Genomic organization and mapping of transcription and translation products of the NADL-2 strain of porcine parvovirus.

The NADL-2 strain of PPV was cloned into pUC19 and independent infectious clones were sequenced. This permitted a correction of published sequences and to predict a cruciform structure as an alternative to the 5'-hairpin of the "-" strand. This 5'-end structural covariance is shared with other parvoviruses of the same group and two alternative sequences ("flip" and "flop") were present in the region of the cruciform. Transcript and translation product mapping allowed the prediction of the location of the different expression signals. The 5'-startpoints of the transcripts were located at nucleotides 225 and 2035, respectively, and the polyadenylation site at nucleotides 4829-4833. This indicated that the TATA boxes at 196-TATA and 2004-AATA and the 4813-AATAAA polyadenylation sequence would be functional. Alternative splicing of capsid gene (VP) transcripts (either 2280-AG/GT or 2313-AG/GT spliced with 2386-AG/GA), to maintain or remove the first AUG (at 2287) in the ORF, yielded two 2.9-kb mRNAs containing a nested set of protein-coding sequences (VP-1 and VP-2 with predicted molecular mass 80.9 and 64.3 kDa, respectively). Three nonstructural (NS) protein gene transcripts were identified. The 4.7-kb transcript was not spliced in the NS gene and was predicted to code for a 75.5-kDa protein (NS-1; published value of phosphorylated form 84 kDa). The splicing sites of two different 3.3-kb NS transcripts were analyzed. These transcripts were predicted to code for the NS-2 protein (18.1 kDa). Of the two NS-2 transcripts, one had also the VP-intron removed downstream of the NS-2 coding sequences. A 2.9-kb transcript would code for an NS-3 protein (12.4 kDa) although such a protein has not been described before. A flow chart of the information from the viral DNA to the viral proteins is presented and several differences, both for the NS and the VP genes, with closely related parvoviruses are noted.