Rescue of very virulent and mosaic infectious bursal disease virus from cloned cDNA: VP2 is not the sole determinant of the very virulent phenotype

Abstract

Many recent outbreaks of infectious bursal disease in commercial chicken flocks worldwide are due to the spread of very virulent strains of infectious bursal disease virus (vvIBDV). The molecular determinants for the enhanced virulence of vvIBDV compared to classical IBDV are unknown. The lack of a reverse genetics system to rescue vvIBDV from its cloned cDNA hampers the identification and study of these determinants. In this report we describe, for the first time, the rescue of vvIBDV from its cloned cDNA. Two plasmids containing a T7 promoter and either the full-length A- or B-segment cDNA of vvIBDV (D6948) were cotransfected into QM5 cells expressing T7 polymerase. The presence of vvIBDV could be detected after passage of the transfection supernatant in either primary bursa cells (in vitro) or embryonated eggs (in vivo), but not QM5 cells. Rescued vvIBDV (rD6948) appeared to have the same virulence as the parental isolate, D6948. Segment-reassorted IBDV, in which one of the two genomic segments originated from cDNA of classical attenuated IBDV CEF94 and the other from D6948, could also be rescued by using this system. Segment-reassorted virus containing the A segment of the classical attenuated isolate (CEF94) and the B segment of the very virulent isolate (D6948) is not released until 15 h after an in vitro infection. This indicates a slightly retarded replication, as the first release of CEF94 is already found at 10 h after infection. Next to segment reassortants, we generated and analyzed mosaic IBDVs (mIBDVs). In these mIBDVs we replaced the region of CEF94 encoding one of the viral proteins (pVP2, VP3, or VP4) by the corresponding region of D6948. Analysis of these mIBDV isolates showed that tropism for non-B-lymphoid cells was exclusively determined by the viral capsid protein VP2. However, the very virulent phenotype was not solely determined by this protein, since mosaic virus containing VP2 of vvIBDV induced neither morbidity nor mortality in young chickens.

FIG. 1

Amino acid comparison between the products of the different ORFs of the cDNAs of wild-type vvIBDV isolate D6948 and the cell culture-adapted classical isolate CEF94. The complete sequence of the D6948 proteins is given, while only those amino acids which differ from the D6948 sequence are given for CEF94. The nucleotide sequences of the A and B segments which were used to deduce the amino acid sequences can be found in the GenBank database under accession numbers AF240686 and AF194428 (A segments of D6948 and CEF94, respectively) and AF240687 and AF194429 (B segments of D6948 and CEF94, respectively). (A) Amino acid sequence encoded by the first ORF (VP5) of the A segments. (B) Amino acid sequence encoded by the second ORF (polyprotein) of the A segment. The VP4 of the polyprotein (underlined) is preceded by pVP2, while VP3 is located at the C terminus (Fig. 2). The putative cleavage sites between pVP2 and VP4 and between VP4 and VP3 (as suggested by Hudson et al. [18]) have been used. Only recently it was shown that the actual cleavage sites are most likely located between amino acids 512 and 513 (pVP2-VP4) and 755 and 756 (VP4-VP3) (27). (C) Amino acid sequence encoded by the single ORF of the B-segment (VP1/VPg). Dashes show where corresponding amino acids are missing. Amino acid changes which are found in all vvIBDV sequences are in boldface. Amino acid residues that are reported to be involved in adaptation to non-B-lymphoid cells are in italics.

FIG. 2

Schematic representation of the plasmids containing the full-length cDNA sequences of the A segment (pHB-60) and B segment (pHB-55) of the wild-type vvIBDV isolate D6948. The cDNA sequence is preceded by a T7 promoter sequence and is followed by the HDV ribozyme (HDR) and a T7 terminator. The different ORFs are represented by open boxes. UTR, untranslated region.

FIG. 3

Autoradiogram of an SDS-PAGE analysis of a coupled in vitro transcription/translation reaction. (A) Full-length A-segment plasmids of attenuated classical IBDV isolate CEF94 (pHB-36W, lane 1) and wild-type vvIBDV isolate D6948 (pHB-60, lane 2). (B) Full-length B-segment plasmids of CEF94 (pHB-34Z, lane 1) and D6948 (pHB-55, lane 2). (C) Full-length A-segment plasmid of CEF94 (pHB-36W, lane 1) and full-length A-segment plasmid of the classical attenuated IBDV isolate in which either pVP2 (pHB36-vvVP2; lane 2), VP3 (pHB36-vvVP3; lane 3), or VP4 (pHB36-vvVP4; lane 4) has been exchanged. The positions of the viral proteins are on the right. The sizes of the marker proteins (Rainbow marker; Amersham) are on the left.

FIG. 4

Detection of IBDV using the VP3 and VP4 polyclonal antiserum. Samples of supernatant from different plasmid transfections were used to infect QM5 or primary bursa cells. After infection the QM5 cells were incubated for 24 h, while the primary bursa cells were incubated for 48 h. IBDV proteins were visualized by performing an IPMA.

FIG. 5

Single-step growth curves of CEF94, rCEF94, and srIBDV-CADB. QM5 cells were infected with IBDV (MOI = 5) for 1 h beginning at time zero, washed three times, and covered with fresh medium. At different time points samples were taken from the supernatant and the amount of IBDV (TCID₅₀ per milliliter) was determined. The TCID₅₀ at each time point is the mean of three independent experiments; error bars represent standard deviations.

FIG. 6

Schematic representation of the construction of a mosaic PCR fragment having a middle part originating from D6948 cDNA (shaded boxes) and flanking parts originating from CEF94 cDNA (open boxes). For the construction of pHB-36-vvVP2 we first generated PCR fragments VP2a, VP2b, and VP2c. These PCR fragments were purified and subsequently used as templates in a fusion PCR, yielding PCR fragment VP2d (17). This PCR fragment was subsequently purified and used to replace the corresponding part of the CEF94 A-segment cDNA, by using the indicated EcoRI and SacII restriction sites. pHB36-vvVP3 and -vvVP4 were generated by the same approach, using different primers to generate the PCR fragments and different restriction sites to introduce the PCR fragments into the full-length A-segment clone (see Materials and Methods and Table 1).

FIG. 7

Detection of mosaic IBDV using VP3 and VP4 polyclonal antiserum. Samples of supernatant of different plasmid transfections were used to infect QM5 and primary bursa cells. After infection, the QM5 cells were incubated for 24 h, while the primary bursa cells were incubated for 48 h (Fig. 4). In contrast to what was found for the negative control (mock infected), in which no B-lymphoid cells stained positive (data not shown), we found several stained B-lymphoid cells scattered throughout the culture tissue dish in the IPMA shown in the lower panel.