Co-dependent and Interdigitated: Dual Quorum Sensing Systems Regulate Conjugative Transfer of the Ti Plasmid and the At Megaplasmid in Agrobacterium tumefaciens 15955

Members of the Rhizobiaceae, often carry multiple secondary replicons in addition to the primary chromosome with compatible repABC-based replication systems. Unlike secondary chromosomes and chromids, repABC-based megaplasmids and plasmids can undergo copy number fluctuations and are capable of conjugative transfer in response to environmental signals. Several Agrobacterium tumefaciens lineages harbor three secondary repABC-based replicons, including a secondary chromosome (often linear), the Ti (tumor-inducing) plasmid and the At megaplasmid. The Ti plasmid is required for virulence and encodes a conjugative transfer (tra) system that is strictly regulated by a subset of plant-tumor released opines and a well-described acyl-homoserine lactone (AHL)-based quorum-sensing mechanism. The At plasmids are generally not required for virulence, but carry genes that enhance rhizosphere survival, and these plasmids are often conjugatively proficient. We report that the At megaplasmid of the octopine-type strain A. tumefaciens 15955 encodes a quorum-controlled conjugation system that directly interacts with the paralogous quorum sensing system on the co-resident Ti plasmid. Both the pAt15955 and pTi15955 plasmids carry homologs of a TraI-type AHL synthase, a TraR-type AHL-responsive transcription activator, and a TraM-type anti-activator. The traI genes from both pTi15955 and pAt15955 can direct production of the inducing AHL (3-octanoyl-L-homoserine lactone) and together contribute to the overall AHL pool. The TraR protein encoded on each plasmid activates AHL-responsive transcription of target tra gene promoters. The pAt15955 TraR can cross-activate tra genes on the Ti plasmid as strongly as its cognate tra genes, whereas the pTi15955 TraR is preferentially biased toward its own tra genes. Putative tra box elements are located upstream of target promoters, and comparing between plasmids, they are in similar locations and share an inverted repeat structure, but have distinct consensus sequences. The two AHL quorum sensing systems have a combinatorial effect on conjugative transfer of both plasmids. Overall, the interactions described here have implications for the horizontal transfer and evolutionary stability of both plasmids and, in a broad sense, are consistent with other repABC systems that often have multiple quorum-sensing controlled secondary replicons.

Members of the Rhizobiaceae, often carry multiple secondary replicons in addition to the primary chromosome with compatible repABC-based replication systems. Unlike secondary chromosomes and chromids, repABC-based megaplasmids and plasmids can undergo copy number fluctuations and are capable of conjugative transfer in response to environmental signals. Several Agrobacterium tumefaciens lineages harbor three secondary repABC-based replicons, including a secondary chromosome (often linear), the Ti (tumor-inducing) plasmid and the At megaplasmid. The Ti plasmid is required for virulence and encodes a conjugative transfer (tra) system that is strictly regulated by a subset of plant-tumor released opines and a well-described acyl-homoserine lactone (AHL)-based quorum-sensing mechanism. The At plasmids are generally not required for virulence, but carry genes that enhance rhizosphere survival, and these plasmids are often conjugatively proficient. We report that the At megaplasmid of the octopine-type strain A. tumefaciens 15955 encodes a quorum-controlled conjugation system that directly interacts with the paralogous quorum sensing system on the co-resident Ti plasmid. Both the pAt15955 and pTi15955 plasmids carry homologs of a TraI-type AHL synthase, a TraR-type AHL-responsive transcription activator, and a TraM-type anti-activator. The traI genes from both pTi15955 and pAt15955 can direct production of the inducing AHL (3-octanoyl-L-homoserine lactone) and together contribute to the overall AHL pool. The TraR protein encoded on each plasmid activates AHL-responsive transcription of target tra gene promoters. The pAt15955 TraR can cross-activate tra genes on the Ti plasmid as strongly as its cognate tra genes, whereas the pTi15955 TraR is preferentially biased toward its own tra genes. Putative tra box elements are located upstream of target promoters, and comparing between plasmids, they are in similar locations and share an inverted repeat structure, but have distinct consensus sequences. The two AHL quorum sensing systems have a combinatorial effect on conjugative transfer INTRODUCTION Many members of the Alphaproteobacteria (APB) group contain large secondary replicons, in addition to the primary chromosomal replicon with the majority of core genes. These secondary replicons which include megaplasmids, chromids (large replicons, >350 kb, that carry essential genes but are thought to have arisen by horizontal gene transfer), and secondary chromosomes often have repABC-based replication origins (Pinto et al., 2012;Fournes et al., 2018). Several APB, notably within the Family Rhizobiaceae, harbor multiple large secondary replicons that are also stably maintained by compatible repABC systems. These plasmids have extensive systems for their retention and proliferation, including copy number control, efficient partitioning with the chromosomes, and conjugative transfer (Pinto et al., 2012).
The Ti plasmid from the plant pathogen Agrobacterium tumefaciens is one of the most well-studied repABC replicons (Gordon and Christie, 2014), in which control of plasmid copy number and horizontal transfer is directly linked to bacterial population dynamics resulting from interactions with host plants. Octopine-type and nopaline-type Ti plasmids of A. tumefaciens have been most extensively studied for this regulation, but multiple other agrobacterial and rhizobial plasmids exhibit similar regulatory patterns (He and Fuqua, 2006;Wetzel et al., 2015). A. tumefaciens causes the plant neoplastic disease crown gall, by driving cross-kingdom horizontal gene transfer to host plants, delivering a segment of the Ti plasmid called the T-DNA, into host cells (Chilton et al., 1977). This process is strictly regulated by plant-released signals via the VirA-VirG two-component system (Winans, 1991) which activates virulence (vir) gene expression including the VirB Type IV secretion system (T4SS) that delivers the T-DNA to plant cells (Christie, 2001). In the plant, the T-DNA is ushered into the nucleus and inserted into the host genome (Gelvin, 2003). Stable integration of the T-DNA and expression of the transferred genes results in neoplastic growth of the infected tissue, and visible galls. These genes also encode proteins which drive production of unusual metabolites called opines that are utilized as semi-exclusive nutrients by agrobacteria in the rhizosphere via catabolic functions primarily encoded on the Ti plasmids (Dessaux et al., 1998). A subset of these opines, called conjugal opines, stimulate interbacterial conjugative transfer of the Ti plasmid, and increased Ti plasmid copy number (Klapwijk et al., 1978;Ellis et al., 1982). In most cases, the conjugal opines act indirectly, by stimulating expression of an acylhomoserine lactone (AHL) quorum sensing (QS) mechanism. The Ti plasmid encodes a LuxI-type AHL synthase called TraI that synthesizes the AHL N-3-oxo-octanoyl-L-homoserine lactone (3-oxo-C8-HSL) (Fuqua and Winans, 1994;Hwang et al., 1994;Moré et al., 1996). As with many other AHL quorum sensing systems, the signal can diffuse across the membrane and out of the cell. However, as the population density of AHLproducing agrobacteria increases, the relative AHL concentration is also elevated, eventually signaling a bacterial quorum. The AHL-dependent, LuxR-type protein TraR, also encoded on the Ti plasmid, interacts with the AHL to activate transcription of conjugative transfer (tra and trb) genes and the repABC replication genes of the Ti plasmid, driving conjugation and increased copy number Li and Farrand, 2000;Pappas and Winans, 2003). In addition, TraR activates traI expression, resulting in a positive feedback loop on AHL production. Expression of the Ti plasmid traR gene is under the tight transcriptional control of the conjugal opine, through an opine-responsive regulator (Fuqua and Winans, 1994). Therefore, conjugative transfer and copy number of the Ti plasmid is strictly regulated by the plant-tumor-released conjugal opine via traR expression, and AHL-producing bacterial population density via the QS mechanism. Other Ti plasmidencoded functions modulate these processes, most notably the antiactivator TraM which directly inhibits TraR, and maintains the minimum threshold of the protein required to activate tra/trb and repABC expression (Fuqua et al., 1995;Hwang et al., 1995).
In the best studied A. tumefaciens strains, conjugative transfer and copy number of the co-resident At megaplasmid is not recognized to be strongly affected by the opine-responsive and quorum sensing control systems which regulate these processes for the Ti plasmid. Rather, At plasmid conjugative transfer is commonly regulated through the rctA and rctB repressor cascade, which seems to provide largely homeostatic control of At plasmid conjugation (Perez-Mendoza et al., 2005). However, in certain nopaline-type A. tumefaciens such as strain C58, the conjugal opines that activate quorum sensing can also stimulate transcription of the rctA and rctB genes (Lang et al., 2013), thereby providing a regulatory linkage between these plasmids.
Here, we show that the octopine-inducible, quorum sensing-dependent conjugation system of the Ti plasmid from A. tumefaciens 15955 interacts directly with the conjugation system of the co-resident At megaplasmid. We find that both plasmids contribute to the same AHL pool through independent traI AHL synthases, and that they each also encode AHLresponsive TraR transcription factors that can cross-regulate genes on both replicons, impacting the overall quorum-sensing response as well as conjugative transfer of both the Ti plasmid and the At megaplasmid. The co-regulation of conjugative transfer functions observed here suggests direct coordination between both replicons that may affect their populationlevel stability and increase overall horizontal transfer. Our observations provide another alternate example of complex interactions between conjugative transfer regulatory elements of co-resident repABC plasmids in the rhizobia (He et al., 2003;Castellani et al., 2019;Cervantes et al., 2019). Similar interactions are likely to occur within other systems with multiple repABC-based replicons, several examples of which exhibit quorum-sensing controlled replication and horizontal gene transfer systems.

Reagents, Media, Strains, and Growth Conditions
All strains, plasmids, and oligonucleotides used in this study are described in Supplementary Tables 1, 2. Chemicals, antibiotics, and culture media were obtained from Fisher Scientific (Pittsburgh, PA), Sigma Aldrich (St. Louis, MO), and Goldbio (St. Louis, MO) unless otherwise noted. Plasmid design and verification as well as strain creation was performed as described previously (Morton and Fuqua, 2012a) with specific details in the text. Oligonucleotide primers were ordered from Integrated DNA Technologies (Coralville, IA). Single primer extension DNA sequencing was performed by ACGT, Inc. (Wheeling, IL). Plasmids were introduced into Escherichia coli via transformation with standard chemically competent preparations and into A. tumefaciens via electroporation or conjugation. In-frame deletions of traI and traR were constructed using a previously described allelic replacement method (Morton and Fuqua, 2012a).

Genome Sequence and Bioinformatic Analysis
The genome assembly for A. tumefaciens 15955 was deposited into GenBank by Academia Sinica (Genbank Access. No. 7601828). Multiple and pairwise sequence alignments were performed with Clustal Omega (Sievers et al., 2011) using default parameters. Promoter sequence motifs were generated with WebLogo (Crooks et al., 2004).

Controlled Expression and Reporter Plasmids
Plasmid-borne expression constructs were created using the LacIQ encoding, IPTG-inducible series of pSRK vectors (Khan et al., 2008). Coding sequences of traR and traI from pAt15955 (At15955_50350 and At15955_49060, respectively) and pTi15955 (At15955_53990 and At15955_54620, respectively) were PCR amplified from A. tumefaciens 15955 gDNA using primers for the corresponding gene (Supplementary Table 3) and Phusion DNA polymerase (New England Biolabs, Beverly, MA). PCR products were ligated into pGEM-T easy (Promega, Madison, WI), excised by restriction enzyme digest, and ligated into either pSRKKm or pSRKGm. Final pSRK constructs were confirmed with sequencing and restriction enzyme digestion prior to electroporation into A. tumefaciens.
Promoter expression reporter plasmids were created in pRA301 (Akakura and Winans, 2002a). Promoter regions for traA (At15955_53920 for pTi15955 and At15955_50400 for pAt15955) and traI (∼500 bp) were amplified with PCR (as above) using A. tumefaciens 15955 gDNA and the appropriate primers (Supplementary Table 3). All reporter gene fusions have lacZ translationally fused in frame at the start codon of each gene. PCR products were then assembled into restriction enzyme digests of pRA301 in a two-part NEBuilder isothermal assembly reaction (New England Biolabs, Beverly, MA), verified by PCR and sequencing, and introduced into A. tumefaciens via electroporation.

Promoter Expression Assays
For octopine and AHL induction of traA and traI promoters, A. tumefaciens 15955 carrying the appropriate pRA301-based reporter constructs were grown to an optical density at 600 nm (OD 600 ) of 0.4-0.6 in ATGN supplemented with appropriate antibiotics and spotted onto 0.2 µm sterile filter disks (0.45 µm pore size) placed on small ATGN plates (35 × 10 mm) supplemented with octopine (3.25 mM) and/or AHL (0.1 µM 3oxo-C8-HSL) and incubated for 48 h at 28 • C. A. tumefaciens 15955 strains were then removed from the filter discs, diluted to an OD 600 ∼0.6 in sterile water and used as in standard β-galactosidase assays, as previously described (Miller, 1972).
To determine the effect of traR expression on promoter activities of P traA and P traI from pAt15955 and pTi15955, A. tumefaciens 15955 and A. tumefaciens NTL4 derivatives carrying pSRK-based traR expression constructs and pRA301based lacZ reporter constructs (Supplementary Table 1) were grown in ATGN liquid cultures with appropriate antibiotics and 400 µM IPTG in the presence or absence of AHL to an OD 600 ∼0.6 and used in standard β-galactosidase assays.

AHL Detection Assays
For AHL detection on solid media, we utilized the AHL biosensor strain [A. tumefaciens NTL4 (pCF218)(pCF372)] . This biosensor relies on a high expression, plasmid-borne copy of the AHL-responsive TraR protein from A. tumefaciens R10 (pCF218), and a second plasmid carrying a TraR-dependent promoter fused to lacZ (traI-lacZ, pCF372), both harbored in A. tumefaciens NTL4, a Tiplasmidless strain which lacks an AHL synthase, and therefore does not produce endogenous AHLs. The biosensor is a sensitive detection system for exogenous AHLs, expressing lacZ only when inducing AHLs are present (McLean et al., 1997). Strain derivatives to be tested for AHLs were grown overnight to an OD 600 0.4-0.6 in ATGN supplemented with appropriate antibiotics and then each strain to be tested was struck onto ATGN agar supplemented with 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) ∼1 cm from the AHL biosensor strain. Media used in assays to test for AHL production with A. tumefaciens NTL4-derivatives also had 400 µM IPTG unless otherwise noted. The AHL biosensor was also used to detect AHL in cell-free supernatant from A. tumefaciens strains. Strains to be tested were grown overnight in ATGN supplemented with appropriate antibiotics to an OD 600 of 0.4-0.6 and 1 mL of culture was then pelleted, supernatant removed and recentrifuged to remove additional cell debris, and filter sterilized (0.2 µm) to create cell-free supernatant (CFS). One hundred microliter of CFS was then added to cultures of the AHL biosensor, grown to an OD 600 0.4-0.6, and assayed for promoter activity using standard β-galactosidase assays. Relative AHL levels for CFS were then reported as Miller units normalized to the OD 600 of the tested culture.

Conjugation Experiments
Conjugation assays were performed as described previously (Heckel et al., 2014) with alterations. Briefly, liquid cultures of donors and recipients were grown to mid-logarithmic phase in ATGN at 30 • C and normalized to OD 600 ∼0.6, mixed 1:1, and 100 µL was spotted onto 0.2 µm pore size cellulose acetate filter disks placed on small ATGN plates (35 × 10 mm). Conjugations were performed on ATGN with A. tumefaciens 15955 ectopically expressing plasmidborne P lac -traR fusions with or without 400 µM IPTG and appropriate antibiotics. For AHL rescue of conjugation in the traI double mutant, varying amounts of 3-oxo-C8-HSL were also added. Following incubation (18 h, 28 • C), cells were suspended in liquid ATGN, serially diluted, and spotted onto media supplemented with appropriate antibiotics to enumerate donors and transconjugants. Conjugation efficiencies were determined as the ratio of transconjugants per output donor.

Statistical Analysis
Assays were performed with at least three biological replicates. Standard deviation was determined and presented as error bars. Where appropriate, significance was conservatively evaluated using 2-sided, paired t-tests.

Conjugation Systems of pAt15955 and pTi15955 Are Activated by Octopine
A. tumefaciens 15955 is a member of the Genomospecies Group 1 along with similar derivatives A. tumefaciens Ach5 and A6 (Lassalle et al., 2011). Sequence analysis of pAt15955 identified conjugation genes homologous to the quorum-sensing regulated tra/trb system encoded on pTi15955 and other Ti plasmids (Alt-Mörbe et al., 1996;Farrand et al., 1996), including tra and trb structural genes and homologs of the quorum sensing regulators traI, traR, and traM (Supplementary Figure 1A). The conjugation system encoded on pAt15955 is distinct from those found on many other At plasmids, including those of the type strain A. tumefaciens C58, a nopaline-type strain (Genomospecies 8), in which pAtC58 conjugates via a virB-like type IV secretion system (avhB) (Chen et al., 2002) regulated by rctA/rctB-like regulators commonly found for other rhizobial plasmids (Perez-Mendoza et al., 2005;Nogales et al., 2013) (Supplementary Figure 1). Given the presence of TraI, TraR and TraM homologs on pAt15955 (Supplementary Figures 1B,C), it was conceivable that the conjugative transfer of this plasmid might be regulated by quorum sensing and that pAt15955 and pTi15955 could contribute and respond to a common AHL pool. We previously showed that pAt15955 encodes a functional conjugation system and observed that conjugation of pAt15955 270 (a derivative with a 270 kb deletion on pAt) which lacks cis-encoded trb conjugation functions (Supplementary Figure 1A), could be stimulated by octopine-induction in a pTi15955-dependent manner (Barton et al., 2019) suggesting a direct interaction between the conjugation systems of both plasmids.
To compare the impact of octopine and AHL cues on the expression of pAt15955 and pTi15955 conjugative transfer (tra) genes, we constructed independent translational lacZ fusions to the upstream regions and start codons of the predicted traA nickase/helicase homologs and the traI AHL synthase homologs, carried on the broad host range plasmid pRA301 (Akakura and Winans, 2002b) (Supplementary Table 2, Methods). These fusion plasmids were introduced into wild type A. tumefaciens 15955, and used in β-galactosidase assays to measure gene expression in the presence or absence of 0.1 µM 3-oxo-C8-HSL and/or 3.25 mM octopine (Figure 1, Materials and Methods). The β-galactosidase activity imparted by the traA-lacZ fusions from pTi15955 (green bars) and pAt15955 (blue bars) was low in the absence of octopine or AHL addition ( Figure 1A). Addition of octopine alone (medium-fill bars) or octopine plus AHL (dark-fill bars) significantly increased activity from the traA At -lacZ and traA Ti -lacZ fusions, although overall activity of each was relatively low. The same pattern of regulation was observed for 15955 harboring the traI-lacZ fusions, although the level of activity from both traI Ti -lacZ fusions was considerably higher than the traA-lacZ fusions, and addition of octopine and AHL together had a stronger effect than octopine alone ( Figure 1B). The traI At -lacZ fusion (blue bars) showed a similar pattern although its overall expression was lower than the traI Ti fusion (green bars). Taken together with our prior observation of pTi15955-dependent octopine-induction of pAt15955 conjugation (Barton et al., 2019), this data suggests that pTi15955 is directly involved in activation of pAt15955 tra gene expression in response to octopine Winans, 1994, 1996). Additionally, octopine-induction had differential effects on pAt15955 and pTi15955 gene activation, where octopine addition stimulated comparable modest expression from both traA genes ( Figure 1A), but had a more dramatic effect on traI Ti expression than that of traI At (Figure 1B), suggesting complex cross-regulation between both plasmids and supporting our previous observations for pAt15955 conjugation control, as well as previous work on the activation of octopine-type Ti plasmid conjugation Winans, 1994).

A. tumefaciens 15955 Encodes two Functional Homologs of the traI AHL Synthase
The pAt15955 plasmid has two annotated traI homologs, one of which is encoded as the predicted first gene of a traItrb operon (traI At_1 , At15955_49060) in a position similar to pTi-encoded systems, and another ∼20 Kbp distant from trb in an uncharacterized locus (traI At_2 , At15955_RS25205) (Supplementary Figures 1A,C). Both are within the large segment of pAt15955 that is deleted upon curing of pTi15955 (blue bar in Supplementary Figure 1), and a pAt15955 derivative with this deletion is unable to conjugate independently (Barton et al., 2019). The annotated coding sequence of traI At_2 specifies a protein that has a 25 amino acid C-terminal extension relative to TraI Ti and TraI At_1 (Supplementary Figure 2). TraI At_1 is 100% identical to a TraI homolog we identified earlier from the related Genomospecies 1 strain A. tumefaciens A6, that we found to specify production of 3-oxo-C8-HSL (Wang et al., 2014). To evaluate the functionality of the three traI homologs in A. tumefaciens 15955, each coding sequence was fused in-frame with the lacZ start codon under control of P lac in the broad host range vector, pSRKGm, to create the following constructs: pSRKGm P lac -traI At_1 (pIB302), pSRKGm P lac -traI At_2 (pIB305) and pSRKGm P lac -traI Ti (pIB303) (Supplementary Table 2). These plasmids were introduced individually into A. tumefaciens NTL4, a strain that lacks the Ti plasmid and does not produce AHL (Supplementary Table 1) to create derivatives that were used in agar plate-based AHL crossfeeding assays (McLean et al., 1997) with the A. tumefaciens NTL4(pCF218)(pCF372) AHL biosensor (Figure 2A). With this biosensor, diffusible AHLs in the producing strain cause induction of the AHL-responsive biosensor resulting in blue X-Gal pigmentation via increased β-galactosidase activity. The biosensor responds optimally to 3oxo-C8-HSL, but will also detectably respond at diminishing levels to a broad range of non-cognate AHLs, from acyl chain lengths of C6-C14, and either 3-oxo-, 3-OH or fully reduced at the β-carbon (Zhu et al., 1998). Expression of either normal length traI from pAt15955 or pTi15955 grown in the presence of 400 µM IPTG to induce expression of the P lac -traI fusion was able to generate detectable AHL via the plate-based assay, validating that both genes are functional (Figure 2A). However, no AHL production was detected from the strain harboring the traI At_2 plasmid, suggesting it is non-functional or produces a product not recognized by the AHL biosensor strain.
To determine whether both functional traI genes are expressed in WT A. tumefaciens 15955 and contribute to the overall AHL pool, plate-based AHL detection assays were conducted (as above) with WT A. tumefaciens 15955 and strains isogenic for deletion of traI gene from the Ti plasmid ( traI Ti ), the At plasmid ( traI At_1 ), or both ( traI Ti traI At_1 ) (Supplementary Table 1, Figure 2B). AHL production was FIGURE 1 | Octopine-induced activation of P traA and P traI from pAt and pTi in WT A. tumefaciens 15955. WT A. tumefaciens 15955 carrying a plasmid-borne copy of either P traA (top) or P traI (bottom) from pTi15955 (green) or pAt15955 (blue) translationally fused at their start codons to lacZ were spotted on 0.2 µm cellulose acetate filter discs placed on ATGN with or without 3.25 mM octopine (Oct.) and/or 0.1 µM N-3-oxo-C8-HSL (AHL) and incubated at 28 • C for 48 h. Bars with no, light, medium, or dark fill indicate no addition, addition of AHL, addition of octpine, or addition of both octopine and AHL, respectively. Cells were resuspended and promoter activities were calculated as Miller Units (Methods). Bars under the same bracket are not significantly different, unless they are marked with an asterisk (p < 0.05). Bars under brackets that have the same number above them (1, 2, 3) in each panel (A,B) denote a significant difference in averages (p < 0.05) between the bars under the similarly numbered brackets. detected for WT A. tumefaciens 15955 and the individual deletion mutants, but the double mutant did not produce inducing levels of AHL (Figure 2B), indicating that both traI genes are expressed in A. tumefaciens 15955 and each contribute to detectable levels of AHL. To assess the relative amount of AHL produced by each TraI protein in standard growth media, quantitative β-gal assays were conducted with the AHL reporter (NTL4 pCF218 pCF372) using cell-free supernatant from the isogenic traI mutant strains ( Figure 2C). Cell-free supernatant from turbid cultures from either traI mutant strain provided half of the B-gal activity relative to WT 15955, which indicates roughly equal contribution FIGURE 2 | Qualitative and quantitative AHL assays of A. tumefaciens strains ectopically expressing traI genes, and traI mutant derivatives. (A) AHL production by A. tumefaciens NTL4 (Ti plasmidless, AHL − ) ectopically expressing traI genes from pAt15955 and pTi15955 expressed from P lac (top streaks). These strains were used in a plate-based AHL production diffusion assays (McLean et al., 1997) with NTL4(pCF218)(pCF372), (AHL reporter, bottom streak). The traI designations are listed above. Media was ATGN agar supplemented with IPTG (400 µM) and X-Gal (40 ug/ml). (B) A. tumefaciens 15955 or derivatives with clean deletions in traI from pTi15955 ( traI Ti ), pAt15955 ( traI At_1 ), or both ( traI Ti ; traI At_1 ) (top streak) were used in plate-based AHL production assays with A. tumefaciens NTL4(pCF218)(pCF372) (AHL reporter, bottom streak). (C) Cell-free supernatants from A. tumefaciens 15955 (WT, mixed green/blue bar) and derivatives mutated for traI on pTi15955 ( traI Ti , blue bar), pAt15955 ( traI At_1 , green bar), or both ( traI Ti ; traI At_1 , open bar) were added to subcultures of the AHL biosensor strain, A. tumefaciens NTL4(pCF218)(pCF372), and allowed to grow until mid-exponential phase. Standard β-gal assays were then conducted on reporter strain cultures to determine relative AHL levels in the cell-free supernatant, calculated as β-gal activity (Methods). AHL-indicates cell-free supernatant from the biosensor strain A. tumefaciens NTL4(pCF218)(pCF372) (gray bar). Assays performed as three biological replicates, and bars marked with asterisks are significantly different from wild type (p < 0.05).
from each traI under standard growth conditions in the absence of octopine. The double deletion mutant resulted in background activity equivalent to an AHL-control strain ( Figure 2C).

traI-Dependent Activation of pAt15955 and pTi15955 Conjugation by traR Ti
One explanation for octopine-induced activation of pAt15955 conjugation is direct regulation of pAt15955 tra genes by traR Ti .
To determine the effect of traR Ti on conjugation in A. tumefaciens FIGURE 3 | Stimulation of pTi15955 and pAt15955 conjugation by traR expression is dependent on pAt15955-and pTi15955-encoded traI. A tumefaciens 15955 derivatives with a GmR cassette integrated at a neutral location on pTi15955 (green bars) or a KmR cassette similarly integrated on pAt15955 (blue bars) harboring either a plasmid-borne copy of P lac -traR from pTi15955 (light green and light blue bars) or pAt15955 (dark green and dark blue bars) were mixed 1:1 with a plasmidless recipient (A. tumefaciens ERM52, SpR) and spotted on ATGN agar containing 400 µM IPTG. After 24 h incubation, transconjugants and donors were enumerated and conjugation frequencies were calculated as transconjugants per output donor (Methods). Green and blue bars depict pTi15955 and pAt15955 conjugation frequencies, respectively. Asterisks denote significant difference between averages relative to wild type (p < 0.05).
15955, we generated P lac -traR constructs fused in-frame at the lacZ start codon and carried on pSRKKm (pIB309) or pSRKGm (pIB307). These plasmids were independently introduced into WT A. tumefaciens 15955 derivatives that carry antibiotic resistance marked derivatives of either pTi15955 (IB123, Gm R ) or pAt15955 (IB125, Km R ), respectively, and assayed for conjugation efficiencies of pTi15955 or pAt15955 upon induction of traR Ti with 400 µM IPTG (Figure 3, green and blue bars, respectively). As with other Ti plasmids, pTi15955 plasmid does not conjugate at detectable levels in the absence of octopine, but this requirement can be bypassed with ectopic expression of traR (Fuqua and Winans, 1994). Expression of traR Ti stimulated conjugation of pTi15955 to a level consistent with our previous findings, and also strongly stimulated pAt15955 conjugation (Figure 3).
As with many quorum sensing systems, stimulation of conjugation by traR Ti in previously studied A. tumefaciens strains requires elevated levels of AHL, augmented by positive feedback on AHL production, to fully activate TraR by stabilizing functional dimerization and DNA binding (Hwang et al., 1999;Qin et al., 2000). To test how the two traI genes contribute to AHL-mediated induction of pTi15955 and pAt15955 conjugation, experiments were performed as above in derivatives deleted for traI Ti or traI At_1 , or both. Interestingly, conjugation of pTi15955 and of pAt15955 was activated by expression of traR Ti in the presence of either version of traI, but was extremely low or undetectable (<10 −7 conjugation efficiency) when both were mutated (Figure 3 light green and light blue bars). However, while expression of traR Ti was able to stimulate equivalent levels of conjugation in the wild type or in the traI At_1 mutant, deletion of the traI Ti decreased conjugation frequency of both plasmids by 10-fold or greater, although it was still detectable over the traI Ti traI At_1 double mutant. Conjugation of pAt15955 from the traI Ti traI At mutant could be rescued with exogenous addition of 3-oxo-C8-HSL, confirming that the extremely low conjugation from this mutant was due to differences in AHL levels (Supplementary Figure 3).
traR At Is a Potent Activator of pAt15955 and pTi15955 Conjugation LuxR-type proteins are defined by several well-conserved amino acid positions in the N-terminal domain that mediate AHL recognition, and in their C-terminal domains that usually coordinate DNA binding. TraR At is very similar to TraR Ti (63.2% identity) and both share the majority of these defining conserved residues (Supplementary Figure 4). However, they differ at position 194 where TraR Ti has a well-conserved isoleucine (I194) residue whereas at this position TraR At has an alanine (A194) residue, an amino acid residue not found in most other LuxR homologs (Whitehead et al., 2001). To determine whether the traR homolog from pAt15955 could stimulate conjugation of either pAt15955 or pTi15955, conjugation experiments were performed as above with ectopic expression of P lac -traR At (Figure 3, dark green and dark blue bars). The traR At plasmid was able to stimulate conjugation of pTi15955 and pAt15955 to a magnitude similar for traR Ti in a traI-dependent manner (Figure 3). However, in contrast to traR Ti , traR At was able to effectively stimulate conjugation of pTi15955 and pAt15955 when either traI gene, traI Ti or traI At_1 was present (Figure 3, dark green and dark blue bars), suggesting that traR At can potentiate AHL production from either traI gene.

Differential Activation of AHL Production by TraR Ti and TraR At Depends on Each traI Gene
Our findings have revealed that the traI genes from pAt15955 and pTi15955 are activated by octopine and AHL, presumably by one or both TraR proteins (TraR Ti and TraR At ), but that the traI Ti gene is more strongly expressed than traI At_1 ( Figure 1B). As such, one explanation for the reduced conjugation frequencies in the traI Ti mutant is that although TraI At_1 generates similar basal levels of AHL in the absence of octopine induction (Figures 2B,C), its expression is less strongly activated by TraR Ti and thus does not create the same level of positive feedback to drive maximal AHL production. We performed plate-based AHL detection assays with the isogenic traI mutants harboring the P lac -traR Ti plasmid ( Figure 4A). Ectopic expression of traR Ti clearly drives strong AHL synthesis in wild type, consistent with positive autoregulation of AHL production by activated TraR. This elevated AHL production was not observed in the traI Ti mutant or in the traI Ti traI At_1 double mutant, but was clear for the traI At_1 mutant ( Figure 4A). This observation suggested that the weak TraR Ti activation of traI At_1 expression resulting in less pronounced increases in AHL levels might underlie the diminished conjugation observed in the traI Ti mutant. Conversely, we explored how traR At impacted AHL production with the isogenic traI derivatives expressing traR At (Figure 4B). In contrast to traR Ti , AHL production was enhanced by expression of traR At when either traI gene was present ( Figure 4B).

Mutational Analysis of traR Genes Reveals Overlapping Control of Target Promoters
In frame deletions were created in the traR Ti ( traR Ti ) and traR At ( traR At ) genes of A. tumefaciens 15955 to generate single mutants as well as a double mutant ( traR Ti traR At ). The plasmids carrying lacZ fusions to the traI Ti and traI At genes were independently introduced into these mutants, and β-galatosidase activity was measured following growth on solid medium under a range of different conditions (Figures 5A,B). In wild FIGURE 5 | Mutational analysis of traR gene with traI reporter fusions. WT A. tumefaciens 15955 and mutant derivatives ( traR Ti and traR At ), and ( traR Ti traR At ) carrying a plasmid-borne copy of either P traI from pTi15955 (A, green bars) or from pAt15955 (B, blue bars) translationally fused to lacZ were spotted on 0.2 µm cellulose acetate filter discs placed on ATGN on its own (open bars), with 0.1 µM N-3-oxo-C8-HSL (AHL, light-fill bars), with 3.25 mM octopine (Oct., medium-fill bars), and with both (dark-fill bars) Cells were resuspended and promoter activities were calculated as Miller Units (Methods). Bars are standard deviation. All bars that differ by 10-fold or more are significant (p < 0.05). Bars that differ by <10-fold, that are however significantly different (p < 0.05) are designated with a bracketed asterisk. type A. tumefaciens 15955, the traI Ti -lacZ fusion was strongly induced on media with octopine (3.25 mM) with or without the exogenous AHL 3-oxo-C8-HSL (0.1 µM) (Figure 5A). Octopinedependent induction was abolished in the traR Ti mutant, but retained in the traR At mutant. Basal level activity was also decreased in the traR Ti mutant but a modest increase was observed in the presence of AHL (Figure 5A). In contrast, basal activity from this lacZ fusion was slightly increased in the traR At mutant. The double mutant lacking both traR genes had very low activity that was unaffected by any of the different conditions.
Expression of the traI At_1 -lacZ fusion showed a different pattern of regulation in the mutant backgrounds. In the wild type background AHL alone was able to significantly increase β-galactosidase activity roughly 3-fold, and addition of octopine with and without exogenous AHL increased this even further ( Figure 5B). In the traR Ti mutant the overall level of β-galactosidase decreased about 10-fold relative to wild type, but addition of AHL conversely increased the activity 10-fold from that amount. Octopine addition had no effect in the traR Ti mutant. The activity imparted from the traI At_1 -lacZ fusion also decreased 10-fold in the traR At mutant relative to wild type, but in contrast to the traR Ti mutant exogenous addition of AHL alone had no effect, whereas octopine or AHL plus octopine both resulted in a 10-fold induction. As with the traR Ti -lacZ fusion, low level activity from the traI At_1 -lacZ fusion was unaffected by either octopine or AHL alone or in combination in the traI Ti traI At_1 mutant ( Figure 5B). Introduction of compatible plasmids expressing either the traR Ti or traR At gene from the P lac promoter into wild type 15955 and all three different traR mutants resulted in strong IPTG-inducible β-galactosidase activity from both the traI Ti -lacZ and traR At -lacZ fusions, in the absence of exogenous AHL or octopine (Supplementary Figure 5). This is consistent with other A. tumefaciens strains where ectopic expression of the traR protein, and its positive impact on AHL production, can bypass the need for conjugal opine addition or the effect of exogenous AHL (Fuqua and Winans, 1994). Ectopic expression of the traR Ti gene however stimulated notably lower β-galactosidase activity from the traI At_1 -lacZ fusion than it did for traI Ti -lacZ, and this decreased even further in the traR mutant backgrounds, with the lowest activity in the traR Ti traR At mutant. Ectopic expression of traR At seemed to have an equivalent impact on both promoters, and this was largely unaffected in the traR deletion mutants.

In vivo Target Gene Specificity for TraR Ti and TraR At
Our observations on plasmid conjugation, AHL production, and tra gene expression suggested that TraR Ti and TraR At might have different specificities for target promoters. Wellstudied TraR proteins are known to bind upstream of regulated promoters at conserved tra box elements Zhu and Winans, 1999), sequence motifs that can readily be identified on pTi15955 upstream of tra genes. Sequence analysis of the presumptive promoter regions of conjugative transfer genes on pAt15955 identified similar inverted repeat, tra boxtype elements (Supplementary Figures 6A,B), suggesting direct regulation by one or both TraR proteins. However, the tra box consensus motif for pAt15955-encoded genes is slightly different from that of the pTi15955 elements (Supplementary Figure 6B).
The traI-lacZ and traA-lacZ fusions from pAt15955 and pTi15955 were introduced into the AHL − Ti plasmid-cured A. tumefaciens NTL4 strain harboring either the traR Ti or traR At expression plasmids, that provide IPTG-inducible expression. Very low-β-galactosidase activity was imparted by all of the fusions in the absence of exogenous AHL, even with high level expression of either traR gene in the presence of 800 µM IPTG (Figure 6). In the presence of AHL, the traA At -lacZ was strongly activated by traR At , but about 4-fold more weakly activated by traR Ti , whereas the difference for activation of the traA Ti -lacZ fusion was less pronounced (Figure 6A, compare blue-filled bars relative to green-filled bars). Activation patterns for the traI fusions were similar to the traA fusions, though the magnitude of FIGURE 6 | Differential activation of pAt15955 and pTi15955 P traA and P traI by traR. Plasmid-borne copies of P traA -lacZ or P traI -lacZ from pTi15955 (green) or pAt15955 (blue) and P lac -traR with traR genes from either pAt15955 (At) or pTi15955 (Ti) were moved into A. tumefaciens NTL4 and β-galactosidase activity (MU) of P traA (A) or P traI (B) were determined in the presence or absence of 0.1 µM 3-oxo-C8-HSL (+ AHL and -AHL, respectively). Bars with the same numbers possess significant difference in averages (p < 0.05).
β-galactosidase activity was many fold greater ( Figure 6B). The traI At_1 -lacZ fusion was again more weakly activated by traR Ti than by traR At (blue-filled bars), with the activation of the traI Ti -lacZ closer to equivalency for the two regulators (compare greenfilled bars). Thus, for both gene fusions derived from pAt15955 the Ti-plasmid encoded traR leads to weaker induction, whereas the equivalent gene fusions from pTi15955 are strongly activated by both the TraR Ti and TraR At .

DISCUSSION
Members of the Family Rhizobiaceae often harbor multiple repABC-based secondary replicons that are compatible with each other and are replicated and segregated faithfully during the cell cycle (Cevallos et al., 2008;Castillo-Ramirez et al., 2009). Unlike secondary chromosomes and chromids, megaplasmids, and plasmids within this group often encode conjugation machinery and undergo copy number fluctuations tied with functions such as virulence, symbiosis, and horizontal transfer. The Ti plasmid from A. tumefaciens is one of the most well-studied repABCcontrolled replicons, where virulence and horizontal transfer are tightly controlled by environmental signals. In addition to proper coordination with the cell cycle, tight regulation of copy number and horizontal gene transfer of large secondary replicons serves to limit the costs associated with DNA replication and horizontal transfer.
The opine-dependent quorum sensing system is wellcharacterized in Agrobacterium species harboring the wide host range octopine-type and nopaline-type Ti plasmids, with prototypes pTiR10 and pTiC58, respectively (Gordon and Christie, 2014). Based on extensive genomic sequencing the Agrobacterium genus has been phylogenetically subdivided into over 10 different genomospecies (Lassalle et al., 2011), with pTiR10 found in Genomospecies 4 and pTiC58 found in Genomospecies 8. In these systems there is a single quorum sensing system comprised of the TraR transcription factor, the TraI AHL synthase, and the TraM antiactivator, which together regulate Ti plasmid conjugation and copy number. However, in several members of the Agrobacterium G1 genomospecies with octopine-type Ti plasmids, such as A. tumefaciens 15955, A. tumefaciens A6, and A. tumefaciens Ach5, the At plasmid carries a second TraR-TraI-TraM quorum sensing system, that is closely related to the Ti-encoded regulators. We have demonstrated that this second At-encoded system controls conjugative transfer through regulation of tra genes on the At plasmid, but that it can also have a profound impact on Ti plasmid functions. Likewise, the Ti-encoded system can impact At plasmid conjugative transfer, most notably imparting octopine-inducible control over pAt15955 conjugation (Barton et al., 2019). These co-resident quorum sensing systems interdigitate at multiple levels to regulate horizontal transfer of both plasmids (Figure 7). Both TraI gene products are highly similar to each other, and drive AHL synthesis. Although the chemical identity of the AHL is not formally shown here for A. tumefaciens 15955, it has been reported for the closely related A. tumefaciens A6 that both the Ti and At traI genes specify production of 3-oxo-C8-HSL (Wang et al., 2014). The TraI proteins between A6 and 15955 show 100% amino acid identity. Furthermore, we show here that either traI gene can function to facilitate conjugation of both plasmids, but loss of both traI genes together abolishes conjugation of each plasmid. Therefore, we predict that the two TraI proteins in A. tumefaciens 15955 primarily contribute to the same pool of 3-oxo-C8-HSL (Figure 7). This pool of AHL is required for TraR-mediated activation of tra genes and plasmid conjugation, through protein-ligand interactions with their amino-terminal ligand binding domains.
In the presence of inducing levels of AHL the TraR proteins can bind to DNA and activate expression of target promoters, and stimulate plasmid conjugative transfer. The presence of octopine, naturally produced by plant tissue transformed with T-DNA encoding the octopine synthase genes, has been shown to activate expression of the Ti plasmid traR gene in other octopine-type plasmids (Fuqua and Winans, 1994). The operon structure FIGURE 7 | Model of Dual Quorum Sensing Systems in A. tumefaciens 15955. Green shaded symbols representing proteins and genes from pTi15955, and blue shaded symbols represent proteins and genes from pAt15955. Solid arrows indicate translation of specific genes, dotted arrows represent enzymatic synthesis of AHL, and dashed arrows are regulatory impacts of each system. The thinner arrows from TraR Ti to the pAt15955 target genes reflect the observed weaker effect of this protein on their expression. upstream of the pTi15955 traR is identical to that on these better studied octopine-type Ti plasmids (Supplementary Figure 1B). We posit that octopine increases expression of traR Ti on pTi15955 as well, and that this increase stimulates expression of the traI Ti gene, which in turn elevates AHL levels, creating a positive feedback loop as observed for many LuxR-LuxI-type systems (Fuqua and Greenberg, 2002). Ectopically increasing expression of traR Ti bypasses the need for octopine. However, traR Ti only weakly stimulates the traI gene on the At plasmid, which in turn limits its impact on AHL synthesis, thereby explaining the weaker stimulation of both Ti plasmid and At plasmid conjugation in the traI Ti mutant (Figure 3). The traR At gene product appears to activate expression of both the traI At_1 and traI Ti with rough equivalence. However, there is no evidence that its expression is affected by octopine, and consistent with this, in the absence of the traR Ti gene octopine has no impact on the tra gene expression or conjugation of either plasmid. We hypothesize that under laboratory conditions, traR At expression is low, and insufficient to stimulate significant positive feedback on AHL production, and thus pTi15955 conjugative transfer is extremely infrequent or undetectable in the absence of octopine. The basal expression of the traI Ti gene is in fact modestly elevated in the traR At mutant raising the possibility that TraR At plays a mildly inhibitory role in the absence of octopine. In contrast to pTi15955, low level conjugation of pAt15955 is detectable in the absence of octopine (Barton et al., 2019). The observation that even this minimal conjugation is largely abolished in the traI Ti traI At double mutants suggests that quorum sensing is functioning to regulate the process even at its lowest levels. It also is plausible that expression of the traR At gene is regulated by a condition that we have not yet identified such as has been observed in other rhizobia, and under certain conditions might elevate pAt15955 and thereby indirectly pTi15955 conjugation (Banuelos-Vazquez et al., 2020).
The molecular mechanisms by which the two TraR proteins function in the same cell remains an area for future studies. There are abundant examples of multiple LuxR-type regulators functioning in a single bacterial species, often in some form of hierarchal network. For example, in Pseudomonas aeruginosa there are three distinct LuxR-type regulators, LasR, RhlR, and QscR with overlapping regulons (Ding et al., 2018). In other Alphaproteobacteria multiple LuxR proteins are also common. For example, in the marine bacterium Ruegeria sp. KLH11 the SsaR LuxR-type protein activates motility in response to long chain AHLs generated by the SsaI AHL synthase, and also regulates the activity of a second LuxR-type protein SsbR and two AHL synthases SsbI and SscI (Zan et al., 2012(Zan et al., , 2015. Multiple, co-resident LuxR-LuxI-type proteins are welldocumented in the rhizobia, and their regulatory interactions are notably complex and convoluted. Rhizobium leguminosarum biovar viciae has a cascade of multiple LuxR and LuxI-type proteins including several that control plasmid conjugation (Lithgow et al., 2002;Cervantes et al., 2019). The biocontrol agent known as Rhizobium rhizogenes K84 encodes two TraR homologs (TraR acc and TraR noc ) on the single pAtK84b plasmid that combine to control conjugative transfer of this plasmid (Oger et al., 1998). Remarkably, in K84 expression of each traR gene is regulated by a different conjugal opine, nopaline and agrocinopine A/B, and consequently both opines independently stimulate pAtK84b conjugation through a single conjugative machinery. Recently, two of the repABC plasmids of Sinorhizobium fredii GR64, pSfr64a, and pSfr64b have been reported to encode two separate TraR-TraM-TraI quorum sensing systems, in addition to a quorum sensing regulatory system encoded by the chromosome, NgrR-NgrI (Cervantes et al., 2019). All three of these quorum sensing systems impact each other and contribute to control of pSfr64a and pSfr64b conjugation. Similar to S. fredii GR64, A. tumefaciens 15955 and its immediate relatives have dual, highly conserved TraR-TraI-TraM systems on two separate plasmids that each encode their own conjugative transfer machinery. Although there is no chromosomal LuxR-LuxI system in A. tumefaciens 15955, our findings reveal that the plasmid encoded quorum sensing systems are clearly interdependent.
Dimerization is known to be a requirement for DNA binding among diverse LuxR homologs including TraR. Given the similarity of the TraR Ti and TraR At proteins (63.2% identity), heterodimer formation is plausible, potentially altering the activity of each TraR protein. Interestingly, A. tumefaciens 15955 is known to encode a truncated copy of the traR Ti gene, named TrlR that is 88% identical, but ablated for the DNA binding domain, expression of which is inducible by the non-conjugative opine mannopine (Oger et al., 1998;Zhu et al., 1998). TrlR is a dominant negative homolog of TraR, that forms an inactive heterodimer with TraR Ti and prevents DNA binding (Chai et al., 2001). In the absence of mannopine, TrlR is not expressed. Our genetic analysis revealed that mutation of traR At modestly elevated expression of the traI Ti -lacZ fusion in response to octopine, suggesting that TraR At may also have a mild dominant negative effect on TraR Ti .
Another component of the quorum sensing systems commonly encoded on Ti plasmids is the TraM anti-activator (Fuqua et al., 1995;Hwang et al., 1995). TraM forms a multimeric complex with TraR, abrogating its activity and targeting it for proteolysis (Swiderska et al., 2001;Chen et al., 2007;Costa et al., 2012). A. tumefaciens 15955 encodes TraM proteins on pAt15955 and pTi15955, that are 65% identical. Genetic and biochemical analysis of a second TraM homolog from the close relative A. tumefaciens A6 demonstrated that it can form a complex with TraR encoded by the Ti plasmid and adopts the same three-dimensional structure as TraM Ti (Chen et al., 2006;Wang et al., 2014). The two TraM proteins seemed to be functionally redundant in A. tumefaciens A6, and mutation of both genes was required to disregulate the Ti plasmid TraR. Subsequent studies prior to genome sequencing also identified traR and traI genes linked to this second traM gene, all of which we now recognize to be carried on the pAtA6 plasmid (Wang et al., 2014). Although we did not examine the role for the traM Ti and traM At in the current study on A. tumefaciens 15955, it seems very likely that they play similar roles as those in A6. As such, the orchestration between the pTi15955 and pAt15955 quorum sensing systems to control plasmid transfer, very likely also includes these TraM anti-activators and their ability to inhibit their cognate and non-cognate TraR proteins.
We observed that TraR Ti only weakly activated the pAt15955 gene fusions tested, while it strongly affected the homologous pTi15955 gene fusions. In contrast, TraR At was largely equivalent in activation of both pTi15955 and pAt15955 targets, suggesting that it has a more permissive DNA binding specificity. Comparatively, the tra box elements identified upstream of these and other potential target genes, are similar 18 bp inverted repeats, but there are several differences at specific positions, that may distinguish the pTi15955 targets from the pAt15955 sequences. Amino acid sequence differences between TraR Ti and TraR At within the C-terminal DNA-binding domain are likely to impart the DNA binding site specificity. Although the proteins are quite similar overall in the DNA binding domain (69% identity), there are several differences including a conserved isoleucine residue (I194) in TraR Ti (either an I or V in most LuxR-type proteins, Whitehead et al., 2001), that is an alanine (A194) in TraR At (Supplementary Figure 6). The octopine-type A. tumefaciens R10 TraR protein which is encoded on pTiR10, was structurally characterized in association with the 3-oxo-C8-HSL ligand and bound to a synthetic tra box oligonucleotide (Zhang et al., 2002). Two of the primary contacts with DNA residues within the tra box were Arg206 (R206) and Arg210 (R210) which are conserved in the 15955 TraR Ti protein (Supplementary Figure 4, asterisks). However, these positions are divergent in 15955 TraR At and are Lys206 and Glu210 (Supplementary Figure 4). Thus, it seems the potential mechanism of tra box recognition may be different between these two regulators. Our findings suggest that sequence divergence between the two proteins imparts a less restrictive DNA binding site specificity to TraR At than to TraR Ti . However, testing the binding of each TraR paralogue to pAt15955 and pTi15955 promoters in vitro would provide an effective way to directly evaluate their specificity.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.