Identification of Equid herpesvirus 2 in tissue-engineered equine tendon

Abstract

Background: Incidental findings of virus-like particles were identified following electron microscopy of tissue-engineered tendon constructs (TETC) derived from equine tenocytes. We set out to determine the nature of these particles, as there are few studies which identify virus in tendons per se, and their presence could have implications for tissue-engineering using allogenic grafts. Methods: Virus particles were identified in electron microscopy of TETCs. Virion morphology was used to initially hypothesise the virus identity. Next generation sequencing was implemented to identify the virus. A pan herpesvirus PCR was used to validate the RNASeq findings using an independent platform. Histological analysis and biochemical analysis was undertaken on the TETCs. Results: Morphological features suggested the virus to be either a retrovirus or herpesvirus. Subsequent next generation sequencing mapped reads to Equid herpesvirus 2 (EHV2). Histological examination and biochemical testing for collagen content revealed no significant differences between virally affected TETCs and non-affected TETCs. An independent set of equine superficial digital flexor tendon tissue (n=10) examined using designed primers for specific EHV2 contigs identified at sequencing were negative. These data suggest that EHV is resident in some equine tendon. Conclusions: EHV2 was demonstrated in equine tenocytes for the first time; likely from in vivo infection. The presence of EHV2 could have implications to both tissue-engineering and tendinopathy.

Keywords: Equine herpesvirus 2; equine; next-generation sequencing; superficial digital flexor tendon; tissue-engineered tendon.

Figure 1.

A and B: Electron micrographs of virus-like particles found within TETCs. Tenocyte (T), extracellular matrix viral particles (black arrows). C and D: Electron micrographs showing full capsids, nucleocapsids (NC) and empty capsids (white arrows) within both of the TETCs from which virus were isolated (C; Y1, D; Y6). Scale bars are shown.

Figure 2.. Histograms of histological scoring of TETCs.

Histological scoring of TETCs from virally affected (n=2) and normal (n=8) donors. Graphs A– F represent the mean scores + standard deviation of the following characteristics; mean extracellular matrix organisation ( A), cell shape ( B), cellular distribution ( C), cellular alignment ( D), TETC cellularity ( E) and mean total score ( F). Where error bars are not present, scores for all donors were equal. Further details of the scoring system are available in Supplementary File 2. No significant differences were found between virally affected and normal donors (p≤0.05).

Figure 3.. Histogram of collagen content of young TETCs containing virus (V; n=2) and not containing virus (NV; n=5).

Graphs represent mean± standard error mean of percentage collagen normalised to dry weight. No significant difference was found in collagen content (p≤0.05).

Figure 4.

A and B. Log2-Coverage plot demonstrating read mapping. A. Read mapping plot of NV (O3). B. Read mapping plot of V (Y6). Both samples were mapped against the Equid herpesvirus 2 strain G9/92 (KM924294) genomes. Y-axes; coverage is log2-scaled. Zero coverage bases were assigned a log2-coverage value of -3 for plotting purposes.

Figure 5.. Pan herpesvirus gel image.

Gel image of PCR fragments following restriction by ECOR 1. Virus-infected samples Y1, Y6 and virus negative sample O3 are shown. Amplicon size is 229 bp. Bands were removed and subsequently sequenced to confirm identification of herpes virus.

Figure 6.. EHV2 PCR assay in an additional cohort of SDFT samples.

DNA extractions from ten equine SDFT samples (1–10) were amplified with primers (V1) designed within a EHV-2 contig identified following NGS. Genomic DNA from EHV-2 was used as a PCR positive control (EHV) and water as a negative control (bl). TrackIT 1Kb Plus DNA ladder was used as a marker (1kb ladder). The positive EHV2 control demonstrates a band at 450bp.

Figure 7.. Neighbour-joining trees.

Trees characterised the relationship between virus isolated from V (EHV2 RJW248419) and previously isolated strains using glycoprotein B gene. Bootstrap analysis (1,000 replicates) was used to provide support for individual nodes.