report describing the clinical, pathological, immunological and parasitological findings using a dose range of T. spiralis in experimental infection in rats this wide.
Infection with 10-2 000 ML caused no clinical symptoms in experimentally infected rats. Body weights of rats infected with 4 000 ML or more, declined in a dose-dependent manner until day 8-9 pi. This was probably caused by diarrhea and a lesser food intake during the intestinal phase of T. spiralis infection as observed before by Aulí and Fernández . Although the probability of survival might have been biased by administering soft food, all animals, except the rats in the highest dose group (16 000 ML) were able to overcome illness, regained weight and survived until the end of the experiment.
This was corroborated by the differences in post mortem pathological findings between rats euthanized and examined at day 8 pi (16 000 ML infection group) and rats euthanized at day 42 (200-14 000 ML). During the intestinal phase of T. spiralis (day 8 pi, 16 000 ML), rats showed increased mucus production and intestinal hypermotility. This was in accordance with the model of Khan and Collins , which describes enteritis induced Th2 immune response resulting in intestinal muscle hyperactivity and increased mucus production by goblet cells.
Histological changes were most prominent in the duodenum at day 8 pi, while on day 42 the duodenum was not altered in any of the other infection groups. Also the thymus showed severe atrophy at day 8 pi, but showed no pathological alterations after 42 dpi in the other infection groups. Both features indicate successful regeneration of these organs after regression and inactivity indicative of severe biological stress. Findings in the spleen and (mandibular) lymph nodes seen on day 42 pi were considered indicative of systemic immune stimulation. The finding of wild caught Rattus norvegicus with very high LPG, ranging from 5 720 to 7 692 [5, 7], indicates that also in nature wild rats are able to survive these high infection levels.
Infection with high doses of T. spiralis ML was characterized by a large variability in the numbers of muscle larvae that developed in individual rats. The maximum number of recovered muscle larvae was found in rats infected with 10 000 ML. Rats infected with 12 000 ML developed fewer muscle larvae and infection with 14 000 ML produced the fewest muscle larvae. This was probably due to severe competition for space (in the intestinal epithelium as well as in muscle fibers) and nutrients, in combination with expulsion of T. spiralis intestinal stages by the host. No ML could be recovered on day 8 pi from the muscle tissue of animals infected with 16 000 ML, although some ML could be identified in the diaphragm after histology. This might indicate that the intramuscular larvae might have been destroyed after digestion at this early developmental stage, since the protective nurse cell formation process starts at 8 dpi and is completed only on day 26 pi .
One of our goals was to determine whether ELISA was a suitable tool for detecting T. spiralis in rats. Therefore, it would be necessary to determine (1) the correlation between OD and infection dose in time and (2) the correlation between measured OD and exact numbers of T. spiralis ML in the rat. This is important to be able to evaluate ELISA results of future serological surveys in wild rats.
There was a clear antibody response in all infection groups. Seroconversion in rats infected with 25 ML or more, took place between days 7 and 14 pi, but already at day 7 pi, ODn levels of rats infected with 14 000-16 000 ML were above cut-off. Rats infected with 10 ML showed a gradual linear increase in OD above the cut-off level from 14 to 42 dpi.
We found a positive correlation between ODn values and infection dose, although less pronounced in the 10-400 ML infection groups. Some studies, also performed in rats, focused mainly on immune response with a limited number of infective doses. Salinas-Tobon et al.  performed experiments in rats with infection doses of 700, 2 000, 4 000 and 8 000 ML. Specific antibodies did not increase proportionally with the different ML doses tested, but the seroconversion period that occurred between 10 and 19 dpi, varied according to the infection dose.
We found a positive correlation between the number of recovered ML and serum antibody levels, although the predictive value of measured OD to estimate infection levels is limited, reflecting host variation in the immune response against a Trichinella infection.
The kinetics of anti-T. spiralis newborn larvae (NBL) immunity and its dose effects were studied in vivo by Wang . In that study, rats were infected with 500, 2 000, 5 000 and 6 000 ML and immune response was measured as a reduction in the numbers of recoverable NBL after intravenous challenge with 10 000-100 000 NBL on day 16 pi. One of the results of that study was that infection with 2 000 ML per os induced the strongest immunity and that high dose immunization might induce a suppressive effect on host immunity. Our results demonstrate the exhausting effect on rats for doses above 8 000 ML and the more moderate clinical effects for doses of 6 000 and 8 000 ML. The lower immune effects with 5 000 and 6 000 ML observed by Wang  could reflect this strong impact on rat health rather than a real suppressive effect on the immune system. Serum antibodies were not measured in that study.
In our study, infection dose and OD correlated linearly at all time points after infection with 2 000-10 000 T. spiralis ML. Infection with lower doses resulted in a positive correlation with OD, but only after 14 dpi or more, and at a lower, though increasing level with time. Infection with higher doses (12 000 and 14 000 ML) resulted in lower numbers of recovered larvae as compared to the 10 000 ML infection group (70% and 60% respectively), which explains the observed comparable decline in ODn between 28 and 42 dpi. However, at day 7 pi the ODn values in the 12 000-16 000 ML infection groups are considerably higher than those of the 10 000 ML group.
Rats that were infected with 10, 25 or 50 ML showed a serological response in the T. spiralis western blot 2 weeks post infection. Other rodents like mice have been experimentally infected with low doses of 50, 10, and 5 ML of Trichinella spiralis per animal [20, 21]. Seroconversion was observed in these experiments at 30 dpi with an ES-ELISA and specific antibodies increased until 80 dpi. Measured splenic T-lymphocyte activity increased from day 10 to day 15 pi, even with 10 ML as an infective dose . This implies that for epidemiological studies, low infection levels can be detected by serology, both in rats and mice and we confirmed that low infection doses of T. spiralis larvae induce the production of specific antibodies at detectable levels in rats. These low doses reflect the infection level that we can find in the sylvatic cycle. Hurnikova and Dublinskyl  underlined that Trichinella (britovi, pseudospiralis and spiralis) infection in wild foxes is usually below 20 LPG and far less with wild boar. In The Netherlands, Trichinella britovi infection in wild foxes is even lower with LPG ranging from 0.04 to 0.71 . For confirmation of rat sera from animals with low Trichinella infection levels or higher infection levels early in the time course of infection that exhibit OD values around the cut-off, we showed that western blots are a suitable instrument.
The results of our study confirmed previously conducted studies that were performed with very limited infection dose ranges in other host species like cattle , sheep [25, 26], goats [27, 28], horses [29–31], wild boars, pigs and foxes [32–34]. In these studies, the animals were experimentally infected with T. spiralis or other Trichinella species. As in our study, in most of these experiments, the time point of seroconversion and specific antibody titer were also dose dependent.
In summary, we show that rats, even infected with a low dose of 10 Trichinella ML, develop an immune response, which can be detected by use of serological assays and this immunological response is dose dependent up to an infection level of 10 000 ML. This indicates that the Trichinella ES-ELISA and Western blot are useful instruments for the detection of the presence of T. spiralis in sentinel populations like wild rats, which easily cross sylvatic-domestic borders. We also show that antibody levels can not be used to calculate exact LPG values in rats due to high variation in infection rates.