Both genome segments contribute to the pathogenicity of very virulent infectious bursal disease virus

Abstract

Infectious bursal disease virus (IBDV) causes an economically significant disease of chickens worldwide. Very virulent IBDV (vvIBDV) strains have emerged and induce as much as 60% mortality. The molecular basis for vvIBDV pathogenicity is not understood, and the relative contributions of the two genome segments, A and B, to this phenomenon are not known. Isolate 94432 has been shown previously to be genetically related to vvIBDVs but exhibits atypical antigenicity and does not cause mortality. Here the full-length genome of 94432 was determined, and a reverse genetics system was established. The molecular clone was rescued and exhibited the same antigenicity and reduced pathogenicity as isolate 94432. Genetically modified viruses derived from 94432, whose vvIBDV consensus nucleotide sequence was restored in segment A and/or B, were produced, and their pathogenicity was assessed in specific-pathogen-free chickens. We found that a valine (position 321) that modifies the most exposed part of the capsid protein VP2 critically modified the antigenicity and partially reduced the pathogenicity of 94432. However, a threonine (position 276) located in the finger domain of the virus polymerase (VP1) contributed even more significantly to attenuation. This threonine is partially exposed in a hydrophobic groove on the VP1 surface, suggesting possible interactions between VP1 and another, as yet unidentified molecule at this amino acid position. The restored vvIBDV-like pathogenicity was associated with increased replication and lesions in the thymus and spleen. These results demonstrate that both genome segments influence vvIBDV pathogenicity and may provide new targets for the attenuation of vvIBDVs.

Fig 1

Organization of IBDV segments A and B and construction of recombinant A and B segments derived from pA94 and pB94. Segments A and B of 94432 cloned into the pUC18 plasmid under the control of the T7 promoter were designated pA94 and pB94, respectively. The positions and identities of 94432 nucleotides (NCRs) or amino acids (coding region) that differentiate 94432 segments A and B from those of typical vvIBDVs are given in filled boxes below each segment map. (A) Segment A. The small ORF codes for VP5, the transmembrane domain (aa 73 to 90) of which is shaded. The large ORF codes for VP2, VP4, and VP3. The hypervariable domain of VP2 (aa 206 to 350) (light shading) contains two major hydrophilic domains (domains A and B) (dark shading). Changes introduced into pA94 resulted in five constructs. The positions and identities of the vvIBDV nucleotide(s) and/or amino acid(s) introduced are given in open boxes below each segment map. (B) Segment B. The first double-headed arrow shows the region where the self-guanylylation site (Gs) of IBDV RdRp is predicted to be positioned (72). The second double-headed arrow indicates the domain of VP1 interacting with VP3 (DI-VP3) (18). The 3-dimensional structure of the IBDV RdRp allowed us to determine that the polypeptide chain can be divided into an N-terminal domain (N-ter); the central polymerase domain, which contains the “finger” (F1 and F2), “palm” (P1 and P2), and “thumb” (T) subdomains; and a C-terminal domain (C-ter) (25, 45). Changes introduced into pB94 by site-directed mutagenesis resulted in a “vvIBDV-like” B plasmid (pBvv) and in five mutated derivatives of plasmid pBvv, each retaining 1 aa typical of 94432. The positions and identities of the vvIBDV amino acids introduced are given in open boxes below each segment map.

Fig 2

Location of amino acid 321 in the VP2 crystal structure. (A) (Left) Amino acid 321 (magenta), located in the exposed hydrophilic peak B of the hypervariable domain, is shown in the crystal structure of the VP2 monomer (11) along with aa 222 and 223 (orange), which were identified as belonging to the Mab4 epitope (43, 44). Note that the variation in the number of spheres for a given amino acid is representative of the complexity of its side chain. (Right) The VP2 monomer is shown in a putty representation, where color and width are proportional to the crystallographic temperature values. The colors range from dark blue, corresponding to a B-factor minimum of 34, to red, corresponding to a B-factor maximum of 99. Amino acid 321 is indicated, along with aa 222 and 223. The amino and carboxy termini are labeled N and C, respectively. (B) Locations of aa 321, 222, and 223 in the VP2 trimers.

Fig 3

mc94432 is attenuated compared to typical vvIBDVs. The pathogenicity of mc94432 was first compared with those of typical vvIBDV strains BD3 and 89163 or with that of the parental strain, 94432. For the protocols of the animal experiments, see Table 2, experiments 1 and 2. (A) Percentages of cumulated mortality at 5 days postinfection. (B) Mean individual symptomatic index scores for surviving chickens from 1 to 10 days postinfection. Simplified representations of the recombinant segments A and B (based on those in Fig. 1) used to generate the indicated recombinant viruses are shown. Percentages of mortality were compared to that of the mock-infected group by using Fisher's exact test (*, P < 0.05). Multiple comparisons of symptomatic index scores were performed with the Kruskal-Wallis test. Treatments sharing the same lowercase letter do not differ significantly at a confidence level (P) of ≤0.05.

Fig 4

Segment A of 94432 is partially responsible for the reduced pathogenicity of this virus. The contribution of segment A to the reduced pathogenicity of 94432 was evaluated. For the protocol and genetic makeup of the viruses used in this animal experiment, see Table 2, experiment 2, and Fig. 1A. (A) Percentage of cumulated mortality at 5 days postinfection. (B) Mean individual symptomatic index scores from 1 to 10 days postinfection. Statistical analysis was performed as described for Fig. 3, and statistical differences are represented as in Fig. 3.

Fig 5

Segment B of 94432 is mainly responsible for the reduced pathogenicity of this virus. The contribution of segment B to the reduced pathogenicity of mc94432 was evaluated. For the protocol and genetic makeup of the viruses used in this animal experiment, see Table 2, experiment 4, and Fig. 1B. (A) Percentage of cumulated mortality at 5 days postinfection, compared to that of the mock-infected group by using Fisher's exact test (*, P < 0.05). (B) Mean individual symptomatic index scores from 1 to 10 days postinfection. (C) Quantification of negative B strands in the thymus and spleen at 4 days postinfection (n, 4 to 5 chickens per group). (D) Histological lesion scores in the thymus and spleen at 4 days postinfection (n, 5 chickens per group). The box plots show the median flanked by the upper and lower quartiles. The outer bars show the range of values. Statistical analysis was performed as described for Fig. 3, and statistical differences are represented as in Fig. 3.

Fig 6

The valine residue at VP2 position 321 of segment A contributes to reducing the pathogenicity of 94432. The pathogenicities of mc94432, AvvB94, and A94[V321A]B94 were evaluated in SPF chickens. For the protocol and genetic makeup of the viruses used in this animal experiment, see Table 2, experiment 3, and Fig. 1A. (A) Percentage of cumulated mortality at 5 days postinfection. (B) Mean individual symptomatic index scores for surviving chickens from 1 to 10 days postinfection. See Materials and Methods for the method of calculation. (C) Histological lesion scores in the thymus and spleen at 4 days postinfection (n, 4 to 5 chickens per group). Statistical analysis was performed as for Fig. 3 and 5. Data are presented, and statistical differences are represented, as in Fig. 3 and 5.

Fig 7

The threonine residue at VP1 position 276 of segment B mainly contributes to the reduced pathogenicity of 94432. The pathogenicities of mc94432, AvvBvv, AvvBvv-61, AvvBvv-112, AvvBvv-276, AvvBvv-363, and AvvBvv-763 were assessed in SPF chickens. For the protocol and genetic makeup of the viruses used in this animal experiment, see Table 2, experiment 5, and Fig. 1B. (A) Percentage of cumulated mortality at 5 days postinfection. (B) Mean individual symptomatic index scores for surviving chickens from 1 to 10 days postinfection. (C) Quantification of negative B strands in the thymus and spleen at 4 days postinfection (n, 4 to 5 chickens per group). (D) Histological lesion scores in the thymus and spleen at 4 days postinfection (n, 5 chickens per group). Statistical analysis was performed as for Fig. 3 and 5. Data are presented, and statistical differences are represented, as in Fig. 3 and 5.