Detection of Rotavirus RNA and Antigens in Serum and Cerebrospinal Fluid Samples from Diarrheic Children with Seizures

Transcription

1 Jpn. J. Infect. Dis., 62, , 2009 Original Article Detection of Rotavirus RNA and Antigens in Serum and Cerebrospinal Fluid Samples from Diarrheic Children with Seizures Bisei Liu, Yukihiko Fujita*, Chikako Arakawa, Ryutaro Kohira, Tatsuo Fuchigami, Hideo Mugishima, and Mitsutaka Kuzuya 1 Department of Pediatrics and Child Health, Nihon University School of Medicine, Tokyo , and 1 Department of Virology, Okayama Prefectural Institute for Environmental Science and Public Health, Okayama , Japan (Received March 10, Accepted May 18, 2009) SUMMARY: Group A rotavirus (GARV) genes (the VP7 and genes) in acute-phase cerebrospinal fluid (CSF), sera and stool samples from 6 children with convulsions accompanied by GARV gastroenteritis were investigated by reverse transcription-polymerase chain reaction (RT-PCR). When the VP7 gene was amplified from the samples, the G genotype (G type) of GARV was determined by RT-PCR. GARV genes were detected in the CSF samples of all 6 children, in 2 of the 3 blood samples, and in all of 4 stool samples. The G typing of GARV from 12 of a total of 13 samples indicated that G3 was the predominant G type in all samples. GARV antigens were detected by enzyme-linked immunosorbent assay in all of the 3 tested sera samples, while no GARV antigens were detected in any of the 5 tested CSF samples. We confirmed the presence of GARV genomes in the CSF samples from all of the children with rotavirus-associated seizures, including encephalopathy. However, the relationship between convulsions and the existence of GARV RNA in CSF remains unclear and further study is required. INTRODUCTION Seizures associated with rotavirus gastroenteritis, despite having a relatively good prognosis, often occur in clusters and are difficult to treat effectively. There are seven species of rotavirus, referred to as A, B, C, D, E, F and G. Humans are primarily infected by species A, B and C, and most commonly by species A. Most cases of rotavirus gastroenteritis in children are caused by group A rotavirus (GARV). It has been reported that GARV genes can be detected in the stools, blood and cerebrospinal fluid (CSF) of rotavirus-infected children who suffer seizures (1,2). It was previously believed that the propagation of GARV outside the gastrointestinal system was quite rare. In recent years, however, GARV genes have frequently been detected in the sera of patients with GARV gastroenteritis (3). Furthermore, it has been reported that antigenemia occurrs in most immunocompetent children with GARV gastroenteritis (3,4). However, the pathophysiology of the extragastrointestinal spread of GARV remains to be elucidated. To examine the pathophysiology of GARV infection, we investigated the presence of GARV genes and antigens, and of norovirus genes, in the stool, CSF and serum of patients with watery white diarrhea accompanied by seizures. MATERIALS AND METHODS Patients: We studied six patients with watery white diarrhea complicated by seizures who were hospitalized at the Department of Pediatrics, Nihon University Itabashi Hospital, during a 5-month period from December 2003 to April *Corresponding author: Mailing address: Department of Pediatrics and Child Health, Nihon University School of Medicine, 30-1 Oyaguchi-Kamicho, Itabashi-ku, Tokyo , Japan. Tel: ext. 2442, Fax: , yfujita@med.nihon-u.ac.jp 2004 (Table 1). All patients were clinically suspected of clustering seizures associated with GARV. Informed consent was obtained from the parents of each of children. The patients were two boys and four girls, whose mean age was 20.3 ± 6.1 months (mean ± standard deviation), ranging from 16 to 27 months. Five of the six patients were discharged from the hospital within 1 week, and the patient with the most severe symptoms (Patient 6) was hospitalized for 17 days. All patients, including Patient 6, were discharged without any neurological sequelae. Clinical symptoms other than seizures included vomiting and diarrhea. No patient had fever during the seizure attacks, although some had fever at other times during the clinical course of the illness. All patients except for Patient 6 suffered two or more seizure attacks, each of which lasted about 1 min (none lasted more than 5 min). The treatments of the six patients for seizures are shown in Table 1. No patient had abnormal findings on routine CSF examination. One patient (Patient 6) was severely ill and required intensive care due to respiratory arrest and disseminated intravascular coagulation (DIC); this patient was diagnosed with GARV encephalopathy. The clinical findings of all six patients are summarized in Table 1. Four stool samples from Patients 1 to 4, three sera samples from Patients 1, 4 and 6, and six CSF samples, one from each patient, were subjected to the examinations described below. We did not perform tests to examine the GARV gene and antigen because there were no available samples. Two stool samples and three sera samples were not available for testing (see Table 2). GARV detection by reverse transcription-polymerase chain reaction (RT-PCR): The outer capsid glycoprotein (VP7) and nonstructural protein () genes of GARV were detected in the patients CSF, sera and stool by RT-PCR using the primers listed in Fig. 1. A total of 13 samples (6 CSF samples, 3 sera samples and 4 stool samples) were collected from the 6 patients (see Table 2) and were evaluated for the presence of GARV genes. Viral RNAs were extracted from 279

2 Table 1. Clinical characteristics of the six patients with white diarrhea complicated by seizures Case Age Sex Clinical symptom Stool CSF findings Seizures (no. of GARV Treatment for seizures cell count, protein, attacks, duration) antigen sugar (mg/dl) Stool bacterial culture test 1 16 m M Vomiting, diarrhea GTCS (3 times, Positive DZP (4 mg, suppository) Cell count 3/ l, Pseudomonas aeruginosa 30 s to 1 min) DZP (5 mg, iv) protein 11, sugar 58 (2+) PHT (150 mg, iv) Normal flora 2 18 m M Vomiting, diarrhea GTCS (3 times, Negative DZP (4 mg, suppository) Cell count 2/ l, Normal flora 1 min to 2 min) protein 19, sugar m F Vomiting, diarrhea GTS (2 times, Negative DZP (4 mg, suppository) Cell count 1/ l, Normal flora 3 min to 5 min) protein 10, sugar m F Vomiting, diarrhea GTS (2 times, Positive DZP (3 mg, iv) Cell count 1/ l, Normal flora (fever) 20 s to 1 min) PHT (150 mg, iv) protein 10, sugar m F Vomiting, diarrhea GTCS (2 times, Positive DZP (4 mg, suppository) Cell count 4/ l, Normal flora (fever) 1 min to 3 min) protein 11, sugar m F Vomiting, GTCS (3 times, Positive Respiration control by Cell count 2/ l, Normal flora diarrhea (fever), 20 min to 30 min) ventilator, DZP (20 mg, iv), protein 14, respiratory arrest, Thiopental 60 mg/kg/day, sugar 92 DIC PHT 10 mg/kg/dose days DIC, disseminated intravascular coagulation; GTCS, generalized tonic-clonic seizures; GTS, generalized tonic seizures; DZP, diazepam; PHT, phenytoin; iv, intravenous. Fig. 1. RT-PCR procedure for the detection of GARV. the samples using a QIAmp viral RNA mini kit (Qiagen, Tokyo, Japan) according to the manufacturer s instructions. The extracted RNAs were stored at 80 C until use. Complementary DNAs were synthesized from the RNAs and amplified by PCR. Specific bands were detected by agarose gel electrophoresis. If no band was detected in the first round of PCR, we performed a second round of PCR under the same conditions using nested primers (Fig. 1). As a negative control for the RT-PCR method, we used 10 CSF samples from children who had seizures but no symptoms of gastroenteritis. G genotyping: GARVs are classified into 20 G genotypes (G types) based on the nucleotide sequence of the VP7 gene (5). We determined the G type of GARV using RT-PCR when the VP7 gene was amplified from the sample. G typing was performed by PCR using a mixture of primers specific for G1 to G4, G8 and G9 according to the G-type-specific PCR procedure previously reported by Gouvea et al. (6). When the presence of more than one G type was suspected, we reconfirmed the G type by performing PCR using single G- type-specific primers. Reconfirmation tests were performed twice to rule out the possibility of contamination. GARV antigen detection: GARV antigens in stools were detected using a commercial kit (Rapidtesta Rota-Adeno; Daiichi Pure Chemicals Co., Ltd., Tokyo, Japan). GARV antigens were also detected from serum and CSF samples using Rotaclone (Meridian Diagnostics, Cincinnati, Ohio, USA) according to the manufacturer s instructions. Twofold diluted serum or CSF samples (100 L) were used for the detection of GARV antigen. In this assay, absorbance of 0.3 or greater was considered to be positive for antigen, following the report by Fischer et al. (4). Norovirus detection using real-time PCR: Norovirus genes were also detected in stool and CSF samples from two patients (Patients 2 and 3) using the real-time PCR method previously reported by Kageyama et al. (7). RESULTS Detection of the VP7 and genes: The results of VP7 and gene tests are summarized in Table 2. The 280

3 Table 2. Detection of GARV gene in stool, serum and CSF, and antigen in serum and CSF Stool CSF Serum Case VP7 Day of VP7 ELISA Day of VP7 ELISA (G type) collection 1) (G type) OD 2) collection (G type) OD 1 ++ (G3) (G3) + NT 2 + (G3) (G3) (G3) NT NT NT 3 + (G3) (G3) NT NT NT 4 ++ (G1) (G3, G1) (G3, G1) NT NT 3 + (G3) NT NT NT 6 NT NT 5 + (G3) ) : Day of collection was counted from the onset of the disease. 2) : Cut-off for positive value is , positive by the first-round PCR; +, positive by the second-round PCR;, negative; NT, not tested. Table 3. Results of norovirus detection tests Case Sample Result of real-time PCR method Genogroup I Genogroup II (amount) 2 Stool + (10 9 copies/g<) CSF 3 Stool + (10 9 copies/g<) CSF VP7 and genes were detected in all tested samples, except for the serum of Patient 6 (Table 2). GARV genes were detected in the stool and CSF samples of Patients 2 and 3, whose stool samples were negative for GARV antigens (serum samples were insufficient for examination). PCR tests for herpesviruses that could cause encephalitis and/or encephalopathy (herpes simplex virus, varicella-zoster virus, cytomegalovirus, Epstein-Barr virus and human herpesvirus 6) were all negative. No GARV gene was detected in negativecontrol CSF samples. G typing: The VP7 genes amplified in this study were subjected to G-type determination. As shown in Table 2, while only G3 was detected in Patients 1, 2, 3 and 5, both G1 and G3 were detected in the serum and CSF samples of Patient 4. When the G type of the stool and CFS samples of the same patient were compared, the same G type was detected in both samples in Patients 1, 2 and 3. In Patient 4, only G1 was detected in the stool sample, while the CSF sample contained G3 as well as G1. The G types in the serum samples were consistent with those in the CSF samples. Detection of GARV antigens by enzyme-linked immunosorbent assay (ELISA): We performed ELISA detection of GARV antigens in sera from Patients 1, 4 and 6, and in CSF samples from Patients 2-6. All serum samples were positive for GARV antigen, and virus genes were also detected, indicating that these three patients were suspected to have antigenemia. A large amount of GARV antigen (ELISA absorbance greater than 2.0) was detected in the sera from Patients 1 and 4, in which virus genes were also detected. In Patient 6 a small amount of the antigen was found in the serum and no virus genes were detected. No GARV antigen was detected in any of the CSF samples (Table 2). Detection of norovirus genes: We performed additional tests for norovirus genes to investigate the possibility of the involvement of other gastroenteritis viruses using stool samples from Patients 2 and 3. GARV antigens were negative but the norovirus genes were detected using nested PCR (Table 3). Real-time PCR indicated that both patients showed noroviruses belonging to genogroup II at extremely high levels (10 9 copies or more per 1 g stool). Real-time PCR examination of the CSF samples from these two patients showed no evidence of norovirus genes. Thus, the stool samples of Patients 2 and 3 showed both GARV and norovirus genes, but their CSF samples showed only GARV genes. DISCUSSION In the present study, we detected GARV RNA from the CSF and sera of six patients with watery white diarrhea complicated by seizures. This suggests that GARV-associated seizures might be caused by direct viral invasion of the central nervous system (CNS). Previous studies have suggested that GARV causes infection that is limited to the gastrointestinal tract. However, based on the present results, we propose that this virus may be capable of spreading from the gastrointestinal tract into the blood, and of being further transported into the CSF. We focused our investigation on the VP7 and genes of GARV in the CSF of our six patients. Because Mossel and Raming (8) reported that is the primary determinant for extraintestinal spread in the neonatal mouse, we hypothesized the genes of GARV to be one of the best targets. It was not possible to obtain a large amount of sera or CSF from our patients since they were less than 27 months old. Since 1993, when GARV RNA was first detected in the CSF and blood of patients with GARV infection complicated by seizures (1), there have been various reports confirming these findings (Table 4). G typing of the reported cases has shown that the G1 genotype is the most common (1,2,9-12), but other G types have also been found (13-15,20). In some cases, G typing was not carried out (3,16-19). Autopsy findings were reported for two patients in whom GARV RNA was detected within the microvascular endothelial cells in various tissues, including the CNS (17). Different G types such as G3 and G9 have been detected in stool, as in the present study, but only G3 was previously found in CSF samples (9). GARVs have previously been believed to cause only local infection of the gastrointestinal tract; however, based on the results of the present study together with the previous data, we suggest that GARVs in general, rather than only specific G types of GARV, may have the potential to be transported into blood. It has been reported that 66% of serum samples collected from children infected by GARV contain detectable viral antigens, referred to as antigenemia (18,19). Although the GARV genome in the CSF of patients with rotavirus-associated encephalopathy has been elucidated, there have been few reports to date on antigenemia in patients with rotavirus- 281

4 Table 4. Reports of the detection of GARV RNA from CSF and other samples by PCR Report (Reference no.) No. of cases (age range) Diagnosis Sample G type Nishimura et al (1) 8 (10 m to 3 y) Gastroenteritis, seizure CSF and serum G1 Ushijima et al (2) 14 out of 15 (1 m to 1 y) Seizure Pharyngeal smear, stool, G1, G3, G4 CSF and serum Yoshida et al (13) 1 (2 y) Encephalitis CSF G3, P9 Pang et al (9) 1 (9 m) Seizure Stool, CSF and serum G1 Makino et al (10) 1 (21 m) Encephalopathy Stool, CSF and serum G2 Hongou et al (14) 1 (2 y) Encephalitis CSF G4, P8 Pager et al (16) 1 (0 m) Seizure, DIC Stool and CSF NT Lynch et al (15) 2 (2 y and 6 y) Seizure, encephalitis CSF P4 Goldwater et al (11) 2 (13 m and 32 m) Encephalopathy, CSF G1, G3 encephalitis Morrison et al (17) 2 (1 y and 4 y) Fatal rotavirus Vascular endothelial NT infection cells and brain Blutt et al (18) 3 (1 m to 3 y) Acute gastroenteritis Serum NT Chiappini et al (3) 9 (1 m to 23 m) Acute gastroenteritis Serum NT Nakagomi et al (12) 3 out of 8 (18 m to 9 y) Encephalopathy CSF G1, G3 Ray et al (19) 26 (younger than 3 y) Acute gastroenteritis Serum NT Furuya et al (20) 1 (3 y) Acute gastroenteritis Stool, CSF and throat G3 encephalitis Present study 6 (16 m to 27 m) Encephalopathy, seizure Stool, CSF and serum G3, G1/G3 associated encephalopathy. Nakagomi and Nakagomi recently reported that GARV antigen was detected in the sera of five of eight children with rotavirus-associated encephalopathy (12). In the present study, we detected GARV antigens in the sera of three patients using an ELISA kit. These results indicate that antigenemia may also be associated with encephalopathy. However, consistent with the study (12), the RT-PCR performed in the present study failed to detect virus antigens in the CSF samples that were virus-rna-positive. This suggests that the amount of rotavirus antigen in CSF, if any, is far smaller than the amount present in serum. Other interesting findings of the present study are that two different G types were detected in the CSF or serum of Patients 4 and 6, and that norovirus genes were detected in the stool samples from Patients 2 and 3, who were negative for GARV antigens in stool. Based on the fact that only the G1 genotype was detected in stool, and that both G1 and G3 were detected in the serum and CSF of Patient 4, we postulate that this patient was initially infected by G3, and then later infected with G1 before the G3-type viruses had disappeared from the patient s body. Since noroviruses were detected in the stool and GARVs in the CSF in Patients 2 and 3, it appears unlikely that the noroviruses were the primary cause of the seizures, but noroviruses may nevertheless play some role as a trigger of seizures. To the best of our knowledge, however, there have been no definitive reports to date to support this idea. We will perform further studies to determine the potential involvement of noroviruses in rotavirus-associated encephalopathy, repeatedly analyzing tissue samples throughout the course of the illness to determine the time-course of viral involvement. In conclusion, GARV genes were detected in the CSF samples of all six of our subjects, in two of the three blood samples, and all of four stool samples. The results of G typing of GARV from these 12 samples (6 CSF, 2 serum and 4 stool samples) indicated that G3 was the predominant G type in all samples. We confirmed the presence of GARV genomes in the CSF samples from all of the children who had rotavirusassociated seizures, including encephalopathy. However, the relationship between convulsions and the existence of GARV RNA in CSF remains unclear and further study is required. REFERENCES 1. Nishimura, S., Ushijima, H., Nishimura, S., et al. (1993): Detection of rotavirus in cerebrospinal fluid and blood of patients with convulsions and gastroenteritis by means of the reverse transcription polymerase chain reaction. Brain Dev., 15, Ushijima, H., Xin, K., Nishimura, S., et al. (1994): Detection and sequencing of rotavirus VP7 gene from human materials (stools, sera, cerebral fluids, and throat swabs) by reverse transcription and PCR. J. Clin. Microbiol., 32, Chiappini, E., Azzari, C., Moriondo, M., et al. (2005): Viraemia is a common finding in immunocompetent children with rotavirus infection. J. Med. Virol., 76, Fischer, T.K., Ashley, D., Kerin, T., et al. (2005): Rotavirus antigenemia in patients with acute gastroenteritis. J. Infect. Dis., 192, Solberg, O.D., Hasing, M.E., Trueba, G., et al. (2009): Characterization of novel VP7, VP4, and VP6 genotypes of a previously untypeable group A rotavirus. Virology, 385, Gouvea, V., Glass, R.I., Woods, P., et al. (1990): Polymerase chain reaction amplification and typing of rotavirus nucleic acid from stool specimens. J. Clin. Microbiol., 28, Kageyama, T., Kojima, S., Shinohara, M., et al. (2003): Broadly reactive and highly sensitive assay for Norwalk-like viruses based on real-time quantitative reverse transcription-pcr. J. Clin. Microbiol., 41, Mossel, E.C. and Ramig, R.F. (2002): Rotavirus genome segment 7 () is a determinant of extraintestinal spread in the neonatal mouse. J. Virol., 76, Pang, X.L., Joensuu, J. and Vesikari, T. (1996): Detection of rotavirus RNA in cerebrospinal fluid in a case of rotavirus gastroenteritis with febrile seizures. Pediatr. Infect. Dis. J., 15, Makino, M., Tanabe, Y., Shinozaki, K., et al. (1996): Haemorrhagic shock and encephalopathy associated with rotavirus infection. Acta Pediatr., 85, Goldwater, P.N., Rowland, K., Thesinger, M., et al. (2001): Rotavirus encephalopathy: pathogenesis reviewed. J. Paediatr. Child Health, 37, Nakagomi, T. and Nakagomi, O. (2005): Rotavirus antigenemia in children with encephalopathy accompanied by rotavirus gastroenteritis. Arch. Virol., 150, Yoshida, A., Kawamitu, T., Tanaka, R., et al. (1995): Rotavirus encephalitis: detection of the virus genomic RNA in the cerebrospinal fluid of a child. Pediatr. Infect. Dis. J., 14, Hongou, K., Konishi, T., Yagi, S., et al. (1998): Rotavirus encephalitis mimicking afebrile benign convulsions in infants. Pediatr. Neurol., 18, Lynch, M., Lee, B., Azimi, P., et al. (2001): Rotavirus and central nervous system symptoms: cause or contaminant? Case report and review. Clin. Infect. Dis., 33, Pager, C., Steele, D., Gwamanda, P., et al. (2000): A neonatal death 282

5 associated with rotavirus infection-detection of rotavirus dsrna in the cerebrospinal fluid. S. Afr. Med. J., 90, Morrison, C., Gilson, T. and Nuovo, G. (2001): Histologic distribution of fatal rotaviral infection: an immunohistochemical and reverse transcriptase in situ polymerase chain reaction analysis. Hum. Pathol., 32, Blutt, S.E., Kirkwood, C.D., Parreño, V., et al. (2003): Rotavirus antigenaemia and viraemia: a common event? Lancet, 362, Ray, P., Fenaux, M., Sharma, S., et al. (2006): Quantitative evaluation of rotaviral antigenemia in children with acute rotaviral diarrhea. J. Infect. Dis., 194, Furuya, Y., Katayama, T., Miyahara, K., et al. (2007): Detection of the rotavirus A genome from the cerebrospinal fluid of a gastroenteritis patient: a case report. Jpn. J. Infect. Dis., 60,