* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Polar location and functional domains of the Agrobacterium
Survey
Document related concepts
G protein–coupled receptor wikipedia , lookup
Cell membrane wikipedia , lookup
Cytokinesis wikipedia , lookup
Protein phosphorylation wikipedia , lookup
Signal transduction wikipedia , lookup
Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup
Protein moonlighting wikipedia , lookup
Protein domain wikipedia , lookup
Magnesium transporter wikipedia , lookup
Endomembrane system wikipedia , lookup
Western blot wikipedia , lookup
Trimeric autotransporter adhesin wikipedia , lookup
Transcript
Molecular Microbiology (2002) 43(6), 1523–1532 Polar location and functional domains of the Agrobacterium tumefaciens DNA transfer protein VirD4 Renu B. Kumar1 and Anath Das1,2* 1 Department of Biochemistry, Molecular Biology and Biophysics, and 2Plant Molecular Genetics Institute, University of Minnesota, 1479 Gortner Avenue, St Paul, MN 55108, USA. Summary Agrobacterium tumefaciens VirD4 is essential for DNA transfer to plants. VirD4 presumably functions as a coupling factor that facilitates communication between a substrate and the transport pore. To serve as a coupling protein, VirD4 may be required to localize near the transport apparatus. In a previous study, we observed that several constituents of the transport apparatus localize to the cell membranes. In this study, we demonstrate that VirD4 has a unique cellular location. In immunofluorescence microscopy, cells probed with anti-VirD4 antibodies had foci of fluorescence primarily at the cell poles, indicating that VirD4 localizes to the cell pole. Polar location of VirD4 was not dependent on T-DNA processing, the formation of the transport apparatus and the presence of other Vir proteins. VirD4 is an integral membrane protein with one periplasmic domain. The large cytoplasmic region contains a nucleotide-binding domain. To investigate the role of these domains in DNA transfer, we introduced mutations in virD4 and studied the effect of a mutation on substrate transfer. A deletion of most of the periplasmic domain as well as the alterations of glycine 151 to serine and lysine 152 to alanine led to the complete loss of DNA transfer, indicating that both domains are essential for substrate transfer. Subcellular localization of the mutant proteins indicated that both the periplasmic and the nucleotide-binding domains are required for polar localization of VirD4. The periplasmic domain mutant VirD4D36–61 was distributed throughout the cell membrane, whereas the nucleotide binding site mutant VirD4G151S localized to sites other than the cell poles. Polar location of VirD4 suggests a role for the cell pole in DNA transfer. Accepted 5 December, 2001. *For correspondence. E-mail [email protected]; Tel. (+1) 612 624 3239; Fax (+1) 612 625 5780. © 2002 Blackwell Science Ltd Introduction The transfer of macromolecules across membranes and kingdoms is an essential biological process. Type IV transport is used by bacteria to deliver macromolecules to prokaryotic and eukaryotic cells (Covacci et al., 1999; Christie, 2001). Agrobacterium tumefaciens uses type IV transport to transfer DNA to plant cells. Escherichia coli transfers plasmids by conjugation by the type IV mechanism. Human and animal pathogens, e.g. Helicobacter pylori, Bordetella pertussis, Brucella suis and Rickettsia prowazekii, use a similar method to deliver pathogenesisrelated effector proteins and other molecules. DNA transfer from Agrobacterium to plants causes the crown gall tumour disease (Zupan et al., 2000). The transferred (T-)DNA and the genes essential for its transfer, the virulence (vir) genes, are encoded within the large Ti(tumour-inducing) plasmid. DNA transfer requires the virA, virB, virD, virE and virG genes (Stachel and Nester, 1986). The Vir proteins are primarily responsible for the processing of the DNA, its transfer to the plant cell and its integration into the plant nuclear genome. The T-DNA exits the bacterium through a transport apparatus composed primarily of the VirB proteins. VirB2 and VirB5 form a T-pilus that presumably promotes host–recipient interaction (Fullner et al., 1996; Lai and Kado, 1998; SchmidtEisenlohr et al., 1999). The T-pilus is found primarily at one cell pole (Lai et al., 2000). Five proteins, VirB6– VirB10, are postulated to be the core constituents of the transport apparatus (Das and Xie, 2000). Most of these proteins interact with one another and form various protein complexes (Anderson et al., 1996; Spudich et al., 1996; Baron et al., 1997; Beaupre et al., 1997; Das et al., 1997; Das and Xie, 2000). Subcellular localization studies have indicated that three VirB proteins localize to a few sites on the bacterial membranes (Kumar et al., 2000). These sites may represent the transport apparatus. The VirD proteins catalyse processing of the T-DNA and its transfer. VirDl and VirD2 introduce a nick at the T-DNA border sequences producing the single-stranded T-strand DNA composed of the bottom strand of the T-DNA (Zupan et al., 2000). VirD3 is not required for tumour formation (Vogel and Das, 1992). VirD4, an essential DNA transfer protein, is postulated to mediate communication between a substrate and the transport pore (Cabezon et al., 1997). VirD4 is an inner membrane protein with two membranespanning domains near the N-terminus, and both the N- 1524 R. B. Kumar and A. Das VirD4 by immunofluorescence (IF) and immunoelectron (IE) microscopy. In IFM, wild-type A. tumefaciens labelled with anti-VirD4 antibodies had a punctated appearance (Fig. 1). The bacteria had a single or few foci of high fluorescence on the membranes, indicating that VirD4 localized to very few sites (Fig. 1C). In bacteria with a single or two foci of fluorescence, the protein invariably localized to the cell poles. Quantitative analysis indicated that about 95% of the foci localized to the cell poles (Table 3). About 50% of the cells exhibited multiple foci of fluorescence. In these bacteria, the protein localized to other areas of the membranes including the poles. Strains with mutations in the vir genes were analysed to study the requirement of T-DNA processing and transport apparatus assembly in the polar localization of VirD4. Neither T-strand DNA synthesis nor the formation of a transport apparatus is required for targeting VirD4 to the cell poles, because a non-polar deletion in virD2 and a deletion in virB had no effect on VirD4 localization (Fig. 1E and F). In bacteria that expressed only virD4, the protein localized to the cell pole, indicating that no other Vir protein is essential for its proper subcellular localization (Fig. 1G). In IEM, gold particles representing VirD4 localized to the membranes and were found in clusters (Fig. 1 and Table 1). Clustering of gold particles suggests complex formation. Proteins of the bacterial chemoreceptor complex and the T-DNA transport apparatus exhibit a similar pattern in IEM (Maddock and Shapiro, 1993; Kumar et al., 2000). About 85% of gold particles localized to the membranes, and ª 70% of those were in the clusters, suggesting that VirD4 forms a protein complex (Table 1). High fluorescence at the VirD4 localization sites in IFM (Fig. 1) is consistent with this hypothesis. and C-terminus of the protein are cytoplasmic (Okamoto et al., 1991; Das and Xie, 1998). Amino acid residues 13–30 and 68–86 comprise the membrane-spanning domains. Near the N-terminus of the large cytoplasmic domain, there is a Gly-x-Gly-x-Gly-xn-Lys (x = any residue) sequence at residues 151–174 that bears weak resemblance to the Walker A box sequence, a sequence motif found conserved in nucleotide-binding proteins (Balzer et al., 1994). The role of this sequence in T-DNA transfer is not known. The VirD4 and VirB proteins are essential for DNA transfer (Beijersbergen et al., 1992). These proteins are found conserved in the type IV transport family (Lessl et al., 1992; Christie, 2001). Like VirD4, its homologues in the E. coli plasmids F, RP4 and R388 are essential for conjugal transfer of plasmid DNA (Balzer et al., 1994). These proteins can functionally substitute for one another in the transfer of only a subset of conjugal plasmids (Cabezon et al., 1994; Hamilton et al., 2000). Two regions of the VirD4 homologues that correspond to motifs A and B of the Walker box have the highest sequence conservation. These motifs of the RP4 homologue TraG are essential for the conjugal transfer of the plasmid (Balzer et al., 1994). The C-terminal one-quarter of the proteins exhibits low sequence similarity. Recent analysis of the structure of the cytoplasmic domain of R388 TrwB, a VirD4 homologue, indicated that it is a hexamer with a central 20 A0 channel that can function as a pore at the inner membrane (Gomis-Ruth et al., 2001). The role as a coupling protein and/or a pore-forming function of the VirD4 family of proteins will require them to localize near the transport apparatus. In the present study, we determined the subcellular location of VirD4 and identified its essential domains. We demonstrate that VirD4 localizes to the cell pole and that two domains are essential for targeting it to the proper site on the cell membrane. Functional domains of VirD4 VirD4 is a bitopic membrane protein with a short periplasmic domain near its N-terminus (Das and Xie, 1998). The cytoplasmic region of VirD4 contains a Gly-x-Gly-x-Gly nucleotide-binding motif. To determine the role of the nucleotide-binding motif in DNA transfer, we introduced mutations in virD4 that led to a change in the conserved amino acids. Three residues, glycine at position 151, lysine 152 and lysine 174, were mutated to serine, alanine and threonine respectively. These changes are conservative and are expected to have no major effect on protein structure (Bordo and Argos, 1991). Each mutant was Results VirD4 localizes to the cell pole Transport through the T-DNA transport apparatus will require co-operation between VirD4 and the apparatus. The process will be facilitated by proximal cellular location of the two components. Our previous studies demonstrated that the VirB proteins forms a protein complex at the bacterial membranes (Kumar et al., 2000). In the present study, we determined the subcellular location of Strain vir gene expressed Total no. of gold particles (no. of cells) No. of particles on membranes (%) No. of particles in clusters (%) AD709 A348 virD4 virD4, virB1–11 234 (40) 302 (62) 197 (84) 277 (91) 136 (69) 206 (74) Table 1. Subcellular localization of VirD4 by immunoelectron microscopy. © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 Polar location of Agrobacterium VirD4 1525 Fig. 1. Subcellular localization of VirD4. Subcellular location of VirD4 was determined by immunofluorescence (A–G) and immunoelectron (H and I) microscopy as described previously (Kumar et al., 2000). An arrowhead indicates the location of VirD4. A–C and H. Agrobacterium A348. D. A348DD. E. WR1715. F. A348DB. G and I. AD709. AD709 expresses only the virD4 gene. A348DD, A348DB and WR1715 have a deletion in virD, virB and a non-polar deletion in virD2 respectively. A. Phase-contrast microscopy. B. Fluorescence microscopy. C–G. Composites of phase and fluorescence images. © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 1526 R. B. Kumar and A. Das spacer. A large deletion of the vir gene regulator VirA did not affect its ability to sense plant phenolics (Melchers et al., 1989). By site-specific mutagenesis, we introduced a deletion of 26 residues in the periplasmic domain, amino acids 36–61, and studied the effect of the deletion on VirD4 function by complementation assays. The virD4D36–61 mutant failed to complement the null mutation in virD4 (Fig. 2A), indicating that the periplasmic domain is essential for VirD4 function. Another region of interest is the C-terminal end of VirD4. Sequence comparison with the homologues indicated that this is the most divergent region. We speculated that a low homology suggests that this region probably confers a specificity function. This property will make the C-terminal end indispensable. To test the requirement of the extreme C-terminus in VirD4 function, we introduced a stop codon at residue 553. The mutant will synthesize a truncated protein that lacks the C-terminal 104 amino acids. In complementation experiments, the mutant virD4L553X failed to complement the virD4 null mutant, indicating that the truncated protein is non-functional. Fig. 2. A. Phenotype of the virD4 mutants. The effect of a mutation in virD4 on T-DNA transfer was monitored by complementation assays. Kalanchöe diagremontiana leaves were infected with the indicated strains, and tumour formation was assessed about 4 weeks after infection. Strains used: A136, –Ti plasmid; A348, +pTiA6; wt, G151S, K152A, K174T, L553X and D36-61 –At12506 harbouring a plasmid that expresses wild-type virD4 or its mutant. B. Effect of a mutation on the stability of VirD4. Total proteins of A. tumefaciens strains expressing the virD4 alleles were separated by SDS–PAGE, transferred to a nitrocellulose membrane and probed with anti-VirD4 antibodies. Lane 1, uninduced A. tumefaciens A348; lanes 2–7, induced A. tumefaciens At12506 expressing wild-type virD4 or its mutants. introduced into the virD4 null mutant A. tumefaciens At12506 (Fullner et al., 1994), and its ability to complement the virD4 mutation was tested by tumour formation assays. Wild-type virD4, as expected, complemented the mutation (Fig. 2A). None of the mutants, virD4G151S, virD4K152A and virD4Kl74T, complemented the virD4 null mutant, indicating that glycine 151, lysine 152 and lysine 174 are essential for VirD4 function. As these residues are likely to be involved in nucleotide binding, nucleotide binding (and/or hydrolysis) by VirD4 is probably required for DNA transfer. The short (residues ª 31–67) periplasmic domain of VirD4 can have a structural role, i.e. it serves as a spacer between the two membrane-spanning domains or is required for an additional function. To study its role in DNA transfer, we introduced a deletion in the periplasmic domain and studied the effect of the mutation on VirD4 function. The length of the region is expected to be amenable to alteration if its sole function is to serve as a Alteration of lysine 174 and the loss of the C-terminal end affect protein stability The loss of function of the virD4 mutants described above could result from the effect of a mutation on protein function or protein stability. To study whether a mutation affects protein stability, Western blot assays were used (Anderson et al., 1996). The purified anti-VirD4 antibodies recognized a single protein band in the induced wildtype A. tumefaciens A348 (Fig. 2B, lane 2). Analysis of the mutants by Western blot assays indicated that three mutations, G151S, D36-61 and K152T, do not affect protein stability because strains expressing the mutants accumulated the mutant protein at a level comparable with that of the wild-type protein (Fig. 2B, lanes 3, 6 and 7). The other two mutant proteins, VirD4K174T and VirD4L553X, were unstable, indicating that the mutations destabilized the protein (Fig. 2B, lanes 4 and 5). The labile nature of VirD4K174T is somewhat surprising, as the equivalent lysine in one VirD4 homologue tolerated a similar substitution (Balzer et al., 1994). VirD4 mutants are defective in VirE2 transfer The A. tumefaciens T-DNA transport apparatus can transport both DNA and protein (Beijersbergen et al., 1992; Vergunst et al., 2000). One protein substrate translocated through the transport apparatus is VirE2. We studied whether the VirD4 domains essential for DNA transfer are also essential for VirE2 transfer. The in planta complementation assay of Otten et al. (1984) was used to monitor VirE2 transfer. None of the virD4 mutants com© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 Polar location of Agrobacterium VirD4 1527 Table 2. Effect of a mutation in virD4 on VirE2 transfer. Strain VirE2 donor (VirD4 phenotype) T-DNA donor Tumour formation PC2760 (wt) – PC2760 (wt) AD 1011 (G151S) AD1012 (K152A) AD1015 (D36–61) AD1016 (wt) – MX358 MX358 MX358 MX358 MX358 MX358 – – + – – – + VirE2 transfer was monitored by in planta complementation assays (Otten et al., 1984). –, no tumour; +, tumour. plemented the virE2 mutant, indicating that all mutations in virD4 abolished the ability to transfer VirE2 to plants (Table 2). In dominance assays, all mutants had a recessive phenotype. The C-terminal end encodes a VirD4–VirD4 interaction domain Subcellular localization studies suggested that VirD4 probably forms an oligomeric complex. The formation of the complex will require VirD4 to be involved in homotypic interactions. Therefore, VirD4 should contain selfinteraction domains. We tested this hypothesis using the two-hybrid assay in yeast (Das and Xie, 2000). Owing to its large size (656 amino acids), VirD4 was divided into three smaller overlapping proteins, and interaction between each segment was monitored by the two-hybrid assay. No VirD4–VirD4 interaction was observed with fusions containing the N-terminal (N) and the central (M) regions of the protein. A polypeptide that contained the Cterminal domain (CTD) was functional in the interaction, indicating that the CTD of VirD4 encodes an interaction domain (Fig. 3). A yeast cell expressing a CTD–activator fusion and a CTD–LexA fusion activated expression of the reporter lacZ gene, producing a blue colony colour phenotype on indicator plates containing Xgal. In control Fig. 3. Identification of the VirD4 interaction domains. VirD4–VirD4 interaction was studied by the yeast two-hybrid assay (Das and Xie, 2000). VirD4 was divided into three overlapping segments: the N-terminal, middle and C–terminal segments. Interactions between the N-terminal (N–N), middle (M–M) and C-terminal (C–C) segments were studied. Cells expressing the vector plasmids (V), the CTD–LexA fusion and the activator vector (C+Act) and the CTD–activator fusion and the LexA vector (C+LexA) are shown as controls. experiments, yeast expressing a CTD fusion and the reciprocal non-fusion protein (LexA/activator) exhibited a white colony colour phenotype. Two domains are required for targeting VirD4 to the cell pole To study whether an avirulent mutation is affected in VirD4 localization, we determined the subcellular location of the mutants by IFM (Fig. 4). Two mutants failed to target VirD4 to the cell poles. The VirD4D36–61 protein failed to form foci of fluorescence and was distributed randomly across the cell membrane (Fig. 4A). The subcellular distribution of the mutant protein is similar to that of alkaline phosphatase reported previously (Kumar et al., 2000). A second mutant, virD4G151S, was inefficient in targeting the protein to the cell pole (Fig. 4B). Cells expressing VirD4G151S had foci of high fluorescence; however, these foci were found at sites other than the cell poles at a reasonable frequency (arrowhead). About 68% of the foci localized to sites other than the cell pole (Table 3). In con- Fig. 4. Subcellular location of the VirD4 mutants. Subcellular location of VirD4D36–61 (A), VirD4G151S (B) and VirD4K152A (C) were determined by immunofluorescence microscopy. An arrowhead identifies mislocalized VirD4. © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 1528 R. B. Kumar and A. Das virD4 expressed No. of cells analysed No. of cells with 1–2 foci (total no. of foci) No. of cells with polar foci (no.a) Percentage of polar focib Wild type G151S K152A 171 103 99 81 (122) 49 (68) 50 (69) 75 (116) 9 (22) 37 (56) 95 32 81 Table 3. Effect of mutations in virD4 on polar location of the protein. a. No. of polar foci includes count of foci in cells with exclusively polar foci and cells with both polar and non-polar foci. b. Wild type versus G151S and wild type versus K152A have P-values of 2.2 ¥ 10–16 and 4.6 ¥ 10–3 respectively. A value <0.05 indicates that a difference is statistically significant. trast, in bacteria expressing wild-type VirD4, 95% of the foci localized to the poles. As the G151S mutation maps to the nucleotide-binding domain, the mutant is probably defective in ATP utilization. These results suggest that targeting of VirD4 to the cell pole may be an energyrequiring process. The third mutant virD4K152A, a mutation in the nucleotide binding site motif, had a near wild-type phenotype. The mutant protein was found primarily at the cell pole, although targeting to other areas of the membrane was observed at a very low frequency (Fig. 4C, arrowhead). The mutant probably has a low ATP utilization activity that is nearly sufficient for its targeting to the cell poles but is insufficient for DNA transfer. An alternative hypothesis is that the G151S mutation affects another property of the protein essential for its proper subcellular localization. Discussion This study reports the subcellular location of the A. tumefaciens VirD4 protein and the identification of two of its essential functional domains. The periplasmic domain and the nucleotide-binding domain are required for the substrate transfer function of VirD4. Two other nucleotidebinding proteins, VirB4 and VirB11, are also essential for T-DNA transfer (Christie, 2001). The requirement of multiple nucleotide-binding proteins in T-DNA transfer suggests that several steps in the DNA translocation pathway are energy-dependent processes requiring the participation of a specific protein(s). The present study demonstrates that the short periplasmic domain of VirD4 is essential for VirD4 function. A deletion of most of the domain led to the complete loss of VirD4 activity. The deletion is unlikely to affect the membrane topology of the protein, because it does not affect sequences immediately adjacent to the hydrophobic domains. A similar analysis of the A. tumefaciens VirA protein indicated that the periplasmic domain does not play a role in sensing vir gene inducers (Melchers et al., 1989). Subcellular localization by IFM demonstrated that VirD4 localizes to the cell poles. These studies also suggest that VirD4 forms a large oligomeric complex. The formation of foci of fluorescence in IFM and the clustering of gold particles in IEM are suggestive of oligomerization. A similar analysis of the mutants indicated that both the periplasmic and the nucleotide-binding domains are essential for the proper subcellular localization of VirD4. The avirulent phenotype of the mutants suggests a requirement for polar location of VirD4 in DNA transfer. These studies indicated that the primary sequence of VirD4 encodes information necessary for its proper subcellular localization. Other proteins, e.g. Pseudomonas aeruginosa histidine kinase PilS, are known to encode information on polar localization within the primary sequence (Boyd, 2000). The phenotype of the two mutants suggests a mechanism for the polar localization of VirD4. Random distribution of VirD4D36–61 throughout the cell membranes indicates that the periplasmic domain is essential for the targeting of VirD4 to the cell poles. The lack of foci of fluorescence in the mutant suggests that the protein is probably defective in oligomerization as well. Defects in both oligomerization and polar localization suggest a two-step mechanism for the targeting of VirD4 to a cell pole. The first step is oligomerization of VirD4, which is manifested as foci of fluorescence in IFM. The second step is the targeting of an oligomer to the cell pole. VirD4D36–61 is defective in oligomer formation and therefore cannot localize to the cell pole. As protein–protein interaction is required for oligomerization, the periplasmic domain probably encodes an interaction domain that supports the formation of higher oligomers of VirD4. The other interaction domain we identified, the C-terminal interaction domain (Fig. 3), is probably involved in forming a small oligomer, e.g. a dimer. The phenotype of a second mutant, VirD4G151S, is consistent with the proposed model for VirD4 localization and provides new information on the mechanism. VirD4G151S contains the periplasmic domain and would be able to form the higher oligomer. In IFM, this mutant exhibited foci of fluorescence, and not an overall fluorescence. The mutant, however, failed to localize exclusively to the cell pole. As the mutation is expected to affect nucleotide binding, these results suggest that the second step, the targeting of the oligomer to the cell pole, may be an energy-dependent process. Two bacterial proteins, Listeria monocytogenes IscA and Shigella flexneri ActA, localize to a single pole and use different mechanisms for their unipolar localization (Lybarger and Maddock, 2001). © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 Polar location of Agrobacterium VirD4 1529 The polar location of VirD4 raises interesting questions on the location of the T-DNA transport apparatus. The VirD4 and VirB complexes act together in the transport of substrates across the transport apparatus. To serve their role in DNA transfer, VirD4 and the VirB protein complex need to be adjacent to, if not overlapping, one another. Therefore, a transport apparatus assembled at a cell pole can mediate substrate transfer. Three constituents of the transport apparatus, VirB8, VirB9 and VirB10, localized to sites on the cell membrane including the poles (Kumar et al., 2000). Although the VirB proteins were found at several sites, VirD4 was more localized. One explanation for the difference is that all complexes contain both VirD4 and the VirB proteins. A difference in the quality of the antibodies (titre and/or specificity) affected the sensitivity of the detection system. A second possibility is that there are two types of complexes, a VirD4–VirB complex that transports substrate from the cytoplasm and a VirB complex that transports periplasmic and membrane proteins. A relatively lower level of VirD4 in the cell will enable only a subset of the VirB complexes to associate with VirD4. The excess VirB complexes are used for the transport of the pilin protein from the membrane to the site of pilus assembly in a virD4-independent manner. Two studies suggest that the VirB proteins without VirD4 can form a functional transport apparatus (Craig-Mylius and Weiss, 1999; Lai et al., 2000). The formation of the T-pilus does not require VirD4 but is absolutely dependent on the VirB proteins, and a VirD4 homologue is not necessary for the export of pertussis toxin in B. pertussis. Both VirB2 and the pertussis toxin subunits are transported across the cytoplasmic membrane in a Sec-dependent process. Transport of these substrates can bypass the requirement for VirD4. VirD4 may be the additional necessary component when vir-dependent transport of a cytoplasmic substrate is required, e.g. DNA, VirD2, VirE2 and VirF. We postulate that a polar VirD4–VirB complex functions in substrate transfer from the cytoplasm. Co-localization studies suggesting that both VirD4 and a transport apparatus protein VirB8 localize to the cell pole support the existence of a polar VirD4–VirB complex (R. B. Kumar, unpublished results). The polar location of the VirD4–VirB complexes explains how bacteria can deliver a substrate to the plant cell. Attachment of bacteria to plant cells is essential for the transfer of T-DNA, and bacteria attaches to the plant cell in a polar manner (Matthysse, 1987). The assembly of the transport apparatus at the cell pole will therefore be the most efficient mechanism for substrate transfer. Proteins essential for critical biological processes, e.g. cell division, chromosome partitioning and cell cycle control in bacteria, have been localized to the cell poles (Lybarger and Maddock, 2001). Substrate transfer by the type IV transport mechanism is another process that probably uses the cell pole. © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 Experimental procedures Strains and plasmids Strains and plasmids used in this study are listed in Table 4. Mutagenesis of VirD4 Mutations in virD4 were introduced by deoxyoligonucleotidedirected site-specific mutagenesis using single-stranded pAD1373 DNA as a template (Das and Xie, 2000). To facilitate screening of the mutants, a new restriction site was introduced at the site of mutation whenever possible. The mutation and the mutagenic oligonucleotide used were as follows (base change underlined): glycine 151 to serine, dACGCGTGCTAGCAAAGGCG; lysine 152 to alanine, dGCGTGCTGGCGCCGGCGTCGGCATC; lysine 174 to threonine, dCCCTCGACGTTACGGGAGAATTGTT; and deletion of residues 36–61, dTTTCGCCGGGGTACCGTC TTCTGG. All mutations were confirmed by DNA sequence analysis. Plasmids pAD1373 and the mutant derivatives were linearized with the restriction endonuclease EcoICRI and cloned into the filled-in HindIII site of plasmid pTJS75 to construct the wide-host-range derivatives. Plasmids were introduced into A. tumefaciens At12506 for complementation assays (Das and Xie, 2000). Transfer of VirE2 was monitored by in planta complementation assays (Otten et al., 1984). Proteins were analysed by Western blot assays after SDS–PAGE (Anderson et al., 1996). Subcellular localization Subcellular location of VirD4 and its mutants was determined by immunofluorescence and immunoelectron microscopy as described previously (Kumar et al., 2000). Wild-type Agrobacterium A348 and its derivatives were grown in AB minimal medium. Bacteria were induced by growth in AB Mes, pH 5.8, medium in the presence of 100 mM acetosyringone. Antibodies Antibodies against VirD4 were raised in rabbits using a histidine-tagged VirD4 fusion protein as an antigen. The antibodies were purified by affinity chromatography on immobilized fusion protein–affigel columns. VirD4–VirD4 interaction Interaction of VirD4 was monitored by the yeast two-hybrid assay (Das and Xie, 2000). The large cytoplasmic region of VirD4 was divided into three overlapping regions (leucine 87–glycine 321, proline 264–arginine 501 and proline 461–end) and used for the analysis of protein–protein interaction by the yeast two-hybrid assay. Genes encoding the desired coding region were amplified by polymerase chain reaction (PCR) and cloned into both pJK202 and pJG4-5. The plasmids were introduced into yeast AD842 and tested for interaction by monitoring reporter lacZ gene expression. 1530 R. B. Kumar and A. Das Table 4. Strains and plasmids. Strain or plasmid Relevant characteristics Source or reference recA1 endA1 hsdR17(rk–mk+) supE44 thi-1 gyrA relA1 l– f80d/lacZDM15(DlacZYA-argF)U169 Laboratory stock A. tumefaciens A136 A348 A348DB A348DD At12506 WR1715 MX358 PC2760 AD709 AD1011 AD1012 AD1013 AD1014 AD1015 AD1016 C58 heat cured of pTiC58 A136 containing pTiA6 A348 with a deletion in pTiA6 virB A348 with a deletion in pTiA6 virD A348 virD4:tn3hoho1 A348 with a non-polar deletion in pTiA6 virD2 A348 virE2:tn3hoho1 Ach5 with a deletion of pTiAch5 T-DNA A136 with plasmid pAD1356 At12506 carrying pAD1678 At12506 carrying pAD1679 At12506 carrying pAD1680 At12506 carrying pAD1681 At12506 carrying pAD1682 At12506 carrying pAD1686 Laboratory stock Laboratory stock Anderson et al. (1996) Vogel and Das (1992) Fullner et al. (1994) Shurvinton et al. (1992) Stachel and Nester (1986) Beijersbergen et al. (1992) Das and Xie (1998) This study This study This study This study This study This study Saccharomyces cerevisiae EGY48 AD842 ura3 his3 trp1 LexAop-leu2 EGY48 containing pSH18-34 Finley and Brent (1995) Das and Xie (2000) A virDp–virD4 chimeric gene in a colE1 vector Plasmid pAD1373 in wide-host-range vector pTJS75 Plasmid pAD1373G151S in pTJS75 Plasmid pAD1373K152 A in pTJS75 Plasmid pAD1373K174T in pTJS75 Plasmid pAD1373L553X in pTJS75 Plasmid pAD1373D36-61 in pTJS75 2.3 kb PstI–EcoRI fragment into the PstI site of plasmid pRSET C, expresses a his-tagged fusion protein containing residues 111–656 of VirD4 Cloning vector for the construction of LexA fusions Cloning vector for the construction of activator fusions A plasmid containing Gal1-LexAop-lacZ, a reporter gene in yeast 0.7 kb EcoRI fragment encoding Leu-87/Gly-327 of VirD4 obtained by PCR amplification in pJG4-5 (activator–VirD4 N-terminal fusion) 0.65 kb EcoRI fragment encoding Pro-461/end of VirD4 obtained by PCR amplification in pJG4-5 (activator–VirD4 C-terminal fusion) 0.72 kb XhoI fragment encoding Pro-264/Arg-501 of VirD4 obtained by PCR amplification in pJG4-5 (activator–VirD4 middle fusion) 0.7 kb EcoRI fragment of pAD1580 in pJK202 (LexA–VirD4N fusion) 0.72 kb XhoI fragment of pAD1592 in pJK202 (LexA–VirD4M fusion) 0.65 kb EcoRI fragment of pAD1582 in pJK202 (LexA–VirD4C fusion) Das and Xie (1998) This study This study This study This study This study This study This study E. coli DH5a Plasmids pAD1373 pAD1686 pAD1678 pAD1679 pAD1680 pAD1681 pAD1682 pTJ16 pJK202 pJG4-5 pSH18-34 pAD1580 pAD1582 pAD1592 pAD1693 pAD1694 pAD1695 Das and Xie (2000) Finley and Brent (1995) Finley and Brent (1995) This study This study This study This study This study This study Acknowledgements References We thank T. Johnson for generating VirD4 antibodies, P. Judd for experimental assistance, S. Banerjee for statistical analysis, M. Sanders and G. Ahlstrand for assistance with microscopy, Drs E. Nester and W. Ream for Agrobacterium strains, and Drs B. Barry, A. Binns and G. Dunny for critical reading of an earlier version of the manuscript. This work was supported by a grant from the National Science Foundation (MCB 9982804). Anderson, L.B., Hertzel, A.V., and Das, A. (1996) Agrobacterium tumefaciens VirB7 and VirB9 form a disulfide-linked protein complex. Proc Natl Acad Sci USA 93: 8889– 8894. Balzer, D., Pansegrau, W., and Lanka, E. (1994) Essential motifs of relaxase (TraI) and TraG proteins involved in conjugative transfer of plasmid RP4. J Bacteriol 176: 4285–4295. © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 Polar location of Agrobacterium VirD4 1531 Baron, C., Thorstenson, Y.R., and Zambryski, P.C. (1997) The lipoprotein VirB7 interacts with VirB9 in the membranes of Agrobacterium tumefaciens. J Bacteriol 179: 1211–1218. Beaupre, C.E., Bohne, J., Dale, E.M., and Binns, A.N. (1997) Interactions between VirB9 and VirB10 membrane proteins involved in movement of DNA from Agrobacterium tumefaciens into plant cells. J Bacteriol 179: 78–89. Beijersbergen, A., Dulk-Ras, A.D., Schilperoort, R.A., and Hooykaas, P.J. (1992) Conjugative transfer by the virulence system of Agrobacterium tumefaciens. Science 256: 1324–1327. Bordo, D., and Argos, P. (1991) Suggestions for ‘safe’ residue substitutions in site-specific mutagenesis. J Mol Biol 217: 721–729. Boyd, J.M. (2000) Localization of the histidine kinase PilS to the poles of Pseudomonas aeruginosa and identification of a localization domain. Mol Microbiol 36: 153–162. Cabezon, E., Lanka, E., and de la Cruz, F. (1994) Requirements for mobilization of plasmids RSF1010 and ColE1 by the IncW plasmid R388: trwB and RP4 traG are interchangeable. J Bacteriol 176: 4455–4458. Cabezon, E., Sastre, J.I., and de la Cruz, F. (1997) Genetic evidence of a coupling role for the TraG protein family in bacterial conjugation. Mol Gen Genet 254: 400–406. Christie, P.J. (2001) Type IV secretion: intercellular transfer of macromolecules by systems ancestrally related to conjugation machines. Mol Microbiol 40: 294–305. Covacci, A., Telford, J.L., Del Giudice, G., Parsonnet, J., and Rappuoli, R. (1999) Helicobacter pylori virulence and genetic geography. Science 284: 1328–1333. Craig-Mylius, K., and Weiss, A. (1999) Mutants in the ptlA–H genes of Bordetella pertussis are deficient for pertussis toxin secretion. FEMS Microbiol Lett 179: 479–484. Das, A., and Xie, Y.-H. (1998) Construction of transposon Tn3phoA: Its application in defining the membrane topology of the Agrobacterium tumefaciens DNA transfer proteins. Mol Microbiol 27: 405–414. Das, A., and Xie, Y.-H. (2000) Agrobacterium tumefaciens T-DNA transport pore proteins VirB8, VirB9 and VirB10 interact with one another. J Bacteriol 182: 758–763. Das, A., Anderson, L., and Xie, Y.-H. (1997) Delineation of the interaction domains of Agrobacterium tumefaciens VirB7 and VirB9 by use of the yeast two-hybrid assay. J Bacteriol 179: 3404–3409. Finley, R., and Brent, R. (1995) Interaction trap cloning in yeast. In DNA Cloning, Expression Systems – A Practical Approach. Hames, B., and Glover, D. (eds). Oxford: Oxford University Press, pp. 169–203. Fullner, K., Stephens, K., and Nester, E. (1994) An essential virulence protein of Agrobacterium tumefaciens, VirB4, requires an intact mononucleotide binding domain to function in transfer of T-DNA. Mol Gen Genet 245: 704– 715. Fullner, K., Lara, J.C., and Nester, E. (1996) Pilus assembly by Agrobacterium T-DNA transfer genes. Science 273: 1107–1109. Gomis-Ruth, F., Moncalian, G., Perez-Luque, R., Gonzalez, A., Cabezon, E., de la Cruz, F., and Coll, M. (2001) The bacterial conjugation protein TrwB resembles ring helicases and F1-ATPase. Nature 409: 637–641. © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532 Hamilton, C.M., Lee, H., Li, P., Cook, D., Piper, K., von Bodman, S., et al. (2000) TraG from RP4 and TraG and VirD4 from Ti plasmids confer relaxosome specificity to the conjugal transfer system of pTiC58. J Bacteriol 182: 1541–1548. Kumar, R., Xie, Y.-H., and Das, A. (2000) Subcellular localization of the Agrobacterium tumefaciens T-DNA transport pore proteins: VirB8 is essential for the assembly of the transport pore. Mol Microbiol 36: 608–617. Lai, E.M., and Kado, C.I. (1998) Processed VirB2 is the major subunit of the promiscuous pilus of Agrobacterium tumefaciens. J Bacteriol 180: 2711–2717. Lai, E.-M., Chesnokova, O., Banta, L., and Kado, C. (2000) Genetic and environmental factors affecting T-pilin export and T-pilus biogenesis in relation to flagellation of Agrobacterium tumefaciens. J Bacteriol 182: 3705–3716. Lessl, M., Pansegrau, W., and Lanka, E. (1992) Relationship of DNA-transfer-systems: essential transfer factors of plasmid RP4, Ti and F share common sequences. Nucleic Acids Res 20: 6099–6100. Lybarger, S., and Maddock, J. (2001) Polarity in action: asymmetric protein localization in bacteria. J Bacteriol 183: 3261–3267. Maddock, J.R., and Shapiro, L. (1993) Polar location of the chemoreceptor complex in the Escherichia coli cell. Science 259: 1717–1723. Matthysse, A. (1987) Characterization of nonattaching mutants of Agrobacterium tumefaciens. J Bacteriol 169: 313–323. Melchers, L.S., Regensburg-Tuink, T., Bourret, R., Sedee, J., Schilperoort, R., and Hooykaas, P. (1989) Membrane topology and functional analysis of the sensory protein VirA of Agrobacterium tumefaciens. EMBO J 8: 1919–1925. Okamoto, S., Toyoda-Yamamoto, A., Ito, K., Takebe, I., and Machida, Y. (1991) Localization and orientation of the VirD4 protein of Agrobacterium tumefaciens in the cell membrane. Mol Gen Genet 228: 24–32. Otten, L., DeGreve, H., Leemans, J., Hain, R., Hooykaas, P., and Schell, J. (1984) Restoration of virulence of vir region mutants of Agrobacterium tumefaciens strain B6S3 by coinfection with normal and mutant Agrobacterium strains. Mol Gen Genet 175: 159–163. Schmidt-Eisenlohr, H., Domke, N., Angerer, C., Wanner, G., Zambryski, P., and Baron, C. (1999) Vir proteins stabilize VirB5 and mediate its association with the T pilus of Agrobacterium tumefaciens. J Bacteriol 181: 7485– 7492. Shurvinton, C.E., Hodges, L., and Ream, W. (1992) A nuclear localization signal and the C-terminal omega sequence in the Agrobacterium tumefaciens VirD2 endonuclease are important for tumor formation. Proc Natl Acad Sci USA 89: 11837–11841. Spudich, G., Fernandez, D., Zhou, X.-R., and Christie, P. (1996) Intermolecular disulfide bonds stabilize VirB7 homodimers and VirB7/VirB9 heterodimers during biogenesis of the Agrobacterium tumefaciens T-complex transport pore. Proc Natl Acad Sci USA 93: 7512–7517. Stachel, S., and Nester, E. (1986) The genetic and transcriptional organization of the vir region of the A6 Ti plasmid of Agrobacterium tumefaciens. EMBO J 5: 1445– 1454. 1532 R. B. Kumar and A. Das Vergunst, A.C., Schrammeijer, B., den Dulk-Ras, A., de Vlaam, C.M., Regensburg-Tuink, T., and Hooykaas, P.J. (2000) VirB/D4-dependent protein translocation from Agrobacterium into plant cells. Science 290: 979–982. Vogel, A.M., and Das, A. (1992) The Agrobacterium tumefa- ciens virD3 gene is not essential for tumorigenicity on plants. J Bacteriol 174: 5161–5164. Zupan, J., Muth, T., Draper, O., and Zambryski, P. (2000) The transfer of DNA from Agrobacterium tumefaciens into plants: a feast of fundamental insights. Plant J 23: 11–28. © 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1523–1532