Introduction

Africa is experiencing a worsening hunger trends, with malnutrition affecting 20% of the population in 2023 and projected to reach 25.7% by 20301,2. Sustainable poultry production can help achieve the United Nations Sustainable Development Goals (SDGs 1, 2, and 3) by providing affordable animal protein and improving food security1. Poultry sector is expected to continue growing as demand for meat and eggs is driven by the growing human populations, rise in incomes and urbanization3,4. Nigeria’s population is expected to reach 398 million by 2050, leading to a significant increase in the demand for livestock products5. To meet this growing demand, Nigerian livestock producers are anticipated to invest in strategies that enhance production and productivity. According to estimates from the Food and Agriculture Organization (FAO), between 2012 and 2050, the production of poultry meat and eggs in Nigeria is projected to increase by 532 and 657%, respectively6. However, Newcastle disease (ND) is a major problem for poultry farmers in Nigeria, where the disease spreads rapidly resulting in high mortality rates within a short period of time7. In a participatory community appraisal on factors affecting poultry production in rural households across Nigeria, ND was ranked first as the major constraint to rural poultry production and most of these outbreaks are not reported especially in rural chickens8. These challenges translate into substantial economic losses for farmers, undermining livelihoods and further straining food security and nutrition efforts9,10. These economic losses range from degradation of birds’ value from mortality and culling, to job losses and management costs9,11.

The ND outbreaks in Nigeria are usually attributed to vaccine failure as commonly used vaccines are not effective12,13. Besides, poor management practices and limited awareness of diagnostic services among poultry farmers also contribute to delayed veterinary interventions, and ND outbreaks14. Newcastle disease outbreak is caused by a single serotype of Avian orthoavulavirus 1 (AAvV-1) commonly known as ND virus (NDV), yet controlling the disease has become a big challenge in Nigeria due to genetic diversity and uniqueness of the virulent NDV strains12. The NDV is an enveloped virus with a linear, single-stranded, non-segmented, negative-sense RNA in its genomic makeup belonging to the genus Orthoavulavirus, family Paramyxoviridae and order Mononegavirales15. The virus genome encodes six major proteins among which the fusion (F) and hemagglutinin-neuraminidase (HN) proteins, forming the spikes of the virus, are the most important for viral infectivity16. The cleavage site of the F protein is considered to be the virulent determinant of the virus. Some studies postulate that the HN and large polymerase (L) proteins also play a role in the pathogenicity of the virus16,17. The virus has been shown to downregulate cellular immune response, which explains why viral shedding continues despite the presence of protective antibody titers18. By disrupting interferon signalling and inducing immune cell apoptosis, NDV evades host immunity. However, progress in understanding immune evasion in avian viruses has been challenged by limited reagents and experimental tools19.

The clinical signs and pathological lesions of ND are not absolutely pathognomonic20, therefore, confirmatory diagnosis such as serology and reverse transcription polymerase chain reaction (RT-PCR), are needed to confirm NDV infection. Newcastle disease virus is only said to be virulence if detected with F gene multi-basic cleavage site in poultry while detecting similar multi basic cleavage site in non-poultry birds does not necessarily indicate virulent NDV21. This distinction is critical for precise pathotyping, epidemiological assessments, and the implementation of targeted disease control strategies across different species in endemic areas. As NDV continues to infect a broader range of hosts, including humans, its zoonotic potential points out the importance of such distinctions for both animal and public health22. For instance, most human NDV isolates have 112RRQRKF117 fusion protein cleavage site, particularly the first lethal case (EF555096) had glutamine (Q) at position 114 substituted with lysine (K), and such genetic variations may enhance zoonotic potential23.

Nigeria and to some extent West African region are known for diverse class II NDV genotypes, including the newly emerging genotypes XIV, XVII, and XVIII, in addition to II, V, and VII. Although, sub-genotype XVII.2 is only documented in Nigeria24,25. However, genotype XIV is particularly more prevalent in Nigeria, representing the most frequently isolated NDV strains in the country12. In the recent period, molecular characterization of seven NDV isolates from vaccinated flocks in Northern Nigeria identified all the isolates as sub-genotype XIV.213. And more recently, the same sub-genotype was detected in 10 sequenced isolates from live bird markets across the country, demonstrating active circulation of sub-genotype XIV.2 in Nigeria26. A key finding from NDV sequence analysis revealed a notable K78R mutation within the A2 antigenic epitope from the F protein in all the Nigerian sub-genotype XIV.2 isolates, which was previously linked to evade monoclonal antibodies generated after LaSota vaccination13,27. Studies have associated the high prevalence of NDV to dry climatic conditions, with heat stress potentially compromising poultry immunity and/or increasing viral adaptation26,28,29. This is evident in Northern Nigeria’s hot climate, where a recent virulent NDV outbreak at a major breeder farm in Maiduguri highlighted the growing threat to poultry production, raising more concerns about the efficacy of the vaccine against the virulent NDV. In-silico analysis has proven to be a reliable and efficient method for evaluating vaccine efficacy against such pleomorphic viruses30,31,32. Therefore, this study aimed to characterize the isolated NDV from a recent outbreak in vaccinated birds in Nigeria and employed computational prediction models to elucidate its immunomodulatory strategies in order to inform the rational design of an effective vaccine against the virulent virus.

Materials and methods

Study area

This study was conducted in Maiduguri to identify factors contributing to persistent NDV outbreaks in Nigeria. Temperature remains high throughout the year in the region, especially during the daytime, with daily extreme temperatures reaching up to 47 °C in April33. The poultry sector remains vital to Maiduguri’s food security, providing income and employment, but outbreaks of diseases like ND worsen the vulnerability of smallholder farmers and low-income households already affected by conflict and displacement34,35,36,37. For over three decades, NDV has consistently been the most significant challenge in poultry production in Maiduguri and now worsened by a sharp decline due to the insurgency14,38,39.

Study design

The present study was cross-sectional40, investigating the alarming NDV infection in one of the poultry farms located in Maiduguri, Nigeria. The 522 mixed-sex Cornish Cross breeder chickens were introduced to the farm at 19 weeks of age, with records indicating they had received all necessary vaccines, including vaccines against ND. The outbreak was presented to the Poultry Unit of the University of Maiduguri Veterinary Teaching Hospital (UMVTH) on November 30, 2023, when the birds were 23 weeks old with ND-like clinical signs (Suppl. Figure 1&2). This led to a detailed investigation, including an in-silico analysis using an alternative immunoinformatic approach to examine the potential factors contributing to the vaccine failure. As the first immunoinformatic analysis of NDV sub-genotype XIV.2, the study explores the virulent NDV’s potential to modulate host immune responses using predicted T-cell epitopes.

Sample collection, laboratory analyses, and virus isolation

Tissues from five carcasses, tracheal and cloacal swabs, as well as blood from five sick birds were collected in various test-specific containers/tubes. A comprehensive diagnostic workup was performed, which comprises postmortem examination as previously described41, histopathology as described previously42, bacterial culture as previously described43, and hemagglutination inhibition (HI) test following the method outlined in44. The swab samples (n = 5), and tissues (n = 5) were pooled into two separate 1.8 mL cryotubes and stored at −20 °C before processing at the National Veterinary Research Institute, Vom, Nigeria. Each of the pooled samples was filtered by a 0.22 μm PES syringe filter and 0.2 mL each was inoculated into 10-day-old specific antibody negative (SAN) embryonated chicken eggs and the virus was isolated as previously described45. The isolates were impregnated on a fast technology for analysis of nucleic acids (FTA) cards and transported at room temperature to Konkuk University Zoonotic Disease Research Center in Seoul, Republic of Korea for further analysis.

RNA extraction, amplification, and sequencing

The RNA was extracted from the FTA disks following the QIAamp® Viral RNA Mini Kit’s protocol (QIAGEN, Germany) as described in prior literature46,47. The RNA was amplified using Qiagen One-step RT-PCR kit under a thermocycling condition with primers designed by48 targeting partial F gene (286 bp fragments). For rapid pathotyping of the virus, the purified RT-PCR product was sequenced using Sanger dideoxynucleotide cycle sequencing method (Macrogen, Republic of Korea). The Sequence-Independent, Single-Primer-Amplification (SISPA) was employed to prepare the RNA of one sample for next generation sequencing (NGS) as described previously49. Briefly, the RNA was treated with DNase I and reverse transcribed into cDNA, which was then converted into double-stranded cDNA and purified. The purified dsDNA was amplified using sequence-independent PCR and the final purified product was sent to Bionics, Republic of Korea for sequencing using the Illumina platform.

Phylogenetic analyses

The raw sequence reads were trimmed by removing adapters and low-quality bases using BBDuk (v38.84) with a minimum quality threshold set at 30 and minimum length at 50 bp. De novo and reference-based assemblies of the genome sequences were conducted using Geneious Prime software version 2025.0.3. (https://www.geneious.com/). For reference-based assembly, the trimmed reads were mapped to the reference sequence using Minimap (v2.24) with default options50. Only the full F gene consensus sequence was assembled, likely due to RNA degradation from prolonged transport on FTA cards. The full F gene sequence was analyzed using the BLAST query function of the NCBI Virus database, and the top 200 BLAST hits were retrieved on March 20, 2025. Multiple sequence alignment was carried out using MAFFT (v7.490)51after which the retrieved sequences were trimmed to match the full F gene length. To eliminate redundancy and focus on distinct evolutionary lineages, sequences with over 98% nucleotide identity were removed using ElimDupes (https://www.hiv.lanl.gov/content/sequence/elimdupesv2/elimdupes.html), and the dataset was reduced to 23 unique sequences. Sixteen other sequences of interest (JF950510, AF309418, KT445901, EF201805, AY741404, MH996952, KJ544861, MG018205, AB605247, OR230625, JX519467, JX012096, PP788551, MT409238, MH996921, MH392227) were added and realigned with the 23, making 39 in total. Phylogenetic analysis was performed on the aligned dataset using RAxML (v8.2.12) under the GTR-GAMMA nucleotide substitution model with 1,000 bootstrap replicates52. The resulting phylogenetic tree was visualized using iTOL version 6 (https://itol.embl.de).

Protein sequence retrieval and epitope prediction

Sub-genotype XIV.2 lineage whole genome sequence (OQ920548) was used as reference to extract F, HN, and L proteins collectively for the peptide prediction, based on its close evolutionary relationship (98%) with the isolated sample’s F gene of the study (Fig. 1). Cytotoxic T-lymphocyte (CTL) epitopes were predicted using the ANN 4.0 tool, while helper T-lymphocyte (HTL) epitopes were predicted with IEDB’s NN-align 2.3 available on the Immune Epitope Database (IEDB) server (https://www.iedb.org/). Peptide lengths were set at 8 and 9-mers for CTL epitopes, and 15-mers for HTL epitopes. Due to the limited availability of chicken Major Histocompatibility Complex (MHC) allele data within IEDB, human leukocyte antigen (HLA) alleles were used as a proxy. This approach is supported by previous studies showing that certain structural features, particularly within the peptide-binding groove, are conserved between chicken MHC and human HLA31. Peptides demonstrating strong binding affinity (IC50 < 500 nM) to HLA alleles were shortlisted for further analysis including assessment of conservation and antigenicity. For MHC class I and II epitope selection, peptides with an IC50 value greater than 500 nM were excluded. The remaining peptides were ranked in ascending order based on their rank score. Among the top 100 candidates, 49 conserved epitopes from the F protein, 36 from the HN protein, and 67 from the L protein across sub-genotype XIV.2 strains were further analyzed for antigenicity using the VaxiJen 2.0 server (https://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html).

Mutation analysis between NDV field and vaccine strains

A comparative analysis of the sub-genotype XIV.2 peptide sequences was conducted with other NDV virulent strains and vaccine strains (Genotypes VII, XVII, and XVIII, LaSota, HitchnerB1, Komarov, Mukteswar) using CLC Main Workbench 24v. This analysis aimed to identify amino acid substitutions and mutations across the various genotypes, providing some information about the genetic variations and potential impact on immunogenicity and vaccine efficacy.

Modeling and molecular docking of chicken allele and peptides

The B-F allele (BF2*2101) of chicken MHC class I was retrieved from UniProt (ID: Q95601) and modeled using AlphaFold 3 (Google DeepMind). The predicted structure was validated using PROCHECK (Ramachandran plot analysis), ProSA (z plot) to ensure structural accuracy and reliability. Additionally, the total number of positively (Arg + Lys) and negatively (Asp + Glu) charged residues were assessed using ProtParam tool available on the ExPASy (Expert Protein Analysis System) server (https://web.expasy.org/protparam/) to evaluate the electrostatic properties of the protein, which can influence folding, solubility, and molecular interactions. Predicted epitope structures were generated using PEP-FOLD3, and molecular docking was performed with chicken MHC class I using AutoDock Vina. Binding interactions were assessed based on docking scores. To validate the docking approach, two positive control peptides, IDWFDGKE and RRALREGY known to bind the chicken MHC class I alleles53,54, were included. These control peptides served as reference points to compare binding affinities and interaction patterns with the test peptides.

Immune response simulation

To investigate the immune response dynamics against the predicted NDV peptides, computational simulations were conducted using the C-immsim platform (https://www.violinet.org/c-immstim/). The selected MHC class I epitopes were joined using AAY (Ala-Ala-Tyr) linkers55 to enhance epitope processing and presentation. Two distinct scenarios were modeled to assess the efficacy of the peptides in LaSota-Komarov prime-boost vaccine simulation strategy. In the first scenario, a homologous antigen exposure was simulated, wherein 1000 units of LaSota antigen were introduced at day 0 (prime) and Komarov at day 21 (boost), followed by 100 units of LaSota virus at day 30. In the second scenario, a heterologous antigen exposure was simulated, employing the same prime-boost regimen with 1000 units of LaSota antigen at day 0 and Komarov at day 21, but challenging with 100 units of sub-genotype XIV.2 virus at day 30. For both scenarios, simulation parameters, including a viral multiplication factor of 0.2 and infectivity of 0.6, were consistently applied to reflect moderate viral replication. The simulations ran for 60 days to capture the temporal dynamics of key immune components, including antigen levels, antibody titers (IgM, IgG1, IgG2), cytokine levels (IFN-γ, IL-4, IL-12, TGF-β, TNF-α, IL-10, IL-6, IFN-β, IL-18, IL-23), cytotoxic T cell (TC) populations (active, duplicating, resting, anergic), and B cell populations (total B cells, B memory cells, B isotype IgM, IgG1, IgG2). Data were plotted to visualize immune parameter changes over time, and the two scenarios were compared to assess the effectiveness of the LaSota-Komarov prime-boost against homologous and heterologous simulated antigen exposure. The simulation outputs were plotted to visualize the predicted immune dynamics, and the two antigen exposure scenarios were compared to assess the potential effectiveness of the LaSota-Komarov epitope combination against homologous and heterologous NDV antigenic variants.

Results

History and clinical signs

The outbreak led to 100% morbidity despite prior ND vaccination, with 25 deaths (Suppl. Figure 1 A-B) and 2 birds slaughtered on the first day, and eventually reaching 95% mortality over 14 days (Suppl. Figure 3). Clinical signs included inappetence (Suppl. Figure 1C), dehydration, dyspnea, dullness, greenish diarrhea (Suppl. Figure 1D), raising concerns about vaccine failure or a more virulent strain. The veterinarian who visited the farm also had fever, headache, fatigue and severe conjunctivitis (Suppl. Figure 4&5) four days after coming in contact with the sick birds.

Gross, histopathological and laboratory findings

The postmortem findings showed petechial hemorrhages at the proventricular junction (Suppl. Figure 1E) and congested trachea (Suppl. Figure 1F). The histopathology showed a severe congestion of the pulmonary vessels, necrosis of the cardiac myocytes, intestinal epithelial desquamation and liver vacuolar degeneration (Suppl. Figure 6–9). The bacterial culture yielded mild growth of Escherichia coli and Enterobacter species. The HI result showed NDV antibody titer between 24 and 26.

RNA amplification and sequencing

The RNA was successfully amplified using the NDV pathotyping primers, and the partial F gene sequence identified the epidemic virus as sub-genotype XIV.2. The complete F gene was successfully assembled and submitted to GenBank under accession number PV339969.

Phylogenetic analyses

Phylogenetic analysis of the full F gene (PV339969) showed a close genetic relationship with other Nigerian genotype XIV isolates (98%), demonstrating high sequence identity and clustering within the same lineage (Fig. 1). In contrast, NDV sequences from other genotypes originating from other countries exhibited some genetic divergence, especially with the sub-genotype XIV.2 (Suppl. Figure 10). Genotype XIV was further grouped with NDV genotypes XVII and XVIII from Nigeria, suggesting the co-circulation of diverse NDV lineages within Nigeria. This phylogenetic distinction underscores the regional evolution of sub-genotype XIV.2. Conversely, genotype IV (purple); MH996952.1 (Nigeria, 1973) and AY741404.1 (Herts, UK, 1933) clustered separately from the study sample clade, highlighting distinct evolutionary pathways (Fig. 1).

Fig. 1
figure 1

Midpoint rooted phylogenetic tree showing genotypes and bootstrap values above 70 in different colors. The scale bar represents the rate of nucleotide substitutions per site. A maximum likelihood phylogenetic tree was generated from the sample full F gene alignment with all the retrieved NDV strains in this study. The sample ▲PV339969 of this study clustered closely with genotype XIV sequences (pink) but shows divergence from the commonly used NDV vaccine strains (blue).

Full size image

Prediction and selection of epitope candidates

The predicted peptide 01; RRKRFVGA within the F protein of the sub-genotype XIV.2 exhibited the highest antigenicity score (1.157) with an intermediate-affinity IC50 value of 422.16 nM. Meanwhile, TAFGPQITS also displayed a strong antigenicity score (1.1015) and a more favorable IC50 value of 244.93 nM. For the HN protein, YTDPYPLVF (IC50: 5.69 nM) and MAKSSYKPR (IC50: 7.93 nM) stood out, exhibiting both low IC50 values and high antigenicity, suggesting their strong antigenic potential. Similarly, within the L protein, RAKFAWLHI (IC50: 4.44 nM, antigenicity 2.0126) and SVAVSYSLK (IC50: 12.36 nM, antigenicity 1.4707) emerged as promising candidates (Table 1). Subsequently, following MHC class II epitope prediction, we identified conserved epitopes and compared them with the previously selected MHC class I epitopes. Interestingly, the selected MHC class I epitopes also overlapped with these conserved MHC class II regions, reinforcing their potential immunogenicity. The overlapping MHC class I and class II epitopes are presented in Table 2.

Table 1 Top-performing MHC class I conserved epitopes from the F, HN, and L proteins, predicted using the IEDB server.
Full size table
Table 2 Top performing MHC-2 conserved and overlapped epitopes with selected MHC-1 epitopes from the F, HN, and L proteins.
Full size table

Mutation analysis of selected epitopes against available vaccine strains

Mutation analysis of six selected peptide candidates from sub-genotype XIV.2 against four common NDV vaccine strains (LaSota, Hitchner-B1, Komarov, and Mukteswar) revealed significant amino acid differences (marked as red) (Table 3). Notably, Peptide 01, corresponding to the cleavage site motif, exhibited substantial changes, particularly in LaSota and Hitchner-B1, where the virulent polybasic motif (RRKRFVGA) mutated to the avirulent monobasic motif (RQGRLIGA). Komarov and Mukteswar also showed mutations in this region (RQKRFIGA and RQRRFIGA, respectively). Other peptides displayed single amino acid substitutions across the vaccine strains, including TAFGPQITS to TVFGPQITS (Peptide 02), YTDPYPLVF to YTDPYPLIF (Peptide 03, except for Mukteswar which matched the field strain), MAKSSYKPR to MAKSSYKGR (Peptide 04), RAKFAWLHI to TAKFAWLHI (Peptide 05), and SVAVSYSLK to NVAVSYSLK (Peptide 06, except for Komarov which mutated to DVAVSYSLK). Moreover, the comparative analysis of peptide sequences across Genotypes XIV.2, VII, XVII, and XVIII reveals both similarities and key differences, primarily centered on the cleavage site motif and a single amino acid substitution. Identical sequences were observed for Peptide 05 (RAKFAWLHI), and Peptide 06 (SVAVSYSLK) across all four genotypes. Conversely, differences were identified in four peptides: Peptide 01 (RRKRFVGA) mutated to RQKRFIGA in Genotypes XVII and XVIII, similar to Komarov, while Genotype VII exhibited a distinct mutation (RRKRFIGA). Likewise, Peptide 02 (TAFGPQITS) was mutated to TVFGPQITS in Genotypes VII, XVII, and XVIII, mirroring the vaccine strains. Additional variations were observed in Genotype VII for Peptide 03 (YTDPYPLVF to YTDPYPLIF) and Peptide 04 (MAKSSYKPR to MAKSSYKPQ).

Table 3 Mutation analysis of sub-genotype XIV.2 peptides against common NDV vaccines and genotypes VII, XVII & XVIII. The differences in the amino acid sequences in each strain is highlighted in bold.
Full size table

Modelling and molecular docking of chicken allele and peptides

AlphaFold3 successfully generated a high-quality tertiary structure of the BF2*2101 chicken MHC class I allele (Fig. 2A). The predicted model’s reliability and structural stability were validated by a Z-score of −6.61 (Fig. 2B), and Ramachandran plot analysis with 99% residues within the favored and allowed regions, indicating a well-folded and energetically favorable protein conformation with minimal steric clashes. Only 1% of residues fell into disallowed regions, which are within acceptable limits for modeled structures, supporting the overall accuracy of the predicted model (Fig. 2C). In addition, residue charge distribution was analyzed using the ExPASy ProtParam tool. The protein contains 41 negatively charged residues (Asp + Glu) and 37 positively charged residues (Arg + Lys). This balanced distribution of charged residues suggests no significant electrostatic imbalance that could destabilize the protein structure. Moreover, the presence of a nearly equal number of oppositely charged residues contributes to intramolecular salt bridges and electrostatic interactions that are likely to enhance structural stability and proper peptide-binding orientation in the MHC binding groove. These findings collectively indicate a high-quality, reliable model suitable for downstream molecular docking and immunoinformatic analyses.

Fig. 2
figure 2

Structural validation of the predicted BF2*2101 chicken MHC class I allele model.

Full size image

The molecular docking analysis demonstrated that genotype II (LaSota)-derived peptides exhibited generally stronger binding affinities to the chicken MHC class I allele (BF2*2101) compared to sub-genotype XIV.2 peptides, except for SVAVSYSLK (−7.9 kcal/mol) from XIV.2, which showed the highest affinity among all tested peptides (Table 4). The positive control peptides, IDWFDGKE (−7.0 kcal/mol) and RRALREGY (−6.9 kcal/mol), exhibited moderate binding affinities, suggesting that several LaSota-derived peptides had comparable or superior MHC binding. Among the LaSota peptides, NVAVSYSLK (−7.7 kcal/mol), MAKSSYKGR (−7.5 kcal/mol), and TVFGPQITS (−7.3 kcal/mol) consistently showed strong interactions with BF2*2101, surpassing most XIV.2 peptides. On the contrary, three out of four sub-genotypes XIV.2 peptides; RRKRFVGA (−6.3 kcal/mol), TAFGPQITS (−6.9 kcal/mol), and MAKSSYKPR (−5.9 kcal/mol), demonstrated weaker binding affinities, indicating less favorable interactions with the MHC complex. Notably, MAKSSYKGR (−7.5 kcal/mol) from LaSota exhibited significantly higher affinity than its XIV.2 counterpart MAKSSYKPR (−5.9 kcal/mol), highlighting the impact of single amino acid substitutions on MHC binding. These findings suggest strain-specific variations in epitope presentation, which may influence immune recognition and vaccine efficacy.

Table 4 Molecular Docking scores of genotype II (LaSota) and Sub-genotype XIV.2 peptides with chicken MHC class I (BF2*2101).
Full size table

Immune response simulation

The immune simulation analysis revealed that the LaSota-Komarov prime-boost antigen exposure strategy (LaSota on day 0 and Komarov on day 21) induced a transient antigen peak at day 0, corresponding to the LaSota prime (compare with Suppl. Figure 11), followed by a second peak at day 21 due to the Komarov boost. Following simulated introduction of LaSota virus-derived antigens on day 30, antigen levels showed a slight increase, followed by a sharp rise on day 35, and a significant decline by day 42. The immune response displayed typical features of a primary antibody response, with an initial peak in IgM (compare with Suppl. Figure 11B) followed by a marked increase in IgG1 and IgG2 titers after the Komarov boost (Fig. 3A). Upon subsequent simulated challenge with the LaSota virus, IgG1 and IgG2 levels surged sharply around day 40, indicating a robust secondary immune response. Cytokine analysis (Fig. 3B) revealed an increase in several cytokines, notably IFN-γ and IL-2, after each of the antigen exposure. The IL-2 levels experienced a significant spike following the virus antigen introduction. In parallel, active and proliferating TC increased after each vaccine construct and virus challenge simulation, with a marked rise in active TC cells following the predicted virus antigen exposure (Fig. 3C). B cell dynamics showed a gradual increase in total B cells, B memory cells, and IgM-producing B cells up to day 30, with a subsequent rise in IgG1-producing B cells and memory B cells after virtual virus antigen exposure (Fig. 3D). This suggests active B cell involvement in the immune response. The LaSota-Komarov prime-boost simulation offers effective protection against homologous LaSota (genotype II) predicted virus antigens, as evidenced by the simulated antibody response, cellular activation, and partial viral antigen clearance (Fig. 3A).

Fig. 3
figure 3

Simulated immune response to LaSota-Komarov prime-boost antigen exposure followed by LaSota virus simulated challenge.

Full size image

A second simulation was conducted to evaluate the immune response following the same LaSota-Komarov prime-boost antigen exposure and simulated introduction of sub-genotype XIV.2 virus-derived antigens at day 30. The prime-boost strategy successfully induced significant increases in IgG1 and IgG2 titers (Fig. 4A). However, upon introduction of the sub-genotype XIV.2 simulated virus antigens, no notable increase in IgG1 and IgG2 levels was observed (also supported by Suppl. Figure 12). Instead, there was a sharp and sustained rise in antigen levels, indicating active viral replication. Cytokine profiles (Fig. 4B) showed that both the prime and boost simulated vaccinations, as well as the simulated virus challenge, activated the immune system. While IFN-γ levels decreased after day 37, IL-2 levels exhibited a substantial spike after simulated virus antigen introduction. Additionally, there was an increase in active and proliferating TC cells after each vaccination and virus challenge model, with a marked rise in active TC cells following predicted virus antigen exposure (Fig. 4C). However, there was no significant increase in total B cells, B memory cells, and IgG1 and IgG2-producing B cells after the sub-genotype XIV.2 virus antigen input (Fig. 4D), compared to the response observed in the homologous model (Fig. 3D). These results demonstrate that while the LaSota-Komarov prime-boost regimen induces a robust immune response against the homologous LaSota antigens, it does not confer cross-protective immunity against the heterologous sub-genotype XIV.2 antigens based on the computational analysis.

Fig. 4
figure 4

Simulated immune response to LaSota-Komarov prime-boost antigen exposure followed by sub-genotype XIV.2 virus simulated challenge.

Full size image

Discussion

The severe NDV outbreak in a well-managed broiler breeder farm in Maiduguri, which resulted in huge production losses and disruption of the poultry value chain, has raised concern about vaccine efficacy12. Clinical signs and lesions observed in the affected birds indicated an acute disease condition, while the mild growth of E. coli and Enterobacter species could suggest contamination or early secondary bacterial infection56. Despite having antibody titers between 24 and 26, which generally indicate protective immunity against ND57; the birds still contracted the disease, as confirmed by RT-PCR. The inability of RT-PCR to pinpoint the exact pathotype responsible for the outbreak also emphasizes the urgent need for more advanced, rapid diagnostic tools capable of on-site pathotyping58,59. Practically, genome sequencing remains essential for accurate NDV pathotyping, as ICPI may be unreliable in some strains60. Meanwhile, detection of various NDV strains in mammals reveals their wide host range and genetic diversity, raising concerns about zoonotic risks and public health implications22. Poultry farmers, veterinarians, and laboratory personnel are at high risk of having conjunctivitis and flu-like illness, as observed with the veterinarian in this study. Preventive measures and awareness campaigns are essential to mitigate NDV transmission among animals and humans, especially with frequent undiagnosed respiratory infections in humans61.

Molecular characterization of the isolated virus revealed distinctive features, including a polybasic motif (112RRRKRF117) at the F gene cleavage site, with the Q114K substitution (P3)62. These unique molecular features may influence host immune modulation of the virus and contribute to its pathogenicity. The phylogenetic tree confirmed the NDV isolate’s distinct evolutionary profile, classifying it within sub-genotype XIV.2 with similar Nigerian isolates, which differ from commonly used ND vaccine strains (Fig. 1). This divergence could translate to antigenic mismatch, as field trail of LaSota and VG/GA vaccines have proven ineffective against genotype XVII in commercial chickens in Nigeria63. Yet, Audu and colleagues found that the modified LaSota clone, which shares the same Q residue at the P3 position, provided 100% protection against genotype XVII of NDV. The neutral Q residue at P3 enhances NDV replication and pathogenicity by promoting membrane fusion, and contributes to viscerotropism rather than immune evasion, which remains poorly understood16,19,62,. Relatively, the attenuated yellow fever vaccine strain (17D) infects host cells more efficiently than the virulent wild strain (Asibi), enhancing cytosolic RNA delivery and inducing a robust cytokine response64. In another experiment, sub-genotype XIV.2 viruses exhibited relatively low hemagglutination (1:32) and HI (108.375) titers compared to other NDV genotypes, suggesting potential implications for immune response32. Therefore, virulent NDV strains with Q residue at P3 may promote greater cell-to-cell spread and stronger immune responses compared to strains lacking Q, as seen in this study (sub-genotype XIV.2; 112RRRKRF117). After all, the NDV virulence is modulated by multiple genetic factors60,65,66.

The overlap between MHC class I and II binding sites in sub-genotype XIV.2 epitopes (Table 2) reflects the role of T-cell immunity in clearing paramyxovirus infections3167. Likewise, same peptides from LaSota generated B cell responses (Fig. 3A&D), further supported by AAY linker’s enhancement of antigen presentation and CTL activation. However, the moderate MHC-I binding affinities (Table 1) may impact simulated vaccine efficacy, as peptides with IC₅₀ ≤ 50 nM elicit stronger CTL responses68. Moreover, comparative analysis with other virulent genotypes revealed distinct mutation patterns in sub-genotype XIV.2, which has been recently reported to exhibit the highest antigenic diversity among NDV strains32. The mutation of alanine (A) to valine (V) in Peptide 02 (TAFGPQITS to TVFGPQITS) in both the vaccine strains and other virulent genotypes (Table 3) may influence the peptide’s binding affinity to MHC molecules, impacting immune recognition69. Similarly, the substitution of valine (V) to isoleucine (I) in Peptide 01 (RRKRFVGA to R****IGA) and Peptide 03 (YTDPYPLVF to YTDPYPL*F) may alter the peptide’s conformation and interaction with immune receptors in MHC-I but not necessarily in MHC-II70,71. A study showed that B-cell immunodominant epitopes (IDEs) of HN protein of the LaSota vaccine strain are highly conserved in genotypes I to XIII of class II72 but sub-genotype XIV.2 may differ, as seen in peptides 03 and 04.

Structural analysis validated the BF2*2101 model, confirming stability through proper folding and a balanced electrostatic distribution (Fig. 2). Yet, sub-genotype XIV.2 peptides bound weaker than LaSota’s peptides (Table 4), possibly limiting cross-protection. The observed differences in binding affinities may also imply that LaSota-derived peptides could be more effectively presented to CD8 + T cells, potentially contributing to improved simulated immune response (Fig. 3). But the immune simulation analysis with LaSota-Komarov prime-boost antigen exposure did not confer cross-protective immunity against the heterologous sub-genotype XIV.2 simulated virus antigens (Fig. 4A), despite inducing a robust immune response against the homologous model (Fig. 3A). In the homologous model, the response was characterized by a transient antigen peak, a significant increase in IgG1 and IgG2 titers, and a surge in cytokines, including IFN-γ and IL-2 (Fig. 3) while in the heterogeneous model, there is no increase in the IgG1 and IgG2 levels, leading to a sharp and prolonged increase in antigen levels without adequate B cells response to clear the simulated virus antigens (Fig. 4A). This outcome indicates poor B cell activation, potentially due to weaker antigenic stimulation by the lentogenic LaSota strain or immunosuppressive cytokines like IL-10 and TGF-β (Fig. 4B). The transient increase in anergic CD8⁺ T cells implies T cell exhaustion (Fig. 4C), possibly due to virulent antigen exposure or immune checkpoint upregulation, limiting cytotoxic function. However, the observed increase in total CD8⁺ T cells (TC) and CD4⁺ T helper (Th) cells suggests ongoing immune activation (Fig. 4C), yet this fails to translate into effective viral antigen clearance, potentially due to functional impairment of effector cells or inadequate cytokine signaling (IFN-g, IL-2) (Fig. 4B). Additionally, virulent NDV suppresses immunity by inducing apoptosis in immune cells, impairing cell function, and disrupting interferon responses19,73,74.

Overall, these findings provide an insight for exploring more effective NDV vaccine strategies. Formulating heterologous prime-boost NDV vaccine regimens adjuvanted with cytokines and immune modulation holds promise for enhancing cellular and humoral immunity beyond what is achievable with conventional vaccination. While this study provides strategic guidance for understanding the predicted immune interactions of NDV sub-genotype XIV.2, there are some limitations. First, the use of a single NDV isolate may limit broader applicability of such findings across other NDV genotypes. Second, the analysis is based on in-silico predictions and simulations, which, though insightful, may not capture fully the complexity of in-vivo host-pathogen interactions. Third, the lack of in-vivo or laboratory-based experimental validation prevents definitive conclusions regarding the immunogenicity or protective efficacy of the predicted peptides. Therefore, future work incorporating animal trials and wet-lab immunological assays is essential to validate the immunoinformatic predictions75.

Conclusion

This study presented a striking illustration of NDV’s severity in Nigeria, by characterizing the virulent NDV sub-genotype XIV.2 isolate and identifying antigenic epitopes and structural interactions that may affect immune response. The amino acid substitutions (R114Q, V118I, A220V) in key MHC class I and II epitopes, clearly differentiated sub-genotype XIV.2 from vaccine strains and other NDV. Furthermore, structural modeling and molecular docking showed LaSota peptides had stronger binding affinity to the chicken MHC allele than sub-genotype XIV.2 peptides, potentially affecting cross-immunity. Meanwhile, immune simulation suggested that the LaSota-Komorov prime-boost regimen may not provide effective protection against sub-genotype XIV.2 viral antigen simulation challenge, further indicating possible immune evasion by the virulent sub-genotype XIV.2. These findings stress the need for improved ND vaccine strategies in Nigeria and provides a foundation for future experimental studies.