Schistosoma mansoni infection suppresses the growth of Plasmodium yoelii parasites in the liver and reduces gametocyte infectivity to mosquitoes

Malaria and schistosomiasis are major parasitic diseases causing morbidity and mortality in the tropics. Epidemiological surveys have revealed coinfection rates of up to 30% among children in Sub-Saharan Africa. To investigate the impact of coinfection of these two parasites on disease epidemiology and pathology, we carried out coinfection studies using Plasmodium yoelii and Schistosoma mansoni in mice. Malaria parasite growth in the liver following sporozoite inoculation is significantly inhibited in mice infected with S. mansoni, so that when low numbers of sporozoites are inoculated, there is a large reduction in the percentage of mice that go on to develop blood stage malaria. Furthermore, gametocyte infectivity is much reduced in mice with S. mansoni infections. These results have profound implications for understanding the interactions between Plasmodium and Schistosoma species, and have implications for the control of malaria in schistosome endemic areas.


Author summary
Malaria and schistosomiasis are parasitic infectious diseases that cause severe morbidity and mortality in the tropics. Chronic schistosomiasis causes malnutrition and impaired intellectual development to children while malaria can cause fatal acute infections. Since coinfection of these two parasites is common in the tropics, many studies of both epidemiology and coinfection in animal models have been performed in order to reveal interactions between them. Previous animal studies on the interactions between Plasmodium and Schistosoma parasites have focused on the blood stage pathology of the malaria infection,

Introduction
Malaria and schistosomiasis are two of the most important parasitic diseases in the tropics, and together constitute a severe burden to public health and to the economic development of endemic countries. Malaria is estimated to cause 429,000 deaths per year, 70% of those occurring in children aged under five years old [1]. The WHO has estimated that schistosomiasis causes about 200,000 deaths every year in sub-Saharan Africa and 218 million people were required to undergo preventive chemotherapy against the helminth globally in 2015.
The ranges of Plasmodium and Schistosoma overlap in much of the tropical world, leading to the potential for a great many coinfections of the two parasitic species. It has, for example, been estimated that there may be a greater than 30% coinfection rate among children in Sub-Saharan Africa [2]. Given the importance of such coinfections, interactions between Plasmodium and Schistosoma have been extensively studied both in nature, and using animal models in the laboratory.
Epidemiological studies on coinfections have often produced contrasting results: some reports contend that Schistosoma infection can increase susceptibility to Plasmodium falciparum [3][4][5], whilst others document a protective effect on P. falciparum incidence [6][7][8][9]. Differences in study design, genetic background of host populations and other environmental factors presumably contribute to these conflicting results.
Most laboratory-based animal studies have shown an exacerbation of malaria parasitaemia in Schistosoma infected mice [10][11][12][13] whilst others have revealed a protective effect of Schistosoma infection against experimental cerebral malaria and associated mortality [14][15][16][17]. In experimental S. mansoni infections, it is known that eggs deposited in the liver induce a strong Th2 type immune response [18]. Previous work has suggested that the exacerbation of malaria parasitaemia and protection against experimental cerebral malaria were mediated by a polarized Th2 immune environment which down-modulates inflammatory responses [10,15].
The interactions between Schistosoma and Plasmodium are mainly mediated via host immune responses [19,20]. Previous animal studies have investigated inter-species interactions using experimental infection with Schistosoma cercariae and Plasmodium-parasitized erythrocytes. As both parasites infect the liver at specific stages in their life cycle, we have focused on the immune reactions against those parasites in the liver. The major immunopathology in schistosomiasis is induced by egg-derived antigens in the liver, and hepatic immune cells develop immunity not only against pre-erythrocytic stages but also the blood stages of malaria parasites [21][22][23][24]. We therefore investigated whether an ongoing infection with S. mansoni could affect the rodent malaria parasite Plasmodium yoelii in the livers of mice challenged with sporozoites (SPZ).
We also evaluated whether S. mansoni infection affects malaria parasite gametocyte infectivity to mosquitoes, as it has been shown that the infectivity of malaria gametocytes decreases during the early stage of malaria infection due to host serum-mediated immunity [25][26][27].

Ethical statement
All mouse experiments were approved by the Institutional Animal Research Committee of Nagasaki University (No.1506181240) and performed according to Japanese law for the Humane Treatment and Management of Animals (Law No. 105 dated 19 October 1973 modified on 2 June 2006).

Mice
Six week-old female BALB/cCrSlc (hereafter referred to as BALB/c) and C57BL/6NCrSlc (hereafter referred to as B6) mice were purchased from Japan SLC, Inc. (Shizuoka, Japan). Six week-old female CBA/J mice were purchased from Charles River Laboratories Japan, Inc. (Kanagawa, Japan). IFN-γ-deficient (IFN-γ -/-) mice and IL-4-deficient (IL-4 -/-) mice were produced at RIKEN Yokohama Institute, Yokohama, Japan. All mice were maintained in the animal facilities of Nagasaki University with environmentally controlled, specific pathogen free conditions. Experiments were conducted with BALB/c mice unless otherwise specified.

Parasites and infections
The rodent malaria parasites Plasmodium yoelii yoelii 17x1.1pp [28] (hereafter referred to as P. yoelii) and Plasmodium berghei ANKA (hereafter referred to as P. berghei) were used throughout these experiments. A Puerto Rican strain of Schistosoma mansoni was maintained in the animal facilities of Nagasaki University by passage through Biomphalaria glabrata snails and ICR mice or Meriones unguiculatus (Mongolian jirds). In the coinfection model, experimental mice were percutaneously infected with 50 cercariae. In the intraportal infusion model, eggs were collected from the liver harvested from ICR mice after 7 weeks infection with 250 cercariae and stored at -30˚C until use. The mice were anesthetized and inoculated with 3,000 eggs in a total volume of 100 μL PBS via the portal vein.
Infections with P. yoelii and P. berghei were performed by i.v. inoculation with SPZ collected from Anopheles stephensi mosquitoes, as previously described [29]. 50, 500, or 1,500 SPZ in a total volume of 100 μl PBS were inoculated for assessment of malaria parasitaemia or malaria parasite liver burden. Malaria challenges were performed 10 weeks after S. mansonicercariae infection in the coinfection model and 3 weeks after S. mansoni egg inoculation in the intraportal infusion model.
In transmission experiments, mice were intravenously infected with one million P. yoeliiparasitized erythrocytes. On day 3, 4, and 5 after P. yoelii infection, six-day-post emergence female A. stephensi mosquitoes were allowed to feed on individual mice for 30 minutes and reared in the insectary at 24˚C with 80% humidity for seven days. Midguts were harvested from the mosquitoes and oocysts counted by light microscopy.

Malaria parasite density in the blood and the liver
Ten microliters of blood were collected from the tale-vein of mice at each time point. Livers were harvested 42 hours after malaria challenge infection. DNA was extracted from blood or livers using the EZ1 BioRobot (Qiagen N.V., Hilden, Germany) following the manufacturer's instructions. Real-time quantitative PCR (qPCR) was performed on DNA samples using the Applied Biosystems 7500 Real Time PCR system (Thermo Fisher Scientific, Inc., Massachusetts, USA). A master mix of the following reaction components was prepared: 4.75 μL water, 6.25 μL Power SYBR Green Master Mix (Qiagen N.V., Hilden, Germany), 0.25 μL forward primer (10 μM), and 0.25 μL reverse primer (10 μM). 5 μL DNA samples were added as PCR template. P. yoelii or P. berghei 18s gene was amplified using the primers 18s F1 5' GGAACG ATGTGTGTCTAACACAAGGA 3' and 18s R1 CGCGTGCAGCCTAGTATATCTAAGGAC A 3' (Table 1). Copy numbers of the parasite 18s gene were quantified with reference to a standard curve calculated from known numbers of plasmids containing the same gene sequence, as previously described [30]. The copy number of parasite 18s gene in each sample was standardized by the simultaneous quantification of the mouse glyceraldehyde 3-phosphate dehydrogenase (G3PDH) gene. The mouse G3PDH gene was amplified using the primers MmG3PDHF1 5' CATCTGAGGGCCCACTGAAG 3' and MmG3PDHR1 5' TGCTGTT GAAGTCGCAGGAG 3'( Table 1).

Monitoring of parasitaemia and gametocytaemia
Thin blood films from the tail-vein were prepared daily from day 2 to 8 after SPZ inoculation, and stained with Giemsa's solution. The numbers of infected and non-infected erythrocytes ware counted per 10 microscopic fields to calculate parasitaemia and gametocytaemia.

Flow cytometric analysis
Hepatic nonparenchymal cells were isolated after Schistosoma inoculation as follows. Livers taken from each experimental mouse were homogenized in 5 mL of culture medium (RPMI-1640 supplemented with 10% FCS and 1% penicillin/streptomycin) using gentleMACS Dissociator (Miltenyi Biotec, Bergisch Gladbach, Germany), filtered through a mesh and suspended in HBSS. To remove liver parenchymal cells, the cells were resuspended in 33% Percoll (GE Healthcare UK Ltd., Buckinghamshire, England) containing 2.5mL of 5000U/5mL Heparin (Mochida Pharmaceutical Co., Ltd., Tokyo, Japan) and were centrifuged at 900 g for 20 min at room temperature. The pellet was resuspended in RBC lysis buffer then washed in HBSS, and resuspended in the culture medium.

Data analysis
Data analyses were performed using Microsoft Excel 2010 (Microsoft Corporation, Washington, USA) and GraphPad Prism6 (GraphPad Software, Inc., California, USA). Significance between control and treatment groups was determined with Student's t-tests. Survival and protection rates were statistically examined using the log-rank test. P values less than 0.05 were considered significant.

S. mansoni infection significantly reduces the number of malaria parasites in the liver following sporozoite inoculation
To investigate the impact of chronic S. mansoni infection on the growth of malaria parasites in the liver, BALB/c mice were infected with 50 cercariae subcutaneously 10 weeks prior to i.v. inoculation of 1,500 P. yoelii sporozoites (SPZ). The number of copies of the malaria parasite 18s gene present in the liver of mice was measured by qPCR 42 h after SPZ challenge. Parasitaemia was monitored daily from day 2 to 8 post SPZ challenge. Malaria parasite liver burden was significantly reduced to one-tenth or less in S. mansoniinfected mice compared with non-S. mansoni infected controls ( Fig 1A). This reduction in the numbers of malaria parasites in the liver resulted in a delay in the onset of parasitaemia in the S. mansoni-infected group. All mice in both groups developed blood stage malaria parasite infection, and the peak parasitaemia was not affected by S. mansoni infection (Fig 1B). There was no difference in mortality between S. mansoni-infected and non-infected mice (Fig 1C).

The reduction of liver-stage malaria parasite burden induced by S. mansoni infection was not constrained by malaria species or mouse strains
The genetic background of inbred mouse strains can determine the course of experimentally induced malaria parasite infections [31,32]. We examined the malaria parasite liver burden in three different mouse strains; BALB/c, B6, and CBA/J in coinfections with S. mansoni. Female BALB/c, B6, and CBA/J mice were challenged with two different Plasmodium species: P. yoelii and P. berghei in order to investigate whether the reduction of malaria parasite liver burden occurs across mouse and parasite species. Both B6 and CBA/J mice showed significant reduction in P. yoelii liver burden in S. mansoni-infected mice, consistent with the results observed in BALB/c mice (Fig 2A). The liverstage growth of P. berghei was also reduced in S. mansoni-infected mice in all mouse strains examined (Fig 2B).

S. mansoni infection significantly reduces liver-stage malaria parasite burden from 2 h following SPZ inoculation
There are several potential mechanisms for the reduction in malaria parasite burden at 42 hours post SPZ inoculation in S. mansoni infected mice. It is possible, for example, that antibodies raised against schistosomes may offer protection against sporozoites through crossreactivity [33]. It is also possible that liver fibrosis caused by schistosomiasis may physically impede the invasion of hepatocytes by sporozoites. Alternatively, there may be suppression of P. yoelii hepatocytic stages in the liver mediated by an altered immune environment in the liver caused by the presence of S. mansoni eggs in the organ. In order to investigate at which point S. mansoni-mediated reduction of malaria parasite in the liver occurs, we measured the number of malaria parasites in both the blood and the liver at time points throughout the 48 h growth of parasites in the liver. Ten microliters of blood were sampled from mouse tail-veins at 10, 30, 60, and 120 min, and livers were harvested at 2, 12, 24, and 42 h after i.v. inoculation of 1,500 SPZ. We found that malaria parasite load in the blood was not significantly different in infected compared with non-infected mice at any time point following SPZ inoculation ( Fig  3A). In contrast, the number of malaria parasites in the liver were significantly reduced in S. mansoni infected mice from the first 2 h of development in the liver, with the proliferation rate from 2 to 24 h of liver stage parasites also suppressed in these mice (Fig 3B).

Intraportal inoculation of frozen S. mansoni-eggs significantly reduces liver-stage malaria parasite burden
As eggs are the principle cause of immunopathology in schistosomiasis [34], we assessed whether direct inoculation of S. mansoni eggs to the portal vein of mice could also reduce liver-stage malaria parasite growth. Female B6 mice were inoculated with 3,000 or 10,000 frozen S. mansoni-eggs in a total volume of 100 μL PBS into the portal vein. Control mice were inoculated with 100 μL PBS. Three weeks after egg inoculation, each group was challenged with 1,500 SPZ of P. yoelii. Liver-stage malaria parasite burden was significantly reduced in both groups inoculated with S. mansoni-eggs (Fig 4A).
S. mansoni infection causes severe liver damage associated with eosinophilic granulomas, collagen deposition, and fibrosis due to immunologic reactions to Schistosoma eggs trapped in the tissues. However, livers harvested from the intraportal infusion group did not show conspicuous damage macroscopically. Correspondingly, livers inoculated with Schistosoma eggs developed less and smaller granulomas compared with those harvested from the S. mansonicercariae infection group (Fig 4B and 4C).
Schistosoma infection is known to induce Th2-biased immune responses in mice. However, egg inoculation induced the infiltration of various immune cells into the liver with little pathology. Interferon-mediated innate immune responses are important modulators of malaria parasite growth in the liver [35]. We therefore focused not only on CD4 + T-bet + cells and CD4 + GATA-3 + cells but also NK, NKT, and γδT cells which have previously been implicated as important sources of interferon-gamma (IFN-γ) in the liver during Plasmodium infection. The representative FACS plots are shown in the supplementary graph (S1 Fig). As shown in Fig 4D and 4E, the upregulation of the immune reaction induced by S. mansoni eggs in B6 mice was similar to that induced by S. mansoni-cercariae infection in BALB/c mice. Both CD4 + T-bet + cells and CD4 + GATA-3 + cells significantly increased following intraportal inoculation of S. mansoni-eggs as well as after S. mansoni-cercariae infection. In both cases, the increase of CD4 + GATA-3 + cells was the largest among immune cells. The numbers of NK cells significantly increased only during the early phase, one week after intraportal inoculation of S. mansoni-eggs or 8 weeks after S. mansoni-cercariae infection. The numbers of NKT cells also significantly increased in both cases and this increase was greater in the group inoculated with S. mansoni-eggs. There was no significant increase in the numbers of γδT cells in either group. B6 mice infected with S. mansoni-cercariae showed a similar pattern of upregulation of immune cells (S2 Fig).

The reduction of liver-stage malaria parasite burden was reversed in IFN-γdeficient and IL-4-deficient mice
We hypothesised that the reduction of liver-stage malaria parasite burden is dependent on the immune environment such as IFN-γ production from host immune cells accumulated and activated by the presence S. mansoni-eggs in the liver. IFN-γ is a key cytokine for the control of malaria parasites and is induced in the liver during S. mansoni infection. Interleukin-4 (IL-4) is a major mediator for the induction of the immune response against S. mansoni. Therefore, we examined the impact of IFN-γ and IL-4 on liver-stage malaria parasite burden using IFNγ-deficient (IFN-γ -/-) and IL-4-deficient (IL-4 -/-) mice.
IFN-γ -/and IL-4 -/mice along with wild-type B6 (B6 WT) mice were inoculated 3,000 S. mansoni-eggs 3 weeks prior to SPZ challenge. Liver-stage malaria parasite burden was measured 42 h after i.v. challenge with 1,500 P. yoelii SPZ. Consistent with the result shown in Fig  4A, malaria liver burden was significantly reduced in the intraportal infusion group; however, this reduction was abrogated in IFN-γ -/and IL-4 -/mice (Fig 5).

S. mansoni infection inhibits the development of blood stage malaria following i.v. inoculation of entomologically relevant numbers (50) of SPZ
Through challenge with 1,500 SPZ of P. yoelii, we have demonstrated the impact of S. mansoni infection on liver-stage malaria parasite burden; this number of sporozoites, however, would not be expected to be inoculated naturally during a mosquito bite. In order to mimic the numbers inoculated by a mosquito bite, we reduced the challenge dose from 1,500 to 50 SPZ and examined the outcome of Plasmodium infection.
BALB/c mice (N = 16) infected with 50 S. mansoni-cercariae 10 weeks previously along with non-infected controls were challenged with 50 P. yoelii SPZ. Seven out of 16 S. mansonicoinfected mice did not develop malaria parasitaemia, in contrast to all non-S. mansoni infected control mice developing blood-stage malaria on day 5 after SPZ challenge (Fig 6A). When blood stage malaria parasite infection became patent in coinfected mice, it increased rapidly and matched the peak parasitaemia of non-S. mansoni infected controls (Fig 6B).

Plasmodium yoelii infections in mice infected with S. mansoni were less infectious to mosquitoes
We measured the influence of S. mansoni on the infectivity of malaria parasites to mosquitoes. We allowed Anopheles stephensi mosquitoes to feed on mice infected with S. mansoni + P. yoelii or P. yoelii alone on days 3 and 4 post-infection, when P. yoelii is at its most infectious to mosquitoes. A minimum of eight mosquitoes were allowed to feed on each of 5 mice per group on each day, and the resulting infectivity measured by assessing the proportion of mosquitoes that were infected with oocysts eight days post-feeding, and through quantifying the numbers of oocysts per infected mosquito per group on each day of feeding. Prior to mosquito feeding, the parasitaemia and number of gametocytes circulating in the peripheral blood of mice was calculated via microscopy of thin blood films following i.v. inoculation of P. yoelii-parasitized erythrocytes.
The number of gametocytes circulating in S. mansoni-infected mice was significantly higher than that of mice without S. mansoni infection on days 3 and 4 ( Fig 7B). Despite this, the proportion of mosquitoes infected with oocysts eight days post blood feed was significantly higher for the S. mansoni uninfected mic, compared to mice infected with S. mansoni (Fig 7C). Furthermore, oocyst numbers were significantly reduced in mosquitoes that had fed on mice infected with S. mansoni, compared to uninfected controls (Fig 7D).

Discussion
We have examined the impact of schistosome infection on malaria pathology and transmission capacity in a mouse model. Previous animal studies on the interactions between Plasmodium and Schistosoma have focused exclusively on the blood stage pathology of the malaria parasite infection, and have consistently shown that parasitaemia is enhanced in the presence of the helminth [10][11][12][13]. We have investigated the effect of helminth-malaria parasite coinfection on the growth of rodent malaria parasites in the liver, and show that S. mansoni-coinfection significantly reduces Plasmodium parasite numbers in the liver following sporozoite inoculation. S. mansoni-coinfection with malaria parasites resulted in a large reduction in the percentage of mice developing blood stage malaria parasite infection when mice were challenged with numbers of SPZ similar to those inoculated during natural inoculation through the bites of infected mosquitoes [36][37][38][39].
These results might suggest a possible explanation for the often contradictory results observed in epidemiological studies. Some prospective cohort studies in human populations have suggested a protective effect of Schistosoma infection on Plasmodium infection: Lyke et al. concluded that children infected with Schistosoma haematobium showed increased time to first clinical malaria infection and fewer malaria episodes over the follow-up period compared to children without S. haematobium infection [7]. Doumbo et al. also showed that people coinfected with S. haematobium showed significant delays in time-to-first malaria episode [9]. Hürlimann et al. demonstrated that the Plasmodium parasitaemia incidence rate of children infected with S. mansoni increased after treatment with praziquantel; in contrast, children infected with hookworms had reduced Plasmodium parasitaemia rates following albendazole treatment [40]. Our results, which show that S. mansoni infection reduces the malaria parasite burden in the liver and the number of hosts developed blood stage infection following SPZ inoculation, may account for some of the observed protective effects of Schistosoma infection against malaria.
The intraportal inoculation of S. mansoni eggs reduced malaria parasite burden in the liver to the same extent as inoculation of cercariae. As shown in Fig 4, intraportal inoculation of S. mansoni-eggs induced an increase in the numbers of various lymphocytes, including both Th1 and Th2 cells, NK, NKT, and γδT cells in the liver without apparent fibrotic or granulomatous liver damage. This suggests that the suppression of the growth of malaria parasites in the liver is not mediated by physiological or morphological changes of the liver environment provoked by S. mansoni-eggs, but rather by changes in the hepatic immune microenvironment.
Previous animal studies have shown that interferon-mediated innate immune responses are important modulators of malaria parasite growth in the liver [41]; Miller et al. have shown that IFN-γ production from NKT cells (but not NK cells or T cells) is critical in reducing malaria parasite burden in the liver [35]. We initially hypothesised that malaria parasite liver burden would increase in S. mansoni-coinfected mice as S. mansoni infection has been shown to induce robust Th2 type immune responses with a corresponding downregulation of Th1 responses [42]. However, contrary to our expectations, malaria parasite liver burden was significantly reduced in S. mansoni-coinfected mice. Both IFN-γ and IL-4 are required for this effect, and both the inoculation of S. mansoni eggs to the hepatic portal vein, and the infection of mice with cercariae were shown to increase innate immune cells and both Th1 and Th2 cells in the liver (S2 Fig). Taken together, this suggest that possible sources of IFN-γ and IL-4 include not only innate immune cells such as NKT cells but also helper T cells accumulated and activated by S. mansoni infection in the liver [43].
IFN-γ -/and IL-4 -/mice had lower numbers of malaria parasites in the liver after SPZ challenge compared with B6 WT mice even without S. mansoni infection. The experiment was repeated twice with similar results. These results are in agreement with previous studies with IL-4 -/mice although the mechanisms are not fully understood. IL-4 -/mice are known to be more resistant to sporozoite infection than wild-type mice owing to increased NK cell numbers and expression of inducible nitric oxide synthase in the liver [44]. Since the reduction of malaria parasite burden in the liver in co-infected mice was reversed in IFN-γ -/and IL-4 -/mice, we assume that the mechanism of protection is both IFN-γ and IL-4 dependent. To elucidate the mechanisms underlying this phenomenon, further research into the interactions between Plasmodium and Schistosoma are required.
We demonstrated that the infectivity of Plasmodium gametocytes to Anopheles mosquitoes dropped abruptly from a peak on day 3 to zero on day 4 in control mice not infected with S. mansoni. This phenomenon is known in animal models as malaria infection crisis and has previously been reported to occur in numerous Plasmodium species [45]. The mechanisms behind this loss of infectivity have not been fully explained, but previous animal studies suggest that cytokines such as tumor necrosis factor (TNF), IFN-γ, interleukin-6 (IL-6), and nitric oxide may inhibit the development of gametocytes [25][26][27]. The infectivity of P. yoelii gametocytes in S. mansoni infected mice was significantly reduced on day 3 post-inoculation which is the peak of infectivity in non-S. mansoni infected mice. Observing higher levels of IFN-γ in the sera of S. mansoni infected mice on day 3 and day 4 (S3 Fig), we assume that pre-existing S. mansoni infection induces activation of the host immune environment and this leads to an accelerated infection crisis. Although the mechanisms remain to be elucidated, this experimental transmission model suggests that S. mansoni infection may reduce malaria transmissibility from mosquitoes to mice.
In conclusion, the presence of S. mansoni infection dramatically reduced the number of rodent malaria parasites in the liver. This reduction leads to the inhibition of the development of blood stage malaria parasite infection following inoculation of biologically relevant numbers of sporozoites. We also demonstrate that P. yoelii gametocyte infectivity to mosquitoes is significantly reduced in the presence of S. mansoni in co-infected mice. These results imply that S. mansoni infection can reduce malaria transmission both from mosquitoes to mice and from mice to mosquitoes, and may explain some of the protective effects of Schistosoma infection against malaria described in previous epidemiological studies.