L. Svensson Division of Molecular Virology, Department of Clinical and Experimental Medicine, Medical Faculty, Linköping University, Linköping, Sweden
While rotavirus is well established as a major cause of severe acute gastroenteritis in young children all over the world, our understanding of the mechanisms underlying the clinical symptoms largely remains unresolved. The clinical symptoms are not limited to life threatening diarrhoea and vomiting, but also include systemic effects, collectively called “sickness responses.” In this review we discuss recent progress in exploring the mechanisms of rotavirus-induced diarrhoea and vomiting, delayed gastric emptying, and sickness responses such as fever and abdominal pain.
enteric nervous system
Rotavirus (RV) infections are associated with approximately 450,000 deaths each year worldwide, mainly occurring in developing countries among children under 5 years of age (Tate et al., 2012). The infection can be asymptomatic or symptomatic, and the outcome is affected by both viral and host factors. One host factor is age, with infected neonates rarely responding with symptomatic disease. This protection is thought to result primarily from transplacental transfer of maternal antibodies (Ray et al., 2007). Reductions in these antibodies coincide with the age of maximum susceptibility of infants to severe RV-induced diarrhoea. Another host factor consists of histo-blood group antigens, with the Lewis and secretor antigens contributing to RV susceptibility (Nordgren et al., 2014; Lopman et al., 2015; Imbert-Marcille et al., 2014). Libonati et al. (2014) found that the genomic sequences of RVs infecting symptomatic and asymptomatic neonates are virtually identical, further supporting the notion that host factors contribute to the modulation of the clinical response.
While the clinical importance of human RV disease is well recognized and potent vaccines have been developed, our current understanding of RV disease mechanisms is to a major part derived from in vitro studies and animal models and to a minor part from pathology and treatment interventions in humans (Lundgren and Svensson, 2001; 2003; Michelangeli and Ruiz, 2003; Ramig, 2004). While in vitro and animal studies have generated a significant amount of important information regarding virus–cell interactions, certain issues of human RV illness still remain to be understood. For example, emesis, a hallmark of RV disease, which contributes to dehydration, cannot be studied in rodents as they lack the emetic-gastric reflex (Horn, 2008). However, a recent study (Hagbom et al., 2011) has shown that RV stimulates release of serotonin [5-hydroxytryptamine (5-HT)] from human enterochromaffin (EC) cells and activates structures in the central nervous system (CNS) involved in nausea and vomiting, all suggesting that the enteric nervous system (ENS) and vagal afferent stimuli are involved in RV-induced vomiting and diarrhoea. Furthermore, the mechanisms behind clinical symptoms such as fever, malaise, social withdrawal, fatigue, and anorexia, collectively called “sickness responses” have either scantily or not at all been addressed in respect to RV illness. Four mechanisms have been implicated in RV diarrhoeal disease; increased secretion, decreased absorption (ions and water), altered motility and permeability (Fig. 2.6.1). This review will discuss our current understanding of the mechanisms behind the clinical symptoms associated with RV disease.
2. Clinical symptoms
Human RV infection results in a spectrum of responses that vary and can either be asymptomatic, mild or severe, sometimes resulting in lethal dehydrating illness. The incubation period is less than 48 h with a sudden onset of vomiting, a high frequency of dehydration and a mean duration of diarrhoea lasting 5–6 days (Uhnoo et al., 1986).
2.1. Sickness Response
The clinical symptoms are not limited to life-threatening diarrhoea, but also include systemic responses collectively called “sickness responses.” The acute phase response, or “sickness” refers to an initial response of the innate immune system to a broad range of potentially infectious agents. It comprises an inflammatory reaction mediated by pro-inflammatory factors such as interleukin-1 (IL-1), IL-6, and tumour necrosis factor alpha (TNF-α). In contrast to many invasive bacterial gastrointestinal infections, both C-reactive protein (CRP) and calprotectin, two clinical biomarkers of inflammation, are not elevated during human RV infection (Greenberg and Estes, 2009; Chen et al., 2012; Weh et al., 2013). Uhnoo et al. (1986) found that bloody stools, prolonged diarrhoea, and leukocytosis was significantly associated with pathogenic bacteria. On the other hand viral infections are more associated with nausea and vomiting compared to bacterial infections (Uhnoo et al., 1986; Weh et al., 2013). Collectively this suggests that RV infection in humans, albeit with significant pathology, results in a modest clinical inflammatory response compared to infections with pathogenic bacteria.
Fever, a response of the body’s thermostat, located in hypothalamus, is a part of the acute-phase response by the immune system (Brodal, 2010) and is usually accompanied by sickness behavior, such as inactivity, sleepiness, depression, and reduced intake of food and water (Hart, 1988; Hennessy et al., 2014). IL-1β, TNF-α, and IL-6 have been shown to be the cytokines responsible for the induction of fever, with IL-1β seemingly depending on IL-6 to induce fever (Brodal, 2010; Eskilsson et al., 2014). RV infection is commonly associated with less fever than bacteria (Kutukculer and Caglayan, 1997; Elliott, 2007). Furthermore, children with RV infection have been shown to have elevated levels of cytokines in serum and children with fever had significantly higher levels of IL-6 (Jiang et al., 2003).
Application of noxious stimuli to the gastrointestinal (GI) tract may activate peripheral nerve receptors that are sensitive to chemical, mechanical or inflammatory stimuli and may in turn result in abdominal pain. Abdominal pain is associated with RV infection (Uhnoo et al., 1986) but the underlying mechanisms of how RV induces abdominal pain remain to be determined. In general, abdominal pain is supposed to include alterations in smooth muscle and enteric nerves and is likely related to the altered processing of sensory information from the gut to the CNS.
Intussusception (IS) has been associated with RV infections/vaccines. IS is a process in which a segment of the intestine invaginates into the adjoining intestinal lumen, thereby resulting in bowel obstruction and infarction, which may require clinical or surgical intervention. The first licensed RV vaccine, RRV-TV (Rotashield®) was withdrawn in USA following reports of IS among the vaccinated children (Centers for Disease Control and Prevention, 1999; Murphy et al., 2001). Recent postmarketing surveillance after wide application of Rotarix® and Rotateq® vaccines in Latin America and Australia has reported lower risks of IS in comparison to those observed with the RRV-TV vaccine (Patel et al., 2011; Buttery et al., 2011). Two earlier phase III clinical trials evaluating safety and efficacy of Rotarix® and Rotateq® vaccines did not observe a significant correlation with IS (Ruiz-Palacios et al., 2006; Vesikari et al., 2006). Besides in vaccinated children, IS has also been reported in a few young children after natural RV infections (Konno et al., 1978; Dallar et al., 2009; Mulcahy et al., 1982). Robinson et al. (2004) using ultrasound observed increased thickness of distal ileum and lymphadenopathy in RV infected children compared to controls, which may be the cause of IS. However, data from epidemiological studies did not demonstrate a significant increase of gut IS after natural RV infection (Velazquez et al., 2004; Bahl et al., 2009; Bines et al., 2006).
2.3. Rotavirus Infection Delays Gastric Emptying
The prominence of nausea and vomiting during RV illness suggests abnormal gastric function. A marked delay of gastric emptying has been observed not only after ingestion of norovirus (Meeroff et al., 1980) but also after RV infection. Bardhan et al. (1992) found that RV infection is accompanied by abnormal gastric motor function as manifested by delayed emptying of liquid. Gastric emptying of liquids is believed to be primary a function of the pressure gradient between the stomach and the duodenum. The mechanisms of delay in gastric emptying is proposed to include neural pathways and gastrointestinal hormones such as secretin, gastrin, glucagon (Cooke, 1975), and cholecystokinin (Debas et al., 1975). The neural pathways influencing gastric emptying may include noncholinergic, nonadrenergic, and dopaminergic vagal neurons (Minami and McCallum, 1984). The response is mediated by 5-HT3 receptors and also sodium glucose cotransporter (SGLT-1) expressed by EC cells (Raybould, 2002) further supporting the observation that RV infection stimulates the ENS. It should also be mentioned that gastric emptying and food intake are related and reduction in food intake has been observed in children with acute RV illness (Molla et al., 1983) as part of the sickness response.
3. Nitric oxide in RV illness
Nitric oxide (NO) is synthesized from l-arginine by NO synthases (NOS) and is secreted by cells involved in host defence, homeostatic, and development functions. NOS exists in three isoforms; endothelial NOS (eNOS), neuronal NOS (nNOS), and inducible NOS (iNOS). While eNOS activity is increased in response to an increased blood flow (the augmented NO causing a relaxation of vascular smooth muscles), the iNOS is activated during infection and inflammation. However, immunohistochemical investigations of the intestinal wall have demonstrated nNOS in the myenteric plexus, indicating that NO is involved in the control of intestinal motility and/or as a transmitter in interneurons of local reflexes of the ENS (Furness, 2006). Accumulating evidence suggests that NO plays a role in the modulation of aqueous secretion and the barrier function of intestinal cells. Several recent studies have reported elevated levels of NO in patients with gastroenteritis (Kawashima et al., 2004; Kukuruzovic et al., 2002; Rodriguez-Diaz et al., 2006; Sowmyanarayanan et al., 2009). A study by Rodriguez-Diaz et al. (2006) demonstrated NSP4 to induce rapid release of NO from intestinal epithelial cells (HT-29). Moreover, time kinetics studies showed release of NO in RV-infected mice peaking between days 6 and 9 after infection, thereby suggesting participation of iNOS. The authors further observed elevated levels of NO in RV infected diarrhoeal children, thereby confirming the results of both in vitro and animal experiments. A similar study from India (Sowmyanarayanan et al., 2009) reported elevated levels of NO metabolites among diarrhoeal children infected with RV and norovirus. Borghan et al. (2007) found that ex vivo NSP4 treatment of ileum excised from CD-1 suckling mice resulted in up regulation of ileal iNOS mRNA expression within 4 h. Furthermore, NSP4 was able to induce iNOS expression and NO production in murine peritoneal macrophages (see Chapter 2.4). Kawashima et al. (2004) observed elevated levels of NO metabolites in both serum and cerebrospinal fluid of RV infected gastroenteritis patients having febrile convulsions in comparison with patients with purulent meningitis, encephalitis, and febrile convulsion and of a control group. Additionally, they also observed a relative correlation between IL-6 levels and NO metabolites in some cases. The functional consequences of an increased epithelial NO production is not fully established. Several effects seem possible. Locally produced NO increases the permeability of the intestinal epithelia to hydrophilic solutes. Furthermore, the produced NO may directly influence epithelial transport mechanisms. The control of epithelial sodium and hence water transport by nerve-mediated NO release was recently reviewed (Althaus, 2012).
4. Role of prostaglandins and acetylsalicylic acid in rotavirus diarrhoea
Prostaglandins (PGs) are lipid compounds, enzymatically derived from fatty acids located in the cell lipid bilayer, and can elicit a wide range of physiological responses in the body (Scher and Pillinger, 2009). Cyclooxygenases (COXs) are essential enzymes in the biosynthesis of PGs, converting the arachidonic acid to PGH2; subsequently specific isomerases transform PGH2 to biologically active PGs (Scher and Pillinger, 2009). It has been shown that PGEs are produced under the influence of microorganisms and have immunomodulatory, antiinflammatory as well as pro-inflammatory actions (Scher and Pillinger, 2009). Moreover it has been shown that PGE2 can stimulate water secretion (Sandhu et al., 1981), an effect that can be blocked by drugs attenuating nervous activity, such as hexamethonium (nicotinic receptor blocker) or lidocaine (local anesthetic) (Brunsson et al., 1987) thus indicating that nerves are involved. Studies carried out among children with RV gastroenteritis found elevated levels of PGE2 and PGF2 in both plasma and stool (Yamashiro et al., 1989) and treatment with the COX-inhibitor acetylsalicylic acid (aspirin) (Vane, 1971) reduces the duration of diarrhoea (Yamashiro et al., 1989; Gracey et al., 1984). PGs can be converted to cyclopentenone PGs (cyPGs), which have antiviral properties through NF-kβ activation and inhibit viral replication of both DNA and RNA viruses (Santoro, 1997). Interestingly, in vitro studies have shown that RV replication is inhibited by the cyPGA1 (Superti et al., 1998; Suzuki and Oshitani, 1999).
In conclusion, these observations suggest that RV infection stimulates release of PGs from epithelial cells, which may act directly on the epithelium and/or indirectly via an activation of nerves. The positive therapeutic effect of aspirin on RV diarrhoea is interesting, but needs to be confirmed in larger studies. However, the side effects of aspirin should be considered before giving the drug to small children (Litalien and Jacqz-Aigrain, 2001).
5. Mechanisms of diarrhoea
The intestinal epithelium consists of absorptive cells, EC cells, goblet cells, paneth cells, intraephithelial lymphocytes, and undifferentiated cells. Cell division takes place in the undifferentiated cells of the crypts and as these cells migrate upwards toward the villi, they differentiate into different cell types, for example, absorptive enterocytes. Chloride secreting cells, which are undifferentiated cells, are located in the crypts. The villi of the small intestine are the site where most of the absorption of nutrients such as minerals, sugars, and amino acids occurs. After food has been digested in the stomach by strong hydrochloric acid and enzymes, the pyloric sphincter opens, and food gets pushed into small intestine by peristalsis.