Characterisation of the Vitis vinifera PR10 multigene family

Abstract

Background: Genes belonging to the pathogenesis related 10 (PR10) group have been studied in several plant species, where they form multigene families. Until now, such an analysis has not been performed in Vitis vinifera, although three different PR10 genes were found to be expressed under pathogen attack or abiotic stress, and during somatic embryogenesis induction. We used the complete genome sequence for characterising the whole V. vinifera PR10 gene family. The expression of candidate genes was studied in various non-treated tissues and following somatic embryogenesis induction by the auxin 2,4-D.

Results: In addition to the three V. vinifera PR10 genes already described, namely VvPR10.1, VvPR10.2 and VvPR10.3, fourteen different PR10 related sequences were identified. Showing high similarity, they form a single cluster on the chromosome 5 comprising three pseudogenes. The expression of nine different genes was detected in various tissues. Although differentially expressed in non-treated plant organs, several genes were up-regulated in tissues treated with 2,4-D, as expected for PR genes.

Conclusions: PR10 genes form a multigene family in V. vinifera, as found in birch, apple or peach. Seventeen closely related PR10 sequences are arranged in a tandem array on the chromosome 5, probably reflecting small-scale duplications during evolution. Various expression patterns were found for nine studied genes, highlighting functional diversification. A phylogenetic comparison of deduced proteins with PR10 proteins of other plants showed a characteristic low intraspecific variability. Particularly, a group of seven close tandem duplicates including VvPR10.1, VvPR10.2 and VvPR10.3 showed a very high similarity, suggesting concerted evolution or/and recent duplications.

Figure 1

Organisation of V. vinifera PR10 related sequences in a single cluster on chromosome 5. All sequences are located between the positions 599,000 and 681,000, on the + or the - strand. Nucleotide positions are as referenced on the Genoscope website.

Figure 2

Exon-intron structures of the seventeen PR10 related sequences. A: length of exons (bold dashes) and introns (thin dashes). B: frequency of nucleotides at the 5' and 3' putative splicing sites of introns. The size of letters is proportional to the nucleotide frequency at each position. The numbers indicate nucleotide position relatively to the first nucleotide of intron 1 (number 1 at the 5' end) or to the first nucleotide of exon 2 (number 1 at the 3' end). VvPR10.1, VvPR10.2 and VvPR10.3 respectively correspond to s16, s10 and s12.

Figure 3

Percentages of nucleotide similarity between the CDS of the seventeen sequences. High percentages of nucleotide similarity are highlighted in red. VvPR10.1, VvPR10.2 and VvPR10.3 respectively correspond to s16, s10 and s12. The values were obtained from sequence alignments on ClustalW.

Figure 4

Predicted structures for PR10 related sequences of V. vinifera. Only sequences having canonical transcription signals are shown. The arrow indicates the transcription starting site (+1). Predicted exons are represented as black boxes and deduced introns as dashed boxes. Given positions respectively correspond to the start of the TATA-box, the transcription starting site, the first nucleotide of the CDS, the last nucleotide of exon 1, the first nucleotide of exon 2, the last nucleotide of the CDS and the first nucleotide of the poly-A signal. Predictions were performed using the GeneFinding program. VvPR10.1, VvPR10.2 and VvPR10.3 respectively correspond to s16, s10 and s12.

Figure 5

Sequence alignment of deduced PR10 proteins. The P-loop and Bet v 1 signature are framed. A star (*) marks the amino-acids implied in possible ribonucleasic activity (E102, E149 et Y151). Alignments were performed with ClustalW. A V. vinifera specific Bet v 1 motif was determined: G-[DG]-[VA]-L-x(4)-E-[SY]-[IL]-[CSATV]-[HY]-[ED]-x-[KST]-x-[VE]-x(3)-[GNDS]-G(2)-[CS]-x(2)-K-x(2)-[SK]-X-Y. In PR10.8 and PR10.10, the last aa varies in the P-loop motif (E and T, respectively); in PR10.9, E₁₀₂ is replaced by D₁₀₂, the 4^th aa in the P-loop motif is E, and the Bet v 1 motif presents 4 differences (out of 34 aa) at the positions 1, 6, 23 and 29; in PR10.6, there is one difference (out of 33 aa) in the Bet v1 motif (position 12).

Figure 6

Three-dimensional structure of V. vinifera PR10.1 represented by a ribbon diagram. The structure was predicted on an automated comparative protein modeling server using SWISS-MODEL.

Figure 7

Phylogenetic relationships between V. vinifera PR10 proteins and representative PR10 proteins from Betula pendula, Lupinus luteus, Malus domestica and Prunus persica. GenBank accession numbers are as follows: Betula pendula Betv1.0401 (CAA54482), Betv1.0601 (CAA54484), Betv1.1101 (CAA54694), Betv1.1301 (CAA54696), Betv1.1701 (CAA96539) and Betv1.1801 (CAA96540); Lupinus luteus LlPR10.1A (CAA56298), LlPR10.1B (CAA56299), LlPR10.2A (AAF77633), LlPR10.2B (AAF77634) and LlPR10.2E (AAP37978); Malus domestica Mald1.01 (AAX18288), Mald1.02 (AAX18291), Mald1.03A (AAX18313), Mald1.04 (AAX18294), Mald1.05 (AAX18296), Mald1.06A (AAX18299), Mald1.07 (AAX18307) and Mald1.08 (AAX18310); Prunus persica Prup1.01 (ACE80940), Prup1.02 (ACE80942), Prup1.03 (ACE80944), Prup1.04 (ACE80946), Prup1.05 (ACE80948) and Prup1.06A (ACE80952). The V. vinifera Grip61 gene (CAB85634) was included as an outgroup representative of another Bet v 1 subfamily [3]. The NJ-tree was generated with the Phylo_win program. The bootstrap value is given for each node.

Figure 8

Expression of V. vinifera PR10 genes. A: tissue-specific expression; R = roots, S = stems, L = leaves, F = flowers. B: expression during secondary somatic embryogenesis induction; E = non-treated somatic embryos at the cotyledonary stage, E_2,4-D = calli obtained from embryos treated with 2,4-D. gDNA = genomic DNA. The length of amplified sequences are given in bp.