Role of Polysaccharides in the Infection Process

To invade nodules through infection threads, Sinorhizobium meliloti must synthesize at least one of the following polysaccharides: succinoglycan, EPS II or a symbiotically active form of K-antigen (Figure 1). Mutants defective in the production of these polysaccharides are blocked at the nodule invasion step and induce the formation of small white root nodules devoid of bacteria (Leigh 1985; Reuber et al. 1991; Gonzalez et al. 1996). Many studies have utilized strains, Rml021 and Rm2011, which under non-limiting nutrient conditions, produce only succinoglycan. Rml021 and Rm2011 have a cryptic capacity to synthesize symbiotically active forms of EPS II (Glazebrook et al. 1989).

Succinoglycan is a polymer of an octasaccharide repeated unit, composed of one galactose and seven glucoses residues. Each repeated unit is modified with approximately one acetyl, one succinyl, and one pyruvyl substituent (Reuber et al. 1991; Gonzalez et al. 1998; Wang et al. 1999). Because Calcofluor can bind succinoglycan and then fluoresce under UV light, genetic screens for non-fluorescent mutants were used to identify the exo gene cluster involved in succinoglycan biosynthesis (Leigh et al. 1985; Doherty et al. 1988; Reuber et al. 1993). Many exo mutants were blocked in succinoglycan production and in nodule invasion thereby indicating the importance of extracellular polysaccharides in the nodulation process (Leigh et al. 1985; Reuber et al. 1993). A biochemical approach using radiolabeled sugar precursors was used to construct a detailed model for succinoglycan biosynthesis. Briefly, UDP-glucose and 14C-UDP-galactose were added to different permeabilizing exo mutant backgrounds. Analysis of the lipid-linked intermediates accumulated by various exo mutants showed that galactose is the first sugar added to an undecaprenol carrier, followed by the seven glucoses. The acetyl, succinyl, and pyruvyl modifications are added during the synthesis of the octasaccharide (Reuber et al. 1993).

S. meliloti synthesizes high molecular weight (HMW) forms of succinoglycan, consisting of several hundred or greater octasaccharide repeating units, and low molecular weight (LMW) forms, consisting of monomers, dimers and trimers of the octasaccharide repeating unit (Gonzalez et al. 1998; Wang et al. 1999). The LMW form of succinoglycan, in particular the trimer form, is the symbiotically active species as small added amounts of trimer molecules can partially rescue the symbiotic defect of a non-succinoglycan producing mutant (Wang et al. 1999). Two non-exclusive pathways have been proposed for how S. meliloti produces LMW succinoglycan: (i) a direct synthesis pathway mediated by exoP and exoT in which LMW succinoglycan is produced by limited polymerization of the octasaccharide and (ii) a degradative pathway in which glycanases cleave HMW chains to LMW forms (Gonzalez et al. 1998; York et al. 1998).

| 0-1.6

6I

acetyl

Glc

| ß-1.6

Glc 2—

*

| ß-1.3

Glc —

succinyl

| ß-1,3

&—

acetyl pyruvyl

acetyl pyruvyl

Figure 1. Succinoglycan (A), EPS II (B) and K-antigen (C) repeating units.

Succinoglycan production is needed for the nodule invasion step of symbiosis, but once the rhizobia are inside the plant cell and have differentiated into bacteroids, the exo genes are no longer expressed (Reuber et al. 1991). This suggests that the regulation of succinoglycan production may also be an important factor in nodule development. Several genes have been implicated in the regulation of succinoglycan production: exoR, exoS, exsB, mucR, and exoX (Doherty et al. 1988; Reed et al. 1991a, 1991b; Becker et al. 1993; Bertram-Drogatz et al. 1998). To date, the ways that these genes affect succinoglycan production are unclear. We have shown that ExoS is the sensor of the ExoS/ChvI two-component regulatory system (Cheng et al. 1998a), similar to the ChvG/ChvI system of A. tumefaciens (Charles et al. 1993), which in turn is part of the two-component regulatory system family that includes EnvZ/OmpR. This system is able to stimulate the transcription of the exo genes, but the environmental signal sensed by ExoS that triggers this response remains obscure.

S. meliloti Rml021 is able to produce the EPS II under phosphate-limiting condition or when it carries either the expRlOl allele of the expR locus or a mucR::Jn5 mutation. EPS II is a polymer of a glucose-galactose repeating unit and is modified with acetyl and pyruvyl substituents. Like succinoglycan, it is found in two forms, HMW and LMW EPS II. Low phosphate (Zhan et al. 1991) and mucR mutant (Ruberg et al. 1999; Keller et al. 1995) induce HMW EPSII product, but the expRlOl allele produces also a LMW form, which is symbiotically active and can substitute for succinoglycan during the invasion process (Gonzalez et al. 1996; Glazebrook et al. 1989). We find that the expRlOl mutation results in the functional restoration of the expR gene, which in the parental strain Rml021 is interrupted by an insertion sequence (Pellock et al., unpublished). The cluster of exp genes involved in the synthesis of EPS II has also been identified (Becker et al. 1997), but a model for EPS II synthesis has not yet been proposed. A number of the genes found in the exp cluster have putative functions that are not easily explained in an EPS II biosynthetic model. Some strains of X meliloti (AK631) produce a capsular polysaccharide, structurally analogous to group II K antigens found in Escherichia coli known as the K-antigen (Reuhs et al. 1993). The K-antigen is a polymer of a disaccharide composed of an aminohexulosonic acid and a-keto-3,5,7,9-tetradeoxy-5,7-diaminononulosonic acid, a variant of 3-deoxy-D-manno-2-octulosonic acid (KDO). Some strains produce a strain-specific K-antigen that can substitute exopolysaccharide function in symbiosis in S. meliloti mutants unable to synthesize succinoglycan and/or EPS II. Certain functions are required in common for the synthesis of K antigen and LPS (Campbell et al. 1998).

Interestingly, the ability of these three polysaccharides to promote infection thread initiation and extension shows differences, which can be monitored through strains that constitutively express green fluorescent protein (Cheng et al. 1998b; Pellock et al. 2000). Succinoglycan is the most efficient of all three in promoting infection thread growth. EPS II and K-antigen mediated infection threads frequently exhibit aberrant morphologies. Furthermore, K-antigen is less efficient than succinoglycan in infection thread extension and EPS II mediated infection is more likely to be aborted in the infection thread initiation or extension steps. The latter may explain the stunted host plant growth observed in EPS II mediated symbiosis.

Recent studies in our lab have shown that the proper S. meliloti LPS structure is necessary for the establishment of chronic infection within alfalfa. Different mutants exhibiting anomalies upon entry into plant cells have been identified. One of these mutants, originally termed fix-389, has a disrupted IpsB gene, which encodes a putative glycosyl transferase one family protein (Lagares et al. 2001; Campbell et al. submitted). The loss of LpsB causes striking changes in the carbohydrate core of lipopolysaccharide, including the absence of uronic acids and a 40-fold relative increase in xylose. We have also shown that the lpsB3S9 mutant exhibits sensitivities to the cationic peptides melittin, polymixin B and poly-L-lysine, in a manner that parallels of Brucella abortus Ips mutants (Freer et al. 1999; Ulgade et al. 2000). Sensitivity to these components of the plant's innate immune system may be part of the reason that this mutant is unable to properly sustain a chronic infection within the cells of its host plant alfalfa (Campbell et al. submitted).

3. Involvement of BacA in Intracellular Survival

Upon release from the infection thread into acidic plant membrane-bound compartments, wild-type rhizobia differentiate into nitrogen-fixing bacteroids. In contrast, S. meliloti mutants lacking a functional BacA protein are unable to survive intracellularly, senescing shortly after their release from the infection thread (Le Vier et al. 2000; Glazebrook et al. 1993, 1996; Ichige et al. 1997). The bacA gene was first isolated in a screen that identified S. meliloti mutants with symbiotic deficiencies (Long et al. 1988). This ability to establish a chronic infection into eukaryotic cells is shared with a number of pathogen bacteria including Brucella abortus, a mammalian pathogen, which causes bovine abortions and brucellosis in humans. Intriguingly a bacA mutant of Brucella abortus, a close phylogenetic relative of S. meliloti, is unable to establish a chronic infection in host cells in experimentally infected mice (Le Vier et al. 2000). Replication and survival into murine macrophages is also affected in B. abortus lacking active BacA protein. Moreover, as with S. meliloti, bacA mutants of B. abortus behave like the wild-type bacteria during the initial stages of host infection. Therefore, a parallel can be established between the survival pattern of a plant symbiont, S. meliloti and a mammalian pathogen B. abortus.

This is consistent with common environmental stresses both pathogen and symbiont have to face while invading their host cells. For example, oxidative stress is one of the host defense responses produced both by infected macrophages as well as by plant cells (Lamb et al. 1997; Ugalde 1999). The production of reactive oxygen species by the infected plant host, is an early event of the hypersensitive response against pathogenic bacteria. A prolonged oxidative burst has previously been shown to occur subsequent to the entry of S. meliloti into alfalfa roots (Santos et al.

2001). Non-opsonized B. abortus do not induce this oxidative defense response, while opsonized bacteria lead to a significant production of oxygen reactive species. In this last case, reactive nitrogen is involved in clearing the macrophage of parasites (Ugalde 1999).

It is possible that BacA may be involved in the resistance to at least one type of intracellular stress. The BacA protein is predicted to be an internal cytoplasmic membrane transporter (Glazebrook et al. 1993). Because a bacA mutant has increased resistance to bleomycin (Ichige et al. 1997; Ferguson personal communication; LeVier submitted) it was also suggested that BacA could act as a bleomycin uptake system. However, recent physiological studies (G.P. Ferguson, unpublished) and genetic analyses (LeVier submitted) have highlighted new bacA phenotypes and have led to the hypothesis that BacA can carry out more than one function. BacA could act as a signal receptor, transducing environmental signals or stresses; these signals could be essential to the bacteria adaptation to an intracellular lifestyle. Another possible role for BacA, suggested by the involvement of BacA in deoxycholate, SDS and ethanol resistance, is that BacA could also be involved in the homeostasis of the cell envelope. BacA could also be an efflux pump. We are now attempting to determine the role of BacA and to model its behavior during the cell colonization.

Was this article helpful?

0 0

Post a comment