Abstract
Although many studies have characterized catarrhine and platyrrhine primate herpesviruses, little is known about herpesviruses in prosimians. We aimed to identify and characterize herpesviruses in prosimians with proliferative lymphocytic disease. DNA was extracted from tissues of 9 gray mouse lemurs (Microcebus murinus) and 3 pygmy slow lorises (Nycticebus pygmaeus) with lymphoproliferative lesions, and we performed nested PCR and sequencing for detection of herpesviruses and polyomaviruses. We identified 3 novel herpesviruses and performed phylogenetic analyses to characterize their relationship with other herpesviruses. A gray mouse lemur herpesvirus clustered with other primate herpesviruses within the subfamily Betaherpesvirinae, just basal to the genus Cytomegalovirus. The other gray mouse lemur herpesvirus and the pygmy slow loris herpesvirus clustered within the subfamily Gammaherpesvirinae, although the relationships within the subfamily were less resolved. Quantitative PCR assays were developed for the 2 new gray mouse lemur viruses, providing specific, faster, less expensive, and quantitative detection tools. Further studies are needed to elucidate the relationship between the presence of these viruses and the severity or presence of lymphoproliferative lesions in prosimians.
Prosimians are a paraphyletic group of primates divided into 4 infraorders within the order Primata. Lorisiformes (suborder Strepsirrhini) includes lorises, pottos, and galagos; Lemuriformes (suborder Strepsirrhini) consists of the lemurs; Chiromyiformes (suborder Strepsirrhini) comprises the aye-aye; and Tarsiiformes (suborder Haplorrhini) includes the tarsiers. 22 All prosimians are threatened to various degrees in the wild, and only a few species exist in zoologic institutions.
Viruses in the family Orthoherpesviridae (formerly Herpesviridae) infect amniotes, replicate in the nucleus, and have high host fidelity. 17 Herpesviruses are divided into 3 subfamilies: Alphaherpesvirinae, Betaherpesvirinae, and Gammaherpesvirinae, based on their genetic and biologic properties. 17 Most primate herpesvirus investigations have looked at haplorrhine hosts, with few reports of herpesviral infections in prosimians.
Members of the subfamily Alphaherpesvirinae establish latent infections in sensory ganglia and mononuclear cells, replicate in various host tissues, and have a rapid lytic cycle. 17 Alphaherpesviruses frequently cause asymptomatic or mild infections in their natural hosts. However, they can cause severe illness when transmitted to a new species, as has often been described with fatal herpes simplex virus 1 (Simplexvirus humanalpha1; formerly human alphaherpesvirus 1, HAHV1) infections in New World primates. 29 In prosimians, HAHV1 was identified from a small zoo colony of ring-tailed lemurs (Lemur catta) with fatal illness. 15
Members of the subfamily Betaherpesvirinae are usually latent in lymphoid tissues and have a highly restricted host range, with a slow replicative cycle and delayed cell lysis. 17 Examples of viruses in this subfamily are human betaherpesvirus 5 (HBHV-5; Cytomegalovirus humanbeta5) and human betaherpesvirus 6 (HBHV-6; Roseolovirus humanbeta6a, Roseolovirus humanbeta6b). A study that combined paleovirologic and metagenomic approaches described the first endogenous herpesvirus in prosimians from the genome of the Philippine tarsier (Carlito syrichta), belonging to genus Roseolovirus. 2
Members of the subfamily Gammaherpesvirinae are characterized by their tropism and replication in lymphoid cells, with specificity for B or T lymphocytes depending on the virus. 17 These viruses have a narrow host range, and some are linked to the oncogenic transformation of lymphocytes, for example, Epstein-Barr virus (Lymphocryptovirus humangamma4; formerly human gammaherpesvirus 4, HGHV-4) and Kaposi sarcoma-associated herpesvirus (Rhadinovirus humangamma8; formerly human gammaherpesvirus 8, HGHV-8). 17 In nonhuman primates, ateline gammaherpesvirus 2 (AtGHV-2; Rhadinovirus atelinegamma2) and saimiriine gammaherpesvirus 2 (SaGHV-2; Rhadinovirus saimiriinegamma2) are 2 of the most studied simian lymphotropic herpesviral species.5,6,12,13,19,23,30,31 Aside from their impact on nonhuman primate populations, their relationship to HGHV-4 makes them a valuable model for human disease. 26 In prosimians, complement-fixing antibodies reactive with HGHV-4 antigens were detected in 7 of 13 individuals, indicating potential susceptibility to this virus. 10 Additionally, a rhadinovirus (Gammaherpesvirinae), closely related to HGHV-8, has been described integrated into the aye-aye (Daubentonia madagascariensis) genome. 2
Based on ultrastructural morphology, a herpesvirus was also identified as the etiologic agent of fatal encephalitis in an adult male black-and-white ruffed lemur (Varecia variegata). 16 Furthermore, herpesvirus-like particles have been reported from a slow loris (Nycticebus coucang) with lymphoma, and have also been associated with dental disease, dermatitis, and ocular discharge in this species. 25 In that report, herpesvirus-like particles were observed in cell culture, and size and morphology were indistinguishable from a herpesvirus previously isolated in cultured lymphocytes of slow lorises. Viral serology was positive only for HAHV-1, which may have been an incidental finding. 22
Members of the Orthoherpesviridae family that infect lymphoid cells are defined as lymphotropic herpesviruses. These herpesviruses often cause lytic infections and conditions leading to self-limiting lymphoproliferative disease or malignant lymphoma. 26 The malignant transformation of lymphocytes induced by primate herpesviral species has been extensively studied in haplorrhine primates. SaGHV-2 is regularly found in squirrel monkeys (Saimiri sciureus). Whereas this virus has not been associated with lesions in its natural host, SaGHV-2 can cause lymphoma or lymphocytic leukemia in other New World primate species, including several species of tamarins and marmosets (Saguinus spp. and Callithrix spp.), owl monkeys (Aotus sp.), howler monkeys (Alouatta sp.), and spider monkeys (Ateles sp.) after experimental infection.5,6,12,13,19,23,30,31 Additionally, AtGHV-2 is not known to be pathogenic in its natural host, the spider monkey, but can induce lymphoblastic leukemia in cotton-top tamarins (Saguinus oedipus) and white-lipped tamarins (S. labiatus) after viral inoculation. 9 HGHV-4 and HGHV-8 in humans usually cause latent infections without lesions. However, both can cause lymphoid malignancies, the frequency of which is greatly increased in individuals with immunodeficiency, whether primary or acquired, as seen with HIV infections or therapeutic immunosuppression for organ transplantation. 4
We aimed to identify and characterize herpesviruses in prosimians with proliferative lymphocytic neoplasia and develop tools for investigating a relationship between the lymphocytic diseases and the presence of the virus in these species, as is found with herpesviruses in other primates.
Materials and methods
Samples
All animals tested had been housed at the Duke Lemur Center (Durham, NC, USA) and either died or were euthanized because of declining conditions resulting from lymphoproliferative neoplasia, specifically various forms of leukemia, lymphosarcoma, or both. Tissue samples were collected postmortem from 9 gray mouse lemurs (Microcebus murinus) and 3 pygmy slow lorises (Nycticebus pygmaeus) from 1991–2012. Postmortem examinations were performed on animals within 24 h of death. Samples of all major organs were collected in 10% neutral-buffered formalin, and histologic evaluation was performed by a board-certified veterinary pathologist. Portions of the liver, spleen, kidneys, and skin were banked at −80°C until testing was performed for viruses.
Amplification, sequencing, and gene identification
Molecular tests were performed in 2012–2013. DNA was extracted from tissue samples (16 DNA purification kit, automated extractor; Maxwell, Promega) following the manufacturer’s instructions. Analyzed tissue samples varied between individuals based on which tissue had been stored and was available (Table 1), and included liver (n = 7), spleen (n = 4), and kidney (n = 1) samples. Nested PCR assays were used to detect the herpesviral DNA polymerase gene and the polyomaviral VP1 gene, as described previously.8,27 The herpesvirus protocol used first-round primers DFA, ILK, and KG1, followed by a second round with primers TGV and IYG. 27 The amplification products were separated by agarose gel electrophoresis. Bands of interest were cut from the gel, and DNA was extracted and sequenced. For herpesvirus-positive samples, to obtain longer sequences, the second round was modified to use primers DFA and IYG. Additional primers were designed for a new second round of nested PCR for the positive pygmy slow loris samples. These samples were set up in 2 separate second rounds, 1 with primers DFA and LorisGHVR (5′-GCATTGTTCTTCCCCTCAGA-3′), and the other with LorisGHVF (5′-TGGTCTCTTACCCTGCATCA-3′) and IYG. Homologies to other known proteins were determined using BLASTP. 1
Individual information, histopathology, consensus PCR and quantitative PCR (qPCR) results for 9 gray mouse lemurs (Microcebus murinus) and 3 pygmy slow lorises (Nycticebus pygmaeus).
CBHV-1 = cheirogaleid betaherpesvirus 1; CGHV-2 = cheirogaleid gammaherpesvirus 2; LGHV-1 = lorisid gammaherpesvirus 1; 18S Ct = crossover threshold for control host 18S rRNA gene; MM = gray mouse lemur; ND = not detected; NP = pygmy slow loris; U = unknown.
Phylogenetic analysis
The predicted amino acid sequences of 55 herpesviral DNA polymerases were downloaded from GenBank and aligned using MAFFT. 14 Iguanid herpesvirus 2 (GenBank AY236869) was designated as the outgroup. Bayesian analyses were performed using MrBayes v.3.2.3 11 on the CIPRES 20 server with gamma-distributed rate variation and a proportion of invariant sites, and mixed amino acid substitution models. 24 For the Bayesian analyses, 4 chains were run, each for 2,000,000 generations: 3 hot chains and 1 cold chain with the default heating parameter (temperature = 0.2). Convergence among different runs was evaluated by calculating the average split deviation using a threshold of 0.02%. Chains were sampled every 100 generations; the first 25% were discarded as burn-in.
Quantitative PCR
A quantitative PCR (qPCR) assay was created for the betaherpesvirus and gammaherpesvirus identified in the gray mouse lemur samples. Primers and probes were designed specifically for these novel viruses (Table 2). Each 20-μL solution was run in duplicate and consisted of 1 μL of each primer and probe, 3 μL of water, 4 μL of standard or sample (25 ng/μL), and 10 μL of a commercial universal qPCR mix (TaqMan fast universal PCR master mix 2×; Applied Biosystems).
Quantitative PCR primer and probe sequences (5′–3′) for targeting lemur herpesviruses.
The last letter in the name denotes the type of primer: F = forward, R = reverse. The probe includes reporter and quencher type in sequence: 6-carboxyfluorescein (FAM) reporter and black-hole quencher (BHQ).
The concentration of the amplicons in each PCR product was measured (Nanodrop 8000 spectrophotometer; Thermo Fisher). The qPCR standard curve was made using 10-fold serial dilutions, ranging from 106 to 101 copies of each amplicon containing the target sequence run in triplicate on each plate. The function of the qPCR probes and primers was assessed by evaluating the standard curve R2 values, slopes, and efficiency. For validation of the presence of amplifiable DNA, a eukaryotic 18S rRNA endogenous control kit (VIC/MGB probe; Applied Biosystems) was used in a separate well with each sample. The reaction was amplified (7500 fast real-time PCR system; Thermo Fisher) with the following conditions: initial denaturation at 95°C for 20 s, and 50 cycles of 95°C for 3 s followed by 60°C for 30 s. The samples were run in duplicate on each plate and once with the 18S control. The QuantStudio 7500 software (v.2.0.6; Applied Biosystems) on the real-time PCR system was used to analyze the results and calculate assay slopes, R2 values, and efficiency. The qPCR-specificity was tested using 8 positive samples for 5 other herpesviruses originating from samples submitted to the University of Florida–Zoo Medicine Infectious Disease Laboratory (Gainesville, FL, USA), including marmoset lymphocryptovirus (Lymphocryptovirus callitrichinegamma3), ursid gammaherpesvirus 1, elephantid gammaherpesvirus 4, mustelid gammaherpesvirus 2, and otariid gammaherpesvirus 3.
Results
Samples
All of the animals had some degree of lymphoproliferative neoplasia as determined by histopathology (Table 1). The most affected sites were the liver, spleen, lungs, kidneys, and lymph nodes. Some animals had focal lymphosarcoma; others had diffuse involvement of numerous organs and vessels. Most samples had tissue necrosis associated with neoplastic proliferation.
Amplification, sequencing, and gene identification
Two distinct herpesvirus sequences were identified from the gray mouse lemur samples, 480-bp and 472-bp long after primers were edited out (Table 1). Each sequence was amplified from 2 of 9 samples. Mixed infection was not detected by conventional PCR assay in any animals. BLASTP found that the 480-bp sequence was most similar to members of Betaherpesvirinae and is here referred to as cheirogaleid betaherpesvirus 1 (BHV-1). Two gray mouse lemurs (MM07 and MM09) had 3 nucleotide differences between their obtained sequences that did not result in any predicted amino acid differences. The 472-bp gray mouse lemur sequence was most homologous to members of Gammaherpesvirinae and is here referred to as cheirogaleid gammaherpesvirus 2 (GHV-2). A distinct herpesviral sequence was identified from all 3 pygmy slow loris samples and was 472-bp long after the primers were edited out. BLASTP found that the pygmy slow loris sequence was also closely related to gammaherpesviruses and is here referred to as lorisid gammaherpesvirus 1 (GHV-1). Sequences were submitted to GenBank (KT698104–KT698107). The polyomaviral VP1 gene PCR resulted in no amplification.
Phylogenetic analysis
The WAG model was the most probable amino acid substitution model with a posterior probability of 1.000 (Fig. 1). 28 Phylogenetic analysis shows that cheirogaleid BHV-1, indicated by a red arrow (Fig. 1), is a member of the Betaherpesvirinae subfamily with a posterior probability of 100%. Cheirogaleid BHV-1 is most closely related to genus Cytomegalovirus (Bayesian posterior probability of 98%), although it is basal to the earliest divergence in this genus. Cheirogaleid GHV-2 and lorisid GHV-1, indicated by red and green arrows (Fig. 1), are members of the Gammaherpesvirinae subfamily with a posterior probability of 100%. Cheirogaleid GHV-2 and lorisid GHV-1 did not clearly cluster within other known gammaherpesvirus genera (Fig. 1).
Bayesian phylogenetic tree of homologous partial herpesvirus DNA–dependent DNA polymerase predicted amino acid sequences. Iguanid herpesvirus 2 (GenBank AY236869) was designated as the outgroup. Bayesian posterior probabilities of branching as percentages are shown at nodes. GenBank accessions are given after virus names. Brackets demarcate genera, and arrows indicate viruses identified in our study.
Quantitative PCR
The standard curve generated by the 10-fold serial dilutions was linear in both the cheirogaleid BHV-1 and cheirogaleid GHV-2 PCR assays. The cheirogaleid BHV-1 assay had a slope of −3.392, indicating 97.1% efficiency, and an R2 value of 0.994. The cheirogaleid GHV-2 assay had a slope of −3.425, indicating 95.8% efficiency, and an R2 value of 0.998. Detectable cheirogaleid BHV-1 quantities were found in samples MM07 and MM09, with 12–210,417 copies per reaction (Table 1), whereas results for samples MM01, MM06, MM07, and MM09 were positive for cheirogaleid GHV-2, with viral quantities of 2–3,558 copies per reaction (Table 1). There was no amplification of the non-target herpesviruses with the cheirogaleid BHV-1 and cheirogaleid GHV-2 qPCR protocols, indicating specificity of the primers.
Discussion
We identified 3 novel herpesviruses, cheirogaleid BHV-1 and cheirogaleid GHV-2 in gray mouse lemurs, and lorisid GHV-1 in pygmy slow lorises. However, we could not prove the relationship between the presence of these viruses and lymphoproliferative disease. All samples tested were from subjects with different types of lymphoproliferative neoplasia diagnosed through histologic evaluation, but virus was not detected in all samples. As indicated in Table 1, the animals with the most lymphoproliferative changes did not consistently have high or any virus presence. There was also no relationship between the tissues used for virus detection and virus identification or the year of sample collection. A larger cohort of samples would be needed to evaluate the statistical significance of the findings and potentially highlight correlations between the viruses and lymphoproliferative disease.
As described in other species, gammaherpesviruses have evolved to maintain a delicate balance between the virus and host. The viruses are thought to have co-evolved with their natural hosts and only tend to cause problems when this balance is broken. Cheirogaleid GHV-2 in gray mouse lemurs and lorisid GHV-1 in pygmy slow lorises may not be related to high rates of lymphoproliferative disease because these species could be the endemic hosts to these viruses. Alternatively, the viruses may still have an unclear relationship to lymphoproliferative neoplasia in these species, especially in immunosuppressed animals, given that other gammaherpesviruses can cause disease in their natural hosts under specific conditions. 4 More studies need to be done to evaluate the infectivity of these novel viruses and their impacts on different prosimian populations.
One gray mouse lemur sample (MM09) tested positive for cheirogaleid BHV-1 and cheirogaleid GHV-2. The sample also had the highest amount of virus of all samples tested but was not from the animal with the more severe gross neoplastic lesions. Therefore, no relationship between the presence of the virus and the severity or presence of lymphoproliferative neoplasia can be supported at this time.
Comparing the phylogenetic analysis of the betaherpesvirus clade with the evolution of the placental mammals suggests codivergence. Superorder Euarchontoglires divides into 2 main branches: rodents and lagomorphs on the one hand, and tree shrews, colugos, and primates on the other. Within the primates, the earliest division is between Strepsirrhini and Haplorrhini (containing infraorder Tarsiiformes [tarsiers], parvorder Platyrrhini [New World monkeys], and parvorder Catarrhini [Old World monkeys]).21,22 The same structure can be seen in the betaherpesvirus phylogenetic analysis, implying the relationship between the virus and its hosts. Further investigation into the prosimian betaherpesviruses is expected to provide insight into the origins of the genus Cytomegalovirus.
Gammaherpesviruses were initially divided into 2 genera: Lymphocryptovirus and Rhadinovirus. As the number of viruses in Rhadinovirus became larger and more unwieldy, the genera Macavirus and Percavirus were introduced. 3 At that time, Rhadinovirus was left as a catch-all polyphyletic genus, as has been found in most analyses.7,18 Interestingly, cheirogaleid GHV-2 clusters more significantly with the rhadinoviruses of laurasiatherian host origin than with primate rhadinoviruses. The relationship of lorisid GHV-1 is not well resolved.
All samples that were positive in the consensus PCR for either cheirogaleid BHV-1 or cheirogaleid GHV-2 also tested positive in the qPCR assay. The cheirogaleid GHV-2 qPCR assay was positive in samples that were positive for cheirogaleid BHV-1, but cheirogaleid GHV-2 was not identified in the consensus PCR assay. The amount of gammaherpesvirus in these samples was extremely low (3 and 2 copies per reaction were the mean values), although detected in all replicates. The greater amount of betaherpesvirus present in the sample likely overwhelmed the amplification of the gammaherpesvirus in the consensus PCR. The consensus PCR assay used is also known to have a lower limit of detection for HBHV-5 and HBHV-6 than HGHV-4. 27 Interestingly, both betaherpesvirus-positive samples were also gammaherpesvirus positive, indicating a possible interaction of these viruses. The amplification was considered specific, given that the primer sets did not amplify the other virus.
Although providing definitive detection, PCR assay with sequence identification is a more laborious and slower test than qPCR; qPCR has significantly fewer labor-intensive steps and provides information regarding the agent’s presence and quantitative information. The qPCR protocols developed in our study have greater sensitivity than the coansensus PCR protocols used previously, identifying even low virus concentrations in mixed infections while also remaining specific to the target DNA in the presence of slightly different sequences. Our assays may therefore prove useful for the surveillance of cheirogaleid BHV-1 and cheirogaleid GHV-2 in prosimian populations.
Footnotes
Acknowledgements
We thank Drs. Keith E. Linder, Thomas Cecere, John Cullen, Stuart Hunter, Yumiko Kagawa, Pamela Luther, William MacKenzie, Jeremy Tobias, and Michelle Wells for their work with histopathology in these cases.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
