A Two-Component Regulator Mediates Population-Density-Dependent Expression of the Bradyrhizobium japonicum Nodulation Genes

Bacterial strains and culture conditions.The bacterial strains and plasmids used in this study are shown in Table 1. For routine growth, B. japonicum strains were cultured in RDY medium at 30°C (42). HM salt medium (9) supplemented with 0.1% arabinose was used for growth of B. japonicum for matings and for extracting nucleic acids. Escherichia coli strains were maintained in Luria-Bertani medium at 37°C (35). When appropriate, antibiotics were added to the culture medium at the following concentrations: for B. japonicum we used 100 μg of streptomycin, spectinomycin, kanamycin, or tetracycline/ml and 30 μg of chloramphenicol/ml, and for E. coli we used 50 μg of ampicillin, spectinomycin, or kanamycin/ml and 25 μg of tetracycline/ml.

B. japonicum strain JNWS21, harboring a nwsB interposon mutation, was generated as follows: PCR was used to amplify the coding regions of the B. japonicum nwsB gene. The amplified DNA included regions 5′ and 3′ of the nwsB gene (21). The primers used were 5′-GGAGACCGGGCCGGATCACC-3′ and 5′-CGTTACAACTTGGCTTCATCC-3′. The PCR product (2.2 kb) was ligated into the pCR2.1-TOPO vector (Invitrogen, Carlsbad, Calif.). The resultant plasmid (pCR2.1-NwsB) was digested with SspI and filled in with Klenow. An Sm^r-Sp^r cassette obtained by digesting plasmid pHP45Ω with SmaI was inserted into blunt-ended pCR2.1-NwsB. The resultant plasmid, pCR2.1-NWSBΩ, was digested with EcoRI, and the 4.2-kb fragment containing nwsB was cloned into the EcoRI site of the suicide vector pSUP202 (39). The resultant plasmid was transformed into E. coli S17-1 (39) and mobilized by biparental mating into B. japonicum USDA110. Transconjugants were selected based on Sp^r-Sm^r and Tc^s, the latter being an indication of a double-crossover event. Confirmation of these mutations was verified by Southern blot analyses (data not shown). To generate an nwsB mutation in strain JLA6 (an Sm^r-Sp^r USDA110 strain harboring a chromosomal nolA-lacZ fusion [27]), pCR2.1-NWSBΩ was digested with EcoRI and the 4.2-kb nwsB fragment cloned into the vector pGEM-T (Promega, Madison, Wis.), creating pGEMT-NwsBΩ. pGEMT-NwsBΩ was then digested with HindIII to release the Sm^r-Sp^r cassette, and the nwsB-containing fragment was filled in with Klenow. A Km^r cassette from pBSL15 (1) was then blunt-end cloned into the nwsB gene, yielding pGEMT-NWSB-Km. The nwsB fragment was digested with EcoRI, cloned into pSUP202, and conjugally transferred to strain JLA6 as described above. Transconjugants were selected on the basis of kanamycin resistance and tetracycline sensitivity. The mutations generated were confirmed by Southern hybridization (data not shown).

β-Galactosidase activity assays.β-Galactosidase activity was assayed as described by Banfalvi et al. (2). The induction of nod-lacZ expression by genistein was measured as described previously (41, 47). For induction with conditioned medium (CM), assays were performed as described by Loh et al. (27). CM was obtained from supernatants of B. japonicum cultures grown to an optical density at 600 nm (OD₆₀₀) of 2.0. Under the conditions used, an OD₆₀₀ unit corresponded to ca. 10⁹ cells/ml. The CM was concentrated 100-fold by rotary evaporation and added to induction assays (10 μl of concentrated CM per ml of induction medium [27]). Induction assays were performed in minimal medium (6). β-Galactosidase activity was measured 14 h after induction as detailed by Banfalvi et al. (2).

Plant nodulation assays.Seeds of Glycine max (soybean) cv. Essex and Vigna unguiculata (cowpea) cv. Caloona were surface sterilized as described by Nieuwkoop et al. (31). After germination for 48 h, the seedlings were transferred to either plastic growth pouches (Mega International, Minneapolis, Minn.) or plastic Leonard jars containing vermiculite. The seedlings were then inoculated with 5 × 10⁷ cells of the appropriate bacterial culture. Inoculation of the bacterial cultures was performed in the presence or absence of CM (10 μl of concentrated CM per ml of culture). Plants were grown in a growth chamber under a 16-h photoperiod. The plants were watered with a nitrogen-free nutrient solution (45).

GUS assays.For GUS (β-glucuronidase) assays, 10⁷B. japonicum cells harboring the plasmid pVKNodYGUS4 or pVKGUS (27) were inoculated onto soybean or cowpea roots. At 24 days postinoculation, nodules were harvested and examined for GUS expression. GUS assays were performed as described by Jefferson et al. (23). Nodules were sectioned and prepared as described in Loh et al. (27). Digital images were taken on a stereomicroscope (Olympus SZX12; Olympus, Tokyo, Japan) fitted with an Olympus DP10 camera or a Nikon E600 microscope fitted with Nomarski optics.

Expression of nolA and nodD2 is significantly reduced in an nwsB mutant.Previous work (27) demonstrated that the nolA and nodD2 genes were expressed in a population density fashion. Consistent with this, the expression of both nolA-lacZ and nodD2-lacZ (i.e., in strains Bj110-42 [18] and Bj110-1248 [19], respectively) were found to increase with culture density (Fig. 1A). In contrast, analyses of the same fusions in an nwsB mutant strain revealed a significant reduction in both nolA-lacZ (i.e., in strain JNWS24) and nodD2-lacZ (i.e., strain JNWS41) expression, even at high population densities (Fig. 1A). Since a quorum factor found in CM was previously shown to modulate the population density expression of nolA and nodD2, CM derived from cultures of either USDA110 or the nwsB mutant (i.e., JNWS21) were tested for their ability to induce nolA-lacZ and nodD2-lacZ expression (Table 2). While the addition of CM from USDA110 cultures resulted in induction of nolA-lacZ and nodD2-lacZ expression in the wild type (i.e., Bj110-42 and Bj110-1248, respectively), significantly less expression was detected in the nwsB mutant strains JNWS24 (i.e., nolA-lacZ) and JNWS44 (nodD2-lacZ). Similarly, the addition of CM derived from the nwsB mutant strain JNWS21 resulted in strong induction of nolA-lacZ and nodD2-lacZ expression in the wild type, with reduced expression in both strains JNWS24 and JNWS41. This latter result suggests that the nwsB mutant strain retains the capacity to synthesize the quorum signal but is defective in its response to this signal as measured by nolA and nodD2 induction.

• Open in new tab
• Download powerpoint

FIG. 1.

(A) Population-density-dependent expression of nolA-lacZ and nodD2-lacZ. B. japonicum wild type or NwsB mutant harboring either nolA-lacZ or nodD2-lacZ was grown in minimal medium to various ODs, and the β-galactosidase activities of these fusions were determined (2). (B) Analysis of chromosomal nolA-lacZ fusion expression in nwsB transconjugants harboring either pTE3, pTE3WWT3, or pTE3Wmut10. pTE3WWT3 expresses NodW from the trp promoter of pTE3, while pTE3Wmut10 expresses NodWD70N.

Population density expression of nolA is restored by overexpression of NodW.A closer examination of the expression of nolA-lacZ expression in strain JNWS24 revealed that, while nolA expression was reduced in this strain, slightly elevated levels of nolA expression could still be detected in this strain at high culture densities (cf. wild type [400 U] and JNWS24 [75 U]) (Fig. 1A). Since cross talk between NwsB and NodW had been previously reported (21), we suspected that NodW could account for the residual nolA activity in strain JNWS24. Based on this assumption, we reasoned that overexpression of NodW would complement the quorum-defective phenotype observed with the nwsB mutant. As shown in Fig. 1B, the expression of nodW in strain JNWS82, containing an nwsB mutation and a chromosomal nolA-lacZ fusion, restored the ability of this mutant to express nolA in a population density manner. Interestingly, complementation was also observed in the nwsB mutant (i.e., strain JNWS83) expressing the nonphosphorylated form of NodW (i.e., NodWD70N) (28). Control transconjugants with the vector pTE3 (i.e., JNWS81) did not exhibit population density expression of nolA-lacZ (Fig. 1B).

Effects of nwsB mutation on nodY expression in vitro and in planta.To determine the effects of the nwsB mutation on nodY expression, the ability of genistein to induce a nodY-lacZ fusion at low and high ODs was assayed. At a low OD (i.e., OD₆₀₀ = 0.1), genistein induction of nodY-lacZ expression in the nwsB mutant (i.e., JNWS31) was reduced compared to that of the wild type (i.e., ZB977) (Fig. 2A). However, unlike the wild type, nodY-lacZ induction in the nwsB mutant was not affected by cell population density (cf. OD₆₀₀ = 0.1 versus OD₆₀₀ = 1.8). In agreement with this, the addition of CM derived from B. japonicum cultures grown to an OD₆₀₀ of 2.0 had little effect on nod gene expression in JNWS31 (Fig. 2B). In contrast, addition of CM drastically reduced the ability of genistein to induce nodY-lacZ expression in the wild type. To examine whether the nwsB mutation would affect nodY expression in planta, we inoculated soybean and cowpea plants with nwsB mutant transconjugants harboring a nodY-GUS fusion (i.e., strain JNN41). As shown in Fig. 3, little or no nodY-GUS expression was observed in nodules infected with the wild-type nodY-GUS transconjugants. This result is similar to that reported previously (27), where nodY-GUS expression was not detected in nodules harvested at time intervals ranging from 6 to 30 days postinoculation. In contrast, nodules infected with strain JNN41 exhibited significant GUS staining at 24 days postinoculation (Fig. 3). As a positive control, plants were infected with the nolA mutant BjB3 harboring the nodY-GUS fusion. Nodules derived from this sample showed strong GUS staining, as previously reported (27).

• Open in new tab
• Download powerpoint

FIG. 2.

(A) Comparison of nodY-lacZ expression in B. japonicum USDA110 or JNWS21 (nwsB) mutant at a low or high population density. B. japonicum cells were grown to an OD₆₀₀ of 0.1 or 1.8, and the ability of 1 μM genistein to induce the nodY-lacZ expression was determined after 12 h of incubation. (B) Effect of CM on nodY-lacZ expression in USDA110 or JNWS21 (nwsB mutant). B. japonicum cultures (OD₆₀₀ = 0.1) were induced for 5 h with CM. The CM was obtained from B. japonicum cultures grown to an OD₆₀₀ of 2.0. The results represent the means of three separate experiments ± the standard deviation. Under the conditions used, an OD₆₀₀ unit corresponded to ca. 10⁹ cells/ml.

• Open in new tab
• Download powerpoint

FIG. 3.

Comparison of nodY-GUS expression in cowpea and soybean nodules infected with the wild type, nwsB mutant JNWS21, and nolA mutant BjB3. nodY-GUS staining was determined in nodules 24 days postinoculation. (A to C) Soybean nodules; (D to F) cowpea nodules. In panels A and D, the nodules were infected with USDA110 (nodY-GUS) transconjugants. In panels B and E, the nodules were infected with JNWS21 (nodY-GUS) transconjugants. In panel C and F, the nodules were infected with BjB3 (nodY-GUS). Nodules were hand sectioned and stained with X-Gluc (5-bromo-4-chloro-3-indolyl-β-d-glucuronic acid). Digital images were obtained with a stereomicroscope.

Effect of CM on nodulation.We next sought to determine the effect of CM addition on the nodulation ability of an nwsB mutant. The nodulation tests were performed in plastic growth pouches that allowed us to monitor nodule development spatially and temporally. As a reference mark, the root tip at the time of inoculation was noted on the outside of the plastic pouches. In previous studies, it was shown that the zone of emergent root hairs (just above the root tip) was the region of the root most susceptible to rhizobial infection (see, for example, reference 7). This mark also allowed us to distinguish between root tissue that was present at the time of bacterial inoculation and root tissue that formed after inoculation. Using this method, we were able to separate nodules that formed on root tissue already present (i.e., no delay in nodulation) versus nodules that were colonized by bacteria that were inhibited or delayed in their ability to infect the soybean root (i.e., nodules formed below the root tip mark). In samples where no CM was added, the nwsB mutant showed a slight delay in its ability to nodulate soybean (Fig. 4). This is reflected in fewer nodules found above the root tip mark and a corresponding increase in nodules formed below the root tip mark. The addition of CM to wild-type cultures drastically reduced the ability to nodulate soybean, resulting in very few nodules formed above the root tip mark. This result is consistent with that reported by Loh et al. (27). In contrast to what was observed with the wild type, the addition of CM had little or no effect on the ability of the nwsB mutant (i.e., JNWS21) to nodulate soybean.

• Open in new tab
• Download powerpoint

FIG. 4.

Effect of CM on the ability of B. japonicum wild type and nwsB mutant JNWS21 to nodulate soybean plants. B. japonicum cells, untreated or incubated with CM for 1 h, were inoculated onto soybean plants (10⁷ cells per root). At 21 days postinoculation, the numbers of nodules were determined. The standard error is also presented. Nodules were scored based on their location, either above or below the root tip mark at the time of inoculation (see Materials and Methods).

To further characterize the ability of the JNWS21 to nodulate, we compared the ability of this strain to nodulate soybean in competition with the wild type. JNWS21 and wild-type cells were mixed in different ratios and inoculated onto soybean or cowpea plants in the presence or absence of CM. The nodules formed above the root tip mark were selected and streaked on agar plates. Single colonies were picked and tested for streptomycin resistance, the antibiotic marker for mutant JNWS21. Table 3 clearly shows that the addition of CM to the bacterial inoculant samples increased the ability of the nwsB mutant strain to nodulate both soybean and cowpea.