Prevalence and molecular characterisation of human adenovirus in diarrhoeic children in Tanzania; a case control study

This case control study was conducted between August 2010 and July 2011. Details of the study population and data collection have been previously described [29]. Briefly, sample collection was performed during two seasons, starting in August 2010 and in March 2011, aiming for minimum 300 diarrhoeic children and 300 non-diarrhoeic children in each period. The target for diarrhoeic was reached in January 2011 and in June 2011, while enrolment of non-diarrhoeic children continued in February 2011 and July 2011, thus enrolment continued for one complete year. A total of 1266 samples were collected and of these 1235 samples were adequate for detection of adenovirus using ELISA (690 diarrhoeic and 545 non-diarrhoeic children).

Diarrhoeic children were children hospitalised due to diarrhoea (n = 690) at three major hospitals of Dar es Salaam; Muhimbili National Hospital (MNH), Amana and Temeke Municipal hospitals.

Non-diarrhoeic children (n = 545) were enrolled during the recruitment of diarrhoeic children. These were either children outpatient children attending child health clinics for immunisation and growth monitoring (n = 310) or children admitted to hospital due to diseases other than diarrhoea (n = 235).

Diarrhoeic children included in the study were hospitalised in the diarrhoea wards, with acute or persistent diarrhoea. Diarrhoea was defined as three or more watery stools within 24 hours. An episode of diarrhoea was considered over when two consecutive days pass without diarrhoea. An episode of acute diarrhoea was defined as diarrhoea with duration between 24 hours and less than 14 days. Persistent diarrhoea was defined as diarrhoea for 14 days or more. Non diarrhoeic children included in the study were children without history of diarrhoea for one month prior to enrolment. Children above 24 months of age and cases that could not provide stool sample on the day of admission were excluded from the study. Furthermore diarrhoeic and non-diarrhoeic children whose parent or guardian did not consent to participate in the study were excluded.

Two standardized questionnaires for diarrhoeic and non-diarrhoeic children were used to collect demographic and clinical information, including age (date of birth), sex, place of residence, parent/guardian level of education and history of antibiotic use prior to admission. Consistency of stool and duration of diarrhoea was also recorded. The child’s length and weight measurements were recorded. For diarrhoeic children, additional clinical information was obtained from patient files. This was information on hydration status which was assessed on the day of admission by the attending clinician and HIV testing results. Weight and height measurement was done as previously described [28]. A single stool specimen was collected on inclusion from each child using wide mouthed sterile plastic containers.

Adenovirus antigen was detected using the commercially available ELISA Kit, ProSpecT™Adenovirus kit designed to detect 51 HAdV types (Oxoid, Hants, UK), with 10% faecal suspensions according to the manufacturer’s instructions.

All HAdV positive samples from ELISA screening underwent real-time PCR for genotyping. DNA extraction has been described before by Moyo et al. [29]. DNA was extracted using Magna Pure LC (Roche Diagnostics). Stool samples (50 mg) were mixed 1:10 with Bacterial Lysis buffer (Roche Diagnostics, Mannheim, Germany), and centrifuged at 13.000g for 3 minutes. DNA was extracted from 200μl supernatant using the Magna Pure LC High Performance Total Nucleic Acid Isolation Kit (Roche Applied Science, Mannheim, Germany). Total nucleic acid was eluted and stored at −70°C until analysis.

The primers used for real-time PCR were Ad F: 5′-GAC GCY TCG GAG TAC CTG AG-3′ (nt10-29, AF161559) and Ad R: 5′-CCY ACR GCC AGI GTR WAI CGM RCY TTG TA-3′ (nt 221–193, AF161559). The TaqMan probe used was Ad TM: 5′-CTG GTG CAG TTY GCC CGC-3′ (nt 37–54, AF161559) with FAM labeled as a fluorescent dye on the 5′ end and TAMRA as a fluorescence quencher dye labeled to the 3′ end (TIB Molbiol GmbH, Berlin). In some cases when a larger PCR-product (322 bp) was needed for exact genotyping; an alternative reverse primer (5′-AGCSAGIGARTTGTAAGC-3′) was used. Primers and TaqMan probes were chosen within the conserved region of the hexon-coding gene of adenoviruses [30].

Real time PCR was carried out in 15μL final reaction volume containing; five μL of the DNA-template which was transferred to PerfeCTa Multiplex qPCR SuperMix with uracil-N-glycosidase [12] (Quanta Biosciences), Adeno forward primer (300nM final cons.), Adeno reverse primer (300nm final cons), Adeno TaqMan probe (200nM final cons.) and PCR-grade water. All amplifications were performed using a CFX96 real-time PCR System (Bio-Rad Laboratories, Inc). Amplification conditions were 1 cycle at 45°C for 5 minutes (UNG activation), 1 cycle at 95°C for 3 minutes (polymerase activation), followed by 40 cycles at 95°C for 10 seconds (denaturation) , 55°C for 10 seconds (annealing) and 72°C for 10 seconds (extension). Fluorescence emission was set to be measured at the end of extension step. Positive and negative controls were also included. The described real-time-PCR is able to detect adenovirus from type 1 to type 51. The detection limit of the PCR is approximately 300 copies/ml.

The PCR-product was purified and sequenced using the forward Ad F-primer. Sequencing was performed on an ABI 3130xl DNA Sequencer (Applied Biosystems, Foster city, CA, USA). Nucleotide sequences were analysed using BLAST service at NCBI. The results were compared to known adenovirus sequences in the GenBank by pairwise comparison from multiple alignments using the Genius software package (Biomatters). The phylogenetic tree was constructed using UPGMA and Kimura two-parameter methods [31]. Twenty six sequences with more than 200bp length obtained in this study have been submitted to GenBank with accession numbers: KM000013-KM000038.

Data was analysed using Statistical Package for the Social Sciences (SPSS for IBM-PC, release 18.0; SPSS Inc., Chicago, IL, US). The prevalence and the median age of adenovirus infection in diarrhoeic and non-diarrhoeic children were compared using chi-square (χ²) test and Mann–Whitney U test respectively. To test for significant associations between adenovirus positivity and season, demographic or clinical characteristics, a univariate analysis was performed Weight-for age, and weight for length Z- scores were calculated using EPI Info (USD, Inc., Stone Mountain, GA). A cut off P-value of < 0.05 was considered significant.

The study received ethical approval from the Senate Research and Publications Committee of Muhimbili University of Health and Allied Sciences (MUHAS) and from the Regional Committee for Medical and Health Research Ethics (REK) in Western Norway. Permission was obtained from the respective hospital authorities where recruitment of study participants took place (i.e. MNH, Amana and Temeke Hospitals). Written informed consent was also obtained from parent/guardian of the child.

Human adenovirus was detected in 37 out of 1235 children. The prevalence of HAdV in diarrhoeic and non-diarrhoeic children did not differ significantly, (3.5%, 24/690 vs. 2.4%, 13/545, P = 0.26). There was no significant difference in prevalences of HAdV in non-diarrhoeic children attending child health clinics (3.2%, 10/310) and those admitted to hospital due to diseases other than diarrhoea (1.3%, 3/235, P = 0.14).

In both diarrhoeic children and non-diarrhoeic children, the median age was significantly higher in HAdV infected than HAdV non-infected children (10 vs 9 months, P = 0.032 and 14.1 vs. 10.9 months, P = 0.04). As shown in Table 1, the highest proportion of HAdV infection was in the age group of 7–12 months in diarrhoeic children, P = 0.04. In the age group of zero to six months, only one HAdV was detected in diarrhoeic children while no HAdV was detected in non-diarrhoeic children.

The HIV status was known for 421 children. Twenty six diarrhoeic children and seven non-diarrhoeic children tested positive for HIV. HAdV was not detected in any of these 33 children who tested positive for HIV. HAdV was detected in 2.6% (2/78) HIV negative diarrhoeic children and 3.2% (10/310) HIV negative non-diarrhoeic children.

Table 1 shows that more than half of the adenovirus- infected children with diarrhoea were dehydrated (P = 0.013) and that most of them presented with acute symptoms. There was no significant association between HAdV infection and sex of the child or parent/guardian education level and nutritional status of the child (Table 1).

We analysed data on a monthly basis in order to determine the seasonal distribution of HAdV infection (Figure 1). HAdV was not detected in the month of September 2010 and January 2011 in diarrhoeic children. In non-diarrhoeic children, HAdV was not detected in August 2010, February, March and April 2011.

We divided the months of the study according to the season of the year i.e. short rains season, October through December; long rains season, March through May; and the rest were dry months of the year [32]. The prevalence of HAdV was significantly higher during the short rainy season compared to other months (4.5%, 18/396 vs. 2.3%, 19/839, P = 0.028, OR 2.0 (95% CI: 1.06 to 3.78). However, when data was combined for both rainy seasons, there was no significant difference between rainy and dry seasons (27/799, 3.4% vs. 10/436, 2.3%, P = 0.29). The different types of adenovirus did not show pattern of seasonality.

The nucleotide sequences of the 37 adenoviruses detected were compared to reference adenovirus strains available in the GenBank by BLAST. A phylogenetic tree was also constructed (Figure 2) from the 26. HAdV nucleotide sequences obtained in this study which were submitted to GenBank (18 diarrhoeic and 8 non-diarrhoeic children). This study found different species and types of HAdV. Among the 24 HAdV detected in diarrhoeic children, seven types were defined, and out of these, 50% (12/24) were enteric adenovirus types 40 and 41 which occurred with equal prevalence (Figure 2). Seven different types were also found among the 13 HAdV in non-diarrhoeic children. Of these 46.2% (6/13) were enteric adenovirus types, and type 40 was more prevalent than type 41 (30.8% vs. 15.4%), as shown in Figure 3. The proportions of enteric adenoviruses (type 40 and 41) were not significantly different in diarrhoeic and non-diarrhoeic children (50%, 12/24 vs. 46%, 6/13, P = 0.82). Similarly the proportion of non-enteric adenovirus did not differ between diarrhoeic and non-diarrhoeic children (50%, 12/24 vs. 53.85%, 7/13, P = 0.82).