Virulence attenuation of intestinal pathogenicity via combined gene deletion in duck enteritis vaccine strain restores gut microbiota balance and enhances safety

Abstract

There is an increasing need for a new generation of effective and safe vaccines, in the context of large-scale poultry farming and the prevalence of infectious diseases. With this in mind, we developed, for the first time, a duck enteritis virus (DEV) mutant, ΔTK-ΔgI/gE-ΔgG/gJ, through the deletion of multiple virulence genes. The resulting gene-deletion strain exhibited replication kinetics similar to those of the parent strain and was found to be safe in various animal models, offering a strategy for rapidly generating attenuated DEV strains. Previously, our team reported that DEV infection leads to intestinal dysbiosis; however, the impact of DEV vaccines on the gut microbiota remains unclear. This study aimed to characterize the gut microbiota of ducks, chicks, and mice immunized with DEV strains using microbiome analysis, assess the effects on microbial composition, and compare the outcomes. Both two strains caused significant shifts in gut microbiota diversity. Both strains restored the diversity of the microbiota, whereas the parental vaccine caused the enrichment of potential pathogens in chicks. Moreover, the conventional DEV vaccine disrupted gut microbiota and morphology, but the gene-deleted strain largely reversed these changes. These findings may improve the safety of vaccine through gene editing, thereby enhancing the protection of target animals.

Introduction

Duck enteritis virus (DEV), also known as duck plague virus (DPV), is a typical member of the Herpesviridae family and has caused significant economic losses to the waterfowl industry worldwide (Dhama, et al., 2017). From a temporal perspective, DEV has become widely distributed globally and has shown a trend toward younger hosts in recent years. The pathogen has been successively isolated from commercial ducks in various regions, including India, France, and Poland, among others, confirming the ability of the virus to persist in nature and transmit intermittently (Dandapat, et al., 2022; Islam, et al., 2021; Sarmah, et al., 2020). Owing to the large host range and genome of the DEV, it is possible to create a vector vaccine by inserting one or more foreign genes while ensuring replication of the DEV strain. Clinically, it can serve as a vaccine vector for various infectious diseases in poultry and waterfowl, to prevent infection by DEV. Many recombinant duck enteritis viruses have been demonstrated to effectively induce humoral and cell-mediated immune responses with protective effects (Apinda, et al., 2022; Zhao, et al., 2024). Therefore, a fast-acting, safe, immunogenic vaccine for poultry is still needed.

In previous studies, we reported that DEV infection in two-week-old ducks leads to intestinal damage, affecting the expression of inflammatory factors and disrupting barrier functions (Kong, et al., 2023). Therefore, immunization with a vaccine, as an effective tool for preventing virus infection, may alter the impact of the virus on the intestinal microbiota, with its metabolic products potentially modulating the host immune system (Lynn, et al., 2022; Ou, et al., 2023; Yuan, et al., 2022). When the permeability of the intestinal barrier is compromised, flora and metabolites can further influence target organs, influencing the immune system (Metzger, et al., 2018; Sardinha-Silva, et al., 2022; Yang, et al., 2023). The investigation of DEV infection from the perspective of the intestinal microbiota and its characteristics is of paramount importance for prevention and therapeutic strategies.

DEV, a large DNA virus with typical herpesvirus morphology, is composed of viral particles encompassing capsids, teguments, and envelopes (Dhama, Kumar, Saminathan, Tiwari, Karthik, Kumar, Palanivelu, Shabbir, Malik and Singh, 2017; Zhang, et al., 2020). Thymidine kinase (TK) gene encodes an enzyme that catalyzes thymidine phosphorylation (Li, et al., 2006), thereby supplying purine nucleotides for DNA synthesis during virus infection of host cells. TK also serves as a key determinant of neurovirulence and represents an important target for antiviral therapy and attenuated vaccine development. Glycoprotein E (gE), a nonstructural protein that forms heterodimers with glycoprotein I (gI), plays multifaceted roles in viral pathogenesis by mediating cell-to-cell fusion, intercellular viral spread, neurotropism, and virion release. Functionally, the gE-gI complex exhibits Fc receptor activity by binding to the Fc portion of host immunoglobulin G (IgG), thereby interfering with antibody-mediated neutralization and antibody-dependent cellular cytotoxicity, while also modulating antigen presentation pathways; these functions collectively contribute to immune evasion and enhanced viral virulence (Li, et al., 2011; Ning, et al., 2024). Glycoprotein G (gG) is a membrane glycoprotein among alphaherpesviruses, featuring a C-terminal transmembrane anchoring domain (Devlin, et al., 2006; Smith and Enquist, 1999). Functionally, gG exhibits chemokine-binding activity; however, the molecular mechanisms by which DEV gG mediates intercellular transmission, immune evasion, and virulence enhancement remain to be fully elucidated. Glycoprotein J (gJ), encoded by the US5 gene in herpesviruses. In other alphaherpesviruses, gJ has been demonstrated to contributes to virion morphogenesis and egress, potentially interacting with tegument proteins to stabilize viral particles during assembly. However, the specific functional domains and virulence-associated regions of the gJ gene in DEV have not been systematically investigated. This study utilized bacterial artificial chromosome (BAC) cloning technology combined with Red/ET homologous recombination technology to construct the recombinant plasmid of the BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ, by knocking out the TK, gI, gE, gG, and gJ genes from the BAC-DEV strain. To evaluate the in vitro biological characteristics and safety of the gene deletion strain in various animals, poultry, waterfowl, and mammals. We therefore performed a comprehensive analysis of conducting to examine the diversity and characteristics of the intestinal microbiota post-immunization.

Materials and methods

Cells and viruses

Specific-pathogen-free (SPF) duck embryos were procured from the Harbin Veterinary Research Institute, Chinese Academy of Agricultural Sciences. Duck primary fibroblast (DEF) cells (Tian, et al., 2019) were isolated from 12-day-old SPF duck embryos and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin (both from Thermo Fisher), and 10% fetal calf serum (FBS) at 37°C in a 5% CO₂ incubator. The DEV gene deletion viruses were rescued, propagated and titrated in DEF cells at 37°C. Growth curves were generated with various samples (n ≥ 3) examined at each time point.

Animals

The SPF chickens were sourced from Xinxing DHN Egg Co., Ltd., Guangdong. Young male/female BALB/c mice (4 weeks old) were procured from Zhuhai BesTest Bio-Tech Co., Ltd. Healthy one-day-old sheldrake ducks were obtained from a farm in Yunfu, Guangdong, China. Healthy one-week-old blackmaned geese and four-week-old pigeons were purchased from a breeding facility in Qingyuan, Guangdong, China. Four-week-old white-rumped munia and two-month-old quail were sourced from a hatching farm in Foshan, Guangdong, China. All experimental animals tested negative for DEV antibodies. They were raised in the animal facility at the Animal Laboratory, South China Agricultural University, under specific pathogen-free (SPF) conditions with controlled temperature (22 ± 2°C), humidity (50–60%), and a 12-hour light/dark cycle. Animals had ad libitum access to standard laboratory chow and filtered water. Ethical approval for the studies was obtained from the Institutional Animal Care and Use Committee of South China Agricultural University, and all procedures were conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals (National Institutes of Health, USA).

Construction and generation of the gene deletion strain

The DEV C-KCE strain was saved in our laboratory, and its whole genome was assembled into the BAC. The construction of the gene deletion viruses was based on Red/ET homologous recombination technology, combined with ccdB counterselection. In brief, the linear fragment "a-BamHI-amp-ccdB-a-b" was initially amplified via PCR, where 'a' and 'b' represent 50 bp homologous arms upstream and downstream of the TK gene, respectively. The PCR amplification primers used are detailed in Table 1. BAC-C KCE and the "a- BamHI -amp-ccdB-a-b" fragment were coelectroporated into the GB05-dir strain expressing the stable recombinase Red α/β. Through homologous recombination, the intermediate vector BAC-C KCE-TK/amp-ccdB was obtained. The BAC-C KCE-ΔTK/amp-ccdB plasmid was subsequently linearized by the BamHI restriction endonuclease, producing a linear vector. The linear vector was then electroporated into the GB2005 E.-coli strain, and correct clones of BAC-C KCE-ΔTK were selected through ccdB counterselection. Similarly, via the aforementioned method, the gI, gE, gG, and gJ genes were sequentially deleted. Ultimately, recombinant BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ was obtained.

Table 1.

Sequences of the primers used in construction.

Primer	Forward Sequence (5’-3’)	Reverse Sequence (5’-3’)
a- BamHI -amp- ccdb (TK) -a-b	acaggaaagaaaagctcgaggtgacataGGATCC tttgttcaaaaaaaagcc	atatgtcacctcgagcttttctttGGATCC TGCAGTGGGCTTACATG
a- BamHI -amp- ccdb (gI/gE) -a-b	tcaaaaatagaaatgggaacgacacgacGGATCC gcagcgtatgacagctctgggga	acctcaacaattggtacgatgatggttacGGATCC cgcatctgatcatggatgttga
a- BamHI -amp- ccdb (gG/gJ) -a-b	agaggcgcttgtgtagctggtGGATCC tttgttcaaaaaaaagc	tacacaagcgcctctatGGATCCTGCAG TGGGC

Verification of the gene deletion viruses

After the growth of DEF cells on day 2, the culture medium of DMEM containing 10% FBS was replaced with 2% FBS -containing DMEM without any supplements. Transfection was carried out via Lipofectamine 3000 (Thermo Fisher, US) (Rahimi, et al., 2018). After 48 hours of transfection, the virus was frozen, and the cell suspension was collected to obtain the virus, which served as a seed stock for subsequent experiments.

Viral DNA/RNA was extracted via the Viral DNA/RNA Miniprep Kit (Vazyme Biotech Co., Ltd) (O'Brien, et al., 2021). For PCR identification, the primer pairs gG-F/R and gJ-F/R were used to confirm the deletion of the DEV gG or gJ gene, and the primer pairs gI-F/R and gE-F/R were used to identify the presence of the gI or gE gene. Additionally, the primers TK-F/R were used to identify deletions in the TK gene. The detection primers are detailed in Table 2. The genetic stability of rDEV and the positive control at passages 10, 15, 20, 25, and 30 was assessed by PCR analysis, with the corresponding primers (F/R) listed in Table 2. Following infection of DEF cells with the parent strain or the DEV gene deletion strain at the multiplicity of infection (MOI) of 0.1, resulting in 80% lesions, the cells were fixed with 4% paraformaldehyde. Immunofluorescence assay (IFA) analysis (Zhang, et al., 2024a) was performed using rabbit anti-goat IgG antibodies to further confirm the success of virus rescue.

Table 2.

Primers used to detect stable genes subjected to knockout.

Gene	Total Length	Knockout Site	Post-deletion Length	Forward Primer	Reverse Primer
gG	1380 bp	1145-1380	1144 bp	Atagaggcgcttgtgtagctggt	aaaatagctgtttaaacagatgatt
gJ	1620 bp	445-1183	881 bp	tcgtagtccttatctcatgcagg	catacgcgcatatacatattgccg
gI	1089 bp	1-689	400 bp	atattgagtttcaaaaatagaa	tgttttatgatccccagagctg
gE	1473 bp	544-754	1262 bp	cggaacctcaacaattggtacg	gtcattagttcaacatccatga
TK	1077 bp	457-1017	516 bp	tgcgccaatacatttccgatcaa	cagagctttatttaaaacaaatatattta
F/R	449 bp	/	/	tgtgacaaattgcccttaaccctg	gccggcatcctcttcagggcgata

Viral growth kinetics

The growth kinetics of the parent strain and the gene deletion strain were assessed by generating growth curves at a low MOI (Ning, et al., 2022). In brief, DEF cells were cultured as monolayers in 24-well plates and infected with the virus at an MOI of 0.1 for 2 hours to promote virus replication. After the cells were incubated with the virus at 37°C and 5% CO₂, the culture medium was removed, and the cells were washed with Phosphate Buffered Saline (PBS). The medium was subsequently replaced with DMEM containing 2% FBS to support further viral growth. The infected cells were harvested at various time points (12, 24, 48, 60, 72, and 96 hours post-infection), and the collected samples were frozen and thawed repeatedly. The virus titers were determined via a 50% tissue culture infectious dose (TCID₅₀) assay (Erturk, et al., 1991; LaBarre and Lowy, 2001). All the experiments were conducted in triplicate (n=3) to ensure the reliability of the results.

Safety evaluation of the gene deletion strain

In this study, the parent strain and the gene deletion strain were utilized as experimental subjects on a diverse array of animals from different age and breed groups.

For the SPF chicks' study, 39 1-day-old SPF chicks were divided into three groups. Group Ⅰ received intramuscular (i.m.) (Wang, et al., 2015) of one dose of the parent strain. Group Ⅱ was inoculated with the gene deletion strain via intramuscular injection at the same dosage. Group Ⅲ, which served as the control, received intramuscular injections of an equivalent volume of PBS.

In the case of mice, groups comprising 13 BALB/c mice (4-week-old males/females) were immunized intraperitoneally (i.p.) (Sindi, 2023) with one dose of the parent strain, the DEV gene deletion strain, or PBS, each of which was administered in a final volume of 100 μl under light anesthesia.

Additionally, other experimental animals, including 1-day-old ducks, 4-week-old pigeons, 1-week-old geese, 4-week-old white-rumped munias, and 2-week-old quails, were immunized with the parent strain or DEV gene deletion strain at a dose of 200 μl to ensure that there were 13 animals in each group.

Safety testing was conducted on all the animals inoculated intramuscularly (i.m.) except for the mice (i.p.). On the fourth day postvaccination, three animals from each group were euthanized to collect gut samples and contents. All experimental animals were closely monitored for signs of disease and mortality for a duration of 14 days. All procedures involving normal immunocompetent animals were carried out under light or general anesthesia (Berendt, et al., 1977; Sivula and Suckow, 2018).

Chickens were euthanized by exposure to carbon dioxide in accordance with the UK Animals (Scientific Procedures) Act 1986. Before injection, mild anesthesia was induced with lidocaine (maximum dose 17.5mg/kg). All animals except mice were euthanized while under this anesthesia by CO₂ asphyxiation. Mice were deeply anesthetized with pentobarbital sodium (50∼90 mg kg⁻¹) and subsequently euthanized by cervical dislocation.

The real-time assay revealed the replication of the gene deletion strain in intestine

Next, we implemented a real-time PCR assay (Yu, et al., 2020) to detect the replication of DEV in the intestinal tissues of the animals. The DEV genome was employed as the template for this assay. Initially, the recycled product was ligated with the pMD-19T vector, generating a standard plasmid that functioned as the basis for our standard curve. In the initial step of the assay, the intestinal tissue of different animals was extracted via the SteadyPure Viral DNA/RNA Kit (Axygen, US). Subsequently, 1 μg of DNA was added to each sample, and paired primers were utilized in separate standard curve reaction mixtures within a 20 μl reaction volume. Standard curves were incorporated into every experiment to ensure accuracy and reliability. The PCR conditions were as follows: an initial denaturation at 95°C for 30 seconds, followed by 34 cycles of amplification at 95°C for 10 seconds and 60°C for 30 seconds (Kong, et al., 2022). This was succeeded by a dissociation curve analysis step, ensuring the exact evaluation of the obtained results.

DNA extraction and 16S rRNA amplicon sequencing and analysis

Fresh intestinal samples were promptly collected and swiftly frozen in liquid nitrogen within an hour of sampling (Li, et al., 2023). The ducks were randomly divided into the following groups. Group I: D1 (negative control, n =3), Group II: D5 (gene deletion strain, n =3), and Group III: D9 (parent strain, n =3). The chicks in the SPF chick groups were randomly divided into the following groups. Group I: R1 (negative control, n =3), Group II: R5 (gene deletion strain, n =3), Group III: R9 (parent strain, n =3). The mice were randomly divided into the following groups. Group I: M1 (negative control, n =3), Group II: M5 (gene deletion strain, n =3), and Group III: M9 (parent strain, n =3). The total genomic DNA of ducks, SPF chicks and mice was extracted via the CTAB/SDS method. The hypervariable region V34 of the bacterial 16S rRNA gene was amplified and sequenced using Illumina NovaSeq, with the forward primer 515F (5′-CCTAYGGGRBGCASCAG-3′) and the reverse primer 806R (5′-GGACTACNNGGGTATCTAAT-3′) (Siddiqui, et al., 2023). The amplification products were subsequently extracted and quantified via the Qiagen Gel Extraction Kit (Qiagen, Germany) (Zhang and Cahalan, 2007). The purified amplicons were then pooled in equimolar concentrations and sequenced on an Illumina NovaSeq platform (Sui, et al., 2023) to generate 250 bp paired-end reads. The 16S rRNA sequencing data were processed via Quantitative Insights Into Microbial Ecology (QIIME2) software (Hall and Beiko, 2018). After paired-end merging and error correction, DADA2 (Callahan, et al., 2016) was further applied for quality filtration. In this study, alpha diversity was used to assess the abundance and diversity of the gut microbiota in each animal, whereas beta diversity analysis highlighted differences in microbial communities among different groups (Chen, et al., 2024). The analysis of beta diversity included methods such as Venn diagrams, principal component analysis (PCA) (Rudi, et al., 2007), and Bray-Curtis dissimilarity matrix (Vogtmann, et al., 2024). Furthermore, the optimized sequences from the original data were spliced and filtered. Amplicon sequence variants (ASVs) (García-López, et al., 2021) were obtained via noise reduction methods, and species annotations of the ASVs were analyzed. The classification results were annotated at various taxonomic levels, including phyla, class, order, family, and genus.

Statistical analysis

All data analysis was performed with GraphPad Prism 8.0 software. Two-way repeated measures analysis of variance (ANOVA) (Chatzi and Doody, 2023) was used to test for significant differences in alpha diversity, and phyla, family, etc. One-way or ANOVA with Tukey’s post hoc test or Student's t test was used for data analysis (Mishra, et al., 2019). Significant changes in beta diversity were calculated via PERMANOVA on the basis of the Bray–Curtis distance matrix (Cao, et al., 2023; Liang, et al., 2023). For the animal studies, P values of ≤0.05 (Analytical Methods Committee Amctb, 2020) were considered statistically significant for all the statistical tests used. The number of animals per group is described in each figure or figure legend.

Results

Generation and identification of BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ

Through the utilization of Red/ET homologous recombination technology, we incorporated the linear fragment a-BamHI-amp-ccdB-a-b into the GB05-dir strain. In the initial step, we acquired an intermediate fragment, and subsequently, in the second step, we introduced the linearized fragment into the GB2005 E.coli strain. Positive clones, marked as BAC-C KCE-ΔTK, were successfully constructed. Expanding on this, we sequentially deleted the gI, gE, gG, and gJ genes, ultimately yielding BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ (Fig. 1A).

Fig. 1.

Genetic construction and biological assessment of BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ. (A) Schematic diagram of the construction of the gene-deletion virus BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ via the RedET recombination system. (B) Infection of DEFs with the gene-deletion virus BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ resulted in numerous fluorescent spots and cytopathic effects, and the mutant virus DEV-ΔTK-ΔgI/gE-ΔgG/gJ was rescued. (C) Identification of the genetic stability of the mutant virus. (a) 1-5: the gene-deletion strain, (b) 6-10: the positive control. (D) PCR identification of the mutant virus: (a) 1-2: gG gene identification primers, (b) 3-4: gJ gene identification primer, (c) 5-6: gI gene identification primers, (d) 7-8: gE gene deletion identification primers, (e) 9-10: TK gene deletion identification primers. (E) Determination of the effects of viral titer on growth kinetics. Samples were collected at the indicated time points, and viral titers were determined. The MOI for both the gene deletion strain and the parent strain was 0.1. Data are presented as the mean ± standard deviation (SD) of three independent experiments. (F) Safety assessment in diverse animal models, including ducks, SPF chicks, SPF BALB/c mice, pigeons, geese, white-rumped munias, and quail. Created with BioRender.com.

The BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ plasmid was transfected into a monolayer of DEF cells. Approximately 72 hours later, both cytopathic effects (CPEs) and fluorescence were observed. Compared with normal DEF cells, infected cells displayed aggregation and lesions, with green fluorescence distributed throughout the field of view (Fig. 1B). The gene deletion strain was passaged several times in DEF cells, and different passages of BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ strains were amplified via BCA-specific primers. PCR confirmed the stable replication of the gene deletion strain in DEF cells (Fig. 1C).

Viral DNA was extracted from DEF cells infected with the BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ strain. Using primers specific to the TK, gI, gE, gG, and gJ genes, PCR was performed to amplify gene fragments from both the parent strain and the gene deletion strain. The amplified fragments for the gG, gJ, gI, gE, and TK genes were 1380 bp, 1144 bp, 1620 bp, 881 bp, 1089 bp, 400 bp, 1473 bp, 1262 bp, 1077 bp, and 516 bp, respectively (Fig. 1D). The lengths of the amplified target segments were as expected, indicating the successful deletion of partial segments of the genes.

To evaluate the growth and proliferation of the gene deletion strain, we evaluated the replication of the DEV gene deletion strain, and the parent strain in vivo via DEF cells. As shown in Fig. 1E, both the DEV gene deletion strain and the parent strain presented no detectable virus titers within cells in the initial 12 h after virus adsorption, as determined by observing the number of CPE plaques (Aravind, et al., 2015; Dandapat, et al., 2024). The virus titers subsequently gradually increased, reaching a peak at approximately72 h. Both viral strains were in the replication and release phases. After 96 hpi, the virus titers of both strains tended to decrease. Notably, the DEV gene deletion strain (MOI=0.1) exhibited slightly lower replication in extracellular cells than did the parent strain, but the difference was minimal.

Clinical signs following inoculation of the gene deletion strain

Next, we assessed the safety of the DEV gene deletion strain in various animals. One-day-old SPF chicks were injected with the DEV gene deletion strain, resulting in the absence of clinical symptoms. Conversely, chicks inoculated with an equivalent dose of parent strain presented symptoms such as lethargy on the second day and the presence of greenish feces (Abdulrahim, et al., 2024), leading to three deaths on the third day. The peak mortality occurred between days 4 and 5, with all chicks immunized with the parent strain experiencing mortality (Fig. 2A). The mortality rate displayed a gradient distribution, indicating the lethality of the parent strain to one-day-old SPF chicks. In contrast, the gene deletion strain did not induce mortality in the vaccinated chickens.

Fig. 2.

Survival rates of diverse animals and virus replication within their guts. (A-G) Survival curves of SPF chicks, ducks, quail, SPF BALB/c mice, geese, white-rumped munias, and pigeons are plotted and marked in red, yellow, blue, black, purple, orange, and green, respectively. The gray color represents the parental virus (DEV) inoculated, and the red color represents the negative control. (H) Determination of virus titers in diverse groups of immunized animals, including ducks, SPF chicks, SPF BALB/c mice, pigeons, geese, white-rumped munias, and quail. The X-axis represents different groups, and the Y-axis represents Virus Titer (copies/μL). Data are presented as the mean ± standard deviation (SD) of three independent experiments. *P < 0.1, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Next, we immunized ducks, geese, quails, pigeons, and mice via both intramuscular and intraperitoneal injection methods. Clinical observations were conducted over a 14-day period, during which all immunized animals displayed no abnormal symptoms, and no deaths were reported (Fig. 2B-E, G). Conversely, the immunization of white-rumped munia with the parent strain resulted in a substantial number of deaths on the second day (60%), reaching a 100% mortality rate by the fourth day (Fig. 2F). In contrast, the DEV gene deletion strain group did not record any bird fatalities, and the animals in the negative control group remained under normal conditions. These findings suggest that the DEV gene deletion strain effectively diminishes the pathogenicity of the virus across various animal species, whereas the parent strain exhibits significant pathogenicity in SPF chicks and white-rumped munia at a young age.

Virus replication levels of the gene deletion strain in the intestinal tract

Immunization through inoculation has been utilized to investigate gene deletion strain responses in animals. Owing to the peak replication period of the DEV virus occurring 3–5 days postimmunization in the infected target organs (He, et al., 2012; Yin, et al., 2017), data from the fourth day were specifically analyzed to compare replication levels (Fig. 2H). One-day-old SPF chicks were intramuscularly inoculated with either the gene deletion strain or the DEV parent strain. Three chicks from each group were euthanized on the fourth day post-inoculation, and their intestinal organs were collected for virus titration. Compared with those in the negative control group, there was a significant increase in virus titers in the intestines after immunization with the parent strain. A similar increasing trend was observed in the gene deletion strain group, with the viral load stabilizing at 10^3.00 copies/μL.

Similarly, at the same time point (4 day post inoculation, dpi), the intestinal tissues of ducks, geese, and quails that were immunized with either the DEV gene deletion strain or the parent strain were extracted from three experimental animals in each group. Quantitative analysis revealed virus replication levels of approximately 10^2.50 to 10^4.00 copies/μL, which is consistent with the virus titer levels reported in previous studies in SPF chicks, and there was a general increase in virus titer 4 days after immunization.

In the study of SPF chicks, in four-week-old white-rumped munia, the virus titer of the gene deletion strain was lower than that of the parent strain (P < 0.0001). In both types of animals, the virus titers in the intestinal organs of the parent strain were greater than those detected for the gene deletion strain. In contrast, DEV replication was limited to the species mentioned in this study, while no virus replication of DEV was detected in the intestinal organs of pigeons or mice.

Safety of clinical observation

The subsequent step involves observing whether inoculation with gene deletion strain or parent strain would result in intestinal lesions in different animals (Fig. 3). After immunization, the intestines of one-day-old SPF chicks infected with the parent strain exhibited extensive congestion and bleeding in the intestinal mucosa (Cai, et al., 2023; El-Tholoth, et al., 2019). Similar to those in the negative control group, no pathological changes were observed in the intestinal organs of the gene deletion strain group. Compared with those in the negative control group, similar lesions were still observed in white-rumped munia inoculated with the parent strain. Similarly, the quail intestines presented small areas of congestion and a slightly reddish color. Compared with those in the negative control group and the gene deletion strain group, the blood vessels were more prominent. However, in the cases of ducks, geese, pigeons, and mice, no pathological changes were observed in the intestinal tissue.

Fig. 3.

Observations of clinical symptoms and pathological changes in the gut of animal post-immunization. The animals exhibited varying degrees of clinical signs of disease after inoculation with the parent strain of DEV. Photomicrographs of gut tissues from animals that developed pathology after inoculation with the parent strain of DEV are shown. From left to right, the guts of the samples are ducks, SPF chicks, SPF BALB/c mice, pigeons, geese, white-rumped munias, and quail.

To further observe the type, extent, and severity of lesions under a microscope, we collected intestinal tissues from various animals for histopathological examination (Fig. 3). The intestinal tract is susceptible to DEV. In line with autopsy lesions, one-day-old SPF chicks infected with the parent strain exhibited extensive villous rupture and shedding in the intestinal mucosal layer, revealing evident histopathological changes (Dhama, Kumar, Saminathan, Tiwari, Karthik, Kumar, Palanivelu, Shabbir, Malik and Singh, 2017; Qi, et al., 2009). Similarly, in white-rumped munias immunized with the parent strain, the intestinal mucosal layer displayed significant villous rupture and shedding, along with extensive loss of epithelial cells, partial capillary stasis, and dilation. The parent strain exacerbated damage to intestinal tissues, with no lesions observed in the negative control group. In the safety experiment involving quails, the gene deletion strain did not induce noticeable histopathological changes, and cell morphology remained normal. However, in the parent strain group, the intestinal tissues exhibited self-dissolution of intestinal cells and nuclear condensation. Other animals (ducks, geese, pigeons, and mice) did not exhibit histopathological changes in intestinal morphology. In the next step, we selected ducks, SPF chicks, and mice for further investigation of the microbial communities in the intestines.

The microbiota composition is altered following intestinal infection in ducks

In subsequent studies, the intestinal contents were utilized to monitor changes in the microbiota. A total of 1461552, 734243, and 814404 raw reads were obtained from stool samples of ducks, SPF chicks, and mice, respectively, via the Illumina NovaSeq platform. Nochime reflects the tag sequences that are ultimately employed for subsequent analysis after the chimeras are filtered out, i.e., effective tags (Table 3). Ultimately, 1144749, 583837, and 663549 optimized sequences were acquired and used to profile the intestinal microbiome ASVs as proxies for bacterial species. Effective tags were harvested for subsequent analysis.

Table 3.

Statistical results of sequencing.

Groups	RawPE	Combined	Qualified	Nochime	Base(nt)	Avglen(nt)
D1	490,996	404,734	395,503	342,533	135,426,139	1,185
D5	435,991	386,186	375,991	322,161	129,086,983	1,211
D9	534,565	527,426	514,903	480,055	194,686,834	1,217
R1	257,833	221,528	217,290	188,311	80,126,022	1,276
R5	249,028	248,214	243,310	205,916	87,026,390	1,268
R9	227,382	208,710	204,861	189,610	80,397,077	1,270
M1	263,014	257,558	252,754	190,773	81,399,402	1,280
M5	277,620	264,150	260,269	234,617	91,463,694	1,177
M9	273,770	257,208	252,322	238,159	97,472,672	1,228

RawPE: raw reads; Combined: concatenated tag sequence; Qualified: filtered sequence; Nochime: effective tags; Base: number of effective tags; AvgLen: average length of effective tags.

We extensively discussed the differences in microbiota abundance at the Phylum, Family, and Genus levels. The predominant phylum in all duck intestines, both before and after infection, was Firmicutes (Fig. 4A, D, G). Other detected phyla included Actinobacteria, Proteobacteria, and Bacteroidetes. We observed a significant increase in the relative abundance of Firmicutes and Actinobacteria, along with a corresponding decrease in Proteobacteria in Group II compared with the negative control group. The relative abundance of Proteobacteria was almost undetectable in Group III. Additionally, we found that the Firmicutes abundance increased at the Phylum, Family, and Genus levels after immunization compared with that in the negative control groups. Conversely, the abundance of Proteobacteria in Group II or Group III decreased at the Phylum, Family, and Genus levels compared with that in the negative control group (Group I). Compared with that of the negative control group (Group III), the situation could be altered after the inoculation of the gene deletion strain.

Fig. 4.

Relative abundance of the duck gut microbiota and differences in dominant species. (A) Composition of the bacterial community at the Phylum level. (B) Composition of the bacterial community at the Family level. (C) Composition of the bacterial community at the Genus level. In these figures, each bar on the horizontal axis represents a sample, whereas the vertical axis indicates the relative abundance of sequences at that taxonomic level. Consistent colors across the figures denote the same taxonomic level (A-C). (D) Comparison of OTU abundances categorized at the Phylum level. (E) Comparison of OTU abundances at the Family level. (F) Comparison of OTU abundances at the Genus level. In this figure, the top, left, and right vertices represent the three experimental groups: the parent strain group, the negative control group, and the gene-deletion strain group. Each dot in the figure corresponds to an OTU, with dot size indicating the relative abundance of the OTU, calculated as the average abundance across all samples in the three groups. The dot colors signify OTUs with significant differences (D-F). (G) Heatmap illustrating the correlation between Phylum-related taxa and the intestinal flora. (H) Heatmap illustrating the correlation between Family-related taxa and the intestinal flora. (I) Heatmap illustrating the correlation between Genus-related taxa and intestinal flora. In these heatmaps, each color block represents the abundance of a group, with groups arranged horizontally and taxonomic levels arranged vertically (G-I).

As depicted in Fig. 4B, C, E, F, H, and I, the dominant families in Group II included Clostridiaceae, Micrococcaceae, Corynebacteriaceae, Lactobacillaceae, and Streptococcaceae, and the dominant genus included Candidatus Arthromitus, Rothia, Corynebacterium, Aerococcus, and Streptococcus. Therefore, the composition of the intestinal microbiota was significantly altered following DEV infection, with specific enrichment of Bacteroidetes, Clostridiaceae, and Candidatus Arthromitus in ducks. Compared with that in Group III, some of the intestinal microbiota in Group III was altered by the administration of the gene deletion strain.

DEV induced variations in the diversity and richness of intestinal microbiota in ducks

The Group I samples had 1180 unique OTUs, the Group II samples had 676 unique OTUs, the Group III samples had 286 unique OTUs, and the Group I and Group II samples had 115 OTUs. The Group I and Group III samples had a total of 63 OTUs, indicating a significant recovery in species richness in the intestine after vaccination with the gene deletion strain compared with the parent strain (Fig. 5A). To further investigate the phylogenetic relationships of species at the genus level, representative sequences of the top 100 genera were obtained through multisequence alignment. The abundances of Firmicutes, Proteobacteria, and Actinobacteria were highly distributed (Fig. 5B).

Fig. 5.

Diversity of the duck gut microbiota was significantly altered following vaccination. The ducks were intramuscularly inoculated with 1 × 10⁴ Embryo Lethal Dose 50 (ELD₅₀)/ml of the parent strain, PBS or the gene deletion strain. Sterile fecal samples were collected before vaccination (D0) and on day 4 post-vaccination (D4). (A) Venn diagram outlining the OTUs associated with the parent strain and the gene deletion strain. D1, negative control; D5, the gene-deletion strain; D9, the parent strain. (B) Map depicting the community distribution of species across an evolutionary tree was constructed. The size of each sector represents the relative abundance of species at different taxonomic levels. (C) The Shannon index was employed to assess community diversity, with higher values indicating greater diversity. Diversity was notably high in group D1, moderate in group D5, and low in group D9. (D) The Shannon-Wiener curves representing the diversity of microorganisms in the samples. When the curves of the three datasets tend to flatten, the sequencing data volume is sufficiently large. The horizontal axis represents the number of randomly selected sequences, whereas the vertical axis represents the Shannon index, which reflects species diversity. Group D1 had the highest Shannon index, suggesting the highest species diversity in its samples, whereas Group D9 had the lowest. (E) The dilution curves of intestinal microbial diversity in ducks. The curves of the three sample groups tended to flatten at the end, indicating that the sequencing depth covered all species present in the samples. The samples positioned higher up (Group D1) presented greater species richness than those situated lower down (Group D9). (F-H) Feces were collected before and after vaccination, and the beta diversity of the duck fecal microbiota was assessed. Non-metric multidimensional scaling (NMDS), principal coordinates analysis (PCoA), and PCA were employed on a Bray-Curtis distance matrix to visualize the diversity and facilitate comparisons among the D5, D9, and D1 groups. These analyses demonstrated that following gene-deletion strain vaccination, the gut microbiota structure of the D5 group had significantly diverged from both its baseline and that of the D9 group.

Intestinal microbiota complexity was assessed via alpha-diversity indices (Yuan, et al., 2018). Greater Shannon indices (Rebout, et al., 2021) indicate a more varied and diverse microbial population. Compared with those of the negative control group, the Shannon indices of Group II and Group III were significantly greater (Fig. 5C). These results suggest that species richness and diversity in duck intestines might decrease following inoculation with the parent strain.

Analysis of the rarefaction and rank abundance curves revealed that these curves tended to flatten as the number of sample sequences increased, indicating sufficient richness and uniformity in the samples, and satisfactory sequencing depth (Fig. 5D, E). β-diversity analysis revealed the similarities and differences in the intestinal microbiome of each duck. To confirm that fecal samples were reasonable, we compared the microbiota compositions of intestine samples taken from the same ducks and found no significant differences. The samples with high community structure similarity tended to cluster (Fig. 5F-H).

Overall microbiota distribution and abundance in SPF chicks

Relative abundance analysis was conducted on the basis of the phyla, families, and genera detected in all SPF chick fecal samples. The relative abundances of the phyla were compared among the three groups. Several phyla (Firmicutes, Proteobacteria, Bacteroidota, Actinobacteria, Synergistota, etc.) were enriched in the negative control group, with Firmicutes constituting more than 90% of the relative abundance. Compared with the parent strain group, there was a significant increase in the relative abundance of Proteobacteria. After gene deletion, the relative abundance of Proteobacteria clearly decreased, indicating a negative correlation with the degree of deletion (Fig. 6A, D, G). Conversely, the abundance of Bacteroidota was negatively correlated with the DEV gene deletion strain (Fig. 6A, D, G).

Fig. 6.

Relative abundance of the chick gut microbiota and differences in dominant species. (A) Composition of the bacterial community at the Phylum level. (B) Composition of the bacterial community at the Family level. (C) Composition of the bacterial community at the Genus level. In these figures, each bar on the horizontal axis represents a sample, whereas the vertical axis indicates the relative abundance of sequences at that taxonomic level. Consistent colors across the figures denote the same taxonomic level (A-C). (D) Comparison of OTU abundances at the Phylum level. (E) Comparison of OTU abundances at the Family level. (F) Comparison of OTU abundances at the Genus level. In this figure, the top, left, and right vertices represent the three experimental groups: the parent strain group, the negative control group, and the gene-deletion strain group. Each dot in the figure corresponds to an OTU, with dot size indicating the relative abundance of the OTU, calculated as the average abundance across all samples in the three groups. Dot colors signify OTUs with significant differences (D-F). (G) Heatmap illustrating the correlation between Phylum-related taxa and intestinal flora. (H) Heatmap illustrating the correlation between Family-related taxa and intestinal flora. (I) Heatmap illustrating the correlation between Genus-related taxa and the intestinal flora. In these heatmaps, each color block represents the abundance of a group, with groups arranged horizontally and taxonomic levels arranged vertically (G-I).

Additionally, the abundance of Lactobacillaceae at the family level, regardless of gene deletion, was similar to that of the negative control group (Fig. 6B, E, H). Notably, the relative abundance of Pseudomonadaceae remained similar between the R5 and R9 groups of chicks receiving vaccine treatment, significantly differing from those without treatment. Simultaneously, we assessed the differences in relative abundance at the genus level. In summary, the changes in Lactobacillus, Muribaculaceae, and Pseudomonas were partially reversed by immunization with the gene deletion strain (Fig. 6 C, F, I).

Variations in the diversity and richness of intestinal microbiota induced by DEV in SPF chicks

The Group I samples presented 340 unique Operational Taxonomic Units (OTUs), whereas the Group II samples presented 1028 distinct OTUs, and the Group III samples presented 287 unique OTUs (Fig. 7A, B). The collective data revealed a notable increase in species richness within the duck intestine subsequent to vaccination in the negative control group, which was juxtaposed with the parent strain. Compared with the negative control group and the parent strain group, Group II presented the highest Shannon index, indicating greater richness of the microbial populations. This suggests that, following the establishment of a secure immune response, the species richness and diversity within the duck intestine remain intact (Fig. 7C). Similarly, in comparison with the negative control group, the second group presented a greater Shannon index. This finding indicates that, following the immunization of SPF chicks with the gene deletion strain, the diversity of the intestinal microbial communities peaked, indicating a notable alteration in the microbial population (Fig. 7D, E). Qualitative PCoA, PCA or NMDS analyses were performed to evaluate β diversity. β diversity analysis revealed differences among the various groups of SPF chicks (Fig. 7F, G, H). Samples within the same group are closely related, with a similar species composition structure and high similarity in community structure.

Fig. 7.

Diversity of the chick gut microbiota was significantly altered following vaccination. The chicks were intramuscularly inoculated with 1 × 10⁴ ELD₅₀/ml of the parent strain, PBS or the gene deletion strain. Sterile fecal samples were collected before vaccination (D0) and on day 4 post-vaccination (D4). (A) Venn diagram outlining the OTUs associated with the parent strain and the gene-deletion strain. R1, negative control; R5, the gene deletion strain; R9, the parent strain. (B) Map depicting the community distribution of species across an evolutionary tree was constructed. The size of each sector represents the relative abundance of species at different taxonomic levels. (C) The Shannon index was employed to assess community diversity, with higher values indicating greater diversity. Diversity was notably high in group R5, moderate in group R1, and low in group R9. (D) The Shannon-Wiener curve represents the diversity of microorganisms in the samples. When the curves of the three datasets tend to flatten, the sequencing data volume is sufficiently large. The horizontal axis represents the number of randomly selected sequences, whereas the vertical axis represents the Shannon index, which reflects species diversity. Group R5 had the highest Shannon index, suggesting the highest species diversity in its samples, whereas Group R9 had the lowest. (E) Rarefaction curves of intestinal microbial diversity in chicks. The curves of the three sample groups tended to flatten at the end, indicating that the sequencing depth covered all species present in the samples. The samples positioned higher up (Group R5) presented greater species richness than those situated lower down (Group R9). (F-H) Fecal samples were collected before and after vaccination, and the beta diversity of the chick fecal microbiota was assessed. Nonmetric multidimensional scaling (NMDS), principal coordinates analysis (PCoA), and PCA were employed on a Bray-Curtis distance matrix to visualize the diversity and facilitate comparisons among the R5, R9, and R1 groups. These analyses demonstrated that following gene-deletion strain vaccination, the gut microbiota structure of the R5 group had significantly diverged from both its baseline and that of the R9 group.

The microbiota of the intestines is altered after infection in mice

The third group of mice was subjected to testing and consisted of BALB/c mice. Fecal samples from the mice were analyzed to determine the relative abundance, and a comparison was made among the three groups in terms of the relative abundance of phylum, family, and genus (Fig. 8). The negative control group was enriched primarily in Firmicutes, Bacteroidota, and other phylum, with Firmicutes accounting for the highest proportion (Fig. 8 A, D, G). Conversely, Bacteroidota was scarcely detectable in the parent strain group, and the proportion of Firmicutes was relatively small. The relative abundance of Bacteroidota in the gene deletion strain group was almost identical to that in the negative control group. Furthermore, Lactobacillus, Ligilactobacillus, and other genera were found in nearly all the intestinal tissues.

Fig. 8.

Relative abundance of the mice gut microbiota and differences in dominant species. (A) Composition of the bacterial community at the Phylum level. (B) Composition of the bacterial community at the Family level. (C) Composition of the bacterial community at the Genus level. In these figures, each bar on the horizontal axis represents a sample, whereas the vertical axis indicates the relative abundance of sequences at that taxonomic level. Consistent colors across the figures denote the same taxonomic level (A-C). (D) Comparison of OTU abundances at the Phylum level. (E) Comparison of OTU abundances at the Family level. (F) Comparison of OTU abundances at the Genus level. In the figure, the top, left, and right vertices represent the three experimental groups: the parent strain group, the negative control group, and the gene-deletion strain group. Each dot in the figure corresponds to an OTU, with dot size indicating the relative abundance of the OTU, calculated as the average abundance across all samples in the three groups. Dot colors signify OTUs with significant differences (D-F). (G) Heatmap illustrating the correlation between Phylum-related taxa and intestinal flora. (H) Heatmap illustrating the correlation between Family-related taxa and intestinal flora. (I) Heatmap illustrating the correlation between Genus-related taxa and the intestinal flora. In these heatmaps, each color block represents the abundance of a group, with groups arranged horizontally and taxonomic levels arranged vertically (G-I).

At the family level, the abundance of Lachnospiraceae increased following immunization with the gene deletion strain, in contrast to the other two groups. The results revealed a low abundance of intestinal tissue in the group immunized with the parent strain, which was particularly evident in the phylum and family classifications (Fig. 8B, E, H). Fig. 8C, Fig. 8, Fig. 8 show that, following gene deletion, the overall relative abundance of the genus Lactobacillus decreased. The abundance of Streptococcus in the parent strain group remained constant at the genus level. In summary, significant alterations in gut microbial homeostasis were confirmed during the progression of the DEV gene deletion strain.

Alterations in the diversity and richness of the intestinal microbiota induced by DEV in mice

Unlike the first two animal groups, Groups II and III contained 787 and 774 unique characteristic sequences, respectively. The combined number of shared and unique characteristic sequences between these two groups was 128 (Fig. 9 A, B).

Fig. 9.

Diversity of the mice gut microbiota was significantly altered following vaccination. The mice were intraperitoneally inoculated with 1 × 10⁴ ELD₅₀/ml the parent strain, PBS or the gene deletion strain. Sterile fecal samples were collected before vaccination (D0) and on day 4 post-vaccination (D4). (A) Venn diagram outlining the OTUs associated with the parent strain and the gene deletion strain. M1, negative control; M5, the gene-deletion strain; M9, the parent strain. (B) Map depicting the community distribution of species across an evolutionary tree was constructed. The size of each sector represents the relative abundance of species at different taxonomic levels. (C) The Shannon index was employed to assess community diversity, with higher values indicating greater diversity. Diversity was notably high in group M5, moderate in group M1, and low in group M9. (D) The Shannon-Wiener curve represents the diversity of microorganisms in the samples. When the curves of the three datasets tend to flatten, the sequencing data volume is sufficiently large. The horizontal axis represents the number of randomly selected sequences, whereas the vertical axis represents the Shannon index, which reflects species diversity. Group M5 had the highest Shannon index, suggesting the highest species diversity in its samples, whereas Group M9 had the lowest. (E) Rarefaction curves of intestinal microbial diversity in mice. The curves of the three sample groups tended to flatten at the end, indicating that the sequencing depth covered all species present in the samples. The samples positioned higher up (Group M5) presented greater species richness than those situated lower down (Group M1). (F-H) Fecal samples were collected before and after vaccination, and the beta diversity of mice fecal microbiota was assessed. Non-metric multidimensional scaling (NMDS), principal coordinates analysis (PCoA), and PCA were employed on a Bray-Curtis distance matrix to visualize the diversity and facilitate comparisons among the M5, M9, and M1 groups. As the gene deletion treatment progressed, the gut microbiota showed no additional changes.

The Shannon index reflects the diversity of the microbiota, whereas the Chao1 index reflects the richness of the microbiota (Fig. 9 C, D, E). Our findings indicated that gene deletion led to a decrease in species diversity and an increase in species richness compared with those of the negative control group. β-Diversity analysis revealed the similarities and differences in the gut microbiome among individual mice (Fig. 9 F, G, H). Samples within the same group presented high structural similarity, clustering together, indicative of a similar composition structure among the samples. Consequently, the results of PCoA and PCA, which are based on the Bray‒Curtis distance, demonstrated statistically significant differences among the samples from the three groups (P< 0.01).

Discussion

We innovatively leveraged reverse genetics to rescue and create the first five-gene-deleted strain of DEV. We evaluated its biological characteristics as an attenuated virus strain candidate in vitro. This virus retained the characteristics of the parent strain C-KCE of DEV and demonstrated attenuation in chicks, ducks, geese, mice, pigeons, quails, and white-rumped munias. Most importantly, following immunization, the gene-deleted strain did not induce intestinal pathogenicity in low-age SPF chicks or white-rumped munias, with no clinical observations of mortality, whereas the parent strain exhibited significant intestinal hemorrhage and necrosis. Our results provide a theoretical basis for the safety attenuation of the gene-deleted strain BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ. We report here for the first time an intestinal in vivo evaluation conducted in both target and non-target animal species.

BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ replicates in the intestinal organs of chicks, ducks, geese, quails, and white-rumped munias after immunization (Fig. 2F); however, the virus fails to replicate in the intestinal organs of mice and pigeons following intramuscular and intraperitoneal inoculation. According to the relevant literature, DEV is an acute, febrile, and septicemic infectious disease affecting poultry, waterfowl, and other birds of the order Anseriformes (Chen, et al., 2013; Wang, et al., 2022).

In this study, the data generated from microbiome analysis comprehensively demonstrated changes in the composition of the intestinal community following immunization with the DEV gene-deleted strain BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ. This study extensively investigated the correlations between the parent strain and the gene-deleted strain with respect to microbial community characteristics, providing new insights into the role of the gut microbiota in the immunization mechanism of the DEV gene-deleted strain.

In our previous study, we assessed the characteristic changes in the gut microbiota following DEV infection (Kong, Wu, Liao, Xie, Feng, Chen, Zhang and Xie, 2023). In alphaherpesviruses, TK, gI, and gE are significant virulence genes that influence viral pathogenicity (Ma, et al., 2024; Suzutani, 2000; Tang, et al., 2023). The use of BAC cloning technology in conjunction with Red/ET recombination technology (Chen, et al., 2019; Wang, Ge, Xu, Wang, Qiao, Gu, Liu, Liu and Hou, 2015) has proven to be an effective tool for studying DEV gene deletion strains. For the first time, we employed gene editing techniques to simultaneously knock out a combination of these virulence genes. By measuring the in vitro growth curves of the BAC-DEV-ΔTK-ΔgI/gE-ΔgG/gJ mutant virus in DEF cells and observing the diversity or composition of the gut microbiota in different animal post-immunization, we gained insights into the impact of these genetic modifications.

Research on ducks has indicated that the immune response in the gut to the DEV gene deletion strain shapes the composition of the gut microbiota. We found that the relative abundances of Firmicutes and Actinobacteria in the gut microbiota were positively correlated with immunization against the DEV gene deletion strain. In the parent strain immunization group, we observed that the relative abundance of Proteobacteria was almost undetectable, suggesting that the parent strain of DEV, after immunization, specifically targets certain members of the microbiota, leading to changes in the intestinal environment rather than specific groups. We speculate that one factor contributing to these changes is virulence genes, the gene-deleted strain altered the composition of the duck gut microbiota to some extent, indicating its beneficial role in immunity (Behera, et al., 2023; Menezes-Garcia, et al., 2020; Sun, et al., 2022). Regarding the impact on the diversity of the duck gut microbiota, we discovered a depletion of microbial diversity in the duck gut following immunization with the parent strain. In contrast, compared with ducks immunized with the gene-deleted strains, the diversity of the duck gut microbiota recovered, with significantly greater species richness.

During the assessment of the safety of the DEV gene-deleted strain, we enriched the data on the gut microbiota of non-target animals, including chicks and mice. In the animal experiments, we observed a significant increase in the relative abundance of Proteobacteria following immunization with the gene-deleted strains. Proteobacteria predominantly colonize microecological systems such as the animal gut and oral cavity, where they participate in the maintenance of host physiological balance and immune function (Kitamoto, et al., 2020; Sun, et al., 2021). In contrast, after immunization with the parent strain, the richness of the gut microbial population was lower than that with the gene-deleted strains, which, on the one hand, exacerbated animal mortality. Changes in the gut microbiota are associated with disease and, in some cases, may have contributed to increased animal mortality (Bäumler and Sperandio, 2016; Sánchez, et al., 2017; Zhang, et al., 2024b). In mouse gut tissues, no viral replication was detected. Lactobacillus and related genera are found in almost all gut tissues (Shen, et al., 2023; Xiao, et al., 2020). Interestingly, after immunization with the parent strain, the abundance of the gut microbiota in mice decreased, whereas gut microbial homeostasis in the gene-deleted strain group changed significantly. Our research has several limitations that warrant mention. The small sample size for gut microbiota analysis may limit the richness of our findings. Long-term sampling in animal experiments is essential to further enrich characterize the gut microbiota.

Conclusion

In conclusion, significant alterations in both the gut microbiota and morphological phenotypes were observed in various animals immunized with the DEV vaccine strain. However, the ability of DEV gene-deleted strains to change and restore the gut microbiota of animals represents an important step in the study of attenuated DEV strains. The DEV gene-deleted strains, developed using bacterial artificial chromosomes in conjunction with Red/ET recombination technology, may offer a secure strategy for vaccine development. Understanding the impact of DEV vaccine strains on the gut microbiota and morphology of animals can provide valuable insights into the mechanisms underlying vaccine efficacy and potential side effects.

Data availability

All data generated or analyzed during this study are included in this manuscript, and the sequencing data are provided in the additional files. The sequence data have been uploaded to the Science Data Bank and received the DOI (https://doi.org/10.57760/sciencedb.27645, https://doi.org/10.57760/sciencedb.27640, https://doi.org/10.57760/sciencedb.27595).

Ethics statement

All experiments were carried out in strict accordance with Guidance on the operation of the Animals (Scientific Procedures) Act 1986 and associated guidelines. The use of animals in this study was approved by the South China Agricultural University Committee of Animal Experiments (approval ID: 2021b020).

Funding

This study was supported by the construction project of modern agricultural science and technology innovation alliance in Guangdong province (2026CXTD20), the China Agriculture Research System of MOF and MARA (CARS-42-13), the Talent Introduction Project of Anhui Science and Technology University (DKYJ202404), the Veterinary Science Peak Discipline Project of Anhui Science and Technology University (XK-XJGF002), the National Natural Science Foundation of China (32200760), the Shandong Provincial Natural Science Foundation (ZR2022QC160).

Consent for publication

Not applicable.

Disclosures

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

CRediT authorship contribution statement

Jie Kong: Writing – original draft, Visualization, Validation, Methodology, Investigation, Data curation, Conceptualization. Chenxin Han: Writing – review & editing, Validation, Investigation, Formal analysis. Guanming Shao: Supervision, Conceptualization. Keyu Feng: Supervision, Conceptualization. Chaoyi Song: Writing – review & editing, Conceptualization. Qingmei Xie: Writing – review & editing, Supervision, Project administration, Data curation.