Nuclear localization signal binding protein from Arabidopsis mediates nuclear import of Agrobacterium VirD2 protein

Abstract

T-DNA nuclear import is a central event in genetic transformation of plant cells by Agrobacterium. Presumably, the T-DNA transport intermediate is a single-stranded DNA molecule associated with two bacterial proteins, VirD2 and VirE2, which most likely mediate the transport process. While VirE2 cooperatively coats the transported single-stranded DNA, VirD2 is covalently attached to its 5' end. To better understand the mechanism of VirD2 action, a cellular receptor for VirD2 was identified and its encoding gene cloned from Arabidopsis. The identified protein, designated AtKAPalpha, specifically bound VirD2 in vivo and in vitro. VirD2-AtKAPalpha interaction was absolutely dependent on the carboxyl-terminal bipartite nuclear localization signal sequence of VirD2. The deduced amino acid sequence of AtKAPalpha was homologous to yeast and animal nuclear localization signal-binding proteins belonging to the karyopherin alpha family. Indeed, AtKAPalpha efficiently rescued a yeast mutant defective for nuclear import. Furthermore, AtKAPalpha specifically mediated transport of VirD2 into the nuclei of permeabilized yeast cells.

Figure 1

VirD2–AtKAPα interaction in the two-hybrid system. The indicated combinations bait/prey were achieved by introducing into the yeast strain Y153 the combinations of the following plasmids: VirD2/AtKAPα, pGBTD2 and pGADAtKAPα; VirD2ΔNLS/AtKAPα, pGBTD2ΔNLS and pGADAtKAPα; lamin C/AtKAPα, pLAM5 and pGADAtKAPα. No prey combinations included the corresponding bait and pGAD424 plasmid without insert. Protein–protein interaction was determined by the β-galactosidase assay on a nitrocellulose filter as described (21). All other experimental conditions were as described for the two-hybrid screen of the cDNA library.

Figure 2

VirD2–AtKAPα interaction in vitro. Protein–protein interaction was assayed by coimmunoprecipitation with anti-VirD2 antibody as described. Lane 1, VirD2+AtKAPα; lane 2, VirD2ΔNLS+AtKAPα; lane 3, VirD2 alone; lane 4, AtKAPα alone; lane 5, VirD2+AtKAPα+preimmune serum. Solid and open arrowheads indicate the positions of the radioactively labeled AtKAPα and VirD2 proteins, respectively. Protein molecular mass standards are indicated on right in kDa. Note that the electrophoretic mobility of VirD2 is 56 kDa as described (39), although its molecular mass calculated on the basis of the known DNA sequence is 49,579 Da (40).

Figure 3

Nucleotide sequence of the AtKAPα cDNA (A) and alignment of its deduced amino acid sequence with the yeast SRP1 protein (B). The GenBank accession numbers for AtKAPα and SRP1 are U69533 and M75849, respectively. Alignment was performed by the clustal algorithm (41). Regions of identity are indicated by narrow boxes, gaps introduced for alignment are indicated by dashes. Wide open boxes indicate the amino-terminal basic regions and wide shaded boxes indicate eight “arm” motifs (32). Arrowhead indicates the position of the cdc2 kinase site absent in AtKAPα and SRP1 but found in animal karyopherin α proteins (38).

Figure 4

Complementation of the srp1-31 mutation by AtKAPα. The yeast srp1-31 strain was transformed with either pAtKAPα, pGADΔG, or pSRP1. The transformants were plated on selective medium and grown for 2 days at either 25°C (Upper) or 37°C (Lower).

Figure 5

AtKAPα promotes nuclear import of VirD2 in permeabilized yeast cells. Nuclear import assays were performed as described in Materials and Methods. The following combinations of cytosolic fractions and the fluorescently labeled protein were added to the permeabilized srp1-31 cells: A and B, VirD2+cytosol from cells expressing pGADΔG; C and D, VirD2+cytosol from cells expressing pAtKAPα; E and F, VirD2ΔNLS+cytosol from cells expressing pAtKAPα; G and H, VirD2+cytosol from cells expressing pAtKAPα+ 2 μg of a synthetic VirD2 NLS peptide; I and J, VirD2+cytosol from cells expressing pSRP1. A, C, E, G, and I, fluorescein-labeled protein; B, D, F, H, and J, 4′,6-diamidino-2-phenylindole staining. The one-letter code amino acid sequence of the VirD2 NLS is KRpredddgepseRKReR with basic residues of the first and second domains of the bipartite signal indicated in uppercase (11). (Bar = 50 μm.)