Identification of Enteroviruses in Naturally Infected Captive Primates

Abstract

In a recent study, we investigated cases of diarrheal disease among monkeys at a U.S. primate center. In that study, enteroviruses were detected in a high proportion of the fecal specimens tested. To determine whether the enterovirus detections represented the circulation of one or more simian enteroviruses within the colony or the transmission of human enteroviruses from animal handlers, we determined in the present study the serotype identity of each virus by reverse transcription-PCR and sequencing of a portion of the VP1 gene, a region whose sequence corresponds to antigenic type. Enteroviruses were identified in 37 of 56 specimens (66%), 30 of 40 rhesus macaques, 5 of 11 pigtail macaques, 2 of 4 sooty mangabeys, and 0 of 1 chimpanzee. No previously known human viruses were detected. Three previously known simian enterovirus serotypes—SV6, SV19, and SV46—were among the viruses identified, but more than half of the identified viruses were previously unknown; these have been assigned as new types: EV92 and EV103.

The viral flora of nonhuman primates includes examples of virtually all virus groups that infect and cause disease in humans (5, 12), making these viruses attractive models for human disease, as well as a potential zoonotic disease threat (8, 9, 22). Some of the well-characterized simian viruses include adenoviruses, enteroviruses, herpesviruses, papovaviruses, paramyxoviruses, reoviruses, and retroviruses (11). The simian enteroviruses were originally isolated from primary cell cultures or clinical specimens from captive and wild-caught primate species that were commonly used in biomedical research, including Macaca mulatta (rhesus macaque), M. fascicularis (cynomolgous monkey), Cercopithecus aethiops (African green monkey [vervet]), and Papio species (baboon) (1, 3, 4, 7, 14). A number of these viruses are related to human enteroviruses, with up to 72% amino acid identity in the complete capsid protein (17); as a point of reference, closely related human enterovirus serotypes are ca. 65 to 85% identical to one another in this same region (20) Serologic studies have suggested that some of the simian enteroviruses may be transmissible from primates to humans, primarily animal handlers, and that human enteroviruses may also be transmissible to certain primate species (11). Despite their prevalence, enteroviruses are not commonly associated with disease in monkeys (10).

In a recent study, we investigated cases of diarrheal disease among monkeys at a U.S. primate center (25). In that study, fecal specimens were tested for enteric viruses, including rotavirus, adenovirus, and enterovirus. Of 56 animals tested, enteroviruses were detected in 75% of rhesus macaques (30/40) and 45% of pigtail macaques (5/11), while adenovirus was detected only in rhesus macaques and there were no rotaviruses detected. To determine whether the enterovirus detections represented circulation of one or more simian enteroviruses within the colony or transmission of human enteroviruses from animal handlers, we determined here the serotype identity of each virus by reverse transcription-PCR (RT-PCR) and sequencing of a portion of the VP1 gene, a region whose sequence corresponds to antigenic type (16, 19).

MATERIALS AND METHODS

Primate colony and specimen collection.

The colony was comprised of approximately 2,700 rhesus macaques (M. mulatta), 250 pigtailed macaques (M. nemestrina), and 250 sooty mangabeys (Cercocebus atys). All animals studied were laboratory-born, and all except one rhesus male was born at the Yerkes colony. Fecal specimens were collected from rhesus and pigtailed macaques, sooty mangabeys, and a chimpanzee (Pan troglodytes) housed at the Yerkes National Primate Research Center field station, from February to April 1999, as described previously (24). The pigtailed macaques and rhesus macaques normally reside in natal groups of the same species in the field station outdoor corrals, but they are sometimes housed within the same rooms indoors. The chimpanzees and the sooty mangabeys are housed in separate areas from the other species, regardless of whether they are outdoors or indoors. The animals ranged in age from 6 months to 25 years. All had diarrhea at the time of specimen collection, except two pigtail macaques and three rhesus macaques, which were sampled as normal controls.

Enterovirus identification by RT-PCR and sequencing.

Of the original 59 fecal specimens (25), 56 were available for further virologic investigation. RNA was extracted from stool specimens by using guanidine isothiocyanate and silica, as described previously (2). Enterovirus VP1-specific RT-PCR and sequencing were carried out as previously described (15). Briefly, cDNA was synthesized using primers in VP1 and PCR was performed using primers in VP3 and VP1 to produce a first-round product of ∼700 bp. One microliter of the first PCR product was added to a second PCR containing nested primers within VP1 to produce a product of approximately 320 to 350 bp (∼35% of the VP1 region). Reaction products were separated and visualized on an agarose gel, and purified by using the QIAquick gel extraction kit (Qiagen, Inc., Valencia, CA). The resulting DNA templates were sequenced on both strands with the second-reaction primers using the BigDye Terminator v1.1 ready-reaction cycle sequencing kit on an ABI Prism 3100 automated sequencer (both from Applied Biosystems, Foster City, CA). Amplicon sequences were compared to the VP1 sequences of enterovirus reference strains, including at least one representative of each recognized serotype, by script-driven sequential pairwise comparison using the program Gap (Wisconsin Sequence Analysis Package, version 10.3; Accelrys, Inc., San Diego, CA), as described previously (15, 19). In this scheme, a VP1 nucleotide sequence identity of greater than 75% confirms the enterovirus serotype present in the specimen. This approach has been shown to be suitable for complete VP1 sequences and for partial sequences of at least 300 nucleotides (nt) (15-16, 18-19).

A portion of the 3D (polymerase) region was amplified by RT-nested PCR. 3D-specific cDNA was synthesized with Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA) using primer AN368 (AAACAACTAGACTTRTCNGCNGGNGTCAT, nt 6957 to 6929 relative to SV19) in a 10-μl reaction. The reaction was incubated at 42°C for 1 h, followed by 5 min at 95°C. The first-round PCR used primers AN392 (TTCCCAACCAAAGCNGGNCARTG, nt 5735 to 5757 relative to SV19) and AN368, Taq polymerase (Roche Applied Science, Indianapolis, IN), and the entire cDNA reaction in a 50-μl reaction. The reaction was carried out by performing 40 cycles of 95°C for 30 s, 55°C for 20 s, and 60°C for 50 s. One microliter of first-round PCR product was added to a 50-μl reaction containing primers AN366 (GGCAATGGAAGCCARGGNTTYTGYGC, nt 5807 to 5832 relative to SV19) and AN367 (CCAGAAAGTATCTGGGTTRCANCCNACNGC, nt 6520 to 6491 relative to SV19), and FastStart Taq (Roche Applied Science), with initial denaturation at 95°C for 6 min, followed by 40 cycles of 95°C for 30 s, 60°C for 15 s, and 70°C for 40 s. Amplicon sequences were determined as described above using the primers AN366 and AN367.

Sequence analysis.

The partial VP1 sequences were compared to a database of complete enterovirus VP1 sequences of all serotypes to determine whether the isolates were genetically related to any known enterovirus serotype. Multiple alignments and phylogenetic relationships of VP1 and 3D sequences, including the Yerkes sequences and those of EV reference strains, were constructed by using the neighbor-joining method implemented in CLUSTAL W (23). Phylogenetic analysis of the 3D sequences was conducted on 26 of the 37 VP1-positive samples. Eleven specimens were excluded due to the lack of sufficient sample, failed 3D sequencing, or because they were mixtures of viruses and the 3D sequence could not unambiguously be assigned to the correct VP1 sequence.

Nucleotide sequences.

The sequences described here have been deposited in the GenBank database, accession no. EU194488 to EU194550.

RESULTS

Enterovirus sequences were amplified directly from stool extracts by an RT-snPCR targeting the VP1 region of the genome, using primers that amplify all known enterovirus types (15, 18). Sequencing of the ∼320-nt amplicon and comparison of the sequence to a database of reference sequences yields a type identification (15, 19). The VP1 assay was positive for 37 of 56 (66%) specimens tested (Table 1). The sensitivity and specificity of the 5′-nontranslated region (5′NTR) PCR/hybridization assay (25) and the VP1 assay were approximately equal, as the 5′NTR assay detected enterovirus genome in 36 of 53 (68%) specimens tested (Table 1). One specimen was positive only in the PCR-hybridization assay (25), and one specimen was positive only in the VP1 assay. Of the three specimens not tested by the 5′NTR PCR-hybridization assay, one was positive in the VP1 assay, and two were negative. The VP1-positive animals included 30 of 40 rhesus macaques (75%), 5 of 11 pigtails (45%), and 2 of 4 mangabeys (50%). There were no detections of known human enteroviruses, indicating that human-to-primate transmission was unlikely to have occurred during the study period. All five control animals were among those that were negative by the VP1 assay (data not shown).

TABLE 1.

Detection of enteroviruses by two different RT-PCR assays^a

VP1	No. of virus strains detected by PCR/hybridizationb
	Pos	Neg	Not tested
Pos	35	1c	1d
Neg	1	16	2

^aTwo RT-PCR assays were tested: one using RT-PCR and probe hybridization to target the 5′NTR (PCR/hybridization) and the other using a seminested RT-PCR assay targeting VP1, followed by amplicon sequencing to identify the detected virus (VP1).

^bPos, positive; Neg, negative.

^cSV46 was identified in this specimen using the VP1 assay.

^dEV92 was identified in this specimen using the VP1 assay.

The viruses identified represent five different types, including three known simian enteroviruses (SV6, SV19, and SV46), as well as two apparently new enterovirus types (Table 2). The new types have been registered with the Picornavirus Study Group of the International Committee on the Taxonomy of Viruses as EV92 and EV103. SV19 and SV46 were relatively common, detected in five and eight animals, respectively, while SV6 was detected in only one animal. Surprisingly, EV92 was detected more frequently than any other type, accounting for more than half of all detections (21/37). EV103 was detected in three animals. The initial EV92 VP1 nucleotide sequence, from animal RJg-7, is <66% identical to those of reference strains representing all known enterovirus types, including the other simian enteroviruses. The initial EV103 VP1 nucleotide sequence, from animal POo-1, is <62% identical to those of reference strains representing all known enterovirus types, including the other simian enteroviruses. SV19, SV46, EV92, and EV103 were identified in both rhesus and pigtail macaques, while SV6 was detected only in a mangabey (Table 2). Mixed infections were detected in four animals. Two pigtail macaques had EV92-EV103 and SV46-EV103 coinfections, respectively. An SV46-EV103 mixture was detected in one rhesus macaque, and an SV6-EV92 mixture was detected in a sooty mangabey.

TABLE 2.

Enteroviruses identified in captive primates grouped by host species^a

Virus category	No. of virus strainsb identified in:				Total no. of strains
	Rhesus	Pigtail	Mangabey	Chimpanzee
SV6	-	-	-	-	-
SV19	4	1	-	-	5
SV46	6	-	-	-	6
EV92	19	1	1	-	21
EV103	-	1	-	-	1
Virus mixture	1c	2d	1e	-	4
Negative	11	6	2	1	20
No. positive/total no.	30/40	5/11	2/4	0/1	37/56

^aRhesus, rhesus macaque (M. mulatta); pigtail, pigtail macaque (M. nemestrina); mangabey, sooty mangabey (Cercocebus atys); chimpanzee (Pan troglodytes).

^b-, The indicated virus was not identified in this species.

^cMixture of SV46 and EV103.

^dOne SV46-EV103 mixture and one EV92-EV103 mixture.

^eMixture of SV6 and EV92.

To determine whether the detected viruses represent a single strain of each type or multiple cocirculating strains of a given type, we constructed a phylogenetic tree using the partial VP1 nucleotide sequences of the Yerkes viruses and reference strains of other simian and human enteroviruses (Fig. 1). The SV19 and SV46 strains clustered with their respective reference strains, within the species Human enterovirus (HEV) A, while EV92 formed a separate cluster (as expected for a new enterovirus type), also in HEV-A (Fig. 1). All of the Yerkes HEV-A strains and their respective reference strains were on a branch that also includes the human viruses EV76, EV89, EV90, and EV91. EV103 formed a distinct cluster that is most closely related to SV6 and N125 (Fig. 1). For each of the types with multiple detections (SV19, SV46, EV92, and EV103), the range of diversity within the type (1 to 4% within SV19, 0 to 6% within SV46, 1 to 8% within EV92, and up to 7% within EV103) suggests that multiple strains cocirculated within the colony or that single founder strains of each type have circulated in the colony for several years, with diversification over time.

FIG. 1.

Phylogenetic relationships in the VP1 region based on analysis of partial VP1 sequences of Yerkes enteroviruses and reference strains of other enteroviruses. Phylogenetic trees were generated by the neighbor-joining algorithm implemented in CLUSTAL W (23), with 1,000 bootstrap replicates, based on analysis of partial VP1 sequences of Yerkes enteroviruses and reference strains of other enteroviruses. Bootstrap values over 60% are indicated at the respective nodes on the tree. For clarity, subtrees containing only nonsimian sequences (e.g., other HEV-A) have been collapsed and are indicated by elongated triangles, the length of which depicts the overall diversity within the subtree. Sequences determined in the present study are indicated by an animal identifier (e.g., RHp-4). Reference sequences are indicated by serotype and GenBank accession number. HEV, human enterovirus species A, B, C, or D; HRV, human rhinovirus A or B; PEV, porcine enterovirus; BEV, bovine enterovirus.

Partial 3D amino acid sequences were compared to one another and to those of enterovirus reference strains by phylogenetic reconstruction. The 3D sequences of the Yerkes strains formed three clusters, two of which were related to sequences of the human and simian viruses in HEV-A (Fig. 2). Within the HEV-A group, 11 Yerkes sequences (5 SV19 and 6 EV92) clustered with SV19, and 14 sequences (3 SV46 and 11 EV92) formed an independent cluster within the species cluster. The partial 3D sequence of the EV103 prototype strain, POo-1, was most closely related to that of the SV6 reference strain (Fig. 2).

FIG. 2.

Phylogenetic relationships in the 3D region based on analysis of partial VP1 sequences of Yerkes enteroviruses and reference strains of other enteroviruses. Phylogenetic trees were generated by the neighbor-joining algorithm implemented in CLUSTAL W (23), with 1,000 bootstrap replicates, based on analysis of partial 3D sequences of Yerkes enteroviruses and reference strains of other enteroviruses. Bootstrap values over 60% are indicated at the respective nodes on the tree. For clarity, subtrees containing only nonsimian sequences (e.g., other HEV-A) have been collapsed and are indicated by elongated triangles, the length of which depicts the overall diversity within the subtree. Sequences determined in the present study are indicated by an animal identifier (e.g., RHp-4). Reference sequences are indicated by serotype and GenBank accession number. HEV, human enterovirus species A, B, C, or D; HRV, human rhinovirus A or B; PEV, porcine enterovirus; BEV, bovine enterovirus.

DISCUSSION

In this study, we identified five simian enterovirus types circulating among three primate species in a large primate research colony. Three of the types—SV6, SV19, and SV46—were originally described during early studies of simian viruses in the 1950s and 1960s. The other two types—EV92 and EV103—were not closely related to any known enterovirus type. Based on VP1 sequence comparisons, EV92 and EV103 have been accepted as new enterovirus types (13). The VP1 and 3D sequences of Yerkes SV19 and SV46 strains were most closely related to those of viruses in the species HEV-A, in agreement with previous studies on the SV19 and SV46 reference strains, supporting the classification of these serotypes in HEV-A (17). The EV92 VP1 and 3D sequences also clustered in HEV-A, supporting its classification in that species. Analysis of complete genome sequences for EV92 and EV103 strains will allow definitive classification of these two new viruses.

SV19, EV92, and EV103 were found in both rhesus and pigtail macaques (Table 2). The close genetic relationships between viruses of a given type in the two monkey species (e.g., EV92 strains RFk-5 [rhesus] and PJf-1 [pigtail]) suggests that these viruses have been transmitted between the two host species and that the two monkey populations probably represent a single virus reservoir. Interspecies transmission is considered likely because the two species are often housed together when they are inside. Interestingly, SV19, EV92, and EV103 were not detected in mangabeys, a species that is kept separate from the macaques at the Yerkes field station; however, the number of mangabeys in the study (n = 4) was too small to allow us to determine whether SV19, EV92, and EV103 circulated exclusively in the macaque populations. On the other hand, one can conclude that SV6 was probably present only in the mangabeys and not in the macaques.

There are few data to associate simian enterovirus infection with specific diseases in primates (10). There have been reports of minor central nervous system lesions in experimentally infected animals (10), but most simian enterovirus isolations have been from specimens derived from healthy animals (1, 3, 4, 6, 7, 14). For example, SV4 and SV6 were identified by the observation of cytopathic effect in normal primary monkey kidney cells that were being prepared for the production of polio vaccine (7). On the other hand, SV6 and SV19 have also been isolated from monkeys with acute gastroenteritis, but there was no clear association between infection and disease (4). The vast majority of human enterovirus infections are asymptomatic (21), and the same may be true for the simian enteroviruses. If this is the case, it may be difficult to link infection with disease in primates, particularly since the infections appear to be relatively common.

The origin of the enterovirus infections in the Yerkes colony remains unclear. Since no human enteroviruses were detected, it is unlikely that the viruses were introduced by animal handlers in the facility. No wild-caught animals have been introduced into the colony in approximately 25 years, so it is also unlikely that the viruses were recently introduced from a wild reservoir. Introduction from another primate facility is possible, since animals are sometimes received from other primate colonies, but those facilities would also be expected to have ceased primate importation many years ago. It is more likely that the viruses have been circulating at Yerkes and/or another colony for many years, probably since the time when importation of wild-caught primates from South Asia was common (i.e., the 1950s or 1960s). The range of VP1 sequence diversity within a given type suggests that individual founder strains of the different virus types were introduced once and have subsequently circulated and evolved within the colony for many years. Alternatively, multiple strains of each type may have been introduced at different times, resulting in cocirculation of several strains of a given type. Unfortunately, there is insufficient information about natural simian enterovirus transmission and diversity to distinguish between these possibilities.

Additional studies are needed to assess the prevalence of simian enterovirus infections in wild and captive primate populations to determine whether the viruses cause disease in primates and to investigate the potential for their zoonotic transmission to humans. Studies are under way to survey picornaviruses excreted by wild and captive primates to better understand the diversity, natural history, and ecology of simian enteroviruses in their native environment.