Mutation in a d-alanine–d-alanine ligase of Azospirillum brasilense Cd results in an overproduction of exopolysaccharides and a decreased tolerance to saline stress

Authors

  • Edgardo Jofré,

    1. Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • Sonia Fischer,

    1. Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • Analía Príncipe,

    1. Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • Marina Castro,

    1. Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • Walter Ferrari,

    1. Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Río Cuarto, Córdoba, Argentina
    Search for more papers by this author
  • Antonio Lagares,

    1. IBBM – Instituto de Biotecnología y Biología Molecular, CCT-CONICET-La Plata, Departamento de Ciencias Biológicas, Facultad de Ciencias Exactas, Universidad Nacional de La Plata, La Plata, Argentina
    Search for more papers by this author
  • Gladys Mori

    1. Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Río Cuarto, Córdoba, Argentina
    Search for more papers by this author

  • Editor: Yaacov Okon

Correspondence: Edgardo Jofré, Departamento de Ciencias Naturales, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto, Ruta 36 Km 601 (X5804BYA) Río Cuarto, Córdoba, Argentina. Tel.: +54 358 4676438; fax: +54 358 4676230; e-mail: ejofre@exa.unrc.edu.ar

Abstract

Bacteria of the genus Azospirillum are free-living nitrogen-fixing, rhizobacteria that are found in close association with plant roots, where they exert beneficial effects on plant growth and yield in many crops of agronomic importance. Unlike other bacteria, little is known about the genetics and biochemistry of exopolysaccharides in Azospirillum brasilense. In an attempt to characterize genes associated with exopolysaccharides production, we generated an A. brasilense Cd Tn5 mutant that showed exopolysaccharides overproduction, decreased tolerance to saline conditions, altered cell morphology, and increased sensitivity to detergents. Genetic characterization showed that the Tn5 was inserted within a ddlB gene encoding for a d-alanine–d-alanine ligase, and located upstream of the ftsQAZ gene cluster responsible for cell division in different bacteria. Heterologous complementation of the ddlB Tn5 mutant restored the exopolysaccharides production to wild-type levels and the ability to grow in the presence of detergents, but not the morphology and growth characteristics of the wild-type bacteria, suggesting a polar effect of Tn5 on the fts genes. This result and the construction of a nonpolar ddlB mutant provide solid evidence of the presence of transcriptional coupling between a gene associated with peptidoglycan biosynthesis and the fts genes required to control cell division.

Introduction

Bacteria of the genus Azospirillum are free-living, nitrogen-fixing rhizobacteria that are found in close association with plant roots, where they exert beneficial effects on plant growth and yield in many crops of agronomic importance. Upon inoculation, these bacteria attach to plant roots and proliferate, and some strains subsequently invade and colonize the internal plant tissues (Okon & Vanderleyden, 1997). As a result of this association, several greenhouse- and field-inoculation experiments with Azospirillum, carried out at different locations over the last 25 years, have demonstrated that these bacteria have beneficial effects on plant yields (Döbereiner & Pedrosa, 1987; Okon & Labandera-González, 1994; Burdman et al., 2000b; Dobbelaere et al., 2001, 2002; El Zemrany et al., 2006). These benefits are mainly derived from morphological and physiological changes in the roots of the inoculated plant that lead to an enhancement of water and mineral uptake. Secretion of plant growth-promoting substances such as auxins, gibberellins, and cytokinins by the bacteria seems to be at least partially responsible for these effects (Steenhoudt & Vanderleyden, 2000). Although other mechanisms – such as biological nitrogen fixation and nitrate reduction by the microorganisms – have been reported (Kapulnik et al., 1985; Boddey & Döbereiner, 1994; Fallik et al., 1994), the contribution of these mechanisms to plant growth has been shown to be of little significance in these systems.

Even though considerable evidence has been accumulated over many years indicating the involvement of extracellular polysaccharides and proteins in both the cell-aggregation and root-attachment processes, the precise mechanisms of these phenomena remain unexplained. The difficulty in elucidating these processes derives from their complexity because it seems that they are mediated by several different cell-surface components.

Although the Azospirillum–plant interaction is not yet fully understood, partially owing to the lack of an easily detectable plant phenotype following inoculation, some bacterial surface components have been demonstrated to be involved in the early stages of the interaction with the plant (Croes et al., 1993; Katupitiya et al., 1995; De Troch & Vanderleyden, 1996; Pereg-Gerk et al., 1998, 2000; Vande Broek et al., 1998; Burdman et al., 2000a; Jofréet al., 2004). Several authors have shown that the attachment of Azospirillum to plant roots involves a two-step process, starting with a rapid and reversible bacterial adsorption onto the root system, followed by an irreversible anchoring of the bacteria to the root surface (Michiels et al., 1991). The adsorption of Azospirillum brasilense to wheat roots was shown to be highly dependent on the presence of polar flagella, whereas the anchoring (the second step) was associated with the biosynthesis of a polysaccharide-containing fibrillar material. The lipopolysaccharide produced by A. brasilense Cd also plays a role during early interaction with maize roots, because a mutant affected in the lipopolysaccharide core showed impaired attachment to these roots as well as diminished plant–root colonization (Jofréet al., 2004).

Azospirillum brasilense Cd produces at least one exopolysaccharide, which consists of a heteropolymer containing glucose (47%), fucose (28%), galactose (11%), rhamnose (6%), mannose (3%), xylose (4%), and arabinose (1%) as the main sugars (Burdman et al., 2000a; Fischer et al., 2003). Two molecular forms of this exopolysaccharide have been described, one of high molecular weight and another of low molecular weight (Michiels et al., 1988). The biosynthetic pathway of the A. brasilense exopolysaccharide remains uncharacterized and the genes encoding the enzymes involved in the synthesis of its saccharide repeating unit as well as in the polymerization and export of the molecule have yet to be determined.

In an attempt to determine the genes associated with exopolysaccharide biosynthesis and regulation, we generated an A. brasilense Cd Tn5-mutant library and searched for changes in subsequent exopolysaccharide production. One of the Tn5 mutants displayed exopolysaccharide overproduction along with increased sensitivity to both detergents and saline conditions as well as altered cell morphology. The genetic characterization of this mutant allowed us to identify a gene cluster that includes a ddlB-homolog encoding for a d-alanine–d-alanine ligase as well as the ftsQAZ genes responsible for the septum formation in different bacterial species, suggesting that the integrity of bacterial cell envelopes is required for adaptation to saline environments.

Materials and methods

Bacterial strains, plasmids, and growth conditions

The bacterial strains and plasmids used in this study are listed in Table 1. Azospirillum brasilense Cd was grown at 33 °C in Luria–Bertani (LB) medium supplemented with 2.5 mM CaCl2 and 2.5 mM MgSO4 (Miller, 1972) or in minimal medium lactate (MML; Dreyfus et al., 1983). Escherichia coli strains were grown at 37 °C in LB medium. Antibiotics were added at the following concentrations when required: ampicillin, 25 μg mL−1; chloramphenicol, 25 μg mL−1; kanamycin (Km), 25 μg mL−1; gentamicin, 20 μg mL−1; and tetracycline, 5 μg mL−1.

Table 1.   Strains and plasmids used in this study
 Relevant characteristicsReference or source
  1. Apr, ampicillin resistance; Cmr, chloramphenicol resistance; Gmr, gentamicin resistance.

Strains
 A. brasilense CdATCC 29710 wild-type.Tarrand et al. (1978)
 A. brasilense PG1ddlB∷Tn5 mutant of A. brasilense Cd.This work
 A. brasilense AL1A. brasilense nonpolar ddlB mutantThis work
 E. coli DH5αendA1 recA1 hsdR17 supE44 thi-1 gyrA96 relA1
Δ(lacZYA-argF)U169ϕ80d lacZΔM15.
Hanahan (1983)
 E. coli JM109recA endA1 gyrA96 thi hsdR17 supE44
Δ(lac-proAB) relA1.
Sambrook et al. (1989)
 E. coli S17-1pro recA hsdR, thi RP4-2-Tc∷Mu-Km∷Tn7
integrated into the chromosome.
Simon et al. (1983)
Plasmids
 pK18mob2pK18mob derivative with unique KpnI and SacI sites in polylinker, KmrTauch et al. (1998)
 pGS9Cmr, Kmr, (Tn5) p15A replicon N-tra Tn5 donor.Selveraj & Iyer (1983)
 pSUP102Tcr, Cmr, vector movilizable pACYC184 derivative.Simon et al. (1989)
 pBluescript SKCloning and sequencing vector, Apr, phagemid,
M13 derivative, f1 origin of replication.
Stratagene, La Jolla,
California
 pJN105araC-PBAD cassette cloned in pBBR1MCS-5, Gmr.Newman & Fuqua (1999)
 pPG120-kb EcoRI fragment containing Tn5 insertion
and flanking DNA from A. brasilense PG1 cloned
into pSUP102; Kmr Tcr.
This work
 pPG23.8-kb EcoRI–BamHI fragment, containing
flanking DNA, IS50L and nptII, from pPG1
cloned into pBluescript SK; Kmr Apr.
This work
 pPG316-kb BamHI fragment containing IS50R and
flanking DNA, from pPG1, cloned into
pBluescript SK; Aprlac.
This work
 pKddlB249-bp internal fragment of ddlB from
A. brasilense cloned into pK18mob2, Kmr.
This work
 pJNddlB1000-bp fragment containing ddlB from
S. meliloti cloned into pJN105, Gmr.
This work

DNA manipulations

Plasmid and total DNA preparations, agarose gel electrophoresis, restriction-endonuclease digestion, and cloning were performed according to standard protocols (Sambrook et al., 1989). Escherichia coli transformation was effected by electroporation with a Gene Pulser apparatus (Bio-Rad Laboratories Ltd). Southern hybridizations were carried out with a nonradioactive detection kit, and a chemiluminescence method was used to detect hybridization bands according to the instructions of the manufacturer (Gibco BRL, Life Technologies). The probe was obtained after EcoRI digestion of pPG1: the resulting EcoRI fragment containing Tn5 and flanking DNA from A. brasilense Cd PG1 was then isolated from an agarose gel and purified by means of the Wizard® DNA Clean-Up System kit (Promega).

Mutagenesis

Random transposon Tn5 mutagenesis of A. brasilense Cd was carried out using the mobilizable pGS9 suicide plasmid (Vanstockem et al., 1987). The kanamycin-resistant (Kmr) transconjugants were selected on MML medium, restreaked on LB medium supplemented with calcofluor (200 μg mL−1), and screened for exopolysaccharide production by fluorescence microscopy as described below.

Fluorescence microscopy

Cells grown on LB agar plates supplemented with calcofluor (Fluorescent Brightener 28, Sigma) were suspended in 0.88% (w/v) NaCl solution and then observed under an epifluorescent microscope (Zeiss) at 365 nm.

For 4′,6-diamidino-2-phenylindole (DAPI) staining, strains were grown in MML medium until an OD600 nm of 0.3. Cells from 0.2 mL of culture were harvested by centrifugation at 3500 g for 4 min, suspended in 15 μL of 0.2 mg mL−1 DAPI, and incubated during 10 min at room temperature. DAPI was removed by centrifugation at 3500 g for 4 min and cells were suspended in 0.88% (w/v) NaCl and observed under a Zeiss Axiophot microscope equipped with epifluorescence optic sets and the corresponding filter sets. Images were recorded with an AxioCam HRc camera (Zeiss) and processed using the Adobe photoshop program.

Test of sensitivity to detergents

Azospirillum brasilense Cd strains were grown in MML medium until an OD600 nm of 0.3. Cells from 1 mL of culture were collected by centrifugation – 2 min at 3500 g– suspended in 0.88% (w/v) NaCl solution, and streaked on MML-agar plates containing 0.01% (w/v) sodium dodecyl sulfate (SDS). Plates were incubated at 33 °C for 3 days.

Exopolysaccharides isolation and quantification

For exopolysaccharide isolation, A. brasilense Cd strains were grown in MML containing 10 mM NH4Cl at 33 °C for 2 days. Exopolysaccharides were precipitated from culture supernatants with three volumes of cold ethanol. Pellets were air-dried, suspended in distilled water, and dialyzed (MWCO 12 000 Da) against distilled water for 2 days at 4 °C. Exopolysaccharides were quantified by the anthrone method (Dische, 1962).

Protein determination and growth curves

Protein determination was carried out according to Bradford (1976). The time course of cell growth in the selected Tn5 mutant was examined in standard MML medium. Growth under saline conditions was performed in MML medium supplemented with 300 mM NaCl. Cultures were grown at 33 °C on a rotary shaker at 100 r.p.m. Bacterial growth was monitored spectrophotometrically at 600 nm, and at early-, mid-, and late-exponential-phase growth an aliquot was used to estimate the viable cell number by plate counting in MML medium.

Lipopolysaccharide isolation and analysis

Cultures were grown in LB broth al 33 °C. Lipopolysaccharides were isolated in agarose-immobilized polymyxin B (Affi-prep polymyxin support, Bio-Rad Laboratories Ltd) according to Valverde et al. (1997). Lipopolysaccharides were separated by 20% (w/v) polyacrylamide-gel electrophoresis (PAGE) in an SDS-Tricine buffer system and visualized by silver staining (Tsai & Frasch, 1982).

Cloning of the Tn5-interrupted region

Total DNA from A. brasilense Cd PG1 was digested with EcoRI and ligated to pSUP102. The ligation mixture was used to transform electrocompetent E. coli S17-1 cells. Transformant clones were selected on LB medium containing kanamycin and tetracycline and screened for the presence of IS50 sequences by a PCR assay with the primers IS1 (5′-GCTCGATCTAGAACGTTCATGATAACTTCTGC-3′) and IS2 (5′-TCTGCGGACTGGCTTTCTAC-3′), which leads to a PCR product of 1.2 kb. One of the plasmid isolates [tetracycline-resistant (Tcr), Kmr] was designated pPG1.

Subcloning and outward sequencing from the transposon in both directions

Plasmid pPG1 was digested with BamHI–EcoRI and BamHI and the products were ligated to pBluescript SK. The recombinant plasmids were used to transform electrocompetent E. coli JM109 cells. Clones were selected on LB medium containing ampicillin and kanamycin and LB medium containing ampicillin and X-gal. Plasmids were isolated and purified by means of a Wizard DNA Clean-Up System kit (Promega). DNA was sequenced using Taq FS DNA polymerase and fluorescent-dideoxy terminators in a cycle sequencing method and the resultant DNA fragments were electrophoresed and analyzed using an automated Applied Biosystems 377 DNA sequencer. Sequencing service was carried out by HHMI/Keck DNA Sequencing Facility. The amino acid sequences deduced from the nucleotide sequence were compared with the GenBank database, through the use of the blast algorithm (Altschul et al., 1997).

Site-directed mutagenesis by plasmid integration

An internal 249-bp fragment of ddlB was amplified by PCR using A. brasilense Cd DNA template and primers ddlB f [5′-AAAAGGTACCTACATCGTCAAGCCGGTGGA-3′ (KpnI restriction site underlined)] and ddlB r [5′-AAAAAAGCTTATGACCGGCGCTGTATTTA-3′ (HindIII restriction site underlined)]. The PCR product was purified by means of the QIAquick®/™ PCR-purification kit (Qiagen), restricted with KpnI and HindIII, and then ligated to the corresponding sites in the pK18mob2 vector. The recombinant plasmid, designated pKddlB, was mobilized by mating from E. coli S17-1 to A. brasilense. Transconjugants were selected in MML medium supplemented with kanamycin.

Heterologous complementation

A 1000-bp fragment containing the coding region of ddlB, 47-bp upstream from ddlB start codon and 26-bp downstream from ddlB stop codon, was amplified by PCR using Sinorhizobium meliloti 2011 DNA template and primers SmddlB f [5′-AAAACTGCAGAGGTGCAAGCAAGCTTCG-3′ (PstI restriction site underlined)] and SmddlB r [5′-GCTCTAGAACGAACCCTCTTGCCCCTT-3′ (XbaI restriction site underlined)]. The PCR product was purified by means of the QIAquick®/™ PCR-purification kit (Qiagen), restricted with PstI and XbaI, and then ligated to the corresponding sites in the pJN105 vector. The recombinant plasmid, designated pJNddlB, was mobilized by mating from E. coli S17-1 to A. brasilense. The selection of transconjugants was performed on MML medium supplemented with gentamicin.

Nucleotide sequence accession number

The nucleotide sequence has been deposited in the GenBank database under accession number AF492457.

Statistical analysis

Results were analyzed by one-way anova, with the means evaluated for significance through the least-significant-difference test at P<0.05.

Results

Random Tn5 mutagenesis of A. brasilense Cd and screening for altered exopolysaccharide production

For a better understanding of genes controlling exopolysaccharide biosynthesis in A. brasilense, we generated a collection of Tn5 mutants of A. brasilense Cd as described in Materials and methods and screened for putative alterations in exopolysaccharide production by the fluorescence microscopy of cells growing on LB medium containing calcofluor (Fig. 1). Calcofluor is a fluorescent dye that binds predominantly to β-1,4- and β-1,3-linked glucans (Wood & Fulcher, 1978). One clone, designated PG1, out of 2800 transconjugants showed enhanced fluorescence when compared with the wild-type strain, thus indicating a possible alteration in exopolysaccharide production (Fig. 1a and b).

Figure 1.

 (a–c) Fluorescence microscopy showing the calcofluor phenotype of Azospirillum brasilense cells. (a) Azospirillum brasilense Cd wild-type strain. (b) Azospirillum brasilense Cd PG1 mutant. (c) The resultant clone after the homogenotization (marker exchange) of the wild-type strain with plasmid pPG1. Cells were grown as described in Materials and methods. (d and e) Colony morphology of A. brasilense Cd wild type (d) and A. brasilense Cd PG1 mutant (e) grown on MML-agar medium.

In order to confirm whether the enhanced fluorescence observed in the PG1 mutant is associated with increased exopolysaccharide production, we determined the amount of exopolysaccharides produced both by the mutant and by its wild-type strain. Figure 2a shows that mutant PG1 produced almost three times more exopolysaccharides than the wild-type strain. In spite of the increased exopolysaccharide production, mutant PG1 did not show any alteration in the mucoid-colony phenotype (Fig. 1d and e). Since alterations in exopolysaccharide production are often associated with changes in other polysaccharides such as the lipopolysaccharides, we performed an SDS-PAGE analysis of this latter polysaccharide in the mutant. No differences were observed, however, between the lipopolysaccharide electrophoretic profile of the mutant and that of the wild-type strain (Fig. 2c).

Figure 2.

 (a) Production of exopolysaccharides by wild type Azospirillum brasilense Cd and its mutants PG1 and AL1. WT: A. brasilense Cd wild type; PG1: A. brasilense Cd PG1; AL1: A. brasilense Cd AL1; WT-pJN105: A. brasilense Cd wild type harboring the plasmid pJN105; PG1–pJNddlB: A. brasilense Cd PG1 complemented with plasmid pJNddlB. Data are the average of four determinations ±SE. Different letters represent significant differences (anova, LSD test, P≤0.05). (b) Growth of A. brasilense Cd wild type and mutant PG1 on MML-agar medium supplemented with 0.01% SDS. 1, A. brasilense Cd wild type; 2, A. brasilense Cd PG1 complemented with plasmid pJNddlB; 3, A. brasilense Cd PG1. (c) Lipopolysaccharide profiles from A. brasilense Cd and its mutant PG1 after resolution by SDS-PAGE and silver staining. Lane 1, A. brasilense Cd wild-type; lane 2, A. brasilense Cd PG1.

Bacterial adaptation to saline environments often involves changes in cell-surface components such as exopolysaccharides or lipopolysaccharide (Lloret et al., 1995; Jofréet al., 1998; Chowdhury et al., 2007). For this reason, we examined the ability of the mutant PG1 to grow under conditions of saline stress. The growth of mutant PG1 was thus monitored by reading OD600 nm as well as by estimation of the viable cell number at different growth phases (early-, mid-, and late-exponential-phase growth). When the growth was estimated by reading OD, no appreciable differences were observed between the growth of the mutant PG1 and its wild-type strain under normal conditions (Fig. 3a). However, at any OD, the viable cell number of the mutant PG1 was lower than that of the wild-type strain under normal conditions (Fig. 3b). Such a difference was more evident under saline-stress conditions (Fig. 3b).

Figure 3.

 Effects of saline conditions on the bacterial growth rate of wild-type Azospirillum brasilense and ddlB mutants. (a) The OD increase, while (b) shows the counts of the CFU mL−1. Square: A. brasilense Cd wild type; triangle: A. brasilense Cd PG1 mutant; circle: A. brasilense Cd AL1 mutant; lozenge: A. brasilense Cd wild type harboring plasmid pJN105; star: A. brasilense Cd PG1 complemented with plasmid pJNddlB. Saline conditions were reached adding 300 mM NaCl (black-solid symbols) to MML medium. MML medium without NaCl was used as control (open symbols). Data are the average of three determinations.

Cloning of the Tn5-interrupted DNA region in the PG1 mutant

A 20-kb EcoRI Tn5-containing fragment isolated from mutant PG1 was cloned into vector pSUP102 to yield plasmid pPG1 as indicated in Materials and methods. The 20-kb insert was used as a probe in Southern hybridization against a total EcoRI digestion of genomic DNA from the mutant and from the wild type. Both wild-type and mutant DNA gave a single hybridization signal corresponding to DNA fragments of 14.5 and 20 kb, respectively. The 5.5 kb difference in size was consistent with the presence of a Tn5 insertion in the mutant.

Reverse genetics

The mobilizable nonreplicative plasmid pPG1 was used to homogenotize the Tn5 mutation into the wild-type strain A. brasilense Cd via double crossing-over (i.e. marker exchange). The recombinant obtained (PG2) showed a calcofluor phenotype coincident with that of the original mutant PG1 (Fig. 1c). This result confirmed that the cloned DNA region containing the Tn5 insertion was responsible for the observed changes in the fluorescence of the mutant in the presence of calcofluor.

Genetic characterization of the Tn5-interrupted region present in A. brasilense Cd PG1

The DNA region interrupted by Tn5 in plasmid pPG1 was subcloned and 3.8 kb were sequenced (accession number AF492457) as described in Materials and methods. The proposed genetic organization of the DNA bordering the Tn5 is presented in Fig. 4. The Tn5 insertion mapped at the 3′ end (18 bp upstream from the stop codon) of a 632-bp ORF whose partial deduced amino acid sequence showed significant similarity to d-alanine–d-alanine ligase (ddlB) from Agrobacterium tumefaciens (67%), Gluconobacter oxydans (65%), Mesorhizobium sp. BNC1 (66%), Zymomonas mobilis (62%), S. meliloti (62%), and Bartonella bacilliformis (61%) among others. d-Alanine–d-alanine ligase is an enzyme involved in peptidoglycan biosynthesis (Belanger et al., 2000). Immediately downstream of ddlB and transcribed in the same direction, three complete ORFs could be identified. The translation products of these ORFs shared significant sequence similarity to FtsQ, FtsA, and FtsZ; the cell-division proteins involved in septum formation (Jensen et al., 2005). The amino acid sequence similarity ranged from 47% and 54% for FtsQ, 60% and 68% for FtsA, and 82% and 85% for FtsZ.

Figure 4.

 Schematic representation of the gene organization of the 3.8-kb region containing the Tn5 insertion present in PG1 mutant.

It has been reported that mutations in ftsA are responsible for anomalous filamentous growth. The FtsA protein is a component of the septum which may interact with FtsZ. FtsZ is an essential protein for bacterial cell division that assembles into a dynamic ring on the inner surface of the cytoplasmic membrane at the location of cell division (Karimova et al., 2005). Taking into account the genetic evidence obtained, we examined – using fluorescence microscopy – the cell morphology of the mutant strain PG1 grown in MML medium under normal and saline-stress conditions. The obtained images showed the presence of highly pleomorphic bacilli (Fig. 5c). This abnormal cell morphology was more evident under saline-stress conditions (Fig. 5d).

Figure 5.

 Fluorescence microscopy of DAPI-stained Azospirillum brasilense cells showing the cell morphology of ddlB mutants under normal and saline-stress conditions. Cultures were grown in MML medium until an OD600 nm of 0.3. Saline-stress conditions were achieved supplementing the culture medium with 300 mM NaCl. (a and b) A. brasilense Cd wild-type; (c and d) A. brasilense Cd PG1 mutant; (e and f) A. brasilense Cd AL1 mutant; (g and h) A. brasilense Cd PG1 complemented with plasmid pJNddlB; (i and j) A. brasilense Cd wild type harboring plasmid pJN105. (a, c, e, g, and i) Normal conditions; (b, d, f, h, and j) Saline-stress conditions. The square in the right corner indicates the magnified zone. Scale bar: 5 μm.

Interestingly, the observed cell morphology in mutant PG1 could explain the lower viable cell number achieved in mutant PG1 compared with the wild-type strain (Fig. 3b), because cells from mutant PG1 are longer and wider than those from the wild-type strain (Fig. 5a–d). In addition to cell morphology, we also examined the sensitivity to SDS in the mutant PG1. The results, presented in Fig. 2b, showed that mutant PG1 was more sensitive to SDS than its wild-type strain, suggesting alterations in the integrity of bacterial cell envelopes. The observed phenotype could be attributed to the mutation in ddlB, because such an alteration was abolished by the introduction of plasmid pJNddlB carrying a ddlB ortholog from S. meliloti (Fig. 2b).

Generation of a nonpolar ddlB mutant of A. brasilense Cd

The observed changes in cell morphology in mutant PG1 (Fig. 5c and d) could be the consequence of a polar effect on the transcription of ftsQAZ genes caused by Tn5 insertion. To ascertain whether such a hypothesis was correct, we constructed the nonpolar ddlB mutant A. brasilense Cd AL1 by a site-directed plasmid integration in which the lac promoter of vector pK18mob2 was reading downstream ddlB. Restoration of the wild-type cell morphology was subsequently observed in the nonpolar ddlB mutant AL1 (Fig. 5e and f), thus suggesting that the morphologic alteration in the PG1 mutant may have been caused by a polar effect of Tn5 insertion. Exopolysaccharide production in mutant AL1 was found to be increased to the same extent as that in mutant PG1 (Fig. 2a).

Heterologous complementation of the A. brasilense Cd PG1 mutant with S. meliloti DdlB

To make our assumption of a polar effect of Tn5 insertion in mutant PG1 stronger, we performed heterologous complementation of mutant PG1 with a ddlB ortholog from S. meliloti and evaluated the exopolysaccharide production, cell morphology, and growth of the complemented strain compared with mutant PG1.

The introduction of plasmid pJNddlB into A. brasilense Cd PG1 restored exopolysaccharide production to wild-type levels as well as the ability to grow in the presence of SDS, indicating functional complementation (Fig. 2a and b). As expected, normal exopolysaccharide production was also observed when plasmid pJNddlB was introduced into mutant AL1 (not shown). The effect of high concentrations of sodium chloride on the complemented PG1 mutant was still observed (Fig. 3b, CFU counts), suggesting that changes in cell division – and not exopolysaccharide overproduction – were mainly responsible for mutant growth behavior under salt stress. Similarly, cell morphology was shown to be still impaired in the complemented mutant PG1 as compared with the wild type (Fig. 5g and h). The results are fully consistent with our assumption for a polar effect of Tn5 insertion in the mutant PG1. As a control, plasmid pJN105 was introduced in the wild-type strain to evaluate any possible alterations in exopolysaccharide production and/or cell morphology caused by the presence of plasmid pJN105. No significant differences were observed with the wild type either in the amount of exopolysaccharides (Fig. 2a) or in cell morphology (Fig. 5i and j).

Discussion

In this work, we have characterized an A. brasilense Cd Tn5 mutant displaying an enhanced fluorescence phenotype on medium containing calcofluor. This phenotype was shown to be associated with increased exopolysaccharide production. Furthermore, the mutant was found to be more sensitive to SDS and salinity – a frequent phenotype, in turn, associated with mutants altered in surface components in Azospirillum.

Genetic characterization of the EcoRI DNA fragment that contains Tn5 in mutant PG1 led to the recognition of four ORFs transcribed in the same direction and with sequence similarity to ddlB, ftsQ, ftsA, and ftsZ from A. tumefaciens, G. oxydans, Mesorhizobium sp. BNC1, Z. mobilis, S. meliloti, and B. bacilliformis among others. The genomic organization of the A. brasilense ddlB–ftsQ–ftsA–ftsZ cluster presents perfect synteny with these same six bacterial species. In A. brasilense Cd the ddlB and fstQ ORFs were found to be overlapped by 16 bp, indicating possible transcriptional coupling.

Bacterial peptidoglycan is an important biopolymer layer that protects the bacterial cell from environmental stresses. Its biosynthesis begins with an initial cytoplasmic stage in which an UDP-N-acetyl muramic acid (UDP-MurNAc)-pentapeptide precursor is synthesized. In a second membrane-associated stage, the MurNAc-pentapeptide is transferred to a lipid carrier, derivatized with a N-acetyl glucosamine (GlcNAc) residue, and translocated across the membrane. During a third stage, these lipid-linked precursors are incorporated into the membrane through transglycosylation and transpeptidation reactions (Bugg & Walsh, 1992). The ddlB gene encodes the d-alanine–d-alanine ligase that catalyzes the condensation of two d-alanine residues to form the d-alanine–d-alanine dipeptide that is subsequently ligated to UDP-N-acetylmuramoyl tripeptide during the first cytoplasmic stage of peptidoglycan biosynthesis.

In constrast, FtsQ, FtsA, and FtsZ are proteins involved in the cell-division process. In E. coli, for example, at least 15 proteins have been implicated in cell division. All of these proteins localize to a ring-like structure, at midcell, commonly referred to as the divisome (Buddelmeijer & Beckwith, 2002; Pichoff & Lutkenhaus, 2002, 2005). The structural basis of this divisome is the Z ring, whose structure is the result of the GTP-dependent polymerization of the bacterial-tubulin homolog FtsZ. The cytoplasmic protein FtsA is another divisome-essential protein that is required for the stabilization of the Z ring and its anchoring to the membrane. Once the Z ring is formed, the remaining divisome component proteins are recruited. FtsQ appears to play a central role in the assembly of the divisome. FtsQ is a bitopic membrane protein consisting of a short cytoplasmic domain, a transmembrane segment, and a periplasmic domain (Scheffers et al., 2007). Two-hybrid analysis has suggested that FtsQ interacts with itself as well as with FtsA, FtsK, FtsX, FtsL, FtsB, FtsW, FtsI, FtsN, and YmgF (Di Lallo et al., 2003; Karimova et al., 2005; D'Ulisse et al., 2007). The interaction of FtsQ with FtsA and FtsK requires its N-terminal domain, whereas all other interactions require only its periplasmic domain (Karimova et al., 2005; D'Ulisse et al., 2007). In addition, the FtsQ transmembrane segment plays a role in FtsQ placement at the division site (Scheffers et al., 2007). Several promoters have been identified within the ddlBftsQA coding region in E. coli and B. bacilliformis. In E. coli, the high ratio of FtsZ to FtsA and FtsQ is created by the presence of six strong promoters and one weak promoter in the sequence upstream of the ftsZ gene, while in B. bacilliformis three promoters have been identified in the ddlBftsQA coding region (Fiskus et al., 2003).

Considering that the translation product of ddlB is associated with the biosynthesis of peptidoglycan, the increased exopolysaccharide production in the PG1 mutant is most likely the result of surface-compensation effects accompanying changes in the integrity of the bacterial envelope. Recent results reported by Nagarajan et al. (2007) showed an increased expression of genes related to exopolysaccharides and lipopolysaccharide under saline stress in A. brasilense. In our case, however, the increase in exopolysaccharide production in the nonpolar mutant affected in ddlB did not result in an improved tolerance to salinity.

The A. brasilense Cd PG1 mutant showed changes in its cell morphology that were attributed to a polar effect of Tn5 insertion on the transcription of ftsQAZ genes. Strong evidence for this assumption arose from the fact that the nonpolar mutant AL1 recovered a nearly normal bacterial morphology being still altered in exopolysaccharide production. In addition, complementation of the Tn5 (polar) mutant PG1 with a ddlB ortholog from S. meliloti abolished overproduction of exopolysaccharides, restored wild-type tolerance to SDS, but could not correct the filamentous cell morphology. Likely, Tn5 insertion interrupts the transcription of ftsQAZ from promoters located upstream/within ddlB. Several authors have demonstrated that FtsA and FtsZ are required in an adequate ratio for assembling into a ring. Such a ratio of protein species is maintained from the transcription of different promoters located within the ddlBftsQAZ coding region. Alterations of the FtsZ/FtsA ratio are well known to lead E. coli cells to grow in long filamentous chains (Dai & Lutkenhaus, 1992). In our PG1 mutant, the Tn5 mutation that abolishes promoter activities originated from upstream/within ddlB is expected to result in an imbalance in the FtsA : FtsZ ratio and, as a consequence, in altered cell morphology.

The participation of enzymes related to cell-envelope biosynthesis in the adaptation to saline environments has already been demonstrated by Fulda et al. (2000) and Nagarajan et al. (2007). These authors postulated that alterations in the external cell layers probably establish an enhanced diffusion barrier that reduces the influx of inorganic ions into the periplasm. In S. meliloti 1021, Vriezen (2005) identified a chromosomal transcriptional fusion Tn5luxAB resulting in a rhizobial mutant that was sensitive to desiccation. The Tn5luxAB from this mutant was inserted into an NaCl-inducible ORF encoding a putative d-alanine–d-alanine ligase, supporting a likely active metabolism of peptidoglycan during adaptation of S. meliloti to saline environments. Although we do not yet know how ddlB transcription under saline stress is modulated, in A. brasilense Cd the mutation of the d-alanine–d-alanine ligase in the nonpolar mutant AL1 did not result in changes in the strain tolerance to salinity. The impaired growth under high salt environment in mutant PG1 has thus to be attributed to a deficient expression of fts genes and not directly to the ddlB mutation.

The results presented described a genetic region associated with the envelope structure in A. brasilense, and showed transcriptional coupling between an enzyme of the peptidoglycan biosynthesis (DdlB) and proteins that control cell division (Fts). We have shown that alteration in these activities results in changes in the amount of exopolysaccharides produced, in the bacterial tolerance to saline stress, and in the cell morphology, depending on the kind of genes affected. New studies will be required to understand the molecular basis on which exopolysaccharide production is upregulated in the ddlB mutant, as well as the kind of Fts protein whose expression is dependent on upstream dllB transcription.

Acknowledgements

This investigation was supported by the Secretaría de Ciencia y Técnica de la Universidad Nacional de Río Cuarto, the Agencia Nacional de Promoción Científica y Tecnológica, and the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina. The authors wish to thank Dr Donald F. Haggerty for editing the manuscript. E.J., S.F., and A.L. are members of the Research Career of CONICET.

Ancillary