Efficient vir Gene Induction in Agrobacterium tumefaciens Requires virA, virG, and vir Box from the Same Ti Plasmid

Abstract

The vir genes of octopine, nopaline, and l,l-succinamopine Ti plasmids exhibit structural and functional similarities. However, we observed differences in the interactions between octopine and nopaline vir components. The induction of an octopine virE_A6::lacZ fusion (pSM358cd) was 2.3-fold higher in an octopine strain (A348) than in a nopaline strain (C58). Supplementation of the octopine virG_A6 in a nopaline strain with pSM358 did not completely restore virE_A6 induction. However, addition of the octopine virA_A6 to the above strain increased virE_A6 induction to a level almost comparable to that in octopine strains. In a reciprocal analysis, the induction of a nopaline virE_C58::cat fusion (pUCD1553) was two- to threefold higher in nopaline (C58 and T37) strains than in octopine (A348 and Ach5) and l,l-succinamopine (A281) strains. Supplementation of nopaline virA_C58 and virG_C58 in an octopine strain (A348) harboring pUCD1553 increased induction levels of virE_C58::cat fusion to a level comparable to that in a nopaline strain (C58). Our results suggest that octopine and l,l-succinamopine VirG proteins induce the octopine virE_A6 more efficiently than they do the nopaline virE_C58. Conversely, the nopaline VirG protein induces the nopaline virE_C58 more efficiently than it does the octopine virE_A6. The ability of Bo542 virG to bring about supervirulence in tobacco is observed for an octopine vir helper (LBA4404) but not for a nopaline vir helper (PMP90). Our analyses reveal that quantitative differences exist in the interactions between VirG and vir boxes of different Ti plasmids. Efficient vir gene induction in octopine and nopaline strains requires virA, virG, and vir boxes from the respective Ti plasmids.

Agrobacterium tumefaciens has the unique capability of transferring the T-DNA portion of its Ti plasmid into plant cells at infected wound sites, which results in the formation of crown gall tumors (8, 9, 73). The infection process involves a set of chromosome-encoded genes (chv) involved in attachment of bacteria to plant cells and Ti plasmid-encoded vir genes that function in trans, helping in the generation, transfer, and integration of T strands into the plant genome (reviewed in references 19, 31, and 32).

Sensing of signal molecules released by wounded plant cells is the first step of signal transduction leading to vir gene induction in Agrobacterium (68, 70). VirA, an inner membrane protein, senses the signal molecules (45, 48) and gets autophosphorylated in the His-474 residue (33, 36). The phosphorylated VirA in turn activates the cytoplasmic protein VirG by phosphorylating it at Asp-52 (35). VirG, a DNA binding protein (54, 56), acts as a transcriptional activator of vir genes by binding to vir boxes present upstream of all vir operons (11, 37). Based on protein sequence similarities, VirA and VirG have been assigned to a large group of His-Asp two-component regulatory systems, involving a sensor and a response regulator (45, 76).

The T-DNA portion of the Ti plasmid carries genes that specify the synthesis of tumor-specific compounds called opines (5, 51). Based on the opines utilized, Agrobacterium Ti plasmids are classified as octopine, nopaline, agropine, and succinamopine types (12). High levels of homology exist between the vir regions of octopine and nopaline Ti plasmids (15, 16, 22, 26), between octopine and l,l-succinamopine Ti plasmids (42), and between octopine Ti and Ri plasmids (58, 75). Assays of tumorigenesis in the tomato (30) established the existence of functional similarities among virA, virB, virC, and virE of pTiB6 (octopine type), pTiC58 (nopaline type), and pRi1855 (Ri plasmid). The vir genes, virA, virB, virC, virD, virE, and virG, of pTiBo542 can functionally complement those of pTiA6 to form tumors on Kalanchoe diagremontiana and Nicotiana glauca (42). The octopine vir region (24, 39, 67, 69) and the nopaline vir region (10, 20, 26, 60) have been characterized previously by mutational and complementation analysis (10, 24, 30, 38, 39, 47, 69).

The assay of vir-inducing abilities of various phenolic compounds involved the use of a tester strain, A348 (an octopine type), harboring either pSM358 (a virE_A6::lacZ fusion) or pSM243cd (a virB_A6::lacZ fusion) as a reporter plasmid (4, 49, 65, 66, 68). When four different wild-type Agrobacterium strains harboring the octopine virB_A6::lacZ fusion were induced with a number of vir-inducing signal molecules, the vir genes of KU12 were found to be induced by a group of phenolic compounds which were different from those that induced the vir genes of C58, A6, and Bo542 strains (44). VirA was found to be responsible for the difference in sensing the phenolic compounds. The differences observed among different A. tumefaciens strains for different inducers could be overcome by using sugars such as l-(+)-arabinose along with the inducer (55). This is brought about by the induction of chvE by l-(+)-arabinose through gbpR (14). In the above study (44), acetosyringone (AS)-mediated induction of the octopine virB_A6::lacZ in octopine (A6) and l,l-succinamopine (Bo542) strains was found to be three- and sevenfold higher, respectively, than that in the nopaline strain (C58).

We performed a comparative analysis of an octopine vir gene (virE_A6::lacZ fusion) and a nopaline vir gene (virE_C58::cat fusion) in octopine, nopaline, and l,l-succinamopine Ti plasmid backgrounds to evaluate the quantitative differences between them in vir gene induction. Our results suggest that virG and vir box from the same Ti plasmid interact effectively. Optimal vir induction is observed only when virA, virG, and vir boxes are derived from the same Ti plasmid.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

A. tumefaciens and Escherichia coli strains and plasmids used are listed in Table 1. A. tumefaciens was grown at 28°C either in YEP (7) or in AB minimal medium (7). E. coli was grown in Luria-Bertani medium (50) at 37°C. Antibiotic concentrations (in micrograms per milliliter) used for plates were as follows: for A. tumefaciens, kanamycin monosulfate (100; for all strains except for Ach5, T37, and their derivatives, for which 50 μg/ml was used), rifampin (10), carbenicillin (100; for all strains except for Ach5, T37, and their derivatives, for which 5 μg/ml was used), tetracycline hydrochloride (5), gentamicin sulfate (100; for all strains except for Ach5, T37, and their derivatives, for which 30 μg/ml was used), spectinomycin dihydrochloride (100), and streptomycin sulfate (300); for E. coli, kanamycin monosulfate (50), ampicillin sodium salt (50), tetracycline hydrochloride (20), and gentamicin sulfate (3). All liquid media for A. tumefaciens contained 50% of the specified antibiotic concentrations. All antibiotics were bought from Sigma Chemical Co., St. Louis, Mo.

TABLE 1.

Bacterial strains and plasmids used

Strain or plasmid	Characteristics	Source or reference
Strains
E. coli DH5α	F−recA1 λ− gyrA96 thi-1 hsdR17(rK− mK+) supE44 relA1 deoR Δ(lacZYA-argF)U169 φ80dlacZΔM15 endA1 phoA	2
A. tumefaciens
A136	C58 chromosomal background; pTiC58 cured (Rmr)	62
A348	A136 harboring pTiA6NC; octopine type (Rmr)	18
C58	Nopaline type; pTiC58—wild type (Rmr)	62
Ach5	Octopine type; pTiAch5—wild type	21
T37	Nopaline type; pTiT37—wild type	62
A281	A136 harboring pTiBo542; l,l-succinamopine type (Rmr)	62
pTi237	A virAA6 mutant; pTiA6NC carrying a Tn3HoHo1 insertion in virA, an avirulent strain (Rmr Cbr)	69
LBA4301(pJK107)	Ach5 chromosomal background; pTiAch5 cured; a virAC58 mutant; harbors pTiC58 carrying a Tn5 insertion in virA; an avirulent strain (Rmr Kmr)	59
LBA4301(pJK710)	Ach5 chromosomal background; pTiAch5 cured; a virGC58 mutant; harbors pTiC58 carrying a Tn5 insertion in virG; an avirulent strain (Rmr Kmr)	59
LBA4404	A vir helper; harbors disarmed Ti plasmid pAL4404, a T-DNA deletion derivative of pTiAch5, octopine type (Rmr Smr)	53
EHA105	A vir helper, l,l-succinamopine type, harbors T-DNA deletion derivative of pTiBo542, a supervirulent-type Ti plasmid (Rmr)	29
PMP90	A vir helper, nopaline type, harbors disarmed Ti plasmid pMP90, a T-DNA deletion derivative of pTiC58 (Rmr Gmr)	43
Plasmids
pSM358cd	A virE::lacZ translational fusion derived from pTiA6 (Kmr Apr)	69
pSM358	A virE::lacZ translational fusion derived from pTiA6, also has virG, virC, and virD (Apr Kmr)	67
pAKM2	A 4.4-kb KpnI fragment from pVCK257 having virA of pTiA6 cloned in pUCD2 (Tcr Apr Spr Smr)	This work
pUCD2	IncW, WHR replicon (Tcr Kmr Apr Spr Smr)	10
pUCD1553	virEC58::cat translational fusion, derived from pTiC58, in pUCD1546 (Gmr Apr)	59
pBAL2	pGA472-based binary vector harboring nos-nptII into which CaMV35S-hph from pTRA151 and CaMV35S-gusA with a catalase intron (51) were introduced (Tcr)	This work
pBAL3	The SalI fragment 10 of pTiBo542 vir region having the 3′ end of virB, all of virG and virC, and the 5′ end of virD in pUCD2 (Apr Kmr Spr Smr)	This work
pBAL16	A 3.23-kb virAC58 (coordinates 1621 to 4852) (60) amplicon in pGEM-T Easy vector (Apr)	This work
pBAL17	A 1.53-kb virGC58 (coordinates 14358 to 15888) (60) amplicon in pGEM-T vector (Apr)	This work
pBAL18	A 3.23-kb EcoRI fragment from pBAL16 having virA of pTiC58 was cloned in pBSIIKS+ (Apr)	This work
pBAL19	A 1.53-kb SacII/SalI fragment from pBAL17 having virG of pTiC58 was cloned in pBSIIKS+ (Apr)	This work
pBAL20	A 3.0-kb KpnI/SacI fragment from pBAL18 having virA of pTiC58 was cloned in pUCD2 (Tcr Apr Spr Smr)	This work
pBAL22	A 1.58-kb SacI/KpnI fragment from pBAL19 having virG of pTiC58 was cloned in pMH1002 (Tcr)	This work
pBAL23	A clone of both a 3.0-kb HindIII/SacI fragment from pBAL18 (virAC58) and a 1.58-kb SacI/KpnI fragment from pBAL19 (virGC58) in between HindIII and KpnI sites of pMH1002 (Tcr)	This work
pMH1002	A derivative of pGA472 (1) harboring a multiple cloning site (Tcr)	Lab collection
pSBGA281-G	Harbors the SalI 10 fragment of pTiBo542 containing the 3′ end of virB, all of virG and virC, and the 5′ end of virD in pUC119 (Apr)	S. B. Gelvin

Enzymes and reagents.

Restriction enzymes were purchased from Life Technologies (Gaithersburg, Md.) and Amersham Pharmacia Biotech (Little Chalfont, United Kingdom) and used according to the supplier's recommendations. pGEM-T and pGEM-T Easy vector systems were obtained from Promega Corporation (Madison, Wis.). A high-fidelity PCR system comprising an enzyme blend of Taq and Pwo DNA polymerases was purchased from Roche Molecular Biochemicals and used according to the supplier's recommendations. A random primer DNA labeling kit was bought from Amersham Pharmacia Biotech. A digoxigenin labeling chemiluminescence kit was purchased from Roche Molecular Biochemicals. A Hybond N⁺ nylon membrane was purchased from Amersham Pharmacia Biotech. [α-³²P]dCTP was obtained from BRIT.

Most of the chemicals were purchased from Himedia Laboratories Pvt. Ltd. and Qualigens Fine Chemicals. AS was from Aldrich Chemical Co., Milwaukee, Wis. o-Nitrophenyl-β-d-galactopyranoside and other fine chemicals were bought from Sigma Chemical Co. For tobacco transformation experiments, kanamycin (Kancin; Alembic Chemical Works Co.), carbenicillin (Biopense; Biochem Pharmaceutical Industries), and cefotaxime (Omnatax; Hoechst Marion Roussel Ltd.) were obtained from pharmaceutical suppliers.

Plasmid constructions.

The plasmid pAKM2 (Fig. 1) was constructed by electroeluting a 4.6-kb KpnI fragment of pVCK257 (41) that encodes the pTiA6 virA and placing it in the KpnI site of the kanamycin resistance gene of pUCD2 (10). Clones were screened by insertional inactivation. The nopaline virA and virG genes were cloned by PCR amplification in a Perkin-Elmer GeneAmp PCR 2400 system using a blend of Taq and Pwo DNA polymerases. DNA extracted from A. tumefaciens strain C58 was used as template. The oligonucleotides 5′-AAG GTG GTA CGA ACA CAG-3′ (forward primer) and 5′-TGG CTC TCT AAG ACA ACG-3′ (reverse primer) were used for PCR amplification of a 3,232-bp fragment coding for virA_C58. The cycling parameters were 94°C for 1 min and 30 cycles of 94°C for 1 min, 55°C for 1 min, and 68°C for 2.5 min, followed by a final extension of 72°C for 7 min. Amplified virA_C58 was cloned into pGEM-T Easy vector (Promega) and was designated pBAL16 (Table 1).

FIG. 1.

Linear maps of pAKM2, pBAL20, pBAL22, and pBAL23. The vector (pUCD2) backbone of plasmids pAKM2 and pBAL20 has been linearized at one of its BamHI sites. Similarly, the vector (pMH1002) backbone of plasmids pBAL22 and pBAL23 has been linearized at one of its PstI sites. All maps are drawn to scale. Plasmid pAKM2 has a 4.6-kb KpnI fragment of pVCK257 (41) encoding pTiA6 virA, which was placed into the KpnI site of pUCD2 in its kanamycin resistance gene (10). Plasmid pBAL20 harbors a PCR-amplified 3.2-kb KpnI/SacI fragment encoding pTiC58 virA, placed between the KpnI and SacI sites of pUCD2 in its kanamycin resistance gene. Plasmid pBAL22 harbors a PCR-amplified 1.58-kb SacI/KpnI fragment encoding virG placed between the SacI and KpnI sites of pMH1002. Plasmid pBAL23 is a clone of both virA and virG of pTiC58 in the HindIII/KpnI sites of pMH1002. Sp^r, spectinomycin resistance gene; Km^r, kanamycin resistance gene; Ap^r, ampicillin resistance gene; Tc^r, tetracycline resistance gene; pSa, pSa origin of replication; pMB, pMB origin of replication; ColE1, ColE1 origin of replication; RK2, RK2 origin of replication; LB, left border; RB, right border; B, BamHI; Bg, BglII; E, EcoRI; H, HindIII; K, KpnI; P, PstI; S, SalI; Sa, SacI; X, XhoI. Open arrows indicate the direction of transcription. Black arrowheads indicate the orientation of the left border and the right border repeats.

Oligonucleotides 5′-ATG TCA TCG TAC CCT TCC-3′ (forward primer) and 5′TAC AGT CCT TCC AAG TCG-3′ (reverse primer) were used to PCR amplify a 1,531-bp fragment coding for virG_C58. The cycling parameters were 94°C for 1 min and 30 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 2 min, followed by a final extension of 72°C for 7 min. Amplified virG_C58 was cloned into pGEM-T vector and was designated pBAL17 (Table 1). The 3.2-kb fragment encoding virA_C58 was subcloned as an EcoRI fragment from pBAL16 into pBSIIKS⁺ to obtain pBAL18 (Table 1). Similarly, virG_C58 was subcloned as a 1.5-kb SacII/SalI fragment from pBAL17 into pBSIIKS⁺ to obtain pBAL19 (Table 1).

The plasmid pBAL20 (Fig. 1) was constructed by subcloning a 3.0-kb KpnI/SacI fragment from pBAL18 (which encodes the pTiC58 virA) between the corresponding sites of the kanamycin resistance gene of pUCD2 (10). Recombinant clones were screened by insertional inactivation and confirmed by restriction analysis. The plasmid pBAL22 (Fig. 1) was constructed by cloning a 1.58-kb SacI/KpnI fragment from pBAL19 (encoding the pTiC58 virG) between the corresponding sites of pMH1002 (Table 1). The plasmid pMH1002 is a binary vector derived from pGA472 (1). The plasmid pBAL23 (Fig. 1) was constructed by cloning a 3.0-kb HindIII/KpnI (virA of pTiC58) fragment of pBAL18 and a 1.58-kb SacI/KpnI (virG of pTiC58) fragment from pBAL19 into HindIII/KpnI-digested pMH1002.

The plasmid pBAL2 is a binary vector derived from pGA472 (1) that carries nptII as a plant selection marker. The gusA gene with a catalase intron (52) and the hph gene from the plasmid pTRA151 (77) were introduced into the T-DNA of a pGA472 derivative. The resultant binary plasmid with nptII and hph as plant selection markers and gusA-intron as a reporter is called pBAL2. The plasmid pBAL3 with pTiBo542 virG and virC was constructed as follows: a 4.7-kb SalI fragment 10 representing a part of the pTiBo542 vir region (containing the 3′ end of virB, all of virG and virC, and the 5′ end of virD) (42) was isolated from pSBGA281-G (Table 1) and introduced into pUCD2 (10) at the SalI site in the tetracycline resistance gene. Clones were screened for tetracycline sensitivity.

Introduction of plasmids into A. tumefaciens.

E. coli plasmids were introduced into A. tumefaciens by triparental mating using pRK2013 as mobilization helper (13). The presence of plasmids in transconjugants was confirmed by Southern hybridization analysis (64). The presence of Ti plasmid in transconjugants was confirmed by tumor induction in tobacco leaf disks (57).

Induction of A. tumefaciens vir genes.

Induction of vir genes of A. tumefaciens by AS (68) and by tobacco leaf segments (74) was performed as described earlier (74). Briefly, A. tumefaciens was grown at 28°C in a shaker (200 rpm) to an optical density of 1.0 at A₆₀₀ in AB liquid medium. The culture was centrifuged for 10 min at 5,000 × g. The pellet was resuspended in the same volume of induction medium. To 15 ml of induction medium containing either 60 μM AS or 48-h-preincubated tobacco leaf segments (74), 5 ml of resuspended bacterial cells was added. The standard induction condition used was 120-rpm shaking at 25°C for 24 h with an initial bacterial optical density of 0.25 at A₆₀₀. After 24 h of induction, cells were taken for a β-galactosidase assay (50) or for a chloramphenicol acetyltransferase assay (63). All vir gene induction experiments included A. tumefaciens strain A348(pSM358), a merodiploid strain (67) harboring the pTiA6 virE::lacZ translational fusion to ensure reproducibility in vir induction levels. Representative induction values (Miller units) of this strain are as follows: MS (Murashige and Skoog) medium alone, 95 ± 2; MS plus 60 μM AS, 3,369 ± 150; and MS plus 48-h-preincubated tobacco leaf segments, 3,313 ± 192.

Tobacco transformation.

Tobacco (Nicotiana tabacum L. cv. Wisconsin 38) leaf rings with a 10-mm outer and a 6-mm inner diameter were cut using cork borers from 4- to 6-week-old, axenically grown plants (72). Transformation of leaf rings was performed as described earlier (72). Transformation efficiency is expressed as the number of kanamycin-resistant shoot buds per leaf ring.

RESULTS

Comparison of virE_A6 induction in different Ti plasmid backgrounds.

The vir-inducing ability of a wide range of phenolic compounds varies with their chemical structure (44, 49, 65). Assays used to determine their vir-inducing ability involved the use of vir::lacZ fusion plasmids derived from an octopine strain, A6 (4, 49, 65, 66, 67). Lee et al. (44) performed detailed genetic analysis and reported that VirA is the sensor protein for phenolic compounds. Different wild-type strains, viz., C58, A6, Bo542, and KU12, were found to sense different phenolic compounds as vir inducers to different extents. Lee et al. noted a 3.5-fold-higher virB_A6 induction with AS in the octopine strain (A6) than in the nopaline strain (C58). The pTiA6 virB_A6::lacZ fusion plasmid, pSM243cd (69), was used in their vir gene induction assays.

We performed a detailed analysis by comparing vir gene induction in A. tumefaciens strains harboring different Ti plasmids. The plasmid pSM358cd with the pTiA6 virE_A6::lacZ translational fusion (69) was mobilized by triparental mating into the following Agrobacterium strains: A348 (octopine type, harboring pTiA6), C58 (nopaline type, harboring pTiC58), and A281 (l,l-succinamopine type, harboring pTiBo542). All three strains have the same chromosome background (C58). Levels of induction of the octopine virE_A6::lacZ fusion (pSM358cd) were compared among A6, C58, and Bo542 Ti plasmid backgrounds. Induction was performed with 60 μM AS or with 48-h-preincubated tobacco leaf segments. Induction of virE_A6 in the octopine strain (A348) was 2.3- and 3.2-fold higher than in the nopaline strain (C58) when done with AS and 48-h-preincubated tobacco leaf segments, respectively (Table 2). Our results conform to the observations of Lee et al. (44), who reported a 3.5-fold-higher induction of a virB_A6::lacZ fusion (pSM243cd) in the octopine strain (A6) than in the nopaline strain (C58) when the fusion in both strains was induced with AS. As expected for the supervirulence background of Bo542 (27, 44), induction of virE_A6::lacZ in the l,l-succinamopine strain (A281) was 1.5-fold higher than in the octopine strain (A348) with AS and with 48-h-preincubated tobacco leaf segments (Table 2).

TABLE 2.

Induction of the octopine virE_A6::lacZ fusion (pSM358cd) in different Ti plasmid backgrounds when induced with AS (60 μM) or with 48-h-preincubated tobacco leaf segments

Strain	Type	β-Galactosidase activitya (Miller units)
		MS control	MS + AS (60 μM)	MS + LS-48 hb
A348(pSM358cd)c	Octopine	19 ± 1	416 ± 32	478 ± 17
C58(pSM358cd)	Nopaline	13 ± 1	175 ± 4	149 ± 19
A281(pSM358cd)	l,l-Succinamopine	0.5 ± 0.1	638 ± 18	700 ± 68

^aValues are means (± standard errors) of six assays. 

^bMS induction medium with 48-h-preincubated tobacco leaf segments. 

^cThe plasmid pSM358cd carries the octopine virE_A6::lacZ fusion (69). 

One possible reason for the difference observed in vir induction levels of virE_A6 in octopine and nopaline Ti plasmid backgrounds could be an incomplete interaction between VirG of nopaline Ti plasmid and vir box of the reporter plasmid pSM358cd. To test this possibility, the reporter plasmid pSM358 harboring virE_A6::lacZ and virG_A6 was mobilized into A6, C58, and Bo542 Ti plasmid backgrounds. Now, the interaction between VirG_A6 and the vir box of virE_A6::lacZ was expected to be complete. However, virE_A6 induction in the octopine strain (A348) was higher than in the nopaline strain (C58) by 9-fold with AS and by 12-fold with 48-h-preincubated tobacco leaf segments (Table 3). Similarly, virE_A6 induction in the octopine strain (Ach5) was higher than that in the nopaline strain (T37) by 10-fold with AS and by 6-fold with tobacco leaf segments (Table 3). Thus, the supplementation of octopine virG_A6 along with octopine vir box in the virE_A6::lacZ fusion does not lead to effective completion of the vir gene induction pathway. The virE_A6 induction level in the nopaline strain is not elevated to the corresponding level in the octopine strain. However, the l,l-succinamopine Ti plasmid background (pTiBo542) supported vir gene induction in pSM358 at levels comparable to those obtained with octopine strains (Table 3). Since pSM358 exists in multiple copies in Agrobacterium, the copy number of virG_A6 is high. The high copy number of virG may have led to increased virE_A6::lacZ expression in the octopine strains A348 and Ach5 harboring pSM358 (in comparison to those harboring pSM358cd [Table 2]). Surprisingly, the higher virG_A6 copy number did not bring about a corresponding increase in virE_A6 induction in nopaline strains C58 and T37. It is quite possible that, in addition to a limitation in the interaction between nopaline VirG and the vir box of virE_A6 (Table 2), the compatibility between nopaline VirA and octopine VirG_A6 may also be low.

TABLE 3.

Effect of supplementation of pTiA6 virG in the restoration of induction of the virE_A6::lacZ fusion in the nopaline Ti plasmid backgrounds

Strain	Type	β-Galactosidase activitya (Miller units)
		MS control	MS + AS (60 μM)	MS + LS-48 hb
A348(pSM358)c	Octopine	116 ± 3	3,420 ± 242	3,570 ± 135
C58(pSM358)	Nopaline	77 ± 10	370 ± 9	301 ± 9
Ach5(pSM358)	Octopine	67 ± 4	2,723 ± 50	2,578 ± 417
T37(pSM358)	Nopaline	41 ± 2	272 ± 4	443 ± 60
A281(pSM358)	l,l-Succinamopine	93 ± 10	2,629 ± 136	3,335 ± 88

^aValues are means (± standard errors) of six assays. 

^bMS induction medium with 48-h-preincubated tobacco leaf segments. 

^cThe plasmid pSM358 harbors octopine virG in addition to the virE_A6::lacZ fusion (67). 

Supplementing virA_A6 brings about increased vir gene induction in nopaline strains.

In the nopaline strains, the presence of pTiA6 virG along with virE_A6::lacZ in pSM358 did not restore virE_A6 induction to levels obtained with octopine and l,l-succinamopine strains. We tested whether inclusion of virA_A6 in the above nopaline strains would complete the vir induction pathway and restore efficient vir gene induction, as virA, virG, and virE are derived from the same Ti plasmid, pTiA6. The pTiA6 virA was subcloned into pUCD2 (IncW group) to obtain the plasmid pAKM2 (Fig. 1), which is compatible with pSM358 (IncP group). The functional competence of the cloned virA_A6 in pAKM2 was confirmed by complementing the virA mutation in pTi237 (69). A. tumefaciens strain pTi237(pAKM2) formed tumors on tobacco leaf disks (results not shown).

The plasmid pAKM2 was mobilized into two nopaline strains, C58 and T37, and two octopine strains, A348 and Ach5, all of which harbored pSM358. virE_A6::lacZ was induced with AS or with 48-h-preincubated tobacco leaf segments (Table 4). When induction was done with 48-h-preincubated tobacco leaf segments, virE_A6 induction in T37 was completely (100%) restored to the level of an octopine strain (Ach5) and virE_A6 induction in C58 was restored to 55% of the level of an octopine strain (A348). When virE_A6::lacZ was induced with AS, restoration levels were 65 and 30% in T37 and C58 backgrounds, respectively (Table 4). Restoration in T37 was better than that in C58. We observed similar results in three independent experiments (data not shown). The reason for such a difference in restoration levels is not very clear.

TABLE 4.

Effect of supplementation of pTiA6 virA and virG in restoring the induction of the virE_A6::lacZ fusion in the nopaline Ti plasmid backgrounds

Exptc and strain	Type	β-Galactosidase activitya (Miller units)
		MS control	MS + AS (60 μM)	MS + LS-48 hb
Expt A
A348(pSM358)	Octopine	121 ± 2	3,779 ± 76	3,820 ± 76
A348(pSM358, pAKM2d)	Octopine	92 ± 2	3,354 ± 95	3,286 ± 37
C58(pSM358)	Nopaline	73 ± 1	720 ± 36	512 ± 79
C58(pSM358, pAKM2)	Nopaline	62 ± 1	1,002 ± 33	1,823 ± 76
Expt B
Ach5(pSM358)	Octopine	19 ± 2	2,507 ± 97	2,113 ± 258
Ach5(pSM358, pAKM2)	Octopine	12 ± 1	2,543 ± 112	3,178 ± 477
T37(pSM358)	Nopaline	61 ± 2	698 ± 31	592 ± 47
T37(pSM358, pAKM2)	Nopaline	33 ± 1	1,651 ± 68	2,236 ± 96

^aValues are means (± standard errors) of six assays. 

^bMS induction medium with 48-h-preincubated tobacco leaf segments. 

^cExperiments A and B were done independently. 

^dThe plasmid pAKM2 harbors octopine virA_A6 in pUCD2 (Table 1; Fig. 1). 

Restoration of complete virE_A6 induction by pAKM2 (virA_A6) in nopaline strains may be due to either qualitative differences between octopine (A6) VirA and nopaline (C58) VirA or a quantitative effect of multiple copies of virA_A6 from pAKM2. We evaluated whether multiple copies of nopaline virA_C58 can also restore complete virE_A6 induction in a nopaline strain. We amplified pTiC58 virA and subcloned it into pUCD2 (IncW group) to obtain the plasmid pBAL20 (Fig. 1). The plasmid pBAL20 is compatible with pSM358 (IncP group). Complementation of the virA_C58 insertional mutation in pTiC58 of A. tumefaciens strain LBA4301(pJK107) (59) with pBAL20 was done to check the functional competence of the cloned virA_C58. A. tumefaciens strain LBA4301(pJK107, pBAL20) formed tumors on tobacco leaf disks (results not shown), confirming the functional competence of the PCR-amplified virA_C58.

The plasmid pBAL20 was mobilized into the nopaline strain C58 and the octopine strain A348, both of which harbored pSM358. The virE_A6::lacZ fusion was induced with AS or with 48-h-preincubated tobacco leaf segments (Table 5). Though pAKM2 (virA_A6) restored virE_A6 induction in C58(pSM358) to a higher level, no such restoration was observed upon supplementation with nopaline virA_C58 (pBAL20). Thus, multiple copies of virA_C58 do not help in restoring virE_A6 induction in a nopaline strain (Table 5). The results presented in Tables 4 and 5 clearly indicate that both virA and virG from pTiA6 have to be supplemented for effective restoration of virE_A6 induction in nopaline Ti plasmid backgrounds. This indicates that efficient interaction between vir genes is necessary for optimal vir gene induction. This is possible only if the vir genes are derived from the same Ti plasmid.

TABLE 5.

Effect of supplementation of pTiC58 virA in restoring the induction of the virE_A6::lacZ fusion in the nopaline Ti plasmid background

Strain	Type	β-Galactosidase activitya (Miller units)
		MS control	MS + AS (60 μM)	MS + LS-48 hb
A348(pSM358)	Octopine	97 ± 2	3,121 ± 124	2,560 ± 160
C58(pSM358)	Nopaline	82 ± 10	576 ± 43	365 ± 54
C58(pSM358, pAKM2)	Nopaline	89 ± 5	1,667 ± 121	2,180 ± 239
C58(pSM358, pBAL20c)	Nopaline	46 ± 3	449 ± 5	431 ± 58

^aValues are means (± standard errors) of six assays. 

^bMS induction medium with 48-h-preincubated tobacco leaf segments. 

^cPlasmid pBAL20 harbors virA_C58 in pUCD2 (Table 1; Fig. 1). 

Induction of nopaline virE in different Ti plasmid backgrounds.

Based on our induction experiments with pSM358cd (Table 2) and pSM358 (Table 3), we proposed that the difference in vir gene induction levels between octopine and nopaline Ti plasmid backgrounds is mainly due to incomplete transmission of the vir-inducing signal to virE_A6 via VirA and VirG of nopaline Ti plasmids. We performed a reciprocal experiment to test whether VirA and VirG of octopine Ti plasmids can efficiently transduce the signal to a nopaline vir box. We chose the reporter plasmid pUCD1553 (Table 1) (59), which carries a virE_C58::cat fusion. It was mobilized by triparental mating into two octopine strains (A348 and Ach5), two nopaline strains (C58 and T37), and the l,l-succinamopine strain (A281). Induction of virE_C58 was performed with AS and 48-h-preincubated tobacco leaf segments (Table 6). A chloramphenicol acetyltransferase assay was performed (63). Comparison between octopine and nopaline Ti plasmids was made in two sets (C58-A348 and T37-Ach5). The results of virE_C58 induction experiments are presented in Table 6. The induction of virE_C58 was higher in C58 than in A348 by 3.5-fold with AS and by 2.8-fold with 48-h-preincubated tobacco leaf segments. In a similar pattern, virE_C58 induction was higher in T37 than in the Ach5 Ti plasmid background by 2.2-fold with AS and by 2.8-fold with tobacco leaf segments. These results clearly indicate that nopaline virE_C58 interacts more efficiently with the nopaline VirA-VirG chain than with the octopine VirA-VirG chain.

TABLE 6.

Induction of a nopaline virE_C58::cat fusion in nopaline and octopine Ti plasmid backgrounds when induced with AS (60 μM) or with 48-h-preincubated tobacco leaf segments

Exptc and strain	Type	Sp act of chloramphenicol acetyltransferasea (nmol min−1 mg−1 of protein)
		MS control	MS + AS (60 μM)	MS + LS-48 hb
Expt A
A348(pUCD1553d)	Octopine	188 ± 5	1,228 ± 272	1,680 ± 130
C58(pUCD1553)	Nopaline	283 ± 25	4,358 ± 208	4,757 ± 133
Expt B
Ach5(pUCD1553)	Octopine	222 ± 8	1,883 ± 100	1,591 ± 60
T37(pUCD1553)	Nopaline	243 ± 15	4,084 ± 845	4,500 ± 275
Expt C
C58(pUCD1553)	Nopaline	181 ± 33	4,029 ± 144	3,505 ± 361
A281(pUCD1553)	l,l-Succinamopine	56 ± 8	940 ± 78	929 ± 48

^aValues are means (± standard errors) of six assays. 

^bMS induction medium with 48-h-preincubated tobacco leaf segments. 

^cExperiments A, B, and C were performed independently. 

^dPlasmid pUCD1553 harbors the virE_C58::cat fusion (59). 

Results from Tables 2 and 3 show that VirA and VirG of the l,l-succinamopine Ti plasmid (pTiBo542) transduce the signal efficiently to octopine virE_A6::lacZ. We tested whether pTiBo542 VirA and VirG can transduce the vir induction signal effectively to nopaline virE_C58. The reporter plasmid pUCD1553 with virE_C58::cat was mobilized into A281. Chloramphenicol acetyltransferase assays were performed to evaluate virE_C58 induction. The induction of virE_C58 in a nopaline strain (C58) was higher than in an l,l-succinamopine strain (A281) by 4.3-fold with AS and by 3.8-fold with 48-h-preincubated tobacco leaf segments (Table 6). These results indicate that vir signal transduction between l,l-succinamopine VirA and VirG to nopaline virE_C58 is less efficient.

Supplementation of nopaline virA and virG brings about increased virE_C58::cat induction in an octopine strain.

Results from Tables 4 and 5 show that both virA and virG of pTiA6 have to be supplemented for effective restoration of virE_A6 induction in a nopaline strain (C58) harboring virE_A6::lacZ fusion. We performed a reciprocal experiment to test whether both virA and virG of pTiC58, when supplemented in an octopine strain harboring pUCD1553, would effectively restore virE_C58::cat induction to the levels obtained in nopaline strains. We constructed a plasmid, pBAL23, which harbors PCR-amplified virA and virG of pTiC58 in pMH1002 (Table 1). The plasmid pMH1002 (IncP group) is a derivative of pGA472 (1), which is compatible with pUCD1553 (IncW group). The functional competence of amplified virG_C58 was checked by complementing the virG insertional mutation in A. tumefaciens strain LBA4301(pJK710) (59). Plasmid pBAL22 (Table 1; Fig. 1), which harbors amplified virG_C58 in pMH1002, was used for the complementation analysis. A. tumefaciens strain LBA4301(pJK710, pBAL22) formed tumors on tobacco leaf disks (results not shown). The results confirmed the functional competence of PCR-amplified virG_C58.

The plasmid pBAL23 harboring both virA and virG of pTiC58 was mobilized into the octopine strain A348 and the nopaline strain C58, both of which harbor pUCD1553. The virE_C58::cat fusion was induced with AS or with 48-h-preincubated tobacco leaf segments (Table 7). As observed above (Table 6), virE_C58 induction was higher in C58 than in A348. When virA and virG of pTiC58 were supplemented in an octopine strain (A348) harboring pUCD1553, virE_C58 induction was restored to levels comparable to those of the nopaline strain (C58). Induction of virE_C58 in A348(pUCD1553, pBAL23) was compared to that in C58(pUCD1553, pBAL23) to account for the multiple copies of virA_C58 and virG_C58 from pBAL23. When virE_C58 was induced with AS, virE_C58 induction in A348 was restored to 70% of that obtained with the nopaline strain (C58) (Table 7). Similarly, when virE_C58 was induced with 48-h-preincubated tobacco leaf segments, restoration was to a level of 62% (Table 7). These results clearly indicate that virA and virG of pTiC58 together can effectively restore the virE_C58 induction in the octopine Ti plasmid background.

TABLE 7.

Effect of supplementation of pTiC58 virA and virG in restoring the induction of the virE_C58::cat fusion in the octopine Ti plasmid background

Strain	Type	Sp act of chloramphenicol acetyltransferasea (nmol min−1 mg−1 of protein)
		MS control	MS + AS (60 μM)	MS + LS-48 hb
C58(pUCD1553c)	Nopaline	138 ± 5	3,877 ± 354	4,031 ± 381
A348(pUCD1553)	Octopine	154 ± 31	1,555 ± 234	1,248 ± 56
A348(pUCD1553, pBAL23d)	Octopine	200 ± 12	5,244 ± 222	4,745 ± 215
C58(pUCD1553, pBAL23)	Nopaline	215 ± 12	7,442 ± 45	7,695 ± 467

^aValues are means (± standard errors) of six assays. 

^bMS induction medium with 48-h-preincubated tobacco leaf segments. 

^cPlasmid pUCD1553 harbors the virE_C58::cat fusion (59). 

^dPlasmid pBAL23 harbors virA_C58 and virG_C58 in pMH1002 (Table 1; Fig. 1). 

Results from Tables 4 to 7 indicate that efficient interaction between vir genes is necessary for optimal vir gene induction. This is possible only if the vir genes are derived from the same Ti plasmid. The interaction is less efficient when they are derived from different Ti plasmid backgrounds.

Promotion of tobacco transformation by pTiBo542 (supervirulent) virG in different vir helper backgrounds.

Supplementation of Bo542 virG into octopine strains leads to increased tumorigenesis (supervirulence) in tomato and tobacco plants (27). Such an increase in virulence would be possible only when l,l-succinamopine virG effectively interacts with the vir boxes of vir genes in pTiA6 (A348). We evaluated whether the differences that we observed in the interactions between different vir components in vir gene induction studies get reflected in the transformation efficiencies of different vir helper strains.

We performed transformation of tobacco leaf rings with the following vir helper strains: LBA4404 (an octopine type), EHA105 (an l,l-succinamopine type), and PMP90 (a nopaline type). The plasmid pBAL2, a pGA472 derivative harboring CaMV35S-hph and nos-nptII as plant selection markers, was used as the binary vector. It has CaMV35S-gusA with a catalase intron as a reporter gene. Extra copies of pTiBo542 virG were provided in these three strains in the form of pBAL3. pBAL3 is a pUCD2-based plasmid harboring the SalI fragment 10 of the vir region from pTiBo542. The SalI fragment 10 has complete virG and virC genes. Transformation of tobacco leaf rings was performed with all three vir helpers harboring either pBAL2 or both pBAL2 and pBAL3. Transformation efficiency was expressed as the number of kanamycin-resistant shoot buds per leaf ring. Results reported in Tables 2, 3, and 6 indicate that VirG of pTiBo542 efficiently interacts with the vir box of virE_A6 but not with virE_C58 vir box. We expected increased transformation efficiency in the octopine vir helper background since virG of pTiBo542 can communicate well with vir boxes of pTiA6. The results in Table 8 were obtained after 12 days of selection on kanamycin (100 μg/ml) in an MS shoot-inducing medium (61). LBA4404(pBAL2, pBAL3) showed a significant increase (65%) in transformation efficiency over LBA4404(pBAL2). A moderate increase of 15% (statistically not significant) in transformation efficiency was observed for EHA105(pBAL2, pBAL3) over EHA105(pBAL2). No difference in transformation efficiency was observed in the nopaline strain PMP90(pBAL2), with and without pBAL3 (Table 8).

TABLE 8.

Effect of multiple copies of Bo542 virG on stable transformation efficiencies in tobacco leaf rings when infected with vir helper strains of different Ti plasmid backgrounds^a

Strain	vir helper type	Presence or absence of extra copies of virG from pTiBo542b	Transformation efficiency (no. of Kmr shoot buds/leaf ring)
LBA4404(pBAL2c)	Octopine	−	26 ± 3B
LBA4404(pBAL2, pBAL3d)	Octopine	+	43 ± 4D
EHA105(pBAL2)	l,l-Succinamopine	−	34 ± 3C
EHA105(pBAL2, pBAL3)	l,l-Succinamopine	+	39 ± 4CD
PMP90(pBAL2)	Nopaline	−	19 ± 3A
PMP90(pBAL2, pBAL3)	Nopaline	+	16 ± 2A

^aForty leaf rings were used for each treatment. The Km^r shoot buds were scored after 12 days. MS shoot-inducing medium (61) with 100 μg of kanamycin per ml and 250 μg of cefotaxime per ml was used. Untransformed leaf rings kept on MS shoot-inducing medium with 100 μg of kanamycin per ml did not give rise to any shoots. Different letters (A, B, C, and D) indicate significant differences at P < 0.01 by Student's t test. 

^bPlus and minus signs indicate the presence and absence, respectively, of plasmid pBAL3. 

^cPlasmid pBAL2 is a binary vector harboring nptII and hph as plant selection markers and gusA with an intron as a reporter (Table 1). 

^dPlasmid pBAL3 harbors the SalI fragment 10 of the pTiBo542 vir region having all of virG and virC in pUCD2 (Table 1). 

DISCUSSION

The levels of vir gene induction in different Ti plasmid backgrounds were different. We observed two- to threefold- and four- to fivefold-higher vir gene induction in octopine (A348) and l,l-succinamopine (A281) strains, respectively, than in the nopaline strain (C58), when the virE_A6::lacZ translational fusion plasmid (pSM358cd) was used (Table 2). Earlier, Lee et al. (44) reported three- and sevenfold-higher levels of vir gene induction with AS in an octopine type (A6) and an l,l-succinamopine type (Bo542), respectively, than in C58 (a nopaline-type strain) when pSM243cd (a virB_A6::lacZ translational fusion) was used as a reporter plasmid. A348, A281, and C58 strains used in our induction experiments have the same chromosome (C58) background. Therefore, the differences observed in virE_A6 induction are due to differences in Ti plasmid backgrounds.

Engstrom et al. (17) analyzed induction of Vir protein synthesis by [³⁵S]methionine incorporation. They found that the level of AS-induced Vir protein synthesis in octopine strains carrying pTiA6 and its related plasmids is comparatively lower than in nopaline strains harboring pTiC58 and its derivatives. AS-induced Vir protein synthesis levels were comparable in Bo542 and C58 strains. The observations of Engstrom et al. indicate that induction of Vir protein synthesis is inherently slightly higher in the nopaline (C58) strain than in the octopine (A6) strain. Therefore, the lower level of induction of virE_A6 in the nopaline strain (C58), as shown in Table 2, is possibly due to an incomplete interaction between octopine and nopaline vir components (e.g., VirG of nopaline Ti plasmid pTiC58 and vir box of virE_A6 of pSM358cd). The reciprocal experiments for nopaline virE_C58::cat induction in octopine and nopaline Ti plasmid backgrounds (Table 6) add further evidence to the incomplete interaction between VirG of A348 and the nopaline vir box in the virE_C58::cat fusion of pUCD1553. The induction of the virE_C58::cat fusion was two- to threefold higher in nopaline (C58 and T37) strains than in octopine (A348 and Ach5) and l,l-succinamopine (A281) strains (Table 6). The results obtained from induction studies with the virE_A6::lacZ fusion (Table 2) and the virE_C58::cat fusion (Table 6) indicate that octopine-type (A348 and Ach5) and l,l-succinamopine-type (A281) VirG can interact efficiently with the vir box of virE_A6 but not with the vir box of virE_C58. In contrast, VirG of nopaline-type (C58 and T37) strains interacts more efficiently with the vir box of virE_C58 than with the vir box of virE_A6.

The level of vir gene induction was approximately 1.5-fold higher in A281 than in A348 when pSM358cd was used as a reporter plasmid (Table 2). This could probably indicate that VirG of pTiBo542 interacts more efficiently with the vir box of virE_A6, thereby contributing to the supervirulence of pTiBo542 virG in octopine strains as observed earlier (34).

An incomplete interaction between the VirG encoded by the nopaline Ti plasmid, pTiC58, and the octopine vir box of the virE_A6::lacZ fusion (pSM358cd) is proposed as the cause for lower virE_A6 induction in nopaline strains. This limitation was expected to be overcome by using pSM358, which carries an octopine virG (from pTiA6) along with the virE_A6::lacZ fusion, as the reporter plasmid. Surprisingly, virE_A6::lacZ induction in octopine and l,l-succinamopine strains was found to be 6- to 12-fold higher than in nopaline strains even after supplementation with octopine virG along with the virE_A6::lacZ fusion (Table 3). The induction of virE_A6::lacZ in pSM358 should have been at levels comparable between octopine and nopaline Ti plasmid backgrounds, if nopaline VirA encoded by the Ti plasmid interacts efficiently with octopine VirG encoded in pSM358. The lack of elevation of virE_A6::lacZ induction in nopaline strains to a level comparable to that of octopine strains even after supplementation with octopine virG suggested that interaction between nopaline VirA and octopine VirG may also be incomplete.

Inclusion of virA_A6 in nopaline strains harboring pSM358, which brought octopine virA, virG, and virE together, restored virE_A6::lacZ induction in nopaline strains to 30 to 100% of the level of vir gene induction observed in octopine strains (Tables 4 and 5). Restoration of virE_A6 in C58(pSM358) by virA_A6 (pAKM2) (Table 4) but not by virA_C58 (pBAL20) (Table 5) revealed a qualitative difference between the octopine and nopaline VirA proteins in their interactions with octopine VirG. Thus, the vir gene induction pathway seems to operate efficiently only when the three important components, virA, virG, and vir box, are derived from the same (octopine) Ti plasmid.

In the reciprocal analysis of nopaline virE_C58::cat induction in an octopine strain (A348) harboring pUCD1553, supplementation with both virA_C58 and virG_C58 as pBAL23 resulted in near-complete (62 to 70%) restoration of virE_C58 in the octopine strain (A348) (Table 7). These observations show that efficient nopaline vir gene induction occurs only when all three vir components, virA, virG, and vir box, are derived from the nopaline Ti plasmid.

Belanger et al. (3) reported that compatibility between VirA and ChvE differs between D10B/87 (a biovar 2 nopaline strain) and C58 (a biovar 1 nopaline strain). The wild-type D10B/87 (with pTiD10B/87 Ti plasmid and D10B/87 chromosome background) strain exhibited a high level of vir gene induction when the virB_A6::lacZ fusion (pSM243cd) was used as a reporter plasmid. However, vir gene induction was reduced drastically when the pTiD10B/87 Ti plasmid was mobilized into the C58 chromosome background. Belanger et al. concluded that the reduction in vir gene induction was due to dysfunctional interaction between ChvE of the C58 chromosome and VirA of the D10B/87 Ti plasmid. Our results from Tables 2 and 6 highlight the importance of compatibility between VirG and vir boxes of octopine and nopaline Ti plasmids.

Supplementation of Bo542 virG in an octopine strain (A348) resulted in increased tumorigenesis (supervirulence) in N. glauca leaf disks (34). Such an increase in virulence would be possible only when the Bo542 VirG efficiently interacts with the vir boxes of the octopine strain (pTiA6). Accordingly, we found that supplementation of Bo542 virG (as pBAL3) resulted in a significant increase in tobacco transformation efficiency only in the octopine vir helper background (Table 8). No increase in transformation efficiency was observed in the nopaline vir helper background. Similarly, Liu et al. (46) observed that supplementation of extra copies of octopine (pTiA6) virG or l,l-succinamopine (pTiBo542) virG in nopaline Ti plasmid backgrounds did not bring about a significant increase in transient transformation of celery and rice. This reconfirms our conclusions drawn from Tables 2, 3, and 6 that VirG of Bo542 interacts efficiently with the vir boxes of octopine strains but not with the vir boxes of nopaline strains.

A high level of structural and functional similarity between pTiBo542 and pTiA6 vir regions has been observed (6, 28, 42). The Bo542 virG exhibits 98% homology to virG of pTiA6 at the nucleotide level (6), while virG of C58 is 80% homologous to virG of pTiA6 at the nucleotide level (59). Our results from Tables 2, 3, and 6 also indicate that VirG of Bo542 interacts with the octopine vir box more efficiently than with nopaline VirG. Thus, the strategy of increasing transformation efficiency with additional copies of heterologous vir genes (23) will be feasible only when vir components of the Ti plasmid are compatible with those of the introduced vir genes.

Supplementation of Bo542 virG (pBAL3) in LBA4404 increased transformation efficiency by 65%, but a similar increase was not observed for EHA105 (Table 8). This is understandable, since EHA105 by itself is a vir helper derived from pTiBo542 (29). Jin et al. (34) showed that supplementation of virG of Bo542 increased virulence of A348 on N. glauca leaf disks. However, such an increase was not found for A281. Any further increase in virulence of A281 required the supplementation of the complete virB operon of pTiBo542 in addition to its virG gene.

Hooykaas et al. (30) compared the virulence determinants in an octopine Ti plasmid (pTiB6), a nopaline Ti plasmid (pTiC58), and a Ri plasmid (pRi1855) by complementation analysis in Agrobacterium vir mutants. Functional complementation assays involved the restoration of virulence of vir mutants on tomato plants. Hooykaas et al. (30) initially performed complementation of nopaline vir mutant strains with R prime plasmids that harbor segments of the octopine Ti plasmid (pTiB6) virulence region (24). Nopaline vir mutants LBA2316 (virA) and LBA2362 (virB) were functionally complemented by pAL1813 (octopine virA through virO) and pAL1818 (octopine virA through virE). The corresponding octopine vir genes in pAL1818 similarly complemented LBA2315 (nopaline virC mutant) and LBA2371 (nopaline virE mutant). The above studies indicate a qualitative interaction between the octopine and nopaline vir components, virA, virB, virC, and virE. Surprisingly, there is no mention of complementation of virG mutant strains of either nopaline or octopine background. Similarly, Komari et al. (42) performed functional complementation analysis of pTiA6 vir mutants with vir clones from pTiBo542. Restoration of virulence was checked on K. diagremontiana and N. glauca. Komari et al. observed that pTiA6 virA, virB, virC, virD, virE, and virG mutant strains could be complemented by the corresponding vir clones of pTiBo542. Complementation of vir mutations of one Ti plasmid by the corresponding wild-type vir genes of a second Ti plasmid (25, 30, 40, 42, 59, 71) suggests the existence of qualitative interactions between the vir components of different Ti plasmids. However, our vir gene induction analysis using the virE_A6::lacZ fusion and the virE_C58::cat fusion in different Ti plasmid backgrounds gives an insight into the extent of quantitative differences existing in the interaction of vir genes of different Ti plasmids.