Molecular Basis of the Attenuation Exhibited by Molecularly Cloned Highly Passaged Chicken Anemia Virus Isolates

Abstract

Chimeric virus experiments indicated that the pathogenicity and monoclonal antibody reactivity differences between two molecularly cloned, highly passaged chicken anemia virus isolates could be attributed to the VP1 amino acid change at residue 89. The introduction of this change into a pathogenic cloned low-passage isolate was not sufficient to cause attenuation.

Chicken anemia virus (CAV) has a circular, single-stranded 2.3-kb DNA genome contained within an icosahedral capsid, 25 nm in diameter (9), and is the only member of the genus Gyrovirus of the virus family Circoviridae (6). The virus genome encodes 1 structural (VP1) and 2 nonstructural (VP2 and VP3) proteins (Fig. 1a) (5). To date, all naturally occurring CAV isolates belong to the same serotype, and all are pathogenic when tested experimentally (2). We previously reported that molecularly cloned virus isolates that were selected from the Cuxhaven-1 (Cux) CAV isolate, which had received 310 cell culture passages (P310) in MDCC-MSB1 cells, showed variation with regard to pathogenicity and reactivity with a neutralizing monoclonal antibody (MAb), 2A9 (7, 8). Of these, the attenuated P310-cloned isolate 34, which reacts weakly with MAb 2A9, differed from the pathogenic P310-cloned isolate 33, which reacts strongly with MAb 2A9, at two amino acid residues, namely, VP1 residue 89 and VP3 residue 41. In this study the significance of the VP1 amino acid change at residue 89 as a determinant of pathogenicity was investigated by producing and biologically characterizing chimeric and in-vitro-mutagenized viruses.

FIG. 1.

Production of chimeric and mutated viruses. (a) CAV genome organization showing locations of open reading frames encoding VP1 to VP3 and restriction sites used for constructing chimeric viruses. (b) Chimeric and reconstructed CAVs derived from restriction fragments specified by P310-cloned isolates 33 and 34. (c) Chimeric and mutated CAVs derived from restriction fragments specified by cloned low-passage Cux and P310-cloned isolate 34 or in vitro mutagenesis. The infectivity titers are shown for each of the chimeric, reconstructed, and mutated viruses obtained by transfection.

The cloned low-passage Cux isolate and the P310-cloned isolates 33 and 34 were produced as described previously (4, 7). Indirect immunofluorescence (IIF) was used to determine the reactivities of the cloned, mutated, and chimeric CAV isolates with CAV-specific MAb 2A9 (7). Chimeric CAV replicative form (RF) DNAs were constructed from BamHI-PstI (BP), PstI-StuI (PS), and StuI-BamHI (SB) restriction fragments, which were produced by restricting cloned P310 RF 33 and 34 DNAs and were purified from agarose gel after electrophoretic fractionation. Following ligation, mixtures containing approximately equimolar amounts of the three fragments were used to transfect MDCC-MSB1 cells to generate the chimeric and reconstructed cloned isolate 34 (Fig. 1a and b). An additional chimeric virus isolate was produced by ligating the SB fragment derived from P310 RF 34 to the complementary BS fragment, derived from the recombinant CAV plasmid pCAA5 (4), which specifies the pathogenic cloned low-passage Cux isolate (Fig. 1c). In this case, PCR methods were used to amplify the BS and SB fragments prior to ligation and transfection. The PCR-ligation-PCR method was used to introduce a site-specific mutation by which amino acid 89 in VP1 of the cloned low-passage Cux isolate was changed from threonine to alanine to produce the VP1 aa89 Cux mutant (1). Experimental infections of 1-day-old specific-pathogen-free chicks were used to evaluate the pathogenicities of chimeric and mutated CAV isolates as described previously (3). Nucleotide sequence determinations were largely performed by automated sequencing with the Prism Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems). Manual sequencing, performed with the fmol^R DNA Cycle Sequencing System (Promega) and with gels run at 70°C, was used to resolve GC-rich sequences present in the noncoding regions.

Over the 502-nucleotide noncoding sequence, the cloned RFs specifying P310-cloned isolates 33 and 34 differ at nine nucleotides, and the P310-cloned isolate 34 contains a single nucleotide deletion at position 2232. The effects of the individual VP1 and VP3 amino acid differences and the cumulative nucleotide differences in the noncoding regions that exist between P310-cloned isolates 33 and 34 were investigated by using a chimeric virus approach (Fig. 1b). Pools of each chimeric virus and the reconstructed P310-cloned isolate 34, produced after 5 to 6 cell culture passages after transfection of MDCC-MSB1 cells, had infectivity titers in the range of 10^6.75 to 10^8.0 50% tissue culture infective doses (TCID₅₀s)/0.1 ml (Fig. 1b), and IIF testing showed that only the 34PS:33SB:34BP chimeric virus, which contained the VP1 change at amino acid 89 exhibited by P310-cloned isolate 33, produced positive staining with high dilutions (1:40,000) of MAb 2A9. The parental P310-cloned isolates 33 and 34 differed markedly in their pathogenicities as indicated by differences in the proportions of chicks that were anemic, the mean hematocrit values, and the clinical scores (Table 1). The similarity in the results obtained with the reconstructed P310-cloned isolate 34 and the 33PS:34SB:34BP and 34PS:34SB:33BP chimeric viruses indicated that neither the VP3 amino acid change nor the noncoding nucleotide changes were responsible for the pathogenicity difference between the parental viruses. In contrast, from a comparison of the 34PS:33SB:34BP chimeric virus and the reconstructed 34PS:34SB:34BP virus it was evident that the VP1 amino acid change at residue 89 was largely responsible for the pathogenicity difference between P310-cloned isolates 33 and 34. The infectivity titers of the chimeric CuxBS:34SB and reconstructed CuxBS:CuxBS viruses, produced after 6 passages following transfection, were 10^5.75 TCID₅₀/0.1 ml and 10^6.0 TCID₅₀/0.1 ml, respectively (Fig. 1c), whereas that of the Cux mutant with a change at VP1 amino acid 89, which grew slowly after transfection and required 14 passages to produce a working pool, was 10^6.75 TCID₅₀/0.1 ml. The reconstructed CuxBS:CuxBS virus and the Cux mutant virus with a change at VP1 amino acid 89 both produced strong IIF staining with high MAb 2A9 dilutions (1:40,000), whereas the chimeric CuxBS:34SB virus did not, indicating that the single VP1 change was not sufficient by itself to reduce MAb reactivity. Whereas the reconstructed CuxBS:CuxBS virus was highly pathogenic in terms of its ability to induce anemia and gross lesions, the CuxBS:34SB chimeric virus was markedly attenuated (Table 1). In contrast, the Cux mutant with a change at VP1 amino acid 89 displayed substantial pathogenicity, especially in terms of clinical score.

TABLE 1.

Expt	Virus	No. of chicks	Infectivity titer (Log10 TCID50)	Mean hematocrit value	No. (%) of chicks positive for anemiaa	No. of chickens with pale bone marrow classb		No. of chickens with thymus atrophy classb		Clinical score
						1+	2+	1+	2+
A	None	8	0	32.10	0 (0)	0	0	0	0	0.00
A	Cloned low-passage Cux	11	6.75	16.18	11 (100)	2	9	5	6	3.36
A	P310-cloned isolate 33	12	7.0	26.67	7 (58.3)	5	3	3	0	1.17
A	P310-cloned isolate 34	12	7.0	32.75	0 (0)	1	0	2	0	0.25
A	34PS:34SB:34BP (34 reconstruct)	12	7.0	32.41	0 (0)	0	0	2	0	0.17
A	34PS:33SB:34BP (VP1 change)	11	6.75	23.82	7 (63.63)	4	5	3	1	2.00
A	33PS:34SB:34BP (VP3 change)	9	7.0	31.55	0 (0)	1	0	1	0	0.22
A	34PS:34SB:33BP (noncoding changes)	12	7.0	30.50	1 (8.33)	2	1	2	0	0.50
B	None	12	0	34.50	0 (0)	0	0	0	0	0.00
B	CuxBS:CuxSB (Cux reconstruct)	12	5.75	21.58	11 (91.67)	8	3	5	3	1.92
B	CuxBS:34SB (5 VP1 aa changes)	12	6.0	30.42	0 (0)	0	0	0	0	0.00
B	Cux VP1 aa89 mutant	12	6.0	27.00	4 (33.0)	4	3	2	2	1.33

^aBirds are considered to have anemia if the hematocrit value is less than 27%.

^b+1 and +2, severity of the lesion.

It was concluded that the introduction of the VP1 amino acid 89 change into the pathogenic, cloned low-passage Cux isolate was not sufficient in itself to cause attenuation. However, when this change was introduced in combination with four additional VP1 changes at residues 75, 125, 141, and 144, as present in the SB fragment of P310-cloned isolate 34, pronounced attenuation resulted. This finding indicated that these amino acid changes, which were selected by multiple cell culture passages and which result in reduced reactivity with MAb 2A9, may be responsible for the attenuation exhibited by highly passaged Cux-1 CAV. This research and that by Yamaguchi et al. (10), who showed that VP1 amino acid 394 was a determinant of the pathogenicity exhibited by Japanese CAV isolates, may allow the development of highly attenuated viruses, which may be worthy of evaluation as live attenuated vaccines.

Nucleotide sequence accession numbers. The genome sequences of P310-cloned isolates 33 and 34 have been deposited in GenBank under accession numbers AJ297684 and AJ29785, respectively.