The effect of ivermectin alone and in combination with cobicistat or elacridar in experimental Schistosoma mansoni infection in mice

Schistosoma mansoni is less susceptible to the antiparasitic drug ivermectin than other helminths. By inhibiting the P-glycoprotein or cytochrome P450 3A in mice host or parasites in a murine model, we aimed at increasing the sensitivity of S. mansoni to the drug and thus preventing infection. We assigned 124 BALB/c mice to no treatment, treatment with ivermectin only or a combination of ivermectin with either cobicistat or elacridar once daily for three days before infecting them with 150 S. mansoni cercariae each. The assignment was done by batches without an explicit randomization code. Toxicity was monitored. At eight weeks post-infection, mice were euthanized. We determined number of eggs in intestine and liver, adult worms in portal and mesenteric veins. Disease was assessed by counting granulomas/cm2 of liver and studying organ weight indices and total weight. IgG levels in serum were also considered. No difference between groups treated with ivermectin only or in combination with cobicistat or elacridar compared with untreated, infected controls. Most mice treated with ivermectin and elacridar suffered severe neurological toxicity. In conclusion, systemic treatment with ivermectin, even in the presence of pharmacological inhibition of P-glycoprotein or cytochrome P450 3A, did not result in effective prophylaxis for S. mansoni infection in an experimental murine model.

Experimental design. We used the BALB/c mouse model of infection by Schistosoma mansoni established in the IBSAL-CIETUS of the University of Salamanca, Spain. Mice were divided into five experimental groups: untreated uninfected (G1 Untr, n = 9), infected (G2 Inf, n = 45), treated with 1000 μg/kg of ivermectin daily for three days by oral catheter before infection (G3 Iv, n = 30), treated with 1000 μg/kg of ivermectin and 25 mg/kg of cobicistat daily for three days by oral catheter before infection (G4 Iv + Co, n = 30), and treated with 1000 μg/kg of ivermectin and 2.5 mg/kg of elacridar daily for three days by oral catheter before infection (G5 Iv + El, n = 10) (Fig. 1). In G4 and G5, ivermectin was administered 2 h after cobicistat and elacridar.
After each drug administration, the behavior of the animals was observed for potential signs of toxicity (nervous, neuromuscular, digestive). On the third day of treatment and 4 h after administering the ivermectin dose, the animals in group 2, 3, 4 and 5 were infected with 150 cercariae of S. mansoni each. The infection was performed after anesthesia with a mixture of ketamine (50 mg/kg), diazepam (5 mg/kg) and atropine (1 mg/ kg), in a final volume of 100 µl by intraperitoneal injection in order to immobilize the animals. They were placed supine, a plastic ring was placed on the shaved abdomen, attached with adhesive tape, and the water solution  www.nature.com/scientificreports/ was applied with cercariae. After 45 min the rings were removed from the abdomen of the mice and the water dried. Once the mice recovered mobility after sedation, the mice were returned to their corresponding cages. The animals were weighed and had blood samples taken before starting the experiment, before infection and before the necropsies. Eight weeks post-infection, the mice were euthanized. Serum was obtained from the blood samples by centrifugation at 8000 rpm at 4 °C for 8 min. Samples were stored at − 20 °C until use.
To recover adult worms from the portal vein and mesenteric veins, livers were sectioned and perfused with a 0.9% saline solution with heparin through the left ventricle. Adults worms were recovered with tweezers and the number of Schistosoma couples and single male and female worms was counted.
A part of the intestine and part of the liver of each mouse were also separated and weighed. Subsequently, they were digested to determine the egg load. For this, both liver and intestine samples were incubated at 37 °C for 24 h in 5% KOH while stirring. The digested samples were centrifuged at 800g for 5 min and the supernatant was removed, leaving 5 ml of concentrated egg solution. A McMaster chamber was used to count each sample in triplicate and calculate the number of eggs per gram of intestine or liver. Spleen, bowel and liver indices were obtained from the weights of these organs and the live weight of the mice to determine their degree of inflammation. Photographs were taken where macroscopic lesions of the livers were observed to estimate the number of granulomas per square centimeter. The ImageJ v.1.51k, PowerPoint and Adobe Photoshop CC 2017 programs were used to count granulomas and estimate the affected liver surface.
ELISA for determination of S. mansoni soluble somatic antigen (SoSmAWA)-specific IgG. In order to detect specific antibodies against S. mansoni infection, the mouse sera were analyzed by indirect enzyme immunoassay (ELISA), for specific immunoglobulin G (IgG) levels. We used soluble adult somatic antigen from Schistosoma mansoni (SoSmAWA) according to Abán et al. (1999) 13 .
We coated 96-well polystyrene plates (Costar 3369, Corning Inc.) with 2.5 µg of SoSmAWA per well in a final volume of 100 µl of carbonate buffer at pH 9.6 and incubated them for 18 h at 4 °C. Plates were washed three times with 200 µl of 1X PBS per well at pH 7.2 and 0.05% Tween (PBS-T) to remove residues of unbound antigen, and blocked with 2% BSA in PBS-T (100 µl per well) at 37 °C for 1 h to avoid nonspecific antibody-binding and washed three times. Serum samples from each mouse were incubated for 1 h at 37 °C at a 1:100 dilution in PBS-T, and washed as in the previous steps. The secondary mouse anti-IgG antibody (Sigma Aldrich, anti-mouse IgG-peroxidase, A5906) was added at a 1:1000 dilution in PBS-T, incubated for 1 h at 37 °C, and washed. The plates were then incubated with o-phenylenediamine dihydrochloride (OPD) as a peroxidase substrate, and H 2 O 2 as an oxidizing agent in citrate buffer at pH 5.0. The reaction was stopped with 50 µl per well of 3 N H 2 SO 4 . Absorbance was measured at 492 nm in a spectrophotometer (ThermoScientific Multiskan GO 1510, Finland).

Statistical analysis.
Data are presented as means and standard errors of the mean. A Bartlett test was carried out to verify the homogeneity of the data distribution of all the variables. A t-unpaired two-tailed ANOVA analysis was done and, if statistically significant, the data was analyzed in a post hoc Fisher's Least Significant Difference test to determine the existence of significant differences between the study groups. P-values below 0.05 were considered significant. A multivariate analysis of main variables was also carried out. For these analyses and work chart, we used Simfit V7.3.0 and Statview V5.0 statistical packages.

Effect of Iv and Iv-Co on the weight of BALB/c mice after infection by S. mansoni. Weight loss
after the administration of a xenobiotic can be an indicator of toxicity in the protocol used. Healthy animals gained an average of 3.2 ± 0.6 g during the experimental period, whereas infected animals gained 3.6 ± 0.3 g. Animals treated with ivermectin after infection gained 3.5 ± 0.5 g, whereas those treated with ivermectin + cobicistat only gained 2.5 ± 0.5 g. These differences were not statistically significant (Fig. 2).
Neurotoxicity. After administering the combination of ivermectin and elacridar to G5 as detailed in "Materials and methods", the animals showed seizures, circular movements, and loss of ambulatory capacity. Six animals in the ivermectin and elacridar group died or were euthanized according to the end-point rules of the protocol. Only four out of ten mice survived the administration of the ivermectin and elacridar drug combination. No deaths or signs of neurotoxicity were seen in the group that was administered ivermectin alone or combined with cobicistat.
The four mice surviving in the elacridar and ivermectin group were infected and necropsy performed at eight weeks post infection. In the necropsy, no differences in worms recovered (8.5 ± 3.5 males, 6.0 ± 3.8 females) and egg in tissues were (1992 ± 1082 egg per gram of liver and 1493 ± 801 egg per gram of small intestine) with respect to other groups were seen, no statistical test was performed given the sample size (n = 4).
Effect of Iv and Iv-Co on worm recovery. The parasite load was used to estimate the prophylactic capacity of the drug or drug combinations against infection by S. mansoni. Mice treated prophylactically with ivermectin showed no significant reduction in the total recovery of adult worms. The animals treated with the combination of ivermectin and cobicistat had a non-significant 8% reduction in worm load compared with the control group of infected mice. A similar situation occurred in the separate recovery of male and female worms (6-12% reduction) (Fig. 3).  (Fig. 4).

Effect of Iv and Iv-Co on S. mansoni granulomas and organ indexes. Granuloma formation is an
immune response to the presence of eggs in tissues and it is an indicator of severe disease. There was no reduction in granulomas per liver surface in animals that were administered ivermectin When using the combined ivermectin + cobicistat, the reduction was 3.6% but this was not statistically significant when compared with infected mice (Fig. 5). Hepatic and intestinal indices are indicators of the relative inflammation of each of these organs against the eggs trapped. In schistosome infection, hepato-splenomegaly is described as one of the most prominent signs and is present in severe forms of the disease. Infected mice showed a 350-360% increase in spleen weight (ANOVA F (3.105) 4.439, p = 0.006) and a 45-59% increase in liver weight (ANOVA F (3.105) 16.670, p < 0.001) compared with the untreated infected mice. The intestinal index increased by 64-72%, which was not statistically significant (Fig. 6).

IgG levels against SoSmAWA in infected animals. Immunoglobulin levels during infection with S.
mansoni serve to reveal the status of the infection and the response against the infection of the animals. High levels of specific IgG against the SoSmAWA antigen were observed in all infected groups at eight weeks postinfection. The infected group did not differ significantly from the groups given preventative treatment (Fig. 7).   total recovered worms, eggs in tissue, granulomas in the liver, organ indices, weight variation during the experiment and antibody levels at eight weeks post-infection. Between the principal components 1 and 2 (PC1 and PC2), 74% of the information was collected (Fig. 8). We found no differences between the infected groups previously treated with ivermectin (G3 Iv), ivermectin + cobicistat (G4 Iv-Co) or untreated (G2 Inf). Only the untreated and uninfected group (G1 Untreated) differed (Fig. 5). All variables affected both principal component 1 and 2 (Fig. 9).

Regulatory compliance.
All methods were carried out in accordance with relevant guidelines and regulations.

Discussion
The discovery of glutamate-mediated signaling in S. mansoni raised hopes for new drug targets in this species 14 .
Macrocyclic lactones such as ivermectin target the neuronal glutamate-gated chloride channel that is expressed by arthropods, nematodes and trematodes 15 . Schistosomes express these channels, yet the evidence base for the efficacy of ivermectin against S. mansoni is inconclusive, which might be explained by a relatively low affinity of the S. mansoni glutamate-gated chloride channel for ivermectin 10 . Nevertheless, one study reported a marked reduction in adult worm load after high-dose treatment with ivermectin, which was attributed to the tegumental   www.nature.com/scientificreports/ against schistosomula stages in the lung might allow for higher concentrations against a developmental stage that is not often targeted 17 .
In C. elegans, the susceptibility to ivermectin can be modulated by targeting ABC transporters or the P-gp 18,19 . We previously showed in a pharmaco-enhancement model of mosquitoes feeding on treated pigs that the simultaneous inhibition of cytochrome p450 3A and the P-gp enhanced the effect of ivermectin by two mechanisms: by increasing plasma levels in the pig host and by increasing the susceptibility of the mosquito. This combination resulted in prolonged target insecticidal concentrations of the drug able to kill Anopheles mosquitoes 20 . Given the inconclusive evidence regarding the efficacy of ivermectin treatment in S. mansoni-infected mice, we hypothesized that P-gp and CYP3A inhibitors may also alter the susceptibility of S. mansoni cercariae to ivermectin in a prophylaxis model.
Despite pharmacological inhibition of CYP and P-gp in our study, ivermectin did not have any significant effect on the numbers of S. mansoni eggs or adult worms in the mouse nor on the host immune response as evidenced by a lack of differences in IgG levels or granuloma formation. This finding stands in contrast to Kasinathan et al. who showed a significant effect of P-gp inhibitors in combination with praziquantel on both egg numbers and granuloma size in S. mansoni-infected mice, and suggested effects on parasite immunomodulatory factors to the host 21 . These differences point to different mechanisms involved in the metabolism of praziquantel and ivermectin.
Ivermectin has an excellent safety profile demonstrated over decades in global mass drug administration campaigns. Its safety is partly explained by the activity of P-gp in the capillaries of blood-brain barrier that excludes ivermectin from the mammalian central nervous system. Only a few cases of encephalopathy were described in rare population genotypes (mdr-1 gene) in Loa loa massive drug administration campaigns 22 .
In our study, the combined use of ivermectin and P-gp inhibitor elacridar led to seizures and disturbances in movement defined as final point criteria in most animals in the treated group. As elacridar disrupts the function of P-gp in the murine blood-brain barrier, elevated ivermectin levels in the brain lead to important neurological toxicity 23,24 . It serves as a reminder that in states with an impaired blood-brain barrier, as is the case in hyperinflammatory states, ivermectin could cause significant neurological toxicity.
A limitation of this study pertains to our choice of model animal. Although ivermectin does not affect physiological parameters in mice, they are one of the rodent species most susceptible to ivermectin toxicity 25 . The data on lethal dose included by Merck in the ivermectin label were determined in murine models 26 . This dose potentially does not represent the true toxicity in larger mammals or humans. The vulnerability of mice to ivermectin toxicity excludes the possibility of using higher doses that might produce more pronounced prophylactic effects against Schistosoma infection. Another limitation concerns the standard time span and cercaria dose in which the experiments were conducted to make comparable with other studies. In field conditions people became infected in several times with lower daily doses and measurement time point is difficult to establish. The drug-treated groups were not treated all simultaneously, which might have a small influence on the results.