Herpesviruses in Reptiles

Since the 1970s, several species of herpesviruses have been identified and associated with significant diseases in reptiles. Earlier discoveries placed these viruses into different taxonomic groups on the basis of morphological and biological characteristics, while advancements in molecular methods have led to more recent descriptions of novel reptilian herpesviruses, as well as providing insight into the phylogenetic relationship of these viruses. Herpesvirus infections in reptiles are often characterised by non-pathognomonic signs including stomatitis, encephalitis, conjunctivitis, hepatitis and proliferative lesions. With the exception of fibropapillomatosis in marine turtles, the absence of specific clinical signs has fostered misdiagnosis and underreporting of the actual disease burden in reptilian populations and hampered potential investigations that could lead to the effective control of these diseases. In addition, complex life histories, sampling bias and poor monitoring systems have limited the assessment of the impact of herpesvirus infections in wild populations and captive collections. Here we review the current published knowledge of the taxonomy, pathogenesis, pathology and epidemiology of reptilian herpesviruses.


INTRODUCTION
Reptiles are a group of vertebrates (class Reptilia) that are adapted to a broad range of terrestrial and aquatic environments (1,2). The group comprises over 11,000 extant species placed in four orders, namely: Testudines (turtles, tortoises, and terrapins); Squamata (lizards, snakes, and worm lizards); Crocodilia (crocodiles, alligators, gharials, and caimans); and Rhynchocephalia (tuatara) (3). Reptiles constitute an integral part of the natural ecosystem and play the roles of both pollinators and predators, as well as environmental health indicators (4). In addition to their ecological services, reptiles have become desirable for food, medicinal products, pet trade, leather goods and research applications (5)(6)(7). However, their existence and well-being have constantly been threatened by several factors, such as hunting, environmental pollution, loss of habitat, destructive non-native species, climate change, and infectious diseases (4,(8)(9)(10). Disease surveillance and research in wild populations of reptiles are associated with numerous challenges including difficulties in accessing samples or field data, misleading epidemiological data and missing population data, as well as political and cultural restrictions (11). These challenges well explain the use of captive wildlife as models in many studies to acquire epidemiological information, since diseases are comparable in both wild and captive animals (12-15). Nonetheless, more robust and ideal epidemiological data are obtained when free ranging animals are surveyed. Recently, researchers have taken a renewed interest in reptilian viruses, partly due to the role played by reptiles as reservoir hosts for zoonotic viruses, as well as improvements in viral diagnostic methods that, in turn, have increased understanding of viruses in reptiles (16-20).
Herpesviruses (HVs) are members of the family Herpesviridae, a large taxon of DNA viruses that have been described in most vertebrate animals, including reptiles (16, 21). Herpesviruses are enveloped viruses with an icosahedral nucleocapsid and a linear double-stranded genome of varying length from ∼124 to 259 kbp (22). Generally, HVs replicate within host cell nuclei and are able to remain latent in their natural hosts (17, 23). So far, reptilian HVs that have been identified and characterised all belong to the subfamily Alphaherpesvirinae (17, 24-26).
The occurrence of HV infections among reptiles has been widely documented and associated with stomatitis, tumors, encephalitis, conjunctivitis, hepatitis and mortalities (27, 28). Unfortunately, current treatment options of reptilian HVs are limited and the search for potent vaccines remains a herculean task; therefore, the adoption of preventative strategies is still the most efficient way of controlling these diseases. This review aims to assist in biosecurity planning as well as create a knowledge platform for decision makers and researchers by providing an overview of the taxonomy, pathogenesis, pathology and epidemiology of reptilian HVs.

METHODS
Databases such as Medline (Ovid), PubMed, and Scopus were searched using specific keywords and phrases including Herpesviridae infections, herpesvirus infection, fibropapillomatosis, grey-patch disease, loggerhead genitalrespiratory herpesvirus, herpesvirus disease, reptiles, turtles, tortoise, snakes, lizards, alligators, and crocodiles (Supplementary File 1). To ensure that relevant publications were not missed, each subheading was searched independently on PubMed. Also, an additional literature search was conducted by assessing references of articles selected from previous databases. A summary of the search results is shown in Figure 1. Furthermore, we read the abstracts and full texts of the selected articles, extracted and analysed information on the diagnostic methods used and the reptilian HVs investigated from 1972 to September 8, 2020 (Figures 2, 3

; Supplementary File 2).
Non-English, non-original research, guidelines, and review articles were excluded from the analysis.

BIBLIOMETRICS
We conducted bibliometric analyses of published articles on the topic of reptilian HVs using Vosviewer software (29) and the Web of Science Core Collection database. A total of 245 publications were downloaded from Web of Science Core Collection database using the following search terms; herpesvirus, turtle, lizard, snake, tortoise and crocodile. The strategy involved a combined use of the keywords, tags and Boolean operators to create query sets as follows: ALL= (herpesvirus) AND ALL= (turtle * OR lizard * OR snake * OR tortoise * OR crocodile * ) with no limitations. USA had the highest number of research outputs with 149 (60.8%) articles. This was followed by Germany (28; 11.4%) and Australia FIGURE 1 | Summary of literature search conducted. A total of 1,026 articles were initially screened and 712 articles were later excluded due to duplication. A total of 314 articles were then reviewed for this study. One hundred and thirty studies were then extracted for quantitative analysis.
(26; 10.6%) (Figure 4; Table 1). Using Vosviewer, we visualised the major keywords commonly used in the field of reptilian HVs and the link strengths between collaborating countries (Figure 4; Supplementary File 3). Of note, a low number of records on reptilian HVs were observed for some countries (Indonesia, Mexico and India) that have rich reptile diversity ( Table 1) (3,30,31). Some of these countries also had little or no collaborations with the high research output countries (Figure 4), thus suggesting an under-reporting of reptilian HVs in these countries. Conversely, Germany has less reptile diversity with a higher number of records ( Table 1). This observation could be attributed to the presence of established diagnostic resources or increased monitoring and reporting systems for reptilian diseases in the country. 3 (TeHV-3). The species Chelonid alphaherpesvirus 6 (ChHV-6) belongs to the subfamily Alphaherpesvirinae with unknown generic placement. The Iguanid herpesvirus 2 (IgHV-2) that causes cytopathic infection in iguanids is of unknown generic and subfamilial placement (32)(33)(34)(35).

VIRION AND GENOME ORGANIZATION
All members of the family Herpesviridae share a common virion architecture, comprising a monopartite, linear, double stranded DNA core enclosed within an icosahedral capsid with a T = 16 symmetry (60)(61)(62). The capsid is tightly wrapped by a proteinaceous tegument, which, in turn, is surrounded by an envelope containing polyamines, lipids, and essential antigenic glycoproteins (62,63). Unlike reptilian HVs, the atomic structures of human alphaherpesviruses have mostly been described. For instance, a cryo-electron microscopy (Cryo-EM) resolved the atomic structure of human simplexviruses (HSV-1 and−2), which comprise capsid organisation (hexons, pentons and triplexes), capsid proteins (VP5, VP19C, VP23, and VP26) and tegument proteins (pUL17, pUL25, and pUL36) (64)(65)(66). Although the atomic structures of reptilian HVs have not specifically been resolved, the resolved structures of other alphaherpesviruses provide insights, since their genomes have many similarities (36,67). Consequently, the insights could guide future *Ranking was conducted based on the number of reptile species by countries identified from our bibliometric search and not based on the global ranking by Butler (31).
research in the atomic structure resolutions of reptilian HVs, which in turn could serve as a baseline for reptilian HV vaccinology.
Although the complete nucleotide sequences for most reptilian HVs are yet to be obtained, genomic features can be inferred from other fully sequenced alphaherpesviruses owing to sequence homology (36,67). All alphaherpesvirus genome structures contain unique long (U L ) and short (U S ) sequences and each are flanked by both terminal (TR L , TR S ) and internal (IR L , IR S ,) inverted repeat regions, giving the general configuration TR L -U L -IR L -IR S -U S -TR S (68). The complete genome of two TeHV-3 strains (1976 and 4295) was recently sequenced. The 1976 strain was shown to have a novel inverted repeat (TR T , IR T ) and unique (U T ) regions (69). The genome is approximately 160 kbp, encodes at least 107 open reading frames (ORFs) and consists of U L (107,928 bp) and U S (20,375 bp) regions. The U L is bound to its right by the U S adjoined to inverted repeats (IR S and TR S ; 8,536 bp) and to its left by a third unique region (U T ; 12,595 bp), which is also bordered by inverted repeats (TR T and IR T ; 1,194 bp) to give the overall layout TR T -U T -IR T -U L -IR S -U S -TR S (69). However, this differs from the type D configuration earlier attributed to this species (70). In another study, the complete nucleotide sequence of a Bacterial Artificial Chromosome (BAC) containing the entire genome of ChHV-5 (cloned in pTARBAC2.1) was obtained and it showed a different configuration (U L -IR S -U S -TR S ) from that of TeHV-3, even though they both belong to the genus Scutavirus (71). Moreover, the genome characterisation of strain 4295 identified regions containing genes that could be involved in viral pathogenesis or virulence (69). This is an important finding as these regions could serve as therapeutic or diagnostic targets in future research. Similarly, evidence of recombination among strains of ChHV-5 documented by Morrison et al. could lead to increased virulence and transmission of ChHV-5 variants (72) and these events may remain undetected in sea turtle populations. Therefore, it is pertinent to strengthen current diagnostic approaches to allow for more comprehensive geographical surveys and characterisation of HVs. Also, as new and affordable diagnostic techniques are being developed and improved upon, we expect more novel structures of reptilian HVs to be reported.

TRANSMISSION AND PATHOGENESIS
Several modes of reptilian HV transmission have been postulated including vertical, horizontal and mechanical transmissions (73)(74)(75). Marenzoni et al. reported the first evidence of vertical transmission of TeHV-3 in a captive breeding facility (14). In this study, one hatchling born in isolation from the egg laid by an infected tortoise (Testudo hermanni hermanni) presented with conjunctivitis and tested positive by specific polymerase chain reaction (PCR) targeting the partial sequence of the UL39 gene of . In other studies, Jones and colleagues provided molecular evidence for the horizontal transmission of ChHV-5 in green turtles by demonstrating the molecular link between viral variants and foraging grounds (76,77). Furthermore, experimental studies have revealed the possible transmission of reptilian HVs by direct contact with infectious FIGURE 5 | Midpoint rooted maximum likelihood phylogenetic tree of predicted amino acid sequences of HV-DNA-dependent DNA polymerase. The unassigned reptilian HVs are shown in red and cluster within the subfamily Alphaherpesvirinae. This analysis involved 54 amino acid sequences (37 complete and 17 partial sequences) downloaded from NCBI website (https://www.ncbi.nlm.nih.gov/) and aligned by ClustalW. There were a total of 1,443 positions in the final dataset. The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model. The tree with the highest log likelihood (−60139.09) is shown. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. Evolutionary analyses were conducted in MEGA X (59).
secretions (78)(79)(80) or indirectly via vectors and water (81)(82)(83)(84). By linking viral shedding patterns or frequency to disease occurrence, we could trace the most probable transmission mode of reptilian HVs. For instance, in captive enclosures, HVs could be easily transmitted via contact with secretions or contaminated materials, even at low shedding rates, and the removal of infected animals and infectious materials could stop the spread of the virus. In the wild, different transmission agents such as vectors, fomites and superspreaders could interplay to compensate for the low contact rate and infrequent shedding of some HVs (85). Because of the managerial implications, it is therefore important to further investigate the roles of these factors in the transmission of HVs in wild reptiles.
Four pathogenic mechanisms are highly conserved among all HVs and include: (1) intranuclear replication and capsid assembly; (2) expression of DNA metabolic and synthetic enzymes; (3) cell destruction following the release of viral progenies; and (4) the maintenance of latency in natural host cells (63,86,87). Generally, HV infections begin with viral entry, which is followed either by localisation or systemic spread (87). An experimental transmission study described the systemic dissemination of tortoise HVs (isolates HV 1976 and HV 4295/7R/95) (78). Following experimental infection via intranasal and intramuscular routes, the HVs elicited clinical signs and were detected by PCR in tissue samples from the respiratory, digestive and urogenital tracts, central nervous system (CNS), heart and spleen (78). Fibropapillomatosis (FP) and grey patch disease (GPD) are both associated with clinical signs that could be attributed to local destruction of infected cells due to replication and progeny release (54,80,(88)(89)(90). Evidentially, two studies demonstrated the local replication of ChHV-5 by detecting certain biomarkers (eosinophilic intranuclear inclusions, F-VP26, DNA, and mRNA transcripts) within fibropapillomatous lesions (85,91).
Unfortunately, the specific mechanisms involved in host cell invasion, immune evasion, localisation and spread of reptilian HVs have not been fully elucidated. However, recent molecular studies have provided insights into some virulence factors associated with reptilian HVs (69)(70)(71). Briefly, glycoproteins B (gB), gC, gD, gH, gK gL, gM, and gN have been hypothesised to function in host cell attachment and entry (69)(70)(71). gB and gC are capable of binding to heparan sulphate proteoglycans that are present on the surface of many cells, thus aiding viral adsorption and penetration into different cell types (69,92). The interactions of gB, gD, gH, gK, gL, gM, and gN have also been postulated to mediate membrane fusion and viral entry into the cell (69,70,92). Glycoproteins B, E, H, and L are involved in viral cell to cell spread, which could occur through intercellular bridges or intra-axonal transport, thus circumventing humoral immune responses (69,70). Similarly, the gC can bind to the third complement component (C3b) to block the alternative pathway complement activation (69,93). The gE and gI in HSV-1 inhibit the normal function of antibodies by building up a complex that acts as an Fc-receptor (94). However, the immunosuppression mechanism of the gE homologue in reptilian HVs is not yet clear. Finally, the F-M04 and F-sial proteins were recently identified in ChHV-5 and thought to play a role in FP pathogenesis; however, the specific mechanisms involved are not yet understood (71). Future research should consolidate characterisation of reptilian HVs in order to increase the understanding of host-pathogen interactions and improve clinical interventions.

CLINICAL AND PATHOLOGICAL SIGNS
Herpesvirus infections have been described in reptiles with a range of clinical manifestations (16). To provide an overview, the clinical signs and the gross and histological lesions associated with reptilian HVs are summarised in Table 2. Some of the more detailed descriptions are from sea turtles, tortoises and crocodiles. Grey patch disease and FP, characterised by coalescing greyish papular skin lesions (spreading patches) and branching papillary tumours (Figure 6), respectively, have been reported in sea turtles (42,54,108,109). Lung-eye-trachea disease (LETD) with a clinical course of 2-3 weeks has been seen in green sea turtles (98). Lung-eye-trachea diseased turtles often present with pneumonia, stridor and caseation of the eyes, oropharynx and trachea (98). In freshwater turtles, HV infections are associated with hepatic necrosis, and proliferative and/or ulcerative lesions of the skin and shell (Figure 7) (37,38,110,111). Infections in tortoises result in ulcerative to diphtheroidnecrotizing stomatitis, conjunctivitis, glossitis, rhinitis, dyspnoea, liver disease and neurological disease and could be accompanied by anorexia, regurgitation, neck oedema, lethargy and death (Figure 8) (112)(113)(114)(115). Papillomas, stomatitis, and hepatitis are commonly described in lizards infected with HV (25, 47-49, 116, 117). Recently, five green snakes (Opheodrys vernalis) housed together presented with oropharyngeal squamous cell carcinoma and molecular analysis confirmed the presence of a novel Opheodrys herpesvirus-1 (Alphaherpesvirinae) (52). In another study, a lymphoid follicular cloacal inflammation in juvenile alligators was initially associated with tortoise HV. However, the HV (Genbank accession AY913769.1) was later determined to be a likely laboratory contamination and the actual causative agent is still unknown (118). Similarly, Hyndman et al. identified three novel HVs associated with conjunctivitis and/or pharyngitis (CP), systemic lymphoid proliferation with non-suppurative encephalitis (SLPE), and lymphonodular skin lesions (LNS) in farmed saltwater crocodiles (Crocodylus porosus) and captive freshwater crocodiles (Crocodylus johnstoni) (53). Obviously, HVs can induce significant diseases in both captive and wild reptiles; therefore, there is a need to develop rapid diagnostic tests that will aid disease surveillance and reporting in order to maintain safe biosecurity measures and reduce spread.

EPIDEMIOLOGY
Epidemiological studies of HVs in wild reptiles could be challenging due to a lack of sensitive diagnostics for the detection of unknown or novel species, especially in resource-limited regions. HV infections are commonly characterised by nonspecific clinical signs in most reptiles, thus making diagnoses on the basis of clinical signs alone difficult. An exception to this is FP, in which the presence of cutaneous tumours gives an indication of the disease; hence, more FP-associated HV data have been reported in recent decades (Figure 2; Supplementary File 2). Even so, the complete disease impact on wild populations could be currently underestimated due to the underreporting of outbreaks, sampling bias and poor monitoring systems. For instance, there is a paucity of information for FP epidemiology at the pelagic phase of life in sea turtles as most studies are biassed towards sampling nearshore juveniles and adult females at foraging grounds or nesting beaches. Nevertheless, wildlife workers and researchers who, despite numerous challenges, have provided considerable epidemiological data targeted at conservation efforts towards endangered species should be commended. An overview of some of the epidemiological information including the prevalence and demography of both wild and captive reptilian HVs is discussed in this section.
Herpesviruses are linked to different diseases of marine turtles, including FP, LETD and GPD (54,82). FP is a debilitating disease characterised by the development of tumours (119, 120). Depending on the location of the tumours, FP can have detrimental effects (109,121). On the basis of prevalence and distribution, Tagliolatto et al. reported a prevalence rate of 43% for FP in green turtles captured in a foraging area in south-eastern Brazil (121). Adnyana et al. recorded 22% overall prevalence in green turtles in Indonesia and also observed that the prevalence rate of FP was higher among turtles from waters adjacent to densely populated regions compared to those collected from waters remote from urbanised regions of Indonesia (122). These findings indicate that the epidemiology of FP in marine turtles vary between geographical regions and may be linked to anthropogenic activity. This theory is supported by the findings in another study, which attributed the variation of FP prevalence to environmental cofactors that vary among local habitats (123). A study associated the geographical distribution of FP with the genomic variation of HVs in marine turtles, and observed four forms of the virus corresponding to Atlantic Ocean, west Pacific, mid-Pacific, and east Pacific (124). A similar study conducted in Australian waters identified different genotypes along the east coast of Queensland. Such differences in strains may also effectuate different levels of pathogenicity between strains (76,77) and account for variation in reported  prevalence in different regions. Also, given that the immune system of reptiles is dependent on temperature (125,126), the variation in the prevalence rates of reptilian HVs across regions could be associated with differences in regional climate types. Comprehensive reviews of the epidemiology of FP in marine turtles have been documented elsewhere (73,97,(127)(128)(129). The host immune status influences the clinical course of a disease, as immuno-deficient populations are more likely to succumb to disease outbreaks. Serosurveys have been conducted to determine the immune status of populations and to provide evidence of past and ongoing HV infections (130,131). Seroepidemiological studies in three localities in Florida revealed high anti-ChHV seroprevalences (up to 100%) in both FP and non-FP sea turtles (81,132). Contrastingly, seropositivity to ChHV-5 was dependent on the tumour status in turtles from Hawaii (133). This variation was mainly attributed to differences in the pathogenicity of ChHV-5 subtypes from the two regions (133). In another report, an epizootic of LETD in confined juvenile green sea turtles resulted in 8 to 38% mortality, thus posing significant conservation and management concerns (98,134). The LETD impact on free ranging sea turtles has not been investigated; however, seroprevalence rates of 13% and 22% were reported in two studies, respectively (134,135).
Similarly, HV infections are causing increasingly significant concerns in non-marine chelonians (27,39). Herpesviruses have been implicated as the cause of severe clinical signs and acute death in terrestrial and freshwater turtles ( Table 2) (28, 40,105,110,111,136). Although, HVs have been associated with latent infections in their natural hosts, infections in young, immunosuppressed or non-adapted hosts could result in the development of significant diseases (137). Therefore, monitoring the disease impact on both wild and captive endangered species has become pertinent. In an epidemiological study conducted in Tennessee and Illinois, USA, 128 of 409 free-ranging eastern box turtles (Terrapene Carolina Carolina) tested positive for TeHV-1 using TaqMan quantitative PCR, and the detection rate varied widely between seasons (138). Another study reported 48.3% prevalence of HV infections in endangered populations of bog (Glyptemys muhlenbergii), wood (G. insculpta), and spotted (Clemmys guttata) turtles in the northeastern United States (37). Furthermore, tortoise HVs have been associated with high mortality and morbidity (104,113,(139)(140)(141). Different HV species were identified to cause the death of a large number of pancake (Malacochersus tornieri), Horsfield (Testudo horsfieldii), Hermann's (Testudo hermanni), and Egyptian tortoises (Testudo kleinmanni) during spontaneous outbreaks in Japan, Italy and Germany, respectively (99,102,142). Species dependent susceptibility to HV was reported in a tortoise colony in which T. graeca and T. horsfieldii appeared to be unaffected by the HV species that caused the death of other tortoises in the same colony (113). A possible explanation could be that the causative HV species is well-adapted in these tortoises and they could be transmitting the virus to naïve or non-adapted tortoises. Of the four tortoise HV species (TeHV1-4), TeHV-3 appears to be the most pathogenic and frequently described, causing lethal disease in different tortoise species (14, 142-145). In a recent assessment of the incidence of chelonian HVs in Europe, more than half (54%) of all the detected chelonian viruses were TeHV-3 (146). Again, seroprevalence rates of 27% and 31% were reported for TeHV-1 and TeHV-3, respectively, in different populations of desert tortoises in California (147,148). Despite the significance of HV infections, we observed that the disease is still grossly under-studied in some countries (Table 1). Thus, insufficient data and underreporting have made it difficult to assess the geographical patterns of the HV epidemiology in non-marine chelonians and other reptiles.
Herpesviruses have also been described in various species of squamates and crocodilians (25,47,48,52,106,(116)(117)(118)149). A recent outbreak of a lethal HV infection in a private facility housing 127 snakes resulted in the death of all 71 horned vipers at the premises after a brief illness (150). An earlier study also implicated HV in the death of some boa constrictors within the first year of life (151). Herpesvirus-induced deaths have been reported in different species of lizards with case fatalities nearing 100% (33,49,152). As stated earlier, HV infections in crocodiles are associated with CP, SLPE and LNS syndromes (53). Another study strongly linked HV infection to SLPE and CP syndromes in farmed Australian saltwater crocodiles, with the highest prevalence rates of 94 and 54%, respectively (107). Crocodiles are intensively farmed for commercial purposes in Australia; therefore, the occurrence of HVs in crocodiles has both epidemiological and economic implications (53,153).
Finally, we extracted a total of 130 articles, of which 39% (51 articles) and 32% (41 articles) were studies that investigated HVs in marine turtles and tortoises, respectively. A total of 21 (16%) studies investigated HVs in freshwater turtles. HVs were least studied in lizards (8%; 11 articles), snakes (3%; 4 articles) and crocodiles (2%; 3 articles) (Figure 2). The scant studies of HVs in some reptilian species since the 1970s could be attributed to the unavailability of reagents or sensitive diagnostic assays required to investigate reptilian diseases in remote areas or the lack of interest to investigate HVs in reptiles because of their relatively low socio-economic importance. Therefore, future efforts should be directed towards enhancing collaborations among government agencies, researchers and wildlife workers with a view to creating awareness, increasing access to reagents and sensitive assays, and ultimately conserving endangered reptiles.

DIAGNOSIS
A timeline of reptilian HV diagnosis showed that traditional assays including histopathology, virus isolation (VI) and electron microscopy (EM) have been the mainstays in the diagnosis of reptilian HVs (Supplementary File 4). Many studies have reported the use of these techniques since the 1970s for the investigation of reptilian HVs. A breakdown of the number of studies that have used these methods to detect reptilian HVs is shown in Figure 3. Molecular diagnosis of reptilian HVs started two decades ago and has been used increasingly since then (Figure 3; Supplementary File 4). The advent of molecular diagnostic techniques has provided insight into the genetic characteristics and the phylogenetic relationship of most reptilian HVs. This section highlights some important characteristics of the various techniques used in the diagnosis of reptilian HVs.
Diagnosis of reptilian HVs is tentatively made on the basis of patient history, clinical signs, and gross and histological lesions (16). However, this is not always the case, as host-adapted HVs can cause subclinical, mild or latent infections in their natural hosts, and the demonstration of intranuclear inclusions is not pathognomonic of reptilian HV infections (154). Intranuclear inclusions are frequently associated with other reptilian viruses including adenoviruses and papillomaviruses (155)(156)(157). Earlier researchers used EM to confirm the presence of reptilian HV infections by demonstrating the ultrastructure of the viral particles in fixed, cut and stained sections of tissue samples (51,54,98,106,116,136,140,158). More recently, EM has been used to confirm a necrotic hepatitis associated with HV infection in a tortoise with no clinical signs or lesions in the respiratory tract, oral cavity or other organs (112). The need for high technical capacities and the high cost of electron microscopes limit the use of EM for epidemiological and diagnostic purposes especially in resource-limited areas. Despite these limitations, EM remains a powerful detection tool in most high-class virology laboratories.
Reptilian HVs have been isolated in cell culture and identified on the basis of their cytopathic effects (50,98,107,159). For instance, tortoise HVs were isolated from pharyngeal swabs, trachea, kidney, oesophagus, tongue, stomach, and intestine, and caused cytolysis and rounding of cells in terrapene heart cells (TH-1) (160). In another study, detachment and foci of enlarged, rounded, refractile cells were produced following inoculation of tissue and swab supernatants in turtle heart cells (142). ChHV-5, which historically has been resistant to replication in conventional cultures, produced de novo ballooning degeneration and eosinophilic intranuclear inclusion in plugs and organotypic skin cultures (89). This observation implies that ChHV-5 remains latent in conventional cultures and requires replication of the turtle skin to grow in vitro (89). Aside from the fact that CPE are not obtained for non-cytopathic viruses, cell culture is susceptible to both chemical and biological contaminations, which in turn affect its sensitivity and specificity. Also, diagnostic turnaround could be delayed for slow-growing viruses. Therefore, it should not be solely relied upon for the epidemiological investigations of HVs.
Following primary infections in reptiles, a strong nonspecific (innate) immune response that includes lysozymes, leukocytes, natural antibodies (NAbs), antimicrobial peptides, and the complement pathway, is quickly stimulated (126,161). No specific information is currently documented about adaptive cell mediated immunity to HV infections. Unlike mammals, in reptiles a less robust and slower humoral response (IgA, IgD, IgM, and IgY) is stimulated after the innate immune system is activated (126,154). In tortoises, neutralising antibodies to HV infection were detectable in serum at least 4 weeks postexposure (162). These serum neutralising antibodies did not appear to confer immunity to reinfection or recrudescence (78). Later seroconversion was observed (four months to one year) in green turtles (Chelonia mydas) that were experimentally infected with ChHV (81,132). Generally, the detection of anti-herpesvirus antibodies in a single sample could indicate previous or latent infection, while rising antibody titre in paired samples collected at least 6 weeks apart indicates active infection (154,163). Humoral antibodies are detected by serological assays such as serum neutralisation (SN) tests, ELISA, and immunoperoxidase (IP) assays (132,(164)(165)(166). The SN test is considered the reference test for anti-herpesvirus antibody detection but has limitations such as a delayed turnaround, inherent assay arduity and the requirement for standard isolates (162). ELISAs with high sensitivity and specificity have been developed and deployed in various seroepidemiological studies (44,81,132,147,162,166). However, a high degree of cross-reactivity that potentially affects assay specificity has been demonstrated among different tortoise HV isolates used as antigens in the ELISA (147,162). Cross-reactivity could also occur in other reptilian HVs that share similar antigenic epitopes, giving false positive results and, thus, leading to unnecessary post-exposure interventions. Overall, serological diagnostic techniques are not useful for the early diagnosis of reptilian HVs because of the delay in antibody response and the need for paired serum sample collection weeks apart with accurate timing. However, it can play an important role in retrospective studies and in the diagnosis of latent or asymptomatic patients.
Recent epidemiological studies have largely relied on molecular methods to identify potential genetic and environmental risk factors associated with reptilian HVs (24, 80,138,[167][168][169]. Species-specific PCR-based assays targeting specific gene segments of reptilian HVs have been developed and validated (78,170). Lindeman et al. developed two quantitative PCR assays and recorded a detection limit as low as 1 viral copy per reaction using primers that targeted the EBHV-1 specific segment of DNA polymerase gene (U L 30) (39). In another study, two TaqMan PCR assays developed to target the U L 30 gene of TerHV-1 detected 10 viral copies per reaction (171). Conventional and heminested PCR assays using tortoise HV-specific primers have been developed with assay sensitivity of 10 3 and 10 1 DNA copies, respectively (172). Alternatively, consensus PCR techniques developed by VanDevanter et al. have been employed for the molecular screening and novel detection of reptilian HV species (39,43,100,(173)(174)(175)(176)(177). Although the molecular assays for the diagnosis of reptilian HVs have demonstrated excellent performance, their use still presents a major challenge in remote areas due to high cost, complexity of instrumentation, aseptic technique requirement and the need for electricity to operate PCR machines.
In order to accurately estimate the magnitude and scope of a disease outbreak or occurrence, case definition (that is, standard criteria for categorising diseases) would need to be established. One of the ways to achieve this is to make available rapid, sensitive and affordable assays for confirming the presence of diseases. Rapid diagnostic immunoassays that use lateral flow or chromatographic strategies should be developed for the rapid diagnosis of reptilian HV infections in the field or point of care (POC) settings. This approach could overcome some of the above-mentioned diagnostic challenges, especially in low resource areas. However, the use of lateral flow immunoassays for viral detection in other species have been marred by low and varying sensitivities (178)(179)(180)(181). Sensitive molecular-based rapid assays are relatively expensive and yet to be employed for the diagnosis of reptilian HVs (182)(183)(184)(185). We would propose an ultrasensitive format that combines PCR and immunoassay but then it can be argued that such a laboratory-based system is less rapid and has limited use in low-class laboratories (186,187). Rapid detection techniques such as Microfluidic chip immunoassay and Smartphone-based rapid telemonitoring system (SBRTS) are fast becoming powerful tools in the diagnosis of viral infections (188)(189)(190)(191)(192)(193)(194)(195)(196). Of particular interest, is the SBRTS that combines biosensor and smartphone functionalities to produce a rapid, sensitive and cheap detection system (197). SBRTS has an average turnaround of 30 min, overcomes inherent problems associated with sample handling and preparation, and can remotely monitor and report data on disease occurrence, thus making it suitable for use in resource-limited countries (193,197). This assay if employed could tick all the boxes for the epidemiological investigation and reporting of reptilian HVs.
Herpesvirus diagnostic and epidemiological data should be interpreted with prudence because of the possible influence of coinfection variables that could cause the reactivation of seemingly latent HV infections. For instance, some studies have reported the detection of co-pathogens in reptiles showing clinical signs, some of which are typical of HV infections (27,38,146,168,173,198,199). These observations imply that the detection of HVs may not be the actual cause of the current disease, but because the immune system is compromised by other pathogens, the HVs recrudesce and become easier to detect. Both latency (decreases apparent prevalence and significance) and coinfections (increase apparent prevalence and may also falsely assign the clinical signs to the HV) will have an influence on the disease picture. Therefore, we recommend that biosecurity and conservation measures should include a multiplex pathogen detection model whenever possible in order to fully assess the health of reptilian populations.

TREATMENT, PREVENTION, AND CONTROL
Surgical excision, carbon dioxide (CO 2 ) laser surgery and cryosurgery are some of the commonly used therapeutic strategies for the management of HV-associated tumours (25, 110,111,120,[200][201][202]. High rates of recurrence and the risk of secondary bacterial infections have greatly reduced the efficacy of surgical excision (200,203). CO 2 laser surgery, which combines laser excision and ablation of tumours, has shown improved intraoperative and postoperative outcomes and is therefore the treatment of choice (120,201,204). Nonsurgical approaches including electrochemotherapy (ECT) and photodynamic therapy (PDT) with no known recurrence have recently been employed as alternatives in the treatment of FP (205,206).
Several authors have recommended the use of acyclovir complemented by fluid and antibiotic therapies for the effective treatment of tortoise HV infection (143, [207][208][209]. Marschang et al. showed that acyclovir and ganciclovir effectively inhibited HV replication in vitro at a single dose or repeated daily dose of 25 or 50 µg/mL (142). Similarly, the in vitro activities of acyclovir and ganciclovir were recently tested and shown to be effective against TeHV-3; however, the safety of these drugs is yet to be demonstrated in tortoises (210). Based on the toxicity (on liver and kidney cells) and other biochemical data, this same study showed that eprociclovir is not suitable for use as anti-TeHV-3 in Hermann's tortoises and further in vivo assessment of other potential antiviral drugs was recommended (210).
Recently, an autogenous vaccine therapy was proposed and used for the treatment of HV-associated papillomatosis in Williams' mud turtle (Pelusios williamsi) (111). The autogenous vaccine, which was aseptically prepared from excised fresh tissue induced substantial areas of necrosis of the papillomatous lesions, thus indicating the efficacy of the vaccine (111). Autogenous vaccines potentially contain relevant neoantigens that comparatively improve their efficacy (211). However, their use could be limited by lack of sufficient tumours (in patients) needed to produce adequate vaccine doses. Also, no standard protocol exists for autogenous vaccine production and delivery, and patients' tumours may progress beyond the intervention stage before the vaccine becomes ready for delivery. Allogeneic vaccines on the other hand, can overcome some of the aforementioned challenges; however, they may lack the advantageous self-neoantigens (211). In the past, an inactivated vaccine was evaluated against tortoise HV without success as no significant rise in antibody was detected in vaccinated tortoises after 369 days post vaccination (160). DNA or mRNA based vaccines have the capacity to induce both humoral and cellular immune responses and have shown promising outcomes against some animal and human diseases (212)(213)(214)(215). Although vaccine research and development could be costly, laborious and timeconsuming, the nucleic acid vaccines hold the potential to significantly reduce HV-associated losses in captive collections and wild reptiles of conservation concerns.
Prevention is of utmost importance in the management of reptilian HV infections, since death may still occur following therapeutic interventions and recovered animals remain latent carriers (143, 163). Unfortunately, there are no established preventive or control measures for HV infections in wild populations of reptiles (200), which consequently presents a major conservation challenge. Environmental factors including degraded water quality caused by pollutants, increased water temperature, natural biotoxins, and high dietary arginine concentrations due to microalgae bloom have arguably been linked as cofactors in the development of FP in sea turtles (73,119,(216)(217)(218)(219)(220)(221). Therefore, adopting conservation actions needed to regulate water and species management, as well as regulating human activities leading to climate change, would be sensible.
In captive reptiles, quarantine procedures and adequate testing of new acquisitions are strongly recommended (153,163,167). All previously infected or HV seropositive animals should be treated as latent carriers and potential shedders to naïve populations, as factors including stress, bad husbandry, illness or immunosuppression could reactivate the virus (14, 163,167,200). Generally, strict hygiene practises and adequate biosecurity should be followed in all facilities housing reptiles (162,172,(222)(223)(224).

AUTHOR CONTRIBUTIONS
GO and EA conceived and designed this review. GO wrote the manuscript, analysed the data, and prepared figures and tables. EA, PH, and DW contributed to the concept and reviewed drafts of the manuscript. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
The authors would like to especially thank Professor Elizabeth Tynan for providing editorial support to this review. Authors' special thanks also goes to Dr. Fredric Frye for providing us with the full text of some articles used in this review. Finally, we would like to thank James Cook University librarians for their technical support.