A Novel Screening Method for Isolating Exopolysaccharide-Deficient Mutants

Abstract

A screening method based on differential staining of the wild type and exopolysaccharide-deficient mutants of Rhizobium (Sinorhizobium) meliloti by the lipophilic dye Sudan Black B is described. Mutants defective in the production of either succinoglycan or EPS II (galactoglucan) were isolated by using this method, which might also prove useful for isolating exopolysaccharide-defective derivatives of other bacteria.

Exopolysaccharide synthesis by Rhizobium (Sinorhizobium) meliloti is critical in order for it to invade the nodules it elicits on alfalfa (Medicago sativa) and establish a productive symbiosis (reviewed in references 16 and 21). Nodulation of leguminous plants by rhizobia is a complex developmental process that requires a series of interactions between the bacterium and its plant host (reviewed in references 10, 11, and 31). Failure of R. meliloti Rm1021 to synthesize either of two exopolysaccharides, succinoglycan and EPS II (galactoglucan), results in a very early block in the invasion process (5a, 9, 12, 20, 33).

Succinoglycan, which is also synthesized by several other bacteria (18), is composed of repeating octasaccharide subunits, each of which consists of a backbone of three glucoses and one galactose, a side chain of four glucoses, and 1-carboxyethylidene (“pyruvyl”), acetyl, and succinyl modifications in a ratio of approximately 1:1:1 (1, 6, 18, 19, 28). EPS II consists of a disaccharide subunit of glucose and galactose carrying acetyl and pyruvyl modifications (17). A specific size class of EPS II oligosaccharides has recently been shown to be required for the symbiotic role of EPS II (15). EPS II is a member of a biologically important class of bacterial exopolysaccharides termed galactoglucans, whose backbones consist of alternating glucose and galactose moieties but which differ with respect to the noncarbohydrate modifications they carry. For example, galactoglucans are produced by isolates of Pseudomonas putida and Pseudomonas fluorescens (27), fluorescent Pseudomonas species (8), Agrobacterium radiobacter and Achromobacter strains (35, 36), and Burkholderia cepacia (4). Galactoglucans have been postulated to function in biofilm formation and to serve as pathogenicity determinants.

The discovery that succinoglycan plays a critical role in R. meliloti-alfalfa symbiosis resulted directly from the observation that succinoglycan-producing colonies of R. meliloti fluoresce under UV light when grown on media containing the laundry whitener Calcofluor (20). This made it very easy to isolate mutants of R. meliloti that were defective in production of succinoglycan, even though there were no other observable changes in the morphology of the colonies. The ability of succinoglycan to bind Calcofluor and fluoresce under UV light greatly facilitated subsequent genetic analyses of succinoglycan’s biosynthesis and its symbiotic roles (9, 13, 14, 20, 23, 24), regulation (5, 7, 26), and degradation (34). However, a combination of DNA sequencing, fine-structure genetic analysis, and biochemical studies (13, 14, 29, 34) later revealed that certain classes of mutants affecting succinoglycan production were not easily detected on the basis of their Calcofluor fluorescence phenotypes. The EPS II exopolysaccharide, which is expressed in a symbiotically active form by the Rm1021 expR101 derivative (12, 15), does not bind Calcofluor, making genetic analysis of its synthesis and regulation more difficult than that of succinoglycan.

In this report, we describe a new screening method for exopolysaccharide-deficient mutants that does not depend on Calcofluor binding. This method can be used to detect mutants of R. meliloti that are defective in the production of either succinoglycan or EPS II and even allows the detection of mutants which are altered in succinoglycan production but cannot be detected on the basis of their Calcofluor fluorescence phenotypes. The idea for this method grew out of our efforts to use the lipophilic dye Sudan Black B (3) to analyze poly-β-hydroxybutyrate (PHB) synthesis by R. meliloti in a manner similar to that used to analyze PHB synthesis by Alcaligenes eutrophus (30). In the course of this work, we found that exopolysaccharide-producing strains of R. meliloti excluded the stain Sudan Black B and that exoB mutants, which synthesize neither succinoglycan nor EPS II (12, 20), readily incorporated this stain. On the basis of this observation, we set out to determine whether this differential uptake of Sudan Black B could be used to screen for mutants defective in succinoglycan production.

Culture and Sudan Black B staining conditions.

Culture media and growth conditions for R. meliloti were described previously (14, 15). All the R. meliloti strains described in this report are derivatives of Rm1021 (20). Other bacterial strains and plasmids have been described previously (20, 22). To screen for succinoglycan-producing strains, we used MM1 plates (25) with 0.4% fructose. For EPS II strains, we used the minimal medium described by Amemura et al. (2). A 0.02% solution of Sudan Black B (Sigma, St. Louis, Mo.) dissolved in 96% ethanol was used for the staining. All the bacteria were grown as single colonies on plates containing Luria-Bertani (LB) agar or M9 medium. They were then replica plated onto either MM1-fructose plates (succinoglycan-producing strains) or minimal-medium (2) plates (EPS II-producing strains). Approximately 8 ml of the 0.02% Sudan Black B solution was applied to each replica plate and allowed to remain undisturbed for about 10 min. The dye was then decanted, and the plates were gently rinsed by adding 10 ml of 100% ethanol and swirling for a few minutes. Colonies unable to incorporate the Sudan Black B stain appeared white, while colonies able to incorporate the dye appeared bluish black (Fig. 1).

FIG. 1.

Wild-type strain Rm1021 and exoY mutant strain Rm7210 (arrows) were grown on LB agar and then replica plated onto MM1-fructose medium and stained with Sudan Black B (A) or replica plated onto LB-Calcofluor agar and viewed under UV light (B).

Isolation of mutants altered in succinoglycan production by Sudan Black B staining.

We first checked mutants that have transposon insertions in the exoB, exoA, exoF, and exoY genes and are thus completely defective in succinoglycan synthesis (29). The strains were plated as single colonies on LB agar plates and replica plated onto MM1-fructose plates. After 36 h of growth at 30°C, the plates were stained and examined for Sudan Black B staining. As shown in Table 1, all the exo mutants tested stained with Sudan Black B, while the wild-type Rm1021 strain excluded the dye. The staining of the exo mutants appears to be due to Sudan Black B binding to PHB, since an exoY strain with a disruption in the gene encoding PHB synthase (32) was not stained by the dye (data not shown).

TABLE 1.

Sudan Black B staining of R. meliloti strains

Strain	Relevant genotype	Stained by Sudan Black B	Reference or source
Rm1021	Wild type	No	20
Rm7210	exoY210::Tn5	Yes	20
Rm6024	exoB94::Tn5-233	Yes	12
Rm8049	exoA::Tn5	Yes	20
Rm6089	exoF::Tn5	Yes	20
Rm8545	exoO::Kan	Yes	13
Rm8285	exoW28::Tn5	Yes	13
Rm8600	expR101	No	12
Rm9000	exoY210::Tn5 expR101	No	15
Rm8603	exoF55::Tn5-233 expR101	No	12
Rm8601	exoA32::Tn5-233 expR101	No	12
Rm10000	expA61::Tn5-Tp exoA125::Tn5 expR101	Yes	12
Rm8064	expA125::Tn5	No	12
Rm10001	expA::Tn5 exoA32::Tn5-233	Yes	J. W. Reed and G. C. Walker

Utilizing the same system, we screened about 7,000 Tn5-mutagenized Rm1021 bacteria for a loss in their ability to take up Sudan Black B. The mutagenesis was carried out as described previously (23). This screen yielded 30 colonies that stained black. All of these mutants were analyzed in a plasmid complementation assay in which triparental mating was used to introduce cosmids (pEX312, pEX154, or pD56) that carry part or all of the exo region (20, 23). Each mutant was complemented by one or more of these cosmids, indicating that all mutations mapped to the 28-kb exo region. Twenty-six of the mutants were dark under UV light when grown on LB plates containing Calcofluor, and the corresponding mutations were not mapped further since they corresponded to exoB or other well-characterized mutations that cause a Calcofluor-dark phenotype (13, 14, 23).

The remaining four mutants fluoresced under UV light on Calcofluor medium. Two were shown to be exoU mutants, and two were shown to be exoX mutants. exoU mutants are difficult to isolate on the basis of their Calcofluor phenotypes because exoU mutations cause so slight a change in the color of the fluorescing material that it is difficult to detect (13, 14, 23). Encouraged by this observation, we tested exoW and exoO mutants that also form Calcofluor-bright colonies and had not been detected on the basis of their Calcofluor phenotypes (13, 14, 23). Both of these mutants stained with the Sudan Black B dye, indicating that, even though the strains produce truncated forms of succinoglycan able to bind Calcofluor, they do not efficiently exclude the Sudan Black B dye. The exoX gene product is involved in regulating the production of succinoglycan and apparently acts at a posttranslational level (26, 37).

Isolation of mutants altered in EPS II production by Sudan Black B staining.

We then examined whether the Sudan Black B screening procedure could be used to detect R. meliloti mutants defective in the production of EPS II and obtained the results summarized in Table 1. The expR101 mutation permits Rm1021 to synthesize the otherwise cryptic exopolysaccharide EPS II in a symbiotically active form (12, 15). Strains with mutations in exoA or exoY are unable to synthesize succinoglycan and can be stained dark with Sudan Black B. Introduction of expR101 into these strains prevents them from staining with the dye. Adding an expA mutation to these strains abolishes the production of EPS II. This correlates with the ability of the triple mutants to incorporate Sudan Black B. Members of our laboratory subsequently found that additional mutations that interfere with EPS II production and are genetically unlinked to the exp gene cluster can also be detected by this Sudan Black B staining procedure (2a).

Potential applicability to other systems.

The Sudan Black B screening procedure described herein can be used to isolate mutants of R. meliloti that are defective in the production of either succinoglycan or EPS II. Mutants altered in succinoglycan production have previously been isolated on the basis of their Calcofluor fluorescence phenotypes, but until now mutants altered in EPS II production have been isolated only on the basis of changes in their mucoid phenotypes. Although the Sudan Black B method is less convenient than the Calcofluor-based method because of the requirement for replica plating, it may prove useful for analyzing the ability of other bacteria to produce exopolysaccharides which, like EPS II, do not bind Calcofluor. Even in the case of succinoglycan, the Sudan Black B method enabled the isolation of exo mutants that were not identified on the basis of their Calcofluor fluorescence phenotypes.