|Home | About | Journals | Submit | Contact Us | Français|
Stool samples from two healthy infant siblings collected at about weekly intervals during their first year of life were analyzed by PCR for 15 different enteric viral genera. Adenovirus, Aichi virus, Anellovirus, Astrovirus, Bocavirus, Enterovirus, Parechovirus, Picobirnavirus, and Rotavirus were detected. Not detected were Coronavirus, Cardiovirus, Cosavirus, Salivirus, Sapovirus, and Norovirus. Long-term virus shedding, lasting from one to 12 months, was observed for adenoviruses, anelloviruses, bocaviruses, enteroviruses, parechoviruses, and picobirnaviruses. Repeated administration of oral poliovirus vaccine resulted in progressively shorter periods of poliovirus detection. Four nonpolio enterovirus genotypes were also detected. An average of 1.8 distinct human viruses were found per time point. Ninety-two percent (66/72) of the fecal samples tested contained one to five different human viruses. Two British siblings in the mid-1980s showed nearly constant fecal viral shedding. Our results demonstrate that frequent enteric infections with diverse viruses occur during early childhood in the absence of severe clinical symptoms.
Enteric viral infections can have severe consequences, particularly in malnourished or immunodeficient infants and neonates (40, 48). While capable of inducing diverse symptoms ranging from diarrhea (3, 54) to encephalitis (29), infections with enteric viruses seem to result largely in minor or no symptoms in otherwise healthy children. The rate of viral enteric infections in healthy children has been analyzed in several studies targeting single viral groups. Human enteroviruses (HEVs) are frequently detected in the stool of healthy Norwegian infants by using monthly sampling, with 51% children positive by 28 months of age (68). Also by using monthly sampling, human parechoviruses (HPeVs) were detected in the stool of 48% of healthy Finnish infants by the age of 22 months (41), while 86% of Norwegian infants had been infected by 24 months (60). Sixteen percent of stool samples from children under five from Amsterdam were positive for HPeV and 18% for HEV (6).
Besides being capable of causing acute illnesses, childhood enteric viral infections may also be involved in triggering long-term consequences, such as autoreactivity to islet cells seen in type 1 diabetes (10). Early viral infections also could have beneficial effects, such as prevention of autoimmune responses and providing immunity against subsequent infections caused by related but more pathogenic viral variants (10).
In order to determine the extent of viral infections occurring in early childhood, longitudinally collected stool samples from two siblings who grew up in the mid-1980s in the United Kingdom were tested using an extensive panel of PCR primers against 15 groups of viruses to measure the overall frequency and duration of viral shedding.
Stool samples from two healthy infant siblings collected from October 1983 until September 1984 for child 1 and from November 1985 to January 1987 for child 2 were examined. Samples were taken as swabs and put into virus culture medium without fetal bovine serum. Afterward, the 10%-stool suspensions were transferred to the National Institute for Biological Standards and Control (NIBSC) laboratory, vortexed, and stored at −70°C.
For child 1 (male, born in July 1983), stool samples were examined starting at 116 days of age, with a sampling interval of 2 to 16 days (mean of 6 days). In total, 34 samples were analyzed. Trivalent oral poliovirus vaccine (tOPV) was given on days 107, 218, and 299 after birth.
For child 2 (female, born in August 1985), stool samples were examined starting at day 142 after birth. The sampling intervals were 5 to 9 days, with a mean of 7 days. In total, 38 samples were examined. The tOPV was given on days 129, 252, and 363.
Both infants were breast fed during the entire sampling period. Neither attended a day care facility. No travel outside the United Kingdom was recorded. No clinical signs requiring hospitalization were reported during the sampling period. Flulike symptoms were recorded on day 222 for child 1.
All federal guidelines and institutional guidelines were followed during the course of this study, which was approved by the UCSF committee on human research.
Stool suspensions were mixed with zirconia/silica beads (RPI, Mount Prospect, IL), vigorously vortexed, and centrifuged twice at 6,000 × g for 10 min. Clarified supernatants (140 μl) were used for total nucleic acid extraction by using the RNA minikit (Qiagen, Valencia, CA). The nucleic acids were eluted in 60 μl of elution buffer, immediately mixed with 40 U RiboLock RNase Inhibitor (Fermentas Inc.), and stored at −70°C.
PCR or reverse transcription (RT)-PCR for each of the targeted viral groups was performed using previously published primers and conditions (Table 1). Samples (n = 72) were screened for three DNA viral groups (Adenovirus, Anellovirus, and Bocavirus) and 12 RNA viral groups (Aichi virus, Astrovirus, Cardiovirus, Cosavirus, Coronavirus, Enterovirus Norovirus, Parechovirus, Picobirnavirus, Rotavirus, Salivirus, Sapovirus). All primers were used in 10 μM concentration. The RT step for RNA viruses was performed using random nanomer, oligo(dT) (Eurofins MWG Operon), or specific primers (IDT, Coralville, IA) and added to 10 μl of each extracted viral nucleic acid. Sample was denatured at 72°C for 2 min and then chilled on ice. Reaction mixture (9 μl; containing 4 μl SuperScript buffer [Invitrogen, Carlsbad, CA], 1 μl 100 mM dithiothreitol, 1 μl 10 mM dNTP, 200 U SuperScript III reverse transcriptase [Invitrogen, Carlsbad, CA], and 1 μl 40 U RiboLock RNase inhibitor) was added and incubated at 25°C for 10 min, 50°C for 60 min, and 70°C for 15 min and then chilled. cDNA was stored at −20°C.
Most of the primers selected for diagnostic screening were targeted to conserved regions of viral genomes. The PCRs were carried out either with REDTaq mix (Sigma), NEBTaq (New England BioLabs), Ex Taq (TaKaRa Bio Inc., Shiga, Japan), or PlatinumTaq DNA polymerase (Invitrogen, Carlsbad, CA) enzymes.
The PCR products were visualized in 1.5% agarose gel electrophoresis with ethidium bromide. The amplicons were purified using a QIAquick kit (Qiagen, Valencia, CA). The cleanup of fragments less than 100 bp was done by ExoSAP-IT (USB Corp., Cleveland, OH) and directly sequenced from both directions using second-round PCR primers.
Sequence assembly was done in Sequencher 5.0. Searching for GenBank homology was performed by BLASTn with default settings (http://blast.ncbi.nlm.nih.gov/Blast.cgi).
All sequences were deposited in GenBank with the following accession numbers: JX179277 to JX179299.
Thirty-four fecal samples collected at days 116 to 368 after birth were examined. Nucleic acid was extracted and tested for different viruses using the PCR primers and conditions listed in Table 1. The expected viral sequence of all PCR amplicons was confirmed by direct Sanger sequencing.
Child 1 was positive for 8 viral groups: Adenovirus, Aichi virus, Anellovirus, Astrovirus, Bocavirus, Enterovirus, Parechovirus, and Rotavirus (Table 2). Samples were PCR negative for Cardiovirus, Coronavirus, Cosavirus, Norovirus, Picobirnavirus, Salivirus, and Sapovirus.
Human adenovirus serotype 1 (HAdV-1) shedding was detected at day 174 and lasted for four sampling times examined over 2.7 weeks. A week after the last HAdV-1 infection, human adenovirus 41 was detected in the sample collected on day 200. The serotypes of both these adenoviruses were confirmed by sequencing. Five months later, shedding of the HAdV-1 serotype was again detected over 3 days in the last two samples collected (Table 2). Human Aichi virus, from family Picornaviridae, was detected in 2 sampling times over 1 week (Table 2).
Nested PCR primers were used to distinguish the three anellovirus species: Torque teno virus (TTV), Torque teno midivirus (TTMDV), and Torque teno minivirus (TTMV). TTV was detected during two extended periods, from days 151 to 200 and days 249 to 285. Isolated time point detection was recorded on days 312 and 368 for another anellovirus, TTMV, and again for TTV on day 333 (Table 2).
Sequences similar to human mink-ovine-like astrovirus B (HAstV-HMO-B) isolate NI-196 (GenBank accession no. GQ415661) and the closely related AstV-VA3 isolate (GenBank accession no. GQ502196) were detected at the same time as the Aichi virus over a 1-week period (Table 2). Human bocavirus 1 (HBoV-1) was detected in a single time point (Table 2).
The first trivalent oral poliovirus vaccine (tOPV) was given at day 107. Poliovirus vaccine strain Sabin-2 (HEV-C species) was identified in the first sample analyzed from day 116, and the shedding continued until day 180 (73 days after tOPV administration). During this period consisting of 10 time points, Sabin-2 and Sabin-3 were detected in six and four samples, respectively. Sabin-1 was detected at a single time point on day 222 from a sample collected 4 days after the second tOPV. On the same day, flulike symptoms were recorded for child 1. Following the third tOPV dose on day 252, a sample collected 5 days later was echovirus 9 (HEV-B species) positive. Coxsackievirus A4 (CV-A4) from species HEV-A was also detected over a period of 2 weeks (days 312, 319, 326).
Human parechovirus type 1 (HPeV-1) was continuously detected over a period of 5 weeks, from days 312 to 349. After a 16-day interval, viral HPeV-1 RNA was again detected at a single time point (Table 2). The virus protein 1 (VP1) region of the HPeV-1 was determined for each time point and found to be invariant throughout the sampling period, indicating that the same virus was shed and that no reinfection with a distinct variant took place. The closest VP1 sequence of HPeV-1 in GenBank varied by 4% of nucleotides.
Rotavirus group A was detected at a single time point on day 283. Sequencing of the PCR product indicated the presence of serotype G1P8 (data not shown).
Thirty-eight fecal samples collected at days 142 to 405 after birth were examined. The fecal samples of child 2 were positive for 5 viral groups: Anellovirus, Bocavirus, Enterovirus, Parechovirus, and Picobirnavirus (Table 3). Samples were negative for Adenovirus, Aichi virus, Astrovirus, Cardiovirus, Coronavirus, Cosavirus, Norovirus, Rotavirus, Salivirus, and Sapovirus.
Anelloviruses were detected in the very first sample analyzed, collected on day 142, until the last sampling, on day 405. Specific amplification was observed for both TTV and TTMV in 13 of the 38 TTV-positive samples (Table 3).
Prolonged shedding of human bocavirus type 1 (HBoV-1) was detected during eight samplings from days 192 to 243, for at least 51 days. Sporadic detection of HBoV-1 was also seen in samples from days 284 and 313 (Table 3).
Following tOPV administration on day 129, poliovirus vaccine strain shedding was recorded until day 183, starting with the first collected sample on day 142. Poliovirus (PV) shedding therefore lasted for 54 days after the first tOPV administration, with a dominance of Sabin-3 serotype. After the child received the second tOPV dose on day 252, the Sabin-1 strain was detected from the sample collected 5 days after vaccination. The next sample, collected 19 days after the tOPV boost, was negative, but the following sample, collected 26 days post-tOPV, was positive for PV3. All subsequent samples were negative for polioviruses except one PV3-positive sample collected 75 days after the first tOPV boost. The third tOPV administration did not result in any poliovirus detection. The species HEV-A Coxsackievirus A16 (CV-A16) was found in a total of 6 samples over four distinct time periods.
Human parechovirus type 6 (HPeV-6) was detected from days 299 to 342 for about 6 weeks continuously. After one negative sample, HPeV-6 was again detected at two positive time points spanning 1 week. The VP1 coding region of the HPeV-6 was determined for each time point and found to be invariant throughout the sampling period, indicating that no reinfection took place (the closest HPeV-6 sequence in GenBank varied by >3% of nucleotides).
Human picobirnavirus genogroup I (HPBV-GI) was detected during 23 consecutively collected samples, lasting 185 days, except for two negative samples near the end of the extended period of shedding (Table 3). The PCR amplicons were directly sequenced and were all identical, indicating that no reinfection took place (the closest HPBV sequence in GenBank varied by 24% of nucleotides).
The number of coinfections was high, with 41% of samples from child 1 and 68% of samples from child 2 containing two or more viruses. Six percent of samples from child 1 and 16% from child 2 contained at least 4 distinct human viruses (Table 4). Shedding of attenuated poliovirus vaccine strains was excluded from these calculations.
The availability of frequently collected fecal samples allowed a detailed analysis of viral shedding occurring during the first year of life in two infant sibling from a developed country. A total of 92% of the 72 samples analyzed contained at least one human virus, with some samples containing up to five different viruses. The average fecal samples contained 1.8 viruses. While symptoms requiring hospitalizations were not observed in these two infants, some of the minor signs of infections, such as runny nose and loose stools, frequently seen in infants of that age, may have been caused by these viral infections or coinfections.
Anelloviruses were the most commonly detected viruses in both infants (55/72 [76%] samples were positive), with TTV dominating but TTMV infection or coinfections also detected. The next most common viruses were picobirnaviruses, found over an extended period of time in one infant (25/72 [35%] samples were positive). Human parechoviruses (HPeV types 1 and 6) were the next most common infection (15/72 [21%] samples were positive). HBoV-1 was found in 10 samples of one infant and at a single time point of her sibling (11/72 [15%] samples were positive). Adenovirus groups C and F were detected in only one infant at 7 time points and Aichi virus, astrovirus-HMO-B, and rotavirus were detected in one or two samples.
Prior analyses focusing on enteroviruses reported frequent human enterovirus infection in healthy Norwegian children sampled monthly, with HEV-A, -B, and -C detected in 6.8%, 4.8%, and 0.2% of the samples tested (68). With the weekly sampling used here and excluding the poliovirus vaccine strains, we detected HEV-A in 13.8% and HEV-B in 2.7% of the samples analyzed.
An extended period of viral shedding was detected for adenoviruses, anelloviruses, picobirnaviruses, parechoviruses, and human bocavirus. Human adenoviruses (HAdV) are significant pathogens associated with sporadic cases as well as outbreaks of acute gastroenteritis in humans. HAdV-41 (group F) was detected at one time point for child 1, immediately after extended (19 days) detection of HAdV-1. Human adenovirus type 1 is a group C adenovirus which has been associated with respiratory, gastrointestinal, and ocular diseases among children.
Anelloviruses are found at high prevalence in human blood and generally thought to be a chronic commensal infection (49). While no direct evidence has been reported for anellovirus-induced pathogenesis, their theoretical involvement in carcinogenesis has been discussed (71), and their very wide genetic diversity (8, 30) has the potential to encode a wide range of phenotypes. Anellovirus detection at the first time point tested for child 2 on day 142 and starting on day 151 for child 1 reflects the early infections of these infants. Anelloviruses have been detected in many tissues (49), blood (7, 49), umbilical cord (22, 24), breast milk (24, 46), feces (51, 53), and urine (13), and multiple routes of transmission have been proposed, including trans-placental (22, 24), from the cervix during delivery (14, 19), fecal-oral (51), respiratory (16), breast feeding (24, 46), or community acquisition (42, 44). The early and nearly chronic detection of different anellovirus species in these infants' feces may reflect enteric infections, viruses from plasma, secretion of infected bile (31), or swallowing of infected respiratory secretions (16). The long-term fecal shedding of anelloviruses is likely to reflect chronic infections similar to that observed in blood (49). Coinfection with two or more of the three known species of human anelloviruses (TTV, TTMV, and TTMDV) has been reported in blood (50) and was also detected with TTV and TTMV in a fraction of the stool samples from child 2.
Picobirnaviruses were first identified in 1988 (52) and sequenced in 2005 (66). These viruses are frequently found together with viral pathogens such as rotaviruses or caliciviruses in cases of diarrhea, where they may play a synergistic role (20). Prolonged shedding of HPBV in the stool of AIDS patients lasted between 45 days to 7 months (23, 25) compared to the 185 days seen for child 2.
Human parechovirus types 1 and 6 were also detected over extended periods of time. While detection of HPeVs in the stool of children has not been associated with symptoms (27, 60, 61), neonatal infections can be severe (11, 29, 64). Parechoviruses are secreted through the gastrointestinal and upper respiratory track and have been linked to various central nervous systems as well as other conditions (58). The duration of parechovirus shedding in stool has been calculated to be on average 51 (60) or up to 93 (41) days, in keeping with its detection for 53 and 64 days for child 1 and 2, respectively.
Human bocavirus 1 (HBoV-1) was first genetically characterized in 2005 (1) and generally thought to be a respiratory pathogen, especially when replicating with other respiratory viruses and detectable in blood (33, 59, 62). Seroprevalence to HBoV-1 is high in young children (34, 35). While HBoV-1 has not been associated with diarrhea (15), other closely related species (HBoV-2, -3, and -4) are more commonly detected in feces and may be associated with diarrhea (4, 38). In keeping with our observation of HBoV-1 shedding for 121 days in child 2, extended and intermittent shedding of HBoV-1 has also been reported in asymptomatic children in day care (45).
Shedding of Aichi virus, astrovirus-HMO-B, and rotavirus was of much shorter duration. Consistent with detection of rotavirus at a single time point, asymptomatic shedding of rotavirus has been recognized and found to be of short duration, which is extended in cases of severe diarrhea (56, 67). Aichi virus was first isolated in 1991 in Japan (70) and sequenced in 1998 (69) and is only rarely associated with diarrhea, although seroprevalence is high, indicating that a large fraction of infection may be asymptomatic (17, 55). The duration of Aichi virus shedding is not clear, and our results indicate that it may be short, at least in the absence of severe symptoms. The association of the recently characterized human astrovirus-HMO-B (similar to AstV-VA3) (18, 37) with enteric symptoms and its duration of shedding is still uncertain. Serological testing of a closely related astrovirus species (HMOAstV-C) indicated a high rate of exposure in children, likely reflecting a generally mild infection (12).
Caliciviruses (noroviruses and sapoviruses), picornavirus genera Cardiovirus, Cosavirus, and Salivirus, and coronaviruses were not detected using the RT-PCR primers and conditions used here. The apparent absence of these viruses may also be due to their presence below the sensitivity of these RT-PCR assays, to target sequence mismatches preventing primer annealing, or to nucleic acid degradation during long-term sample storage.
The two infants analyzed here were breast fed during the period of sample collection. The protective effect of maternal antibodies against severe consequences of viral infections may have reduced the diversity or duration of enteric viruses detected in stool specimens (26, 63).
The surprisingly high rate of virus detection reported here may still represent an underestimate of viral shedding due to viruses being below levels of detection of the PCR assays used. The intermittent detections seen for the more persistent infections, including HPBV-GI, HPeV-1, HPeV-6, and HBoV-1, may indeed reflect fluctuation of the viral loads below detection levels. Other missed infections may also have resulted in very transient shedding occurring between the nearly weekly collected sampled analyzed here.
Our study showed that two healthy infant siblings were nearly constantly shedding a wide range of enteric viruses during their first year of life. While only two children were analyzed, their customary upbringing indicates that the diversity and duration of enteric viral shedding observed here may reflect that of typical infants in developed countries. Testing of longitudinally collected samples from a larger number of infants will be required to further substantiate this conclusion.
The high number of different infections and in some cases their long-term persistence detected here by PCR show that as more sensitive methods of viral detection are used, an increasing number of asymptomatic infections can be detected, likely reflecting effective passive and/or active immunity in generally healthy infants. The possibly substantial effect on the education of these infants' immune systems of such frequent and long-lasting viral infections, including protection from subsequent challenges with closely related viruses, remains to be determined.
This work was supported by NHLBI R01HL083254 and BSI to E.D.
Published ahead of print 8 August 2012