Attenuation mechanisms of the P7-P8 live-attenuated cyvirus cyprinidallo2 vaccine potentially involve apoptosis non-inhibition feature: insights into virus pathogenesis

TL;DR

This study explores how a live-attenuated CyHV-2 vaccine induces apoptosis and immune responses in goldfish, potentially reducing virus spread without causing tissue damage.

Contribution

The study identifies apoptosis-related mechanisms in a CyHV-2 vaccine strain, offering insights into vaccine attenuation and virus pathogenesis.

Findings

The P7-P8 vaccine strain triggers apoptosis and immune responses in goldfish without causing tissue damage.

Apoptosis-related genes are upregulated in vaccinated fish, potentially aiding virus clearance.

The vaccine strain shows early-stage apoptosis induction, limiting virus replication.

Abstract

Cyvirus cyprinidallo2 (CyHV-2) is an alloherpesvirus and the causative agent of herpesviral haematopoietic necrosis in goldfish. Whole-genome sequence comparison of the developed live-attenuated vaccine P7-P8 with virulent CyHV-2 strains revealed seven single-nucleotide polymorphisms, five deletions and one inversion in the ORFs, which may be involved in attenuation. A start codon loss in ORF 113, a putative apoptosis-inhibition gene, was observed in the mutations. In vitro assays indicated that apoptosis-related genes were upregulated in cells inoculated with the vaccine or virulent virus compared to uninfected cells. However, the vaccine group showed increased phosphatidylserine externalization and DNA damage, suggesting the apoptosis-inducing properties of P7-P8. In the in vivo experiment, histopathology demonstrated that vaccinated goldfish exhibited immune responses, such as leucocyte aggregation and melanomacrophage centre formation, without marked degeneration. Gene expression analysis showed upregulation of proinflammatory and granzyme B genes in vaccinated fish. In addition, the vaccine strain triggered apoptosis of the infected cells during the early stage of infection, potentially promoting virus clearance and preventing excessive virus replication. The results show that the P7-P8 potentially induces apoptosis and immune responses, contributing to low virus propagation without tissue damage. This study provides insights into CyHV-2 pathogenesis, suggesting that apoptosis-related genes can be the targets for vaccine development against alloherpesviruses in aquaculture species.

Figures

Output of the SnpEff analysis and mutations in ORFs using reference-based (CyHV-2 ST-J1) assembly. The ORF mutations were listed out (vertical marks) relative to the virulent challenge test virus CyHV-2 SaT-1 (a). Bioinformatics analysis and mutation validation. The targeted ORF 25D PCR product (b and c) of vaccinated (lane 1) fish and virulent virus-infected (lanes 2–6). The start codon lost in ORF 113 was confirmed by PCR and sequencing (d). Clustering of reference-based assembled samples. The Roary matrix generated by aligning core genes of sample-annotated protein sequences (e). The SNP distance of core genes generated by Roary (f). The sequence inversion between ORF 24 and ORF 25C detected by de novo assembly was confirmed by PCR and sequencing (g and h). Primers ORF 24 and ORF 25 (lanes 1, 2, 5 and 6) show the original genomic orientation, while ORF 24-Inversion and ORF 25-Inversion (lanes 3, 4, 7 and 8) identify inverted sequence. Red and yellow boxes indicate the position of the expected PCR product size.

Propagation characteristics of the P7-P8 vaccine- and virulent CyHV-2 SaT-1-inoculated cells. The RyuF-2 cell inoculated with the virulent virus (a) and the vaccine (b) showed typical CPE at 4 dpi. The virulent virus sample showed syncytium formation (black arrowhead), but this was not found in the vaccine group. The virus infectious titre of both samples is showing active propagation (c). Comparison of the growth kinetics of P7-P8 and CyHV-2 C2P1 at MOI=1 (e). The CyHV-2 C2P1 (d) showed syncytium formation (black arrowhead). The virus supernatant and sonicated cell pellet were subjected to virus DNA quantification (e). The plaque area of CyHV-2 C2P1 and P7-P8 showed similar sizes (f). Uninfected control, CyHV-2 SaT-1- and P7-P8 vaccine-inoculated cells were tested for apoptosis-related gene expression levels (g) and TUNEL and ANNEXIN V assay (h). Data show mean±se. The significant differences (P<0.05) are denoted with letters or an asterisk.

Histopathological changes in the trunk-kidney and spleen of vaccinated and CyHV-2-infected fish. In the trunk-kidney, no visible histopathological changes were noted at 1 dpi for both groups (a and b). The virus load was 103.53 virus DNA copies mg−1 and 105.06 virus DNA copies mg−1 (h) for the vaccinated and virus-infected groups, respectively. At 3 dpi, vaccinated fish (105.01 virus DNA copies mg−1) did not show signs of necrosis but showed some leucocyte aggregation (white arrowhead) (c). The virus-infected fish (108.41 virus DNA copies mg−1) showed marginated chromatin (black arrowhead) and no visible leucocyte infiltration (d). At 5 dpi, vaccinated fish (105.96 virus DNA copies mg−1) showed increased leucocyte infiltration and melanomacrophage centre formation (white arrowhead) (e). The virus-infected fish (109.66 virus DNA copies mg−1) showed severe necrosis manifested by karyopyknosis and karyorrhexis (black arrowhead) (f). At 7 dpi, the vaccinated fish (105.45 virus DNA copies mg−1) showed leucocyte infiltration and melanomacrophage centre formation (white arrowhead) (g). In the spleen, no visible histopathological changes were noted at 1 dpi for both groups (i and j). The load was 104.51 virus DNA copies mg−1 and 104.61 virus DNA copies mg−1 (p) for the vaccinated and virus-infected groups, respectively. At 3 dpi, vaccinated fish (104.37 virus DNA copies mg−1) did not show signs of necrosis but showed some enlarged ellipsoids (white arrowhead) (k). The virus-infected fish (107.99 virus DNA copies mg−1) showed marginated chromatin (black arrowhead) and melanomacrophage centre formation (l). At 5 dpi, vaccinated fish (105.65 virus DNA copies mg−1) showed more enlarged ellipsoids and melano-macrophage centre formation (white arrowhead) (m). The virus-infected fish (108.83 virus DNA copies mg−1) showed severe necrosis manifested by karyopyknosis and karyorrhexis (black arrowhead) and more melanomacrophage centre formation (white arrowhead) (n). At 7 dpi, the vaccinated fish (105.85 virus DNA copies mg−1) showed enlargement of ellipsoids (white arrowhead) (o). The horizontal line shows the detection limit (102 DNA copies mg−1) (h and p). Haematoxylin and eosin stain. Bar=50 µm.

Gene expression analysis of vaccinated fish. The trunk-kidney, spleen and gills were sampled. IL-1β, TNFα and granzyme-B gene expression relative to the uninfected control (a). Protective effect of the P7-P8 vaccine after virulent CyHV-2 infection (b). Goldfish (n=20) were vaccinated using the showering vaccination method. At 28 dpi, goldfish were infected with the virulent CyHV-2 SaT-1 by immersion. The mortality was recorded up to 21 days post-challenge. The vaccination group exhibited statistical differences (P<0.05) compared with the infected control group.

Detection of TUNEL-positive signals in the trunk-kidney and spleen of vaccinated and CyHV-2-infected fish. Green signals (white arrow) indicate apoptotic nuclei, while the blue signals represent Hoechst 33342-stained nuclei. The TUNEL-positive cell percentage was estimated using ImageJ. Significant differences (P<0.05) at each sampling time point are denoted with different letters. Bar=100 µm.

Tables

Table 1.. The list of mutations found in CyHV-2 C2P1, passaged attenuated viruses and the P7-P8 vaccine relative to CyHV-2 SaT-1. The coordinates, type of mutations and functions of coded proteins found in all samples are listed. Samples with the mutations are marked with a tick.

CyHV-2 ST-J1 (accession no.: NC_019495.1)											Samples with mutation
Coordinate	ORF	ORF comment (location)	Putative function	Anno-tation function	Trans-cription time (IE/E/L)	SnpEff: type of mutation	SnpEff: impact	Refe-rence seque-nce	Mutation	Amino acid variation*	CyHV-2SaT-1	CyHV-2 C2P1	CFS-P4	CFS-P10	CFS-P20	P7-P4	P7-P6	P7-P8
7399	6_1	Protein (in repeat region)	–	ProteinORF6	E	Upstream gene variant	Modifier	T	C	–			✓	✓	✓	✓	✓	✓
9381–9393	7_1	Protein	AJAP1/PANP C-terminus family	ProteinORF7	L	Conservative in-frame deletion	Moderate	GGCTGC-TGCTCTC	G	E249_S256del			✓	✓	✓	✓	✓	✓
10465–10490	6_1	Protein(in repeat region)	–	ProteinORF6	E	Upstream gene variant	Modifier	CGGGAGT-ACGCTAT-TGAGCAA-CACCG	C	–			✓	✓	✓
15115	7_1	Protein(in repeat region)	AJAP1/PANP C-terminus family	ProteinORF7	L	Upstream gene variant	Modifier	C	CA	–					✓	✓	✓	✓
19448	10	Protein	Transcription factor IIA, alpha/beta subunit; condensin II complex subunit CAP-H2 or CNDH2, N-terminal	Telomere-like	E	Missense variant	Moderate	C	T	P934S				✓	✓
25163–25164	20	Protein (non-coding region)	NAD(P)H-binding	–	L	Upstream gene variant	Modifier	TA	T	–			✓	✓	✓	✓	✓	✓
37699–38056	25D	Membrane protein	Orf78 (ac78)	Type 1 membrane protein;ORF25family	IE	Conservative in-frame deletion	Moderate	357 bp	A	T551_T66-9del, V670I			✓	✓	✓	✓	✓	✓
46744–46759	28A	Protein (non-coding region)	–	–	–	Downstream gene variant	Modifier	TTTT-TTCGT-TGT-ATAG	T	–		✓	✓	✓	✓	✓	✓	✓
49059–49060	32	Protein (non-coding region)	–	Containsa signalpeptide;similar to a family of Singapore grouper iridovirus proteins	–	Downstream gene variant	Modifier	GA	G	–		✓	✓	✓	✓	✓	✓	✓
92751	56	Protein (non-coding region: regulatory region)	Keratin, high sulphur B2 protein	–	E	Upstream gene variant	Modifier	T	TA	–			✓	✓	✓	✓	✓	✓
116057	67	Protein (non-coding region)	–	–	L	Start lost and disruptive in-frame insertion	High	C	CATT	–					✓	✓	✓	✓
117308	68	Protein	–	Similarto myosinand related proteins	E	Conservative in-frame insertion	Moderate	A	AGCTGTTGCT	Q1614_L16-15insQQQ				✓	✓
141837	81	Membrane protein (non-coding region)	–	Typethree membrane protein	–	Downstream gene variant	Modifier	T	TGA	–		✓	✓	✓	✓	✓	✓	✓
148257	87	Protein	–	–	–	Missense variant	Moderate	T	G	S1107R								✓
162482	90	Protein Allo37 (non-coding region)	–	–		Upstream gene variant	Modifier	C	CA	–			✓	✓	✓	✓	✓	✓
177000–177002	101	Protein (non-coding region)	–	–	L	Upstream gene variant	Modifier	GTC	G	–					✓	✓	✓	✓
190538	108	Protein	–	Containsa potential transme-mbrane domain	–	Missense variant	Moderate	C	T	T39I		✓	✓	✓	✓	✓	✓	✓
194699–194704	113	Protein	Apoptosis inhibitory protein (API5)	–	–	Start lost and conservative in-frame insertion	High	AGC-CAT	A	p.?			✓	✓	✓	✓	✓	✓
198318	115	Membrane protein	–	Typeone membrane protein	L	Missense variant	Moderate	G	A	D506N			✓	✓	✓	✓	✓	✓
207095	127	Protein	–	Containa signal peptide	–	Missense variant	Moderate	A	G	H80R				✓	✓	✓	✓	✓
216181–216182	132	Membrane protein	Membrane-associated protein117 kDa, PDZK1-interacting protein 1; TMEM154 protein family;Wnt and FGF inhibitory regulator	Type 1 membrane protein	IE	Frameshift variant	High	CG	C	V90Yfs			✓	✓	✓	✓	✓	✓
282653–282679	6_2	Protein (non-coding region)	–	Protein ORF6	E	Upstream genevariant	Modifier	GTCT-TAAT-GGTT-ACTC-TTAAT-GGTTAC	G	–			✓	✓
282670	6_2	Protein (non-coding region)	–	Protein ORF6	E	Upstream genevariant	Modifier	T	C	–					✓	✓	✓	✓
284652–284664	7_2	Protein	AJAP1/PANP C-terminus family	Protein ORF7	L	Conservative in-frame deletion	Moderate	GGC-TGCT-GCT-CTC-	G	S250_E2-53del			✓	✓	✓	✓	✓	✓

Full text

Data Summary

The data have been deposited with links to BioProject accession number PRJDB19095 in the DDBJ BioProject database. The sequenced genomes have been deposited in the DDBJ BioProject database under accession numbers DRR614361–DRR614368 and LC917605–LC917612. The authors confirm that all supporting data have been provided within the article or through supplementary data files.

Introduction

The goldfish (Carassius auratus) is a valuable ornamental species, highly prized for its appearance and shape. However, cyvirus cyprinidallo2 (CyHV-2), previously known as cyprinid herpesvirus 2, is the causative agent of herpesviral haematopoietic necrosis (HVHN), which infects goldfish [1]. This disease was first reported in Japan and was subsequently reported globally, infecting gibel carp (C. auratus gibelio) and crucian carp (C. carassius) [2]. Symptoms of HVHN include pale gills, abdominal distension and lethargy. Infection typically occurs at rearing temperatures of 15–25 °C, with mortality rates of up to 100% [1].

Herpesviruses, including CyHV-2, are part of the Alloherpesviridae family within the order Herpesvirales. This double-stranded DNA virus has an icosahedral nucleocapsid surrounded by a lipid bilayer envelope. CyHV-2 shares genetic similarities with CyHV-1 and CyHV-3 and has a genome size of ~290 kbp, featuring ORFs that encode proteins for virus replication and virulence factors [34]. Alloherpesviruses share similarities with other herpesviruses, such as establishing latency [57], and previous research showed CyHV-2 can inhibit apoptosis [89]. In addition, unannotated proteins of CyHV-2 may play a role in its virulence and create an opportunity for further discoveries.

Virulence factors can be elucidated by examining the virulence phenotype, characterized by specific nonsynonymous protein changes that alter virulence, and by assessing pathogenicity through comparative analysis of the propagation characteristics of virulent and avirulent viruses [10]. Live-attenuated vaccine strains replicate in the host by mimicking natural infection without inducing severe pathogenesis, thereby eliciting a robust immune response against subsequent infections. Consequently, a highly effective live-attenuated vaccine serves as a valuable tool for investigating changes in virulence factors. We developed a live-attenuated vaccine by serially propagating a virulent strain of CyHV-2 in two cell lines derived from non-host species, the CFS cell line for seven passages, followed by eight passages in the KF-1 cell line, producing the P7-P8 vaccine strain [11]. This vaccine demonstrated high efficacy in protecting various goldfish strains against CyHV-2 field isolates [12]. Although the vaccine established infection in the same organs as the virulent virus, it exhibited plateaued growth at a low level of virus load, resulting in no apparent abnormality and eventually cleared from the host by 21 days post-vaccination [1113]. In contrast, fish infected with virulent CyHV-2 exhibited rapid virus replication in haematopoietic tissues, leading to severe necrosis and mortality [1114]. The mechanisms underlying the plateaued growth of the vaccine as opposed to the exponential replication of the virulent virus must be clarified. Host responses within the early time points after exposure to the vaccine or virulent virus most likely determine these diverging outcomes. Therefore, understanding the early host responses against the vaccine strain and virulent virus strain will be crucial for explaining the virus pathogenesis and distinct outcomes of avirulent versus virulent virus infection.

During attenuation, this vaccine strain became non-pathogenic, likely due to genetic adaptations to non-host cell lines. Although the complete genome sequences of virulent CyHV-2 strains are known [315], a whole-genome sequence of live-attenuated CyHV-2 vaccine is lacking. Identifying the accumulated genetic changes of the P7-P8 vaccine could help list important genes involved in CyHV-2 pathogenesis. Hence, the objective of this study was to investigate the attenuation mechanism of the vaccine, providing insights into CyHV-2 pathogenesis. To achieve this, we conducted whole-genome sequencing of the virulent CyHV-2 viruses, the virus cultures in the attenuation process (i.e. passaged attenuated viruses) and the P7-P8 vaccine strain. We then identified the specific mutations found in the passaged attenuated viruses and the vaccine through variant calling. Finally, we studied the propagation characteristics of the vaccine and virulent viruses to understand CyHV-2 pathogenesis and the mechanisms of protective immunity induced by the vaccine strain.

Methods

Fish, cells, virus and vaccine

The wakin variety of goldfish (body weight, 9.8±1.8 g; body length, 7.4±0.5 cm) with no history of CyHV-2 infection was obtained from the Yoshida Field Science Station, Tokyo University of Marine Science and Technology. Sampled fish were euthanized with an overdose of 2-phenoxyethanol (Wako, Japan).

The CFS cell line derived from the ginbuna (C. auratus langsdorfii) [16] and RyuF-2 cell line from the goldfish [17] were maintained in Leibovitz’s L-15 Medium (L-15; Sigma-Aldrich, USA). The KF-1 cell line from the common carp (Cyprinus carpio) [18] was maintained in minimum essential medium (Sigma-Aldrich). Culture media were supplemented with 5% FBS (Sigma-Aldrich), 100 U ml^−1^ penicillin and 100 µg ml^−1^ streptomycin (Sigma-Aldrich). Cells were maintained at 25 °C.

The virulent virus strain CyHV-2 Saitama-1 isolate (CyHV-2 SaT-1), which was reported by Shibata et al. [17], and CyHV-2 C2P1 (SaT-1 strain cloned twice by limiting dilution on RyuF-2 cells and then passaged once on the cells) [11] were used for infection or challenge tests. The live-attenuated vaccine (P7-P8 strain) was used for vaccination [11]. The infectious titre of the CyHV-2 SaT-1, CyHV-2 C2P1 and P7-P8 vaccine virus, measured by titration [1113], was 10^4.1^, 10^5.7^ and 10^7.2^ TCID_50_ ml^−1^, respectively.

Library preparation for whole-genome sequencing of the vaccine and virus strains

The preparation of the virus and vaccine cultures has been described [11], including the virulent CyHV-2 SaT-1 strain. Briefly, the virulent CyHV-2 C2P1 strain (adapted to RyuF-2 cells and the origin of the P7-P8 vaccine) was propagated in the CFS cell line 4, 10 and 20 times (named CFS-P4, CFS-P10 and CFS-P20, respectively). After the virus was propagated seven times in the CFS cell line, the virus was further propagated in the KF-1 cell line four, six and eight times [named P7-P4, P7-P6 and P7-P8 (the vaccine strain), respectively].

After cytopathic effect (CPE) completion, virus cultures were centrifuged at 3,000 r.p.m. at 4 °C for 30 min. The supernatant was collected and centrifuged at 7,000 r.p.m. at 4 °C for 30 min. The supernatant was collected and centrifuged at 12,500 r.p.m. at 4 °C for 30 min to precipitate the virus. The virus pellet was dissolved in TNE buffer [50 mM Tris-HCl, 100 mM NaCl and 0.1 mM EDTA (pH 8.0)], and the DNA was extracted by the method mentioned below.

The DNA samples were pooled to construct paired-end NGS libraries using Nextera XT DNA Library Preparation Kit (Illumina, USA). The libraries were sequenced using Illumina MiSeq Reagent Kit v2 (300 cycles) (Illumina).

Sequence assembly and annotation

Low-quality reads and adaptors were trimmed using fastp, v0.23.2 [19] with the option: -q 30 -n 10 -t 1 -T 1 -l 20 -w 16. FastQC, v0.11.9 [20], was used to confirm the quality of trimmed reads. The trimmed reads were mapped to the reference genome (CyHV-2 ST-J1: accession no. NC_019495.1) using bowtie2 v2.4.4 [21] with default parameters. Alignment sorting and deduplication were performed using SAMtools, v1.13 [22], and GATK MarkDuplicates, v3.1.0 [23], respectively.

Variant calling was performed using Pilon, v1. 24 [24] with the variant option. To annotate each nucleotide variant, sites in the VCF files generated using Pilon with a filter tag other than PASS were removed using VCFtools, v0.1.16 [25]. Then, each variant was annotated using SnpEff, v5.2 [26]. The results were visualized using IGV, v2.16.2 [27].

The gene annotation and protein-coding genes were predicted using Prokka, v1.14.6 [28], with the ‘--kingdom Viruses’ option. Next, the gene alignments were obtained by utilizing Roary, v1.7.8 [29], with the ‘-e --mafft -p 8’ option. Pairwise single-nucleotide polymorphism (SNP) distance matrices of the core genes were analysed using snp-dists v0.8.2 (https://github.com/tseemann/snp-dists). Phylogenies of core genes were aligned using iqtree2, v2.0.7 [30], and the Roary matrix was generated.

De novo assembly of trimmed reads was performed using SPAdes, v3.13.1 [31]. Contig quality was assessed using the quality assessment tool for genome assemblies, v5.2.0 [32]. Contigs related to the CyHV-2 genome were identified using blastn against the National Center for Biotechnology Information (NCBI) database. The scaffolds were assembled, and the orientation of reads was visualized using Mauve multiple alignment [33].

Propagation characteristics of the P7-P8 vaccine in vitro

RyuF-2 cells were seeded into 12-well plates containing L-15 culture medium and incubated at 25 °C. Experiments were prepared in three replicates. After 24 h, the cells were inoculated with CyHV-2 SaT-1 (10^2.1^ TCID_50_ well^−1^) or P7-P8 (10^5.2^ TCID_50_ well^−1^). For the CyHV-2 SaT-1 group, kidney extract (0.02%) was added. The uninfected control group was inoculated with L-15 culture medium. CPE was monitored in all groups. Culture supernatants were collected for virus titration. Concurrently, cells were harvested by trypsinization, centrifuged at 1,000 g for 5 min at 4 °C and prepared for gene expression analysis, terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assay and ANNEXIN V assay.

The virus growth kinetics comparison of P7-P8 and the virulent virus was studied to substantiate the interpretation of the in vitro growth characteristics. In this study, to prepare a multiplicity of infection (MOI)=1, the virulent CyHV-2 C2P1 strain was used for its higher infectious titre [11], and kidney extract was not required [17]. Cells were seeded into 24-well plates. After 24 h, the virus inoculum was adsorbed for 1 h and washed with L-15 medium. Cell monolayers were scraped and centrifuged (1,000 g, 5 min, 4 °C), and supernatants (released virus) were collected. Cell pellets (cell-associated virus) were sonicated to release intracellular virus particles and centrifuged, and the supernatant was collected. Experiments were prepared in three replicates. The collected virus solution was subjected to virus DNA quantification.

To compare the growth properties of the vaccine and virulent virus strain (CyHV-2 C2P1), the plaque assay was performed. Cells per well were seeded into 6-well plates. Following virus adsorption for 1 h, the inoculum was removed, and cells were washed with culture medium before overlaying with 3 ml of plaque medium (a 1 : 1 mixture of 1.2% carboxymethylcellulose and 2× culture medium containing FBS, penicillin and streptomycin). After 6 days, cells were fixed with formalin, stained with 0.5% crystal violet, and the plaque area (mm^2^) (n=20) was estimated using Image J [34].

Propagation characteristics of the P7-P8 vaccine in vivo

Goldfish (n=20) were vaccinated using the showering method, which involved sprinkling 3 ml of the P7-P8 vaccine (10^5.2^ TCID_50_ ml^−1^) onto the fish on a net and holding them for 10 s [12]. The fish were then reared at 25 °C. In addition, goldfish (n=20) were infected with the virulent CyHV-2 SaT-1 by immersion (10^1.1^ TCID_50_ ml^−1^) at 25 °C for 2 h. The uninfected control (n=20) group was treated with L-15 culture medium.

In the vaccinated, uninfected control and virus-infected sampling groups (n=15), the tissue samples (trunk-kidney, spleen or gills) were preserved for virus DNA quantification, gene expression analysis in RNAlater (Thermo Fisher Scientific, USA) and histopathology and TUNEL assay in Davidson’s solution.

At 28 days post-inoculation (dpi), the vaccinated and infected control groups were challenged with the virulent CyHV-2 SaT-1 by immersion as mentioned above. The relative percentage survival (RPS) of vaccinated fish was calculated as: RPS = (1 − cumulative mortality rate of vaccinated group ∕ cumulative mortality rate of the infected control group) × 100 (%).

DNA extraction, PCR and quantitative PCR

DNA extraction of virus pellet, virus solution and sampled organs was performed using the phenol/chloroform/isoamyl alcohol method according to previous studies [1112]. DNA extracted from the virus pellet was further treated with RNase A (Nippon genes, Japan).

DNA samples from randomly selected trunk-kidney samples from virulent virus-infected and vaccinated fish were utilized to validate the genomic alterations in the P7-P8 vaccine strain. The list of primers used is provided in Table S1 (available in the online Supplementary Material). PCR products were visualized, and sequencing was outsourced to Eurofins Genomics (Japan) using Sanger sequencing.

For the quantification of virus DNA, the virus DNA copy number was determined using TaqMan quantitative PCR of the CyHV-2 DNA polymerase gene as described previously [1135]. The absolute quantification of virus in the organs was expressed in DNA copies mg^−1^.

Gene expression analysis

Total RNA was extracted from RyuF-2 cells, and vaccinated fish organs were extracted using a NucleoSpin RNA kit (Macherey-Nagel, Germany) and RNAiso Plus Reagent (Takara, Tokyo, Japan), respectively. First-strand cDNA was synthesized with 500 ng of total RNA using M-MLV Reverse Transcriptase (Thermo Fisher Scientific).

The primers used for gene expression analysis are listed in Table S2. The elongation factor 1α gene was used as the internal control. The THUNDERBIRD SYBR qPCR Mix (Toyobo, Japan) was used for quantitative real-time PCR and performed using QuantStudio 3 (Applied Bioscience, USA), and the 2^−ΔΔCT^ method was used to determine gene expression fold change.

Histopathology

Sampled trunk-kidney and spleen from vaccinated and virulent virus-infected goldfish were fixed in Davidson’s solution. The paraffin-embedded samples were sectioned at 4 µm thickness and subjected to haematoxylin and eosin staining. Histopathological changes in the haematoxylin and eosin-stained samples were observed using a light microscope (ECLIPSE Ci: Nikon, Tokyo, Japan).

TUNEL assay

MEBSTAIN Apoptosis TUNEL Kit Direct (MBL, Japan) was employed to perform TUNEL assays on RyuF-2 cells and paraffin-embedded tissues. RyuF-2 cells were permeabilized with 70% ethanol and stained. The CytoFLEX SRT cell sorter (Beckman Coulter, USA) was used to calculate the percentage of TUNEL-positive cells relative to propidium iodide-stained cells. Data analysis was performed using Kaluza V1.5 software (Beckman Coulter). The gating strategy is shown in Fig. S1.

For paraffin-embedded tissues, antigen recovery was performed before staining. The sections were counterstained with Hoechst 33342 (Thermo Fisher Scientific), and fluorescent images were captured using a fluorescence microscope and processed with NIS Elements software (Nikon, Tokyo, Japan). TUNEL-positive cells were quantified using ImageJ [36] and expressed as a percentage of total cells: TUNEL-positive (%) = (number of positive cells/total number of cells).

Annexin V assay

MEBCYTO Apoptosis Kit (Annexin V-FITC Kit; MBL, Japan) was used for the Annexin V assay to detect apoptosis in RyuF-2 cells. Cells were stained to assess the externalization of phosphatidylserine on the cell membrane. The apoptotic cells were analysed using the CytoFLEX SRT cell sorter (Beckman Coulter), and data were processed using Kaluza v1.5 software (Beckman Coulter). The gating strategy is shown in Fig. S2. Cells positive for Annexin V-FITC were considered apoptotic.

Statistical analysis

In vitro gene expression analysis, TUNEL assay and ANNEXIN V assay were statistically evaluated using one-way ANOVA followed by Tukey’s post hoc test to compare differences between groups. For in vivo gene expression analysis, virus kinetics and plaque area, Student’s t-test was used. Data are presented as mean±standard error (se). The survival rates of vaccinated and infected control groups were analysed using Fisher’s exact test. Statistical significance was considered for P values <0.05, with significance denoted by asterisks or alphabetic markers. Statistical analysis was conducted using R software version 4.3.1 [37].

Results

Whole-genome sequencing and comparative genomics of the P7-P8 vaccine revealed mutations in membrane and functional proteins

The paired-end trimmed reads and GC content generated from Illumina short-read sequencing are presented in Table S3. The genomic distribution of variants among virulent, passaged attenuated and vaccine strains is depicted in Fig. 1a. Variants in CyHV-2 SaT-1 were excluded due to their negligible impact on virulence, whereas those in CyHV-2 C2P1 (origin of the vaccine strain) were included to trace attenuation. Amino acid variations, such as missense mutations, insertions, deletions and frameshift mutations, relative to the reference genome were identified (Table 1). Variant analysis revealed 24 variants, comprising 7 SNPs (5 within ORFs), 11 deletions (5 within ORFs) and 6 insertions (1 within an ORF). Most variations were shared between the attenuated viruses and the vaccine. A notable finding was a 357 bp deletion in membrane protein ORF 25D, exclusive to the attenuated viruses and vaccine strain (Fig. 1b, c). A high-impact start-loss variant in ORF 113 was identified using SnpEff and confirmed through sequencing (Fig. 1d). A previous study has shown that CyHV-2 encodes microRNA miR-C12, which suppresses virus-induced apoptosis and enhances virus replication by targeting caspase-8 [8]. In our analysis, there are no mutations found in the region encoding miR-C12 for the vaccine and virulent virus strains. Phylogenetic analysis revealed 267 gene clusters, including 199 core genes shared across samples (Fig. 1e). The presence and absence of genes are shown in Table S5. The virulent strains clustered together, with CFS-P4 being the closest relative. The continued passage led to the distinct clustering of attenuated viruses and the vaccine away from virulent strains. Notably, the virulent strains exhibited low relative SNP distances of core genes compared with the vaccine (Fig. 1f), suggesting that only a few SNPs underlie the differences between strains.

Output of the SnpEff analysis and mutations in ORFs using reference-based (CyHV-2 ST-J1) assembly. The ORF mutations were listed out (vertical marks) relative to the virulent challenge test virus CyHV-2 SaT-1 (a). Bioinformatics analysis and mutation validation. The targeted ORF 25D PCR product (b and c) of vaccinated (lane 1) fish and virulent virus-infected (lanes 2–6). The start codon lost in ORF 113 was confirmed by PCR and sequencing (d). Clustering of reference-based assembled samples. The Roary matrix generated by aligning core genes of sample-annotated protein sequences (e). The SNP distance of core genes generated by Roary (f). The sequence inversion between ORF 24 and ORF 25C detected by de novo assembly was confirmed by PCR and sequencing (g and h). Primers ORF 24 and ORF 25 (lanes 1, 2, 5 and 6) show the original genomic orientation, while ORF 24-Inversion and ORF 25-Inversion (lanes 3, 4, 7 and 8) identify inverted sequence. Red and yellow boxes indicate the position of the expected PCR product size.

The quality of the de novo assembly contigs was assessed using the QUAST tool (Table S4). The analysis revealed that CyHV-2 SaT-1 retained the same read orientation as the reference genome (Fig. 1g). Notably, CyHV-2 C2P1, along with the passaged attenuated viruses and the vaccine, exhibited a 6,262 bp sequence inversion between ORF 24 and ORF 25C. The inversion at the ORF 25C terminus corresponded to a sequence in-frame inversion, whereas the inversion at the ORF 24 terminus was associated with a 429 bp deletion, resulting in a frameshift. Primers designed to target these inversion sites were evaluated on fish infected with CyHV-2 SaT-1 and those vaccinated with P7-P8, producing distinct PCR bands for each sample (Fig. 1h).

Vaccine propagation characteristics in vitro suggest induction of apoptosis

The propagation characteristics of the P7-P8 vaccine virus in comparison to the CyHV-2 SaT-1 strain were investigated. Both viruses exhibited characteristic CPE, which included cytoplasmic vacuolation and detachment of cells from the culture flask (Fig. 2a, b). Notably, the CyHV-2 SaT-1 strain demonstrated syncytium formation among the infected cells, whereas the P7-P8 vaccine virus exhibited a uniform distribution of CPE without cell fusion. CPE was completed by 5 dpi for both viruses. The infectious titres of both viruses progressively increased from 1 to 4 dpi (Fig. 2c). The CyHV-2 SaT-1 virus and the P7-P8 vaccine virus showed increasing titres from 10^3.63^ TCID_50_ ml^−1^ to 10^5.47^ TCID_50_ ml^−1^ and 10^5.38^ TCID_50_ ml^−1^ to 10^8.47^ TCID_50_ ml^−1^, respectively.

Propagation characteristics of the P7-P8 vaccine- and virulent CyHV-2 SaT-1-inoculated cells. The RyuF-2 cell inoculated with the virulent virus (a) and the vaccine (b) showed typical CPE at 4 dpi. The virulent virus sample showed syncytium formation (black arrowhead), but this was not found in the vaccine group. The virus infectious titre of both samples is showing active propagation (c). Comparison of the growth kinetics of P7-P8 and CyHV-2 C2P1 at MOI=1 (e). The CyHV-2 C2P1 (d) showed syncytium formation (black arrowhead). The virus supernatant and sonicated cell pellet were subjected to virus DNA quantification (e). The plaque area of CyHV-2 C2P1 and P7-P8 showed similar sizes (f). Uninfected control, CyHV-2 SaT-1- and P7-P8 vaccine-inoculated cells were tested for apoptosis-related gene expression levels (g) and TUNEL and ANNEXIN V assay (h). Data show mean±se. The significant differences (P<0.05) are denoted with letters or an asterisk.

Virus replication kinetics of P7-P8 and CyHV-2 C2P1 revealed similar growth patterns while CyHV-2 C2P1 induced syncytium formation (Fig. 2d). Both strains exhibited a latent phase [0 to 6 h post-inoculation (hpi)] characterized by minimal replication activity (Fig. 2e). Exponential replication commenced at 12 hpi, with parallel progression reaching peak titres by 72 hpi, showing ~10^8.8^ and 10^8.3^ DNA copies ml^−1^ from cell pellet (cell-associated virus) and supernatant (released virus), respectively. The CyHV-2 C2P1 pellet exhibited significantly higher virus DNA loads than the P7-P8 pellet at 0 and 12 hpi, while the supernatant showed significantly elevated levels at 2 hpi.

Plaque assay findings showed cell detachment from a radial central area and progress outward. The plaque sizes of the virulent CyHV-2 C2P1 strain (12.79±3.47 mm²) and the vaccine P7-P8 strain (13.577±2.93 mm²) were not significantly different (Fig. 2f). Representative plaque assays are shown in Fig. S4.

The expression levels of apoptosis-related genes were assessed (Fig. 2g). Caspase-8, caspase-3 and Bax genes exhibited comparable expression levels between the vaccine and virulent virus. Although the expression level of the caspase-8 gene did not reveal significant differences, it displayed an increasing trend. The expression level of both caspase-3 and Bax genes demonstrated a significant increase relative to the control group, particularly at 3 and 4 dpi. The expression level of the CyHV-2 SaT-1 cytochrome c gene was significantly elevated compared with that in both the control and virulent virus groups from 1 to 4 dpi, whereas the vaccine group only showed a significant difference from the control group at 3 and 4 dpi.

The proportion of TUNEL-stained positive cells at 2 dpi for both the vaccine and virus groups was significantly higher than that of the uninfected cells (Fig. 2h). At 4 dpi, the percentage of TUNEL-positive cells in the vaccine group was significantly increased compared with both the uninfected and CyHV-2 SaT-1-infected cells. The percentage of ANNEXIN V-stained positive cells showed no significant changes at 1 and 2 dpi (Fig. 2h); however, it was significantly higher in the vaccine-infected cells at 3 and 4 dpi. The proportions of uninfected cells and CyHV-2 SaT-1-infected cells remained similar throughout the study.

Histopathology of vaccinated goldfish revealed leukocyte aggregation and melanomacrophage centre formation, without marked degenerations

Goldfish were vaccinated or infected with a virulent virus, and organ samples were collected. The uninfected control sections are shown in Fig. S3. There is no mortality in the uninfected and vaccinated groups. In contrast, fish infected with the virus exhibited moribund conditions or were dead by 5 dpi, with no survivors in the infected group for sampling at 7 dpi.

At 1 dpi, the trunk-kidney virus load for the vaccinated and infected groups was ~10^3.53^ and 10^5.06^ DNA copies mg^−1^, respectively (Fig. 3h). No signs of necrosis or leucocyte aggregation were detected in any group (Fig. 3a, b). By 3 dpi, a significant divergence in virus load was observed – virulent CyHV-2 SaT-1-infected and vaccinated fish exhibited virus loads of ~10^8.41^ and 10^5.01^ DNA copies mg^−1^, respectively (Fig. 3c, d). Infected cells within the virus-infected group displayed enlarged nuclei in haematopoietic tissues. By contrast, these changes were not observed in the vaccinated group, whereby leucocyte aggregation and melanomacrophage centre development were noted. At 5 dpi, the virus load in the virus-infected fish reached 10^9.66^ DNA copies mg^−1^, and all specimens were in a moribund state, exhibiting severe necrosis manifested by karyopyknosis and karyorrhexis (Fig. 3f). Conversely, vaccinated fish showed a peak virus load of ~10^6^ DNA copies mg^−1^ in the trunk-kidney at 5 and 7 dpi (Fig. 3e, g), without signs of necrosis but with leucocyte aggregation and melanomacrophage centre development.

Histopathological changes in the trunk-kidney and spleen of vaccinated and CyHV-2-infected fish. In the trunk-kidney, no visible histopathological changes were noted at 1 dpi for both groups (a and b). The virus load was 103.53 virus DNA copies mg−1 and 105.06 virus DNA copies mg−1 (h) for the vaccinated and virus-infected groups, respectively. At 3 dpi, vaccinated fish (105.01 virus DNA copies mg−1) did not show signs of necrosis but showed some leucocyte aggregation (white arrowhead) (c). The virus-infected fish (108.41 virus DNA copies mg−1) showed marginated chromatin (black arrowhead) and no visible leucocyte infiltration (d). At 5 dpi, vaccinated fish (105.96 virus DNA copies mg−1) showed increased leucocyte infiltration and melanomacrophage centre formation (white arrowhead) (e). The virus-infected fish (109.66 virus DNA copies mg−1) showed severe necrosis manifested by karyopyknosis and karyorrhexis (black arrowhead) (f). At 7 dpi, the vaccinated fish (105.45 virus DNA copies mg−1) showed leucocyte infiltration and melanomacrophage centre formation (white arrowhead) (g). In the spleen, no visible histopathological changes were noted at 1 dpi for both groups (i and j). The load was 104.51 virus DNA copies mg−1 and 104.61 virus DNA copies mg−1 (p) for the vaccinated and virus-infected groups, respectively. At 3 dpi, vaccinated fish (104.37 virus DNA copies mg−1) did not show signs of necrosis but showed some enlarged ellipsoids (white arrowhead) (k). The virus-infected fish (107.99 virus DNA copies mg−1) showed marginated chromatin (black arrowhead) and melanomacrophage centre formation (l). At 5 dpi, vaccinated fish (105.65 virus DNA copies mg−1) showed more enlarged ellipsoids and melano-macrophage centre formation (white arrowhead) (m). The virus-infected fish (108.83 virus DNA copies mg−1) showed severe necrosis manifested by karyopyknosis and karyorrhexis (black arrowhead) and more melanomacrophage centre formation (white arrowhead) (n). At 7 dpi, the vaccinated fish (105.85 virus DNA copies mg−1) showed enlargement of ellipsoids (white arrowhead) (o). The horizontal line shows the detection limit (102 DNA copies mg−1) (h and p). Haematoxylin and eosin stain. Bar=50 µm.

In the spleen, the virus load showed a similar trend observed in the trunk-kidney samples (Fig. 3p). At 1 dpi, the spleens of both the vaccinated and infected control groups displayed ~10^4.5^ DNA copies mg^−1^, with no signs of necrosis (Fig. 3i, j). By 3 dpi, virus loads diverged, with 10^4.37^ DNA copies mg^−1^ in the vaccinated group and 10^8^ DNA copies mg^−1^ in the virus-infected group (Fig. 3k, l). Both groups showed enlarged ellipsoids; however, the virus-infected group began exhibiting an enlarged nucleus bearing chromatin marginalization and decreased leucocyte infiltration. At 5 dpi, the virus-infected group showed extensive necrosis and melanomacrophage centre formation, with a virus load of 10^8.83^ DNA copies mg^−1^ (Fig. 3n). By contrast, vaccinated fish had a peak virus load of ~10^5.8^ DNA copies mg^−1^ in the spleen at 5 and 7 dpi (Fig. 3m, o), accompanied by enlargement of ellipsoids and formation of melanomacrophage centres, with no indication of necrosis.

Vaccinated goldfish showed upregulation of proinflammatory and granzyme B genes after vaccination and high protective efficacy after virus challenge

Tissues from the trunk-kidney, spleen and gills of vaccinated fish were collected to assess fold-change expression of granzyme B and proinflammatory genes (Fig. 4a). The granzyme B gene exhibited a significant upregulation in the trunk-kidney at 5 dpi, with no notable changes in the other organs. The IL-1β gene demonstrated a significant increase in expression in both the trunk-kidney and gills at 5 dpi. The TNFα gene showed a significant increase in expression in the spleen and gills at 7 dpi.

Gene expression analysis of vaccinated fish. The trunk-kidney, spleen and gills were sampled. IL-1β, TNFα and granzyme-B gene expression relative to the uninfected control (a). Protective effect of the P7-P8 vaccine after virulent CyHV-2 infection (b). Goldfish (n=20) were vaccinated using the showering vaccination method. At 28 dpi, goldfish were infected with the virulent CyHV-2 SaT-1 by immersion. The mortality was recorded up to 21 days post-challenge. The vaccination group exhibited statistical differences (P<0.05) compared with the infected control group.

The infected control group exhibited a survival rate of 15% following a challenge with the virulent virus at 28 dpi (Fig. 4b). By contrast, the vaccinated group achieved a survival rate of 85%, corresponding to an RPS of 82.35%, showing a statistically significant difference between groups.

Vaccinated fish showed signs of early induced apoptosis

The trunk-kidney of both experimental groups exhibited a higher percentage of TUNEL-positive signals than that in the uninfected control fish at 1 dpi (Fig. 5a, b, h). The uninfected control sections are shown in Fig. S3. At 3 dpi, a significant increase was observed in TUNEL signals in the trunk-kidney relative to both the uninfected control and virulent virus-infected fish (Fig. 5c, d), despite the virus DNA copy number in vaccinated fish being ~1,000 times lower than that in the virulent virus-infected group (Fig. 3h). At 5 dpi, moribund infected fish demonstrated a significant increase in TUNEL signals in the trunk-kidney compared with both the uninfected control and vaccinated fish (Fig. 5e, f). However, by 7 dpi, the vaccinated fish did not exhibit a significant increase in TUNEL-positive signals compared with the uninfected control group (Fig. 5g).

Detection of TUNEL-positive signals in the trunk-kidney and spleen of vaccinated and CyHV-2-infected fish. Green signals (white arrow) indicate apoptotic nuclei, while the blue signals represent Hoechst 33342-stained nuclei. The TUNEL-positive cell percentage was estimated using ImageJ. Significant differences (P<0.05) at each sampling time point are denoted with different letters. Bar=100 µm.

In the spleen, vaccinated fish displayed a significantly increased percentage of TUNEL-positive signals relative to both the virulent virus-infected and uninfected control groups at 1 dpi (Fig. 5i, j, p). At 3 dpi, no significant differences in TUNEL signals were observed across groups (Fig. 5k, l). By 5 dpi, the virulent virus group showed a significant increase in TUNEL signals compared with the uninfected control and vaccinated fish (Fig. 5m, n), although the copy number of virus DNA in the vaccinated fish remained ~1,000 times lower than that in the virulent virus-infected group (Fig. 3p). At 7 dpi, no significant increase in TUNEL-positive signal percentage was observed in vaccinated fish compared with uninfected control fish (Fig. 5o).

Discussion

Viruses encode virulence factors that determine pathogenesis in the host. Although research on CyHV-2 virulence factors has advanced, many virus protein functions remain uncharacterized. This research investigated the mechanisms of vaccine attenuation and provided insights into CyHV-2 virulence factors through comparative genomic analysis and studies of the propagation characteristics.

We conducted whole-genome sequencing of virulent viruses, passaged attenuated viruses and the P7-P8 vaccine strain. One of the key features of the vaccine strain was the presence of mutations in membrane proteins, including a deletion in ORF 25D; an inversion between ORF 24 and ORF 25C; and SNPs in ORFs 108, 115 and 132. Studies indicate that DNA or subunit vaccines targeting the CyHV-2 membrane proteins are immunogenic [38]. These proteins are crucial for virus attachment, adsorption and fusion with host cells [39]. Therefore, the cumulative mutations in the membrane proteins of the vaccine may affect virus attachment to host cells. In addition, the vaccine strain exhibited mainly SNPs and deletions, but no mutations in known CyHV-2 virulence genes, such as ORFs 23, 55 (thymidine kinase), 57, 98 or 141 [4042], whereas several mutations were identified in ORFs with undetermined functions (Table 1). Virus pathogenesis is typically expressed through structural and non-structural proteins or non-coding sequences interacting with host responses [10]. ORF 113, which putatively functions as an apoptosis inhibitor [43], lost its start codon. Although the specific role of ORF113 in virus pathogenesis remains to be determined and multiple virus genes may be responsible for anti-apoptosis functions, a previous study suggests that the gene function may be affected, as a loss of start codon can result in gene silencing [44]. In alphaherpesviruses, apoptosis modulation genes can be found in the virus tegument and envelope [34]. Similar strategies have been identified in aquatic viruses [5]. Human simplex virus (HSV-1) glycoproteins mediate both cell entry and apoptosis inhibition, with deletion mutants confirming their essential anti-apoptotic role [45]. These findings appear conserved across herpesviruses, suggesting similar mechanisms may operate in aquatic herpesvirus membrane proteins. In this study, the cumulative mutation of membrane and functional proteins may be responsible for attenuation and apoptosis inhibition. Since the interplay between early apoptosis and progeny virus production proves to be complex, a gene (e.g. ORF 113) knockdown of the wild-type virus might shed light on the infection mechanism.

Bioinformatics analysis guided our focus on the propagation characteristics of both the vaccine and virulent viruses. In cell culture, similar growth kinetics, plaque formation and CPE completion times were observed for both. A mutation in the vaccine membrane proteins did not hinder virus entry or replication. However, syncytium formation was only seen in the virulent virus-infected cells, suggesting that the vaccine strain contains a mutation that prevents syncytium formation in vitro, and this may be related to mutations in the associated membrane protein gene. In CyHV-3, an SNP in the membrane protein ORF 131 contributed to in vitro growth adaptation [46]. In this study, the neighbouring membrane protein (ORF 132) displayed a frameshift mutation, thus potentially linked to growth adaptation in vitro. Studies have focused on deleting CyHV-3 virulence or replication genes, resulting in reduced cell culture infectivity [47]. This finding suggests that the vaccine, which retains intact known virulence virus replication genes, is well-adapted for propagation in the fin cell-derived cell culture, as indicated by the higher virus infectious titre, suggesting that factors other than replication genes play a role in cell culture adaptation.

Apoptosis, as a vital defence mechanism, triggers premature cell death, limiting virus replication and promoting the immune response. However, viruses encode pro-apoptotic or anti-apoptotic genes to generate progeny or promote their release [48]. In this study, gene expression analysis demonstrated elevated levels of host pro-apoptotic genes, including cytochrome c, Bax, caspase-3 and caspase-8, in cell cultures exposed to both the vaccine and virulent viruses. This indicates that both the vaccine and virus activate similar host apoptotic pathways. Previous studies suggest that virulent CyHV-2 upregulates host pro-apoptotic genes [9] and are consistent with our findings. However, only the vaccine-treated group exhibited a marked increase in DNA damage and phosphatidylserine exposure, which are clear indicators of apoptosis. Hence, these findings suggest that the vaccine cannot suppress apoptosis in infected cells.

Analysis of histopathology and apoptosis in vaccinated and virus-infected goldfish highlights key distinctions between the vaccine strain and the virulent virus strain. By 5 dpi, the virulent strain led to a high virus load of ~10^9^ DNA copies mg^−1^ in haematopoietic tissues, whereas the vaccinated fish exhibited slower virus replication, peaking at 10^6^ DNA copies mg^−1^ level, suggesting that differences in propagation occurred at an earlier time point of infection. Necrosis and tissue damage were evident in virus-infected fish by 3 dpi, whereas vaccinated fish displayed early-stage leucocyte infiltration and enhanced apoptosis, while not showing signs of necrosis. Although most of the vaccine-infected cells underwent apoptosis, some vaccine virus particles were released from the infected cells, possibly accounting for the 10-fold increase in virus load between 3 and 5 dpi. By inferring the in vitro results, the externalization of phosphatidylserine likely recruited phagocytes [49], indicated by leucocyte aggregation and melanomacrophage centre development, aiding virus clearance without significant tissue damage. These findings suggest the virulent virus causes cell damage-driven pathology, and vaccine infection promotes an induced immune response clearance with minimal cell damage. Proinflammatory cytokines lead to tissue inflammation and eventually allow tissue regeneration [50]. Granzyme B plays a role in the cytotoxic granzyme B/perforin pathway, which stimulates cytotoxic T lymphocytes to kill virus-infected cells [51]. In this study, the increase in proinflammatory cytokines, such as IL-1β and TNFα, at 5 and 7 dpi, facilitated tissue recovery. The activation of granzyme B further emphasizes the role of apoptosis in controlling virus spread, supporting virus clearance and potentially stimulating adaptive immune responses. Studies suggest that the early apoptosis of infected cells releases fewer virus progeny [52]. In this study, we hypothesize that the early release of vaccine virus DNA and infectious particles triggered the recruitment of immune cells to the infection sites. Therefore, fewer infectious particles were available to propagate the infection cycle. By 7 dpi, most of the detected virus was likely in the form of small numbers of infectious particles and virus DNA, since the peak in virus load did not correspond with increased necrosis or apoptosis.

This research suggests that the P7-P8 vaccine might induce an early apoptosis in the host, particularly during the early infection phases, preventing excessive virus replication and extensive tissue damage, which are characteristics of the virus infection. Virus-induced apoptosis can play a critical role in limiting virus replication and transmission [53]. Contrary to virulent CyHV-2, which employs virus proteins to evade host immunity, vaccine-induced apoptosis and inflammation create an environment that suppresses virus replication, enhancing the host’s ability to clear infection. This immune activation may contribute to vaccine effectiveness. In conclusion, the study of the P7-P8 live-attenuated vaccine highlights the identified genomic mutations that potentially attenuate the virus and demonstrates inflammation and apoptosis in the host, providing a guide for vaccine development, including those for other herpesviruses.

Supplementary material

10.1099/jgv.0.002227Supplementary Material 1.

10.1099/jgv.0.002227Uncited Supplementary Material 2.