INTRODUCTION
Protozoan and helminth infections in pregnant women are as prevalent as in the rest of the community where those individuals live. Many parasitic infections, both protozoa and helminth, are life-long, chronic infections which cannot be cleared by the host immune system. Chronicity of infections during pregnancy such as schistosomiasis, filariasis, malaria and Trypanosoma cruzi mean that the foetus was gestated and breast-fed by a mother with chronic infections. These children have been exposed to parasitic antigens and maternal antibodies in utero, which may have modulated their immune response later in life as was demonstrated for different helminth antigens more than twenty five years ago (Weil et al. Reference Weil, Hussain, Kumaraswami, Phillips and Ottesen1983). A decreased IFNγ response to heterologous antigens has been found in children born to mothers with filariasis and schistosomiasis (Malhotra et al. Reference Malhotra, Mungai, Wamachi, Kioko, Ouma, Kazura and King1999). A common observation in many parasitic infections is long-term antigen persistence resulting in a constant simulation of the immune system, and it has been suggested that the lack of ability to control these infections properly is due, in part, to exhaustion of the CD4+ T cell memory pool resulting in insufficient maintenance of CD8+ effector memory T cells (Brake, Reference Brake2003).
The foetus, neonate and infant are more susceptible to infections than older children and adults. This is considered to be partly due to the immaturity of their immune system especially the ability to generate cell mediated immunity (Adkins, Leclerc and Marshall-Clarke, Reference Adkins, Leclerc and Marshall-Clarke2004). Helminth infections influence the immune system of the host and are associated with a down-regulated immune response typically with polarization towards a Th2 response (Maizels and Yazdanbakhsh, Reference Maizels and Yazdanbakhsh2003; Wilson and Maizels, Reference Wilson, Maizels, Capron and Trottein2006).
This review discusses data on how maternal and congenital infections affect the parasite-specific immune response in the child and considers the protozoan infections malaria, Toxoplasma gondii and Trypanosoma cruzi and the helminth infections Onchocerca volvulus, Wuchereria bancrofti and schistosomiasis; parallels are drawn to congenital or perinatal viral infections with Rubella virus, cytomegalovirus and hepatitis B virus.
MALARIA
It is well known that the immune system is skewed towards a Th2 response in pregnancy in order to minimize the risk of a ‘host versus graft’ reaction. This reduces the ability of the pregnant women to fully control some infections compared to non-pregnant women. This has been well documented for Plasmodium falciparum malaria (Blacklock and Gordon, Reference Blacklock and Gordon1925; Bray and Anderson, Reference Bray and Anderson1979; Brabin, Reference Brabin1983; Bruce-Chwatt, Reference Bruce-Chwatt1983) where a higher parasite density, especially in the second trimester, and a higher spleen rate and enlarged spleen have been found (McGregor, Reference McGregor1984; Brabin et al. Reference Brabin, Brabin, Sapau and Alpers1988). Malaria infections during pregnancy contribute to anaemia in the mother, reduced birth weight and premature birth (Cannon, Reference Cannon1958; Gilles et al. Reference Gilles, Lawson, Sibelas, Voller and Allan1969; Brabin et al. Reference Brabin, Ginny, Sapau, Galme and Paino1990; Nosten et al. Reference Nosten, ter-Kuile, Maelankirri, Decludt and White1991), but immunity to malaria in the pregnant women increases with increasing parity (Desowitz, Reference Desowitz1992).
Congenital malaria infections of around 5% have been reported based on microscopy (Covell, Reference Covell1950; Larkin and Thuma, Reference Larkin and Thuma1991; Tobian et al. Reference Tobian, Mehlotra, Malhotra, Wamachi, Mungai, Koech, Ouma, Zimmerman and King2000), but rates up to 33% have been reported using PCR (Tobian et al. Reference Tobian, Mehlotra, Malhotra, Wamachi, Mungai, Koech, Ouma, Zimmerman and King2000; Xi et al. Reference Xi, Leke, Thuita, Zhou, Leke, Mbu and Taylor2003), and soluble malaria antigens can cross the placenta from infected mothers to the foetus (Druilhe, Monjour and Gentilini, Reference Druilhe, Monjour and Gentilini1976; Desowitz, Reference Desowitz1988; King et al. Reference King, Malhotra, Wamachi, Kioko, Mungai, Wahab, Koech, Zimmerman, Ouma and Kazura2002). A study from Malawi found that placental malaria infection was associated with lower perinatal mortality among normal birth weight (⩾2500 g) infants (OR 0·35, 95% CI: 0·14, 0·92) (McDermott et al. Reference McDermott, Wirima, Steketee, Breman and Heymann1996). This is difficult to explain, but a reduced immune response leading to diminished pathology could be one possibility. The immune response in the neonate is compromised. A study from Cameroon of cord blood samples from 164 newborns found a highly reduced ability of T cells to secrete IFNγ after exposure to malaria antigens as well as leucoagglutinin, but equal levels of IL-2 and IL-4 (Fievet et al. Reference Fievet, Ringwald, Bickii, Dubois, Maubert, Hesran, Cot and Deloron1996). The same study found an increased risk of children having parasites in the blood if the mother had malaria parasites in the placenta when the child was born (Le Hasran et al. Reference Le Hasran, Cot, Personne, Fievet, Dubois, Beyemé, Boudin and Deloron1997; Deloron et al. Reference Deloron, Dubois, Le Hasran, Richie, Fievet, Cornet, Ringwald and Cot1997). A study from Tanzania found that children born to multigravida women with placental malaria had a higher risk of parasitaemia compared to multigravida without placental malaria, but found no difference in children born to primigravida or secundigravida mothers (Mutabingwa et al. Reference Mutabingwa, Bolla, Li, Domingo, Li, Fried and Duffy2005).
A study of the pyrogenic threshold in children and adults found that it was twenty times higher in children compared to adults (Miller, Reference Miller1958); these data were confirmed in a later study from Liberia, which found a six fold decrease in parasite density in adult patients with a body temperature of ⩾38°C (Petersen et al. Reference Petersen, Høgh, Marbiah, David and Hanson1991). This could be explained by an immune tolerance in children which resulted in reduced control over parasite density but also diminished symptomatology.
The immunological mechanisms that confer the relative inability to control malaria in children born to mothers with placental malaria are not completely understood. A study from The Gambia found poor antigen presenting cell, APC, function in newborn infants born to mothers with placental malaria infection (Ismaili et al. Reference Ismaili, van der Sande, Holland, Sambou, Keita, Allsopp, Ota, McAdam and Pinder2003). Another study from Cameroon compared children born to mothers treated for malaria during pregnancy with children born to mothers with malaria during pregnancy and found that IL-10(+) T cells in utero may contribute to suppression of the APC function and Pf Ag-induced Th1 responses were reduced in infants from mothers with placental malaria (Brustoski et al. Reference Brustoski, Moller, Kramer, Petelski, Brenner, Palmer, Bongartz, Kremsner, Luty and Krzych2005, Reference Brustoski, Moller, Kramer, Hartgers, Kremsner, Krzych and Luty2006). A study of newborn infants in Kenya found that neonates were more likely to have a Th2 response to malaria antigens if born to multigravida mothers compared to primi- and secundigravida mothers, and the authors suggested that active suppression or T-cell anergy to malaria specific antigens developed in some newborn infants (Malhotra et al. Reference Malhotra, Mungai, Muchiri, Ouma, Sharma, Kazura and King2005a). Inhibition of co-simulation by dendritic cells has been found in one study (Urban et al. Reference Urban, Ferguson, Pain, Willcox, Plebanski, Austyn and Roberts1999) and insufficient costimulatory signals to neonatal T cells from immature antigen presenting cells has also been described (Broen et al. Reference Broen, Brustoski, Engelmann and Luty2007). It has been proposed that possible refractoriness to toxin-induced Toll-like receptor mediated signalling is due to cross reactivity between endotoxin and malaria glycosylphosphatiylinositol, GPI (Boutlis et al. Reference Boutlis, Yeo and Anstey2006).
The molecular basis for placental malaria is the sequestration of P. falciparum-infected erythrocytes expressing the erythrocyte membrane protein 1, PfEMP1 antigen belonging to the var genes (Baruch et al. Reference Baruch, Ma, Singh, Bi, Pasloske and Howard1997; Su et al. Reference Su, Heatwole, Wertheimer, Guinet, Herrfeldt, Peterson, Ravetch and Wellems1995). PfEMP1 mediates adhesion to a range of host surface receptors which have been identified as targets for P. falciparum binding (Craig and Scherf, Reference Craig and Scherf2001). In the placenta, the P. falciparum-infected erythrocytes accumulate in the intervillous space (Yamada et al. Reference Yamada, Stekete, Abramowsky, Kida, Wirima, Heymann, Rabbege, Breman and Aitkawa1989). Both trophozoites and schizonts adhere to the endothelium in the placenta in contrast to other organs where only schizonts adhere (Pouvelle et al. Reference Pouvelle, Buffet, Lepolard, Scherf and Gysin2000). The observations that primigravidae in particular have a high risk for placental sequestration and low birth weight infants, a particular problem in primigravidae with malaria, have lead to much interest in the development of immunity to var antigens in particular PfEMP1 (Fried et al. Reference Fried, Nosten, Brockman, Brabin and Duffy1998; Riecke et al. Reference Riecke, Staalsoe, Koram, Akanmori, Riley, Theander and Hviid2000). It is, however, unclear how the sequestration of P. falciparum-infected erythrocytes in the placenta results in low birth weight infants. Sequestration of P. falciparum-infected erythrocytes results in an inflammatory response in the placenta (Fried et al. Reference Fried, Nosten, Brockman, Brabin and Duffy1998; Moormann et al. Reference Moormann, Sullivan, Rochford, Chensue, Bock, Nyirenda and Meshnik1999). In particular, monocyte infiltration in the placenta has been identified as an important predictor of low birth weight infants (Ordi et al. Reference Ordi, Ismail, Ventura, Kahigwa, Hirt, Cardesa, Alonso and Menendez1998; Menendez et al. Reference Menendez, Ordi, Ismail, Ventura, Aponte, Kahigwa, Font and Alonso2000; Abrams et al. Reference Abrams, Brown, Chensue, Turner, Tadesse, Lema, Molyneux, Rochford, Meshnick and Rogerson2003; Rogerson et al. Reference Rogerson, Pollina, Getachew, Tadesse, Lema and Molyneux2003). A parallel may be drawn to recent results from congenital CMV infection which is associated with an increased thickness of the placenta probably as a result of placental inflammation. In CMV infections, the placental inflammation is partly responsible for the pathology in the child with congenital CMV (Torre et al. Reference Torre, Nigro, Mazzocco, Best and Adler2006; Schleiss, Reference Schleiss2006). Thus the inflammatory response in the placenta as a result of malaria infection may be important for the intrauterine growth of the foetus and perhaps also be partly responsible for antigen leakage over the placenta.
TOXOPLASMA GONDII
Congenital infection with Toxoplasma gondii is transmitted to about 20% of the offspring. Transmission rates increase with gestation and are below 5% in the first months of pregnancy increasing to approximately 80% in the last few weeks before birth (Dunn et al. Reference Dunn, Wallon, Peyron, Petersen, Peckham and Gilbert1999). Risk of pathology in the foetus is high in the beginning of pregnancy and decreases with increasing gestational age (Dunn et al. Reference Dunn, Wallon, Peyron, Petersen, Peckham and Gilbert1999; Gras et al. Reference Gras, Wallon, Pollak, Cortina-Borja, Evengard, Hayde, Petersen and Gilbert2005). Children with congenital T. gondii infection may develop recurrent retinochoroiditis later in life; by contrast, individuals infected after birth rarely experience recurrent attacks of retinochoroiditis except when immunocompromised. A study from France of 327 children with confirmed congenital T. gondii infection found that 24% (n=79) had had at least one lesion and 7% (n=23) had at least one new lesion diagnosed during the follow-up (Wallon et al. Reference Wallon, Kodjikian, Binquet, Garweg, Fleury, Quantin and Peyron2004). It has also been observed that T. gondii infections can reactivate during pregnancy in otherwise immunocompetent women (Garweg et al. Reference Garweg, Scherrer, Wallon, Kodjikian and Peyron2005). It is not known why congenitally infected children may experience recurrent attacks of retinochoroiditis. Anergy of T cells to T. gondii-specific antigens have been described in a child with congenital toxoplasmosis (McLeod, Beem and Estes, Reference McLeod, Beem and Estes1985) and a reduced number of circulating CD4+ T cells has been found in newborns with congenital T. gondii infection compared to non-infected children (Hohlfeld et al. Reference Hohlfeld, Forestier, Marion, Thulliez, Marcon and Daffos1990). In another study, children with congenital toxoplasmosis had reduced T cell proliferation and impaired production of IL2 and IFNγ to soluble T. gondii antigens compared to those individulals with acquired T. gondii infections, demonstrating that congenital infections can impair the specific T. gondii cellular immune response (Yamamoto et al. Reference Yamamoto, Vallochi, Silveira, Filho, Nussenblatt, Cunha-Neto, Gazzinelli, Belfort and Rizzo2000). Children with congenital T. gondii infections have impaired γδ and αβ T cell response. The γδ T cell response was normalised by the age of one year but the αβ T cell response was still anergic and the children were not followed further (Hara, Reference Hara, Ohashi, Yamashita, Abe, Hisaeda, Himemo, Good and Takeshita1996). A recent study found an impaired stimulation index to TgGRA1 antigen in children with congenital toxoplasmosis and showed that the relative unresponsiveness improved with age (Guglietta et al. Reference Guglietta, Beghetto, Spadoni, Buffolano, del Porto and Gargano2007).
TRYPANOSOMA CRUZI, CHAGAS DISEASE
In a recent study of children born to T. cruzi-infected mothers, 9% (27/302) of newborns were infected (Mora et al. Reference Mora, Sanchez-Negrette, Marco, Barrio, Ciaccio, Segura and Basombrio2005) and previous studies have found transmission in between 2% and 12% of children born to women with chronic T. cruzi infection (Azogue, la Fuente and Darra, Reference Azogue, Fuente and Darra1985). Trypanosoma cruzi infection in the pregnant mother is associated with low Apgar scores, low birth weight, prematurity, respiratory distress syndrome and increased mortality rates in the newborns (Torrico et al. Reference Torrico, Vega, Suarez, Tellez, Brutus, Rodriguez, Torrico, Schneider, Truyens and Carlier2006). It has been estimated that approximately 1 per 1000 newborns in Argentina in 1993 were congenitally infected with T. cruzi (Gürtler, Segura and Cohen, Reference Gürtler, Segura and Cohen2003) and the same number has been found in Brazil (Bittencourt, Reference Bittencourt1992). The outcome of congenital T. cruzi infection in the child is not known in detail, but two case reports of congenitally infected siblings both giving birth to congenitally infected children indicate that children congenitally infected with T. cruzi can become chronically infected and remain infectious for life (Schenone et al. Reference Schenone, Gaggero, Sapunar, Contreras M. C. and Rojas2001).
Maternal antigens are transferred across the placenta and modulate the foetal immune system. NK cells from newborns with congenital T. cruzi infection show an impaired IFNγ production and reduced granzyme B release compared to uninfected newborns (Hermann et al. Reference Hermann, Alonso-Vega, Berthe, Truyens, Flores, Cordova, Moretta, Torrico, Braud and Carlier2006), but develop a fully mature cytotoxic T cell response (Hermann et al. Reference Hermann, Truyens, Alonso-Vega, Even, Rodriguez, Berthe, Gonzalez-Merino, Torrico and Carlier2002).
A study of T. cruzi-uninfected newborn children born to T. cruzi-infected and uninfected mothers showed that at birth, the newborns born to infected mothers had lower CD3+ and CD4+ T cells but higher numbers of MHC class II co-expression, but at six months of age there was no difference in the markers studied between the two groups of children (Neves et al. Reference Neves, Eloi-Santos, Ramos, Rigueirinho, Gazzinelli and Correa-Oliveira1999). Children born to T. cruzi-infected mothers showed up-regulation of the pro-inflamatory cytokines IL1β, IL6 and TNFα (Vekemans et al. Reference Vekemans, Truyens, Torrico, Solano, Torrico, Rodriguez, Alonso-Vega and Carlier2000). Whether this protects the newborn child against transmitted T. cruzi parasites is not known, but a recent study from Argentina found that children with congenital T. cruzi infection were often asymptomatic, although the same study demonstrated second generation transmission from congenitally infected mothers, showing that children with congenital T. cruzi infection, although asymptomatic, may remain infective for life (Negrette, Mora and Basombrio, Reference Negrette, Mora and Basombrio2005).
FILARIASIS, ONCHOCERCA VOLVULUS AND WUCHERERIA BANCROFTI
Children born to mothers infected with O. volvulus during pregnancy had higher parasite densities when infected with O. volvulus later in life and showed detectable parasitaemia at an earlier age compared to children born to mothers without O. volvulus infection during pregnancy eighteen years earlier (Kirch et al. Reference Kirch, Duerr, Boatin, Alley, Hoffmann, Schulz-Key and Soboslay2003). Maternal onchocerciasis was a more important risk factor for childhood infection than the O. volvulus transmission intensity (Kirch et al. Reference Kirch, Duerr, Boatin, Alley, Hoffmann, Schulz-Key and Soboslay2003). Both Th1 and Th2 cytokine production is impaired in children born to mothers with O. volvulus during pregnancy, which suggest that intrauterine exposure to O. volvulus antigens results in hypo-responsiveness (tolerance) to these antigens (Elson et al. Reference Elson, Days, Calvopina, Paredes, Araujo, Guderian, Bradley and Nutman1996). Umbilical cord lymphocytes from children born to O. volvulus-infected mothers showed a reduced response to O. volvulus antigens compared to children born to non-infected mothers (Soboslay et al. Reference Soboslay, Geiger, Drabner, Banla, Batchassi, Kowu, Stadler and Schulz-Key1999). A study from Kenya found a Th2- biased immune response and an increased risk of filarial infection in children born to filarial infected mothers (Malhotra et al. Reference Malhotra, Mungai, Muchiri, Ouma, Sharma, Kazura and King2005b, Reference Malhotra, Mungai, Wamachi, Tisch, Kioko, Ouma, Muchiri, Kazura and King2006).
A study of 21 Polynesian children aged 17 to 19 years showed that children born to mothers who had not had bancroftian filariasis during pregnancy responded to microfilarial antigens but children born to mothers with filariasis during pregnancy had a low IFNγ response (Steel et al. Reference Steel, Guinea, McCarthy and Ottesen1994). The mechanisms behind the hypo-responsiveness of these children to microfilarial antigens even seventeen years after birth are unclear, but demonstrate that intrauterine antigen exposure can have long-lasting effects on the immune response in the offspring (Clark, Reference Clark1994). In a family cluster of bancroftian filariasis, children born to infected mothers were more likely to be infected compared to children born to mothers without microfilariasis during pregnancy (Lammie et al. Reference Lammie, Hitch, Allen, Hightower and Eberhard1991).
SCHISTOSOMIASIS
Administration of soluble Schistosoma mansoni antigens to pregnant rats results in hypo-reponsiveness to granuloma formation in offspring subsequently infected with S. mansoni (Hang, Boros and Warren, Reference Hang, Boros and Warren1974). Children born to mothers infected with Schistosoma mansoni recognised schistosomal antigens in skin tests and macrophage migration tests, demonstrating that a high proportion of these children had been exposed to Schistosoma mansoni antigens during foetal life (Tachon and Borojevic, Reference Tachon and Borojevic1978). Induction of prenatal tolerance to parasite antigens was first demonstrated for schistosomiasis (Lewert and Mandlowitz, Reference Lewert and Mandlowitz1969) and it has been suggested that early life sensitization will induce tolerance and bias the immune response towards a Th2 answer (Ridg, Fuchs and Matzinger, Reference Ridge, Fuchs and Matzinger1996). Children born to mothers with schistosomiasis or filariasis during pregnancy produced significantly less IFNγ to tuberculin after BCG immunization compared to children born to mothers without schistosomiasis or filariasis during pregnancy (Malhotra et al. Reference Malhotra, Mungai, Wamachi, Kioko, Ouma, Kazura and King1999).
CONGENITAL VIRUS INFECTIONS
Congenital infections with viruses like Rubella, cytomegalovirus and hepatitis B result in a prolonged state of virus excretion and, in the case of hepatitis B, life-long inability to clear the virus. Lessons from these viral infections can be applied to parasitic infections and help to estimate the impact of infections during pregnancy and the perinatal period.
It is well documented in congenital Rubella infections, that the cell mediated immune response is impaired and that this impairment is most pronounced when infection takes place at the beginning of pregnancy and less so towards the end of pregnancy (Buimovici and Cooper, Reference Buimovici and Cooper1985; Ou et al. Reference Ou, Chong, Tingle and Gillam1993).
Transmission from mother to foetus is also dependent on gestational age in primary, maternal cytomegalovirus, CMV, infection (Stagno et al. Reference Stagno, Pass, Dworsky and Alford1982). However, a study of children with congenital CMV infections found that the CD8+ T cell response was mature and functional (Marchant et al. Reference Marchant, Appay, Sande M., Dulphy, Liesnard and Kidd2003).
The relationship between the risk of developing a life-long, chronic infection and age is particularly well documented for hepatitis B, HBV, infection, and it is estimated that infection at birth will result in chronic infection in 90% of individuals compared to less than 5% in adults over 15 years of age (Lok et al. Reference Lok and McMahon2001; Ganem and Prince, Reference Ganem and Prince2004). The immunological mechanisms in chronic HBV infection are not known in detail, but a role for T-regulatory cells has been proposed (Accapezzato et al. Reference Accapezzato, Francavilla, Paroli, Casciaro, Chircu, Cividini, Abrignani, Mondelli and Barnaba2004). In adults, control of HBV infection relies primarily on cytotoxic T cells and individuals who fail to clear the infection, including neonates and young children, have a reduced early IFNγ response (Hui and Lau, Reference Hui and Lau2005).
CONCLUSION
Data from both protozoan and helminth infections suggest that intrauterine or perinatal exposure to parasites or viruses, either the intact organism itself or soluble antigens transferred across the placenta, reduce the ability of the newborn infant to respond to these antigens, i.e. induce tolerance or anergy (Eynon and Parker, Reference Eynon and Parker1993; Nossal, Karvelas and Pulendran, Reference Nossal, Karvelas and Pulendran1993). The results from these studies indicate that a common factor is the reduced ability of T cells and NK cells to secrete IFNγ in response to microorganism-specific antigens suggesting a bias towards a Th2 response. In malaria, the induced tolerance to malaria antigens results in high parasite densities but also a higher pyrogenic threshold in infants compared to adults. In some studies this hypo-responsiveness was found even two decades after birth, suggesting that the immune modulation by intrauterine antigen exposure may be long lasting and perhaps life-long as is seen for hepatitis B.
Treatment of protozoan and helminth infections in pregnant women should be given priority and studies are needed to determine how the protective immune responses in newborn infants are influenced by treatment of protozoan and helminth infections in the mother during pregnancy.