Rhizobium Leguminosarum Bv. Trifolii with Increased Nodulation Competitiveness ERIC W

Home , Rhizobium

Construction of a Symbiotically Effective Strain of Rhizobium leguminosarum bv. trifolii with Increased Nodulation Competitiveness ERIC W. TRIPLETT* Department ofAgronomy and Center for the Study of Nitrogen Fixation, 1575 Linden Drive, University of Wisconsin-Madison, Madison, Wisconsin 53706 Received 22 June 1989/Accepted 4 October 1989

Genes involved in nodulation competitiveness (tfx) were inserted by marker exchange into the genome of the effective strain Rhizobium leguminosarum bv. trifolii TAl. Isogenic strains of TAl were constructed which differed only in their ability to produce trifolitoxin, an antirhizobial peptide. Trifolitoxin production by the ineffective strain R. leguminosarum bv. trifolii T24 limited nodulation of clover roots by trifolitoxin-sensitive strains of R. leguminosarum bv. trifolii. The trifolitoxin-producing exconjugant TA1::10-15 was very competitive for nodulation on clover roots when coinoculated with a trifolitoxin-sensitive reference strain. The nonproducing exconjugant TA1::12-10 was not competitive for nodule occupancy when coinoculated with the reference strain. Tetracycline sensitivity and Southern analysis confirmed the loss of vector DNA in the exconjugants. Trifolitoxin production by TA1::10-15 was stable in the absence of selection pressure. Transfer of tfx to TAl did not affect nodule number or nitrogenase activity. These experiments represent the first stable genetic transfer of genes involved in nodulation competitiveness to a symbiotically effective Rhizobium strain.

Inoculation of legume plants with superior strains of Barry (2, 3), who has developed transposon Tn7 vectors for Rhizobium often fails to improve legume productivity. This this purpose. Another technique is marker exchange, in failure is caused by the ability of ineffective or inefficient which the DNA on a vector is homologous with a region on native strains of Rhizobium to prevent nodulation by the the chromosome. An incompatible plasmid and the appro- inoculum strains (9, 11, 17, 24, 25). Several laboratories are priate selection pressure are then used to force the DNA on working on various strategies to address this problem. In this the vector to exchange with the homologous region on the laboratory, we are characterizing the production of a very chromosome. Williams et al. (26) and O'Gara et al. (16) have potent antirhizobial compound, referred to as trifolitoxin, developed integration vectors for this purpose which include which is produced by Rhizobium leguminosarum bv. trifolii a gene from Rhizobium meliloti. The R. meliloti gene is used T24 (20-23). as a source of homology for the marker exchange event. Trifolitoxin production by T24 limits nodulation of clover Genes are chosen for this purpose which do not affect the roots by other strains of R. leguminosarum bv. trifolii (19, symbiosis. A multiple cloning site is included within the R. 21). Strain T24 is not useful as a solution to the Rhizobium meliloti gene. Insert DNA in that cloning site can then be competition problem since it induces ineffective nodules on integrated into the R. meliloti gene. clover. The trifolitoxin genes have been cloned and trans- In this article, the stable insertion and expression of the ferred to effective strains of Rhizobium by conjugation of a trifolitoxin genes into an effective strain of Rhizobium by a recombinant plasmid, pTFX1, which possesses the genes marker exchange event that utilizes homologous DNA necessary for trifolitoxin production and resistance (20). present in pTFX1 are described. However, recombinant plasmids based on pLAFR1 or pLAFR3, such as pTFX1, are commonly unstable in the MATERIALS AND METHODS absence of selection pressure (13, 15). Thus, Rhizobium transconjugants with pTFX1 are not likely to be able to limit Bacteria and plasmids. The bacterial strains and plasmids nodulation by trifolitoxin-sensitive indigenous strains of used in this work are listed in Table 1. Rhizobium under agricultural conditions where tetracycline Bacterial growth conditions. Rhizobium strains were cul- application is impractical. Although pTFX1 has been trans- tured at 28°C on Bergersen synthetic medium (BSM) as ferred to effective strains of Rhizobium (20), the nodulation described by Bergersen (4). Strains of Escherichia coli were competitiveness of the Rhizobium pTFX1-carrying transcon- cultured at 37°C on Luria-Bertani (LB) medium. Antibiotics jugants has not been tested because of the suspected insta- were added as needed at the following final concentrations: bility of pTFX1 following inoculation on legume roots. kanamycin, 50 ,ug/ml; tetracycline, 12.5 ,ug/ml; spectinomy- Experiments to test the nodulation competitiveness of trifo- cin, 50 ,ug/ml; streptomycin, 50 Fig/ml; gentamicin, 25 ,ug/ml; litoxin-producing, effective rhizobia have been delayed until nalidixic acid, 10 ,ug/ml; and neomycin, 75 ,ug/ml. the trifolitoxin production and resistance gene (tfx) has been Bacterial conjugations. Conjugation of the pTFX1::TnS inserted stably into the Rhizobium genome. mutants into Rhizobium strains was performed as described Two methods are now available for the insertion of foreign previously (21) with some modifications. The donor, recipi- genes into the chromosome of rhizobia and other gram- ent, and helper strains were mixed in a 1:1:1 ratio in water, negative bacteria. One such method is that described by each at a cell density of approximately 5 x 107 per ml. After being vortexed, a 5-pl suspension of this mixture was placed on a YM/KB (21) plate with 3% agar. After incubation for 2 * Electronic mail address: triplett@wiscmac3. days at 28°C, each mating mix was suspended in 0.1 ml of 98 VOL. 56, 1990 EFFECTIVE, TRIFOLITOXIN-PRODUCING RHIZOBIUM STRAIN 99

TABLE 1. Bacteria and plasmids Strain or plasmid Genotype or relevant characteristicsa Source or reference R. leguminosarum bv. trifolii T24 Tfx+ Tfxr Fix- Cmp+ 19 2046 Tfxs Fix' USDAb Beltsville Rhizobium collection TAl Tfxs Fix+ A. Gibson, CSIROC TA1::10-15 Tfx+ Tfxr Fix' Cmp+ TnS, Kmr Smr Tcs This work TA1::12-10 Tfx- Tfxr Fix+ Cmp- Tn5, Kmr Smr Tcs This work TA1(pTFX1) Tfx+ Tfxr Tcr TnS, Kmr Smr 20 TA1(pTFX1::10-15) Tfx+ Tfxr Tcr TnS, Kmr Smr This work TA1(pTFX1::12-10) Tfx- Tfxr Tcr Tn5, Kmr Smr This work E. coli DH5a Bethesda Research Laboratories Plasmids pTFX1 pLAFR3 derivative with tfx genes, Tcr 20 pTFX1::10-15 pTFX1 derivative with a Tn5 insertion adjacent to tfx 23 pTFX1::12-10 pTFX1 derivative with a Tn5 insertion inside tfx This work pRK2013 KmrTraT Mob' ColEl replicon 10 pRK2073 pRK2073 Spr::Tn7 14 pPHlJI Tra+ Mob' Gmr Spr Smr Cmr IncP replicon 5 a Tfx+, Trifolitoxin producing; Tfx-, non-trifolitoxin producing; Tfxr, trifolitoxin resistant; Tfxs, trifolitoxin sensitive; Fix', effective symbiotic nitrogen fixation; Fix-, incapable of symbiotic nitrogen fixation (ineffective); Nod, nodule induction; Tra, plasmid transfer function; Mob, plasmid mobilization function; Cmp, competitive for nodulation; Km, kanamycin; Tc, tetracycline; Sp, spectinomycin; Sm, streptomycin; Gm, gentamicin; Nal, nalidixic acid; Nm, neomycin. b USDA, U.S. Department of Agriculture. C CSIRO, Commonwealth Scientific and Industrial Research Organisation. water and spread-plated on a BSM plate prepared with and 5 ,ul of that suspension was spotted in the center of a Noble agar and supplemented with tetracycline and strepto- BSM plate for the assay of trifolitoxin production. A single mycin. The use of Noble agar in the interruption medium colony from the initial plate was used to inoculate a second decreased the background of the parental strains on the plate. After 2 days, confluent growth on the second plate was plates. After 5 days, transconjugants were observed. used to assay trifolitoxin. The assays continued for 10 such Conjugations involving the transfer of plasmid DNA be- purifications to single colonies or until trifolitoxin production tween strains of E. coli were conducted as described above was no longer observed. except that 5 ,ul of the mixture of donor, recipient, and Southern analysis. Total DNA of the TAl exconjugants helper strains was placed on an LB plate and incubated at was digested to completion with EcoRI. The fragments were 37°C overnight. Interruptions were done as described above separated by gel electrophoresis and blotted onto a nylon with the appropriate selective medium on solid LB medium. membrane (Nytran; Schleicher & Schuell, Keene, N.H.). In the transfer of plasmid DNA from E. coli to Rhizobium The blot was then probed with a digoxigenin-labeled strains, E. coli DH5a(pRK2013) was used as the helper pLAFR3 probe as described by the manufacturer (Genius strain. Kit; Boehringer Mannheim Corp., Indianapolis, Ind.). DNA isolation. Large-scale plasmid preparations were Plant culture and inoculation. Clover plants were cultured purified by the boiling method described by Holmes and and inoculated as described previously (21). The following Quigley (12). For restriction analysis of small amounts of plasmid DNA, plasmids were purified by the alkaline lysis method described by Ausubel et al. (1). Total genomic DNA 2046 TAl from strains of Rhizobium was isolated as described by Ausubel et al. (1). The plasmid pTFX1 was mapped by TnS mutagenesis and restriction enzyme analysis as described previously (23). Marker exchange. The method of Ditta (8) was used for the marker exchange of the trifolitoxin genes into the genome of R. leguminosarum bv. trifolii TAl (Fig. 1). The incompatible plasmid pPHlJI was conjugated into two TAl transconjugants with pTFX1::TnS. The conjugation was interrupted on BSM prepared in Noble agar and supplemented with gentamicin, kanamycin, and spectinomycin. The resulting exconjugants were replica-plated on BSM with tetracycline. The tetracycline-resistant colonies were discarded. Determination of the stability of trifolitoxin production in the absence of selection pressure. Strains T24, TA1(pTFX1), and trifolitoxin-producing TA1(pTFX1: :TnS) transconjugants and TAl::TFX-TnS exconjugants were streaked to single colonies on BSM medium in the absence of selective FIG. 1. Rough and smooth colony morphologies of R. legumin- antibiotics. After 2 days of incubation at 28°C, a portion of osarum bv. trifolii strains 2046 (A) and TAl (B) as one of the the confluent growth on the plate was suspended in water, methods used to distinguish the strains for nodule occupancy. 100 TRIPLETT APPL. ENVIRON. MICROBIOL. 10-15 A

T24

TA 1:: 10-15

FIG. 2. Restriction enzyme map of pTFX1, a recombinant plasmid containing the tfx region from strain T24. The symbols + and - represent the ability or inability, respectively, of pTFX1 containing a Tn5 insertion to confer trifolitoxin production in TAl. The locations of the Tn5 insertions in mutants pTFX1::10-15 (10-15) and pTFX1::12-10 (12-10) are shown adjacent to and within the tfx TA1::12-10 region, respectively.

four strains of R. leguminosarum bv. trifolii were used for FIG. 3. Trifolitoxin production by TA1::10-15, TA1::12-10, and T24 before (A) and after (B) four purifications on nonselective inoculation: 2046, TA1, TA1::10-15, and TA1::12-10. Strain medium. TAl and the TAl derivatives were coinoculated with the trifolitoxin-sensitive strain 2046. In treatments requiring coinculation, the two strains were mixed prior to inoculation RESULTS on the plant. Strains TAl and 2046 were chosen for this study because they differ greatly in colony morphology (Fig. Mutants chosen for the marker exchange experiments. Two pTFX1::TnS mutants were 1). This difference in colony morphology allowed us to easily chosen for the construction of distinguish the two strains following their isolation from isogenic, effective strains of Rhizobium which differed only in their ability to produce trifolitoxin. With such excon- nodules. There were five plants per treatment. Inoculum jugants, the contribution of trifolitoxin production to com- ratios were varied in was 10-fold intervals. Inoculum pre- petitiveness was accurately assessed in an effective strain. pared by diluting the cells in water to an A6. of 0.1, 0.01, One of the pTFX1::TnS mutants, pTFX1::10-15, contains a and 0.001. For the inoculation of individual clover strains, TnS insertion immediately adjacent to the tfx region (Fig. 2). 0.1 ml of the 0.01 A6. dilution was inoculated per plant. The This insertion had no effect on the ability of pTFX1::10-15 to actual number of CFU inoculated was determined by dilu- confer trifolitoxin production and resistance following con- tion plating. The number of CFU in 0.1 ml of inoculum at an jugal transfer to Rhizobium spp. The other mutant, A6. of0.01 was 3.6 x 105, 6.7 x 105, 5.6 x 105 and 8.5 x 105 pTFX1: :12-10, contains a TnS insertion at the edge of the tfx for strains 2046, TA1, TA1::10-15, and TA1::12-10, respec- region, 0.5 kilobase (kb) away from the Tn5 insertion in tively. The 0.1 and 0.001 Awo dilutions were used to prepare pTFX1::10-15 (Fig. 2). The insertion in pTFXl::12-10 pre- the 10:1 and 1:10 inoculum treatments, respectively. The vented the expression of trifolitoxin production in Rhizo- ratios in Table 2 reflect the actual ratios of CFU calculated bium spp. following conjugation. The pTFX1::12-10 inser- from dilution plating. tion had no effect on trifolitoxin resistance. Strain nodule occupancy. Four weeks after inoculation, Stability of the Qfx-containing exconjugants. The stability of nodules were harvested and the root nodule bacteria were these exconjugants was demonstrated by purifying the ex- isolated as described previously (21) with the exception that conjugants 10 times to single colonies on nonselective me- kanamycin was not added to the medium. The rhizobia were dium. After each purification, the strains were assayed for isolated from every nodule of every plant in the experiment. trifolitoxin production as well as for kanamycin and specti- Strains in each nodule were identified by colony morphology nomycin resistance. The TAl::10-15 exconjugant maintained (Fig. 1), kanamycin resistance, and trifolitoxin production. trifolitoxin production throughout the experiment (Fig. 3). Each isolate was examined for all three properties. The TAl The stability of trifolitoxin production by TAL::10-15 was exconjugants had the same colony morphology as wild-type very similar to that of wild-type T24 (Fig. 3). Kanamycin TAl. resistance, which is an indication of the presence of TnS, Nitrogenase activity. Nitrogenase was assayed as de- was also maintained. Spectinomycin resistance was lost scribed previously (21). after six purifications. This suggests the instability of the VOL. 56, 1990 EFFECTIVE, TRIFOLITOXIN-PRODUCING RHIZOBIUM STRAIN 101

TABLE 2. Nodule occupancy of clover roots following coinoculation with R. leguminosarum bv. trifolii TAl and A B trifolitoxin-producing and nonproducing exconjugants of TA1l Inoculum ratio, % Nodule occupancy ± SE TA1:2046 Inoculum TAl derivative 2046 1.0:0.0 TAl 100 O A 0O C TA 1) t(pTFX 0.0:1.0 2046 0 0 C 100 0 A 1.0:0.0 TA1::10-15 100 0 A 0 0 C 1.0:0.0 TA1::12-10 100 0 A 0 0 C 19.0:1.0 TA1 + 2046 87 10 A 49 1 B 1.9:1.0 TA1 + 2046 40 + 19 BC 90 7 A 1.0:5.3 TA1 + 2046 0 + 0 C 100 0 A 16.0:1.0 TA1::10-15 + 2046 94 + 4 A 10 ± 5 C 1.6:1.0 TA1::10-15 + 2046 91 ± 9 A 41 ± 14 B TA1(pTFX 1::10-15) 1.0:6.3 TA1::10-15 + 2046 0 ± 0 C 100 ± 0 A 24.0:1.0 TA1::12-10 + 2046 43 ± 13 BC 92 ± 5 A 2.4:1.0 TA1::12-10 + 2046 19 + 9 BC 100 ± 0 A 1.0:4.2 TA1::12-10 + 2046 10 ± 10 BC 100 ± 0 A a Each value is a mean of five replicates with standard errors also listed. FIG. 4. Trifolitoxin production by TA1(pTFX1) and TAl Values followed by the same letter are not statistically different at the 1% level (pTFX1::10-15) before (A) and after (B) 10 purifications on nonse- of confidence. Total nodule occupancy appears to be over 100% in some lective medium. treatments because many nodules were occupied by both inoculum strains. incompatible plasmid, pPHlJI, in the absence of selection pressure. As expected, the TA1::12-10 exconjugant main- Table 2 lists the nodule occupancy data for these treatments tained kanamycin resistance and failed to produce trifolito- as well as the control and other inoculum ratios. xin throughout the experiment (Fig. 3). This suggests that The trifolitoxin-producing exconjugant TA1::10-15 also the TnS insertion present in the tfx region of TA1::12-10 did occupied significantly more nodules than did wild-type TAl. not transpose to another region of the genome. As with the At an inoculum ratio of 1.9:1 in the treatment in which trifolitoxin-producing exconjugant, this exconjugant also wild-type TAl was coinoculated with 2046, 40% of the lost spectinomycin resistance after six purifications. nodules were occupied by TAl while 90% were occupied by The stability of the exconjugants was in sharp contrast to 2046 (Table 2). These numbers are significantly different the stability of TAl transconjugants, which possessed tfx on from those mentioned above for the TA1::10-15 plus 2046 a recombinant plasmid. In these transconjugants, trifolitoxin coinoculation treatment at the 1% level of confidence. Also, production was lost after four purifications in the absence of at approximately equal inoculum ratios, the TAl plus 2046 selection pressure (Fig. 4). treatment results were not significantly different from the Southern analysis of exconjugant DNA. Total DNA from results obtained from the TAl: :12-10 plus 2046 coinoculation the TAl exconjugants was cleaved with EcoRI and hybrid- treatment at the 1% confidence interval (Table 2). ized with a digoxigenin-labeled pLAFR3 probe. No hybrid- The effect of the insertion of pTFX1 DNA into TA1, apart ization was observed (data not shown), which suggests that from trifolitoxin production, on nodulation competitiveness the vector DNA is absent in the exconjugants. was determined by comparing the relative competitiveness Symbiotic effectiveness of the TAl exconjugants. Clover of wild-type TAl and TA1::12-10 versus that of 2046. No plants inoculated with either exconjugant were Fix'. The mean acetylene reduction rates (± standard error) for clover plants inoculated with wild-type TA1, TA1::10-15, and TA1::12-10 were 1.20 (+0.14), 1.29 (+0.37), and 0.96 (+0.08) nmol of C2H4 per h per plant, respectively. These three values were not significantly different at the 1% level of confidence. Nodule number (+ standard error) was also not affected by the transfer of tfx to TAl. Clover roots inoculated with TA1, TA1::10-15, and TA1::12-10 had 6.4 (±0.4), 4.4 (±1.0), and 7.8 (+1.4) nodules per plant, respectively. Nodule number did not differ significantly between the three treatments at the 1% level of confidence. Nodulation competition experiment. When the trifolitoxin- producing exconjugant TAl::10-15 was coinoculated with 2046 at an inoculum ratio of 1.6:1, the trifolitoxin-producing exconjugant occupied 91% of the nodules while the refer- 1:1 10:1 ence strain occupied 41% (Table 2, Fig. 5 and 6). The two Inoculum ratio numbers add up to more than 100% owing to dual-strain exconjugant reference occupancy of 32% of the nodules. When the trifolitoxin- FIG. 5. Clover nodule occupancy by the trifolitoxin-producing minus exconjugant TA1::12-10 was coinoculated with 2046 (Tfx+) exconjugant TA1::10-15 and the trifolitoxin-minus (Tfx-) at an inoculum ratio of 2.4:1, only 19% of the nodules vwere exconjugant TAl::12-10 when coinoculated with the trifolitoxin- occupied by the trifolitoxin-minus exconjugant, with 100% sensitive reference strain 2046. Error bars represent standard errors of the nodules occupied by 2046 (Table 2, Fig. 5 and 6). about the mean. 102 TRIPLETT APPL. ENVIRON. MICROBIOL.

not appear to be the homologous region responsible for the c 100 marker exchange, since stable exconjugants were not obtained when pTFX2 was used in the marker exchange > 80 experiments. The plasmid pTFX2 is a subclone of pTFX1 and contains only 25% of the insert DNA of pTFX1, includ- ing all of tfx and a TnS insertion (23). Thus, pTFX2 seems to * 60 Tfx lack the portion of pTFX1 required for maker exchange. In order to facilitate selection of the exconjugants, deriv- ' 40 atives of pTFX1 were chosen which contained TnS inser- or D I tions within adjacent to tfx. These Tn5 insertions served '0 two purposes in these experiments. First, selection for ': 20 NZ kanamycin resistance during isolation of the exconjugants ensured the exchange of tfx into the genome. Second, the kanamycin and streptomycin resistances expressed by the 1:10 1I1 10r1 20:1 Tn5 sequence in Rhizobium spp. served as markers for the Inoculum ratio exchange event. exconjugant: reference The nodulation competition experiment described here confirms previous observations that trifolitoxin production FIG. 6. Clover nodule occupancy by the trifolitoxin-sensitive the the reference strain 2046 when coinoculated with either a trifolitoxin- increases proportion of nodules occupied by trifoli- producing (Tfx+) exconjugant (TA1::10-15) or a trifolitoxin-minus toxin-producing strain (19, 21). However, the experiments (Tfx-) exconjugant (TA1::12-10). Error bars represent standard described here extend those observations in two important errors about the mean. ways. First, this is the first report of a symbiotically effective, competitive, trifolitoxin-producing strain. These experiments also show that the trifolitoxin-producing exconjugant significant difference in nodule occupancy was observed of TAl is as symbiotically effective as the wild-type TAl when 2046 was coinoculated with either TAl or TA1::12-10 strain. That is, trifolitoxin production does not decrease at nearly equal concentrations. symbiotic effectiveness. The ineffective phenotype of T24 is Dual-strain nodule occupancy was observed in up to 36% due to an as yet unidentified mutation rather than to a of the nodules in certain treatments (Table 2). Dual-strain pleiotropic effect caused by trifolitoxin production. occupancy is commonly observed in nodulation competition All previous competition experiments with trifolitoxin experiments with R. leguminosarum bv. trifolii (6, 7). Dual production have been done with the ineffective strain T24 or occupancy occurred with plants coinoculated TAl and 2046, derivatives of T24. Those experiments have shown that TA1::10-15 and 2046, or TA1::12-10 and 2046. trifolitoxin production benefits nodulation competitiveness, but the extent of that benefit was unclear since the compe- DISCUSSION tition was done by comparing Fix' and Fix- strains. Inef- fective strains commonly induce more nodules on a legume Isolation of nodulation competitiveness genes from a root than do effective strains. Thus, when the percent nodule Rhizobium strain by cosmid cloning and their subsequent occupancy is determined in such an experiment, the Fix- conjugal transfer to superior Rhizobium strains are insuffi- strain will always appear to be more competitive than if two cient to provide a useful inoculum for legume crops. Recom- Fix' strains had been compared. binant plasmids based on pLAFR3, such as pTFX1, are very Thus, with the construction of the TAl::10-15 and unstable in the absence of selection pressure (11, 13) (Fig. 4). TAl::12-10 exconjugants, competitiveness can be assessed Thus, in order to prevent the loss of the competitiveness without the complication of increased nodulation by an genes from the inoculum strains, it is essential to insert the ineffective strain. Nodule number per plant did not change genes into the genome of the inoculum strains. The stability significantly following transfer of tfx to TAl. This implies experiments described here stress the necessity for introduc- that the number of infection events is not altered by the ing tfx into the genome of effective strains rather than relying ability of a Rhizobium strain to produce trifolitoxin. on tfx borne on recombinant plasmids. Also, by constructing both TA1::10-15 and TA1::12-10, Furthermore, stable transfer of competitiveness genes to strains are available which are isogenic in that they differ the chromosome of Rhizobium spp. would prevent conjugal only in the location of the TnS insertion within or adjacent to transfer of the those genes from the inoculum strain to tfx. This allows us to definitively determine the role of inferior, indigenous Rhizobium strains in soil. With the functional tfx genes in the expression of nodulation compet- marker exchange method described above, the tfx region has itiveness. been transferred to the genome of R. leguminosarum bv. The data presented here show that functional tfx genes do trifolii TAl. However, it is not known whether these genes improve the nodulation competitiveness of TAl (Fig. 5 and are inserted into the chromosome or the Sym plasmid of 6, Table 2). However, when the inactive tfx region was TAl. Experiments are in progress to determine the location transferred to TAl for the construction of TAl::12-10, the of the tfx genes. However, since the insert DNA in pTFX1 is competitiveness of TAl declined when present at approxi- of chromosomal origin in T24, it is expected that the region mately 20-fold-higher levels than 2046 (Table 2). This sug- in TAl that is homologous with pTFX1 is also chromosomal. gests that under certain circumstances the region in pTFX1 Thus, the tfx genes are probably chromosomal in the responsible for the marker exchange event does provide a TA1::tfx exconjugants. In any case, data are presented here certain amount of genetic load that may decrease the fitness which illustrate the stability of trifolitoxin production in of TAl. However, when the functional tfx region is used for TA1::10-15 (Fig. 3). the marker exchange, the benefits of trifolitoxin production The region in pTFX1 responsible for the marker exchange in enhancing nodulation competitiveness far outweigh any event has not been identified. However, the tfx region does detriment from the burden of carrying more DNA in the cell. VOL. 56, 1990 EFFECTIVE, TRIFOLITOXIN-PRODUCING RHIZOBIUM STRAIN 103

These experiments represent the first stable genetic trans- 10. Figurski, D. H., and D. R. Helinski. 1979. Replication of an fer of genes involved in nodulation competitiveness to a origin-containing derivative of plasmid RK2 dependent on a symbiotically effective strain of Rhizobium. Future efforts plasmid function provided in trans. Proc. Natl. Acad. Sci. USA must address whether pTFX1::10-15 can be used to produce 76:1648-1652. stable, competitive, effective strains of other rhizobia in R. 11. Franco, A. A., and J. M. Vincent. 1976. Competition amongst leguminosarum bv. phaseoli, R. leguminosarum bv. viceae, rhizobial strains for the colonization and nodulation of two R. fredii, and R. meliloti. The marker exchange of tfx into tropical legumes. Plant Soil 45:27-48. 12. Holmes, D. S., and M. Quigley. 1981. A rapid boiling method for Rhizobium spp. by the method described here may be the preparation of bacterial plasmids. Anal. Biochem. 114: generally useful for providing a competitive phenotype to 193-197. strains of Rhizobium which permit increased productivity of 13. Lambert, G. R., A. R. Harker, M. A. Cantrell, F. J. Hanus, several legume species. S. A. Russell, R. A. Haugland, and H. J. Evans. 1987. Symbiotic Trifolitoxin is the most potent antirhizobial bacteriocin expression of cosmid-bome Bradyrhizobiumjaponicum hydrog- known (18). However, Schwinghamer's (18) assessment of enase genes. Appl. Environ. Microbiol. 53:422-428. bacteriocin-producing rhizobia among natural isolates from 14. Leong, S. A., G. S. Ditta, and D. R. Helinski. 1982. Heme southeastern Australia remains the only example of a sys- biosynthesis in Rhizobium. Identification of a cloned gene tematic search for bacteriocin-producing rhizobia. Among coding for D-aminolevulinic acid synthetase from Rhizobium all of the isolates in Schwinghamer's collection, very few meliloti. J. Biol. Chem. 257:8724-8730. excreted a bacteriocin and none of the bacteriocin producers 15. Long, S. R., W. J. Buikema, and F. M. Ausubel. 1982. Cloning were as potent as T24 (18). Given the few studies on of Rhizobium meliloti nodulation genes by direct complementa- bacteriocin production by strains of Rhizobium, the signifi- tion of Nod mutants. Nature (London) 298:485-488. 16. O'Gara, F., B. Boesten, and S. Fanning. 1988. The development cance of bacteriocin production in regulating nodule occu- and exploitation of "marker genes" suitable for risk evaluation pancy in nature is not known. Although potent bacteriocin studies on the release of genetically engineered microorganisms production by rhizobia may be rare, bacteriocin production in soil, p. 50-64. In W. Klingmuller (ed.), Risk assessment for by an inoculum strain may be useful in limiting nodulation by deliberate releases. Springer-Verlag, Berlin. indigenous strains. Field inoculation with isogenic recombi- 17. Roughley, R. J., W. M. Blowes, and D. F. Herridge. 1976. nant Rhizobium strains which differ only in bacteriocin Nodulation of Trifolium subterraneum by introduced rhizobia in production, such as those strains described here, will be competition with naturalized strains. Soil Biol. Biochem. 8: necessary to determine the agricultural usefulness of bacte- 403-407. riocin-producing Rhizobium strains as part of a solution to 18. Schwinghamer, E. A. 1971. Antagonism between strains of the Rhizobium competition problem. Rhizobium trifolii in culture. Soil Biol. Biochem. 3:355-363. 19. Schwinghamer, E. A., and R. P. Belkengren. 1968. Inhibition of rhizobia by a strain of Rhizobium trifolii: some properties of the ACKNOWLEDGMENTS antibiotic and of the strain. Arch. Mikrobiol. 64:130-145. This work was supported by Department of Agriculture grant 20. Triplett, E. W. 1988. Isolation of genes involved in nodulation 87-CRCR-1-2571 and by the University of Wisconsin-Madison Col- competitiveness from Rhizobium leguminosarum bv. trifolii lege of Agricultural and Life Sciences through Hatch project 3193. T24. Proc. Natl. Acad. Sci. USA 85:3810-3814. 21. Triplett, E. W., and T. M. Barta. 1987. Trifolitoxin production LITERATURE CITED and nodulation are necessary for the expression of superior 1. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G nodulation competitiveness by Rhizobium leguminosarum bv. Seidman, J. A. Smith, and K. Struhl. 1987. Current protocols in trifolii strain T24 on clover. Plant Physiol. 85:335-342. molecular biology. John Wiley & Sons, Inc., New York. 22. Triplett, E. W., B. Lethbridge, S. L. Midland, M. E. Tate, and 2. Barry, G. F. 1986. Permanent insertion of foreign genes into the J. J. Sims. 1988. Cloning of genes from Rhizobium leguminosa- chromosomes of soil bacteria. Bio/Technology 4:446-449. rum bv. trifolii T24 responsible for the production of trifolitoxin, 3. Barry, G. F. 1988. A broad-host-range shuttle system for gene an anti-rhizobial peptide involved in nodulation competitive- insertion into the chromosomes of gram-negative bacteria. Gene ness, p. 788. In H. Bothe, F. J. de Bruijn, and W. E. Newton 71:75-84. (ed.), Nitrogen fixation: hundred years after. Gustav Fischer, 4. Bergersen, F. J. 1961. The growth of Rhizobium in synthetic Stuttgart. media. Aust. J. Biol. Sci. 14:349-360. 23. Triplett, E. W., M. J. Schink, and K. L. Noeldner. 1989. 5. Beringer, J. E., J. L. Beynon, A. V. Buchanan-Wollaston, and A. Mapping and subcloning of the trifolitoxin production and W. B. Johnston. 1978. Transfer of the drug resistance transpo- resistance genes from Rhizobium leguminosarum bv. trifolii son Tn5 to Rhizobium. Nature (London) 276:633-634. T24. Mol. Plant-Microbe Interact. 2:202-208. 6. Bromfield, E. S. P., and D. G. Jones. 1980. Studies on double 24. Vincent, J. M. 1974. Root nodule symbioses with Rhizobium, p. strain occupancy of nodules and the competitive ability of 266-341. In A. Quispel (ed.), The biology of nitrogen fixation. Rhizobium trifolii on red and white clover grown in soil and North-Holland Publishing Co., Amsterdam. agar. Ann. Appl. Biol. 94:51-59. 25. Vincent, J. M., and L. M. Waters. 1953. The influence of the 7. Demezas, D. H., and P. J. Bottomley. 1986. Interstrain compe- host on competition amongst clover root-nodule bacteria. J. tition between representatives of indigenous serotypes ofRhizo- Gen. Microbiol. 9:357-370. bium trifolii. Appl. Environ. Microbiol. 52:1020-1035. 26. Williams, M. K., F. Cannon, P. McLean, and J. Beynon. 1988. 8. Ditta, G. 1986. TnS mapping of Rhizobium nitrogen fixation Vector for the integration of genes into a defined site in the genes. Methods Enzymol. 118:519-528. Rhizobium meliloti genome, p. 198-199. In R. Palacios and D. P. 9. Dowling, D. N., and W. J. Broughton. 1986. Competition for S. Verma (ed.), Molecular genetics of plant-microbe interac- nodulation of legumes. Annu. Rev. Microbiol. 40:131-157. tions. The American Phytopathological Society, St. Paul, Minn.