ACC deaminase genes are conserved among Mesorhizobium species able to nodulate the same host plant

Authors

  • Francisco X. Nascimento,

    1. Laboratório de Microbiologia do Solo, I.C.A.A.M., Instituto de Ciências Agrárias e Ambientais Mediterrânicas, Universidade de Évora, Évora, Portugal
    Search for more papers by this author
  • Clarisse Brígido,

    1. Laboratório de Microbiologia do Solo, I.C.A.A.M., Instituto de Ciências Agrárias e Ambientais Mediterrânicas, Universidade de Évora, Évora, Portugal
    Search for more papers by this author
  • Bernard R. Glick,

    1. Department of Biology, University of Waterloo, Waterloo, ON, Canada
    Search for more papers by this author
  • Solange Oliveira

    Corresponding author
    • Laboratório de Microbiologia do Solo, I.C.A.A.M., Instituto de Ciências Agrárias e Ambientais Mediterrânicas, Universidade de Évora, Évora, Portugal
    Search for more papers by this author

Correspondence: Solange Oliveira, Departamento de Biologia, Universidade de Évora, Apartado 94, 7002-554 Évora, Portugal. Tel.: +351266760878; fax: +351266760914; e-mail: ismo@uevora.pt

Abstract

Rhizobia strains expressing the enzyme 1-aminocyclopropane-1-carboxylate (ACC) deaminase have been reported to display an augmented symbiotic performance as a consequence of lowering the plant ethylene levels that inhibit the nodulation process. Genes encoding ACC deaminase (acdS) have been studied in Rhizobium spp.; however, not much is known about the presence of acdS genes in Mesorhizobium spp. The aim of this study was to assess the prevalence and phylogeny of acdS genes in Mesorhizobium strains including a collection of chickpea-nodulating mesorhizobia from Portugal. ACC deaminase genes were detected in 10 of 12 mesorhizobia type strains as well as in 18 of 18 chickpea Mesorhizobium isolates studied in this work. No ACC deaminase activity was detected in any Mesorhizobium strain tested under free-living conditions. Despite the lack of ACC deaminase activity, it was possible to demonstrate that in Mesorhizobium ciceri UPM-Ca7T, the acdS gene is transcribed under symbiotic conditions. Phylogenetic analysis indicates that strains belonging to different species of Mesorhizobium, but nodulating the same host plant, have similar acdS genes, suggesting that acdS genes are horizontally acquired by transfer of the symbiosis island. This data, together with analysis of the symbiosis islands from completely sequenced Mesorhizobium genomes, suggest the presence of the acdS gene in a Mesorhizobium common ancestor that possessed this gene in a unique symbiosis island.

Introduction

The plant hormone ethylene is known for its inhibitory effects in various aspects of nodule formation and development (Guinel & Geil, 2002) in many different leguminous plants (Goodlass & Smith, 1979; Peters & Crist-Estes, 1989; Penmetsa & Cook, 1997; Tamimi & Timko, 2003). Several authors have suggested that ethylene can inhibit numerous steps of the nodulation process. For example, it has been suggested that ethylene inhibits the calcium spiking process responsible for the perception of bacterial Nod factors in Medicago truncatula (Oldroyd et al., 2001). Ethylene inhibits the formation and development of rhizobial infection threads in Pisum sativum cv. Sparkle (Lee & La Rue, 1992). In Trifolium repens roots, ethylene inhibits cortical cell division, a process that is indispensable for nodule primordia formation (Goodlass & Smith, 1979).

To obviate some of the inhibitory effects of ethylene in nodule formation, development and function, some rhizobial strains utilize different mechanisms for lowering ethylene levels such as the production of the enzyme 1-aminocyclopropane-1-carboxylate (ACC) deaminase; this enzyme is responsible for the cleavage of ACC (the immediate precursor of ethylene in plants) to ammonia and α-ketobutyrate (Honma & Shimomura, 1978), contributing to increase the competitiveness of the strains because of advantages in the processes of nodule formation and occupancy (Ma et al., 2003b, 2004). Other rhizobial strains lower ethylene levels by producing the compound rhizobitoxine, an inhibitor of the plant enzyme ACC synthase (Sugawara et al., 2006).

The prevalence of ACC deaminase genes in rhizobia has been studied primarily in Rhizobium spp. (Ma et al., 2003a; Duan et al., 2009). In these studies, many Rhizobium spp. have been found to possess an acdS gene and produce ACC deaminase under free-living conditions. For example, in a rhizobia collection of isolates from Saskatchewan (Canada), 27 Rhizobium isolates possessed an acdS gene and were able to produce ACC deaminase, thus, showing that acdS genes are present throughout Rhizobium isolates (Duan et al., 2009).

On the other hand, notwithstanding reports documenting the presence of ACC deaminase in Mesorhizobium spp., not much is known about the environmental distribution of acdS genes in this bacterial genus. The first report on acdS gene presence in Mesorhizobium was obtained following the complete sequencing of Mesorhizobium sp. MAFF303099 (Kaneko et al., 2000). Subsequently, the presence of an acdS gene in the symbiosis island of Mesorhizobium loti R7A was also reported (Sullivan et al., 2002). However, when Mesorhizobium sp. MAFF303099 and Mesorhizobium ciceri UPM Ca-7 were tested for ACC deaminase activity and the presence of an acdS gene, no activity was detected and the acdS gene was not found in M. ciceri (Ma et al., 2003b). Recently, the genome sequences of Mesorhizobium opportunistum WSM2075T (Lucas et al., 2011a), Mesorhizobium australicum WSM2073T (Lucas et al., 2011b), and Mesorhizobium ciceri bv. biserrulae WSM1271 (Lucas et al., 2011c), revealed the presence of an acdS gene in these strains.

In some strains of Mesorhizobium, the production of ACC deaminase has been shown to be an important mechanism to promote nodule formation. When compared to the wild-type strain, Mesorhizobium sp. MAFF303099 acdS knockout mutant has a decreased ability to form and occupy nodules, losing both its effectiveness and competitiveness (Uchiumi et al., 2004). Moreover, a chickpea Mesorhizobium expressing an exogenous ACC deaminase significantly increased its ability to nodulate chickpea (Nascimento et al., 2012ab).

Although in some Mesorhizobium strains, no ACC deaminase activity was detected under free-living conditions (Ma et al., 2003b; Nascimento et al., 2012a), it has been shown that Mesorhizobium. sp. MAFF303099 expresses ACC deaminase under symbiotic conditions, in a NifA2-dependent manner (Uchiumi et al., 2004; Nukui et al., 2006). One explanation for these somewhat disparate results includes the possibility that those acdS genes under the transcriptional control of a NifA-regulated promoter are either exclusively or primarily expressed within nodules resulting in a decreased rate of nodule senescence. On the other hand, those acdS genes under the transcriptional control of an Lrp-regulated promoter (Ma et al., 2003a) are primarily involved in facilitating the nodulation process and are not expressed within the nodule itself.

The aim of the present study was to assess the prevalence and phylogeny of acdS genes in Mesorhizobium strains including isolates from a collection of chickpea mesorhizobia from Portuguese soils.

Material and methods

Bacteria and growth conditions

In the present study, 12 Mesorhizobium type strains as well as 18 chickpea Mesorhizobium isolates from Portugal were tested for the presence of acdS genes and ACC deaminase activity under free-living conditions.

The chickpea Mesorhizobium isolates from Portugal were collected from different sites throughout the country, as previously described (Alexandre et al., 2009).

Mesorhizobium strains were grown at 28 °C in TY medium (Beringer, 1974), in YMA medium (Vincent, 1970), and in modified M9 minimal medium (Robertsen et al., 1981) when necessary. The bacterial strains used in this work are presented in Table 1.

Table 1. Bacterial strains and gene accession numbers (GenBank) used in this work
 OriginHostACCD activityacdSnifHnodC16S rRNA
  1. ND, not detected; NT, not tested; NA, not available.

Strain/isolate
 Agrobacterium tumefaciens D3GermanyYesAF315580HM143942
 Azorhizobium caulinodans ORS571SenegalSesbania rostrataNTAP009384AP009384AP009384AP009384
 Bradyrhizobium japonicum USDA110USAGlycine maxNTBA000040BA000040BA000040BA000040
 Phyllobacterium brassicacearum STM 196TGermanyYesEF452620.1AY785319
 Rhizobium leguminosarum bv. viciae 3841UKPisum sativumNTAM236084AM236084AM236084AM236080
 Sinorhizobium medicae WSM419ItalyMedicago lupulinaNTCP000741CP000740CP000741CP000738
 Sinorhizobium meliloti SM11GermanyMedicago sativaYesCP001831CP001831CP001831CP001830
 Mesorhizobium amorphae ACCC 19665TChinaAmorpha fruticosa, LespedezaNDNDEU267714AF217261AF041442
 M. albiziae CCBAU 61158TChinaAlbizia kalkora, Albizia julibrissin, Glycine max, Leucaena leucocephala and Phaseolus vulgarisNDJQ013380DQ311093GQ167236DQ100066
 M. australicum WSM2073TAustraliaBiserrula pelecinusNTAGIX00000000AGIX00000000AGIX00000000AY601516
 M. chacoense LMG19008TArgentinaProsopis albaNDJQ013381DQ450927DQ450937AJ278249
 M. ciceri UPM-Ca7TSpainCicer arietinumNDJQ013382DQ450928DQ407292DQ444456
 M. ciceri bv. biserrulae WSM1271ItalyBiserrula pelecinusNDCP002447CP002447CP002447CP002447
 M. huakuii CCBAU2609TChinaAstragalus sinicus, AcaciaNDNDNANA D13431
 Mesorhizobium sp. MAFF303099Japan Lotus corniculatus ND BA000012 BA000012 BA000012 BA000012
 M. loti LMG6125TNew Zealand Lotus corniculatus ND JQ013383 DQ450929 DQ450939 X67229
 M. loti R7ANew Zealand Lotus corniculatus NT AL672114 AL672114 AL672113 U50166
 M. mediterraneum UPM-Ca36TSpain Cicer arietinum ND JQ013384 DQ450930 DQ450940 L38825
 M. opportunistum WSM2075TAustralia Biserrula pelecinus NT CP002279 CP002279 CP002279 CP002279
 M. plurifarium ORS1032TSenegalAcacia senegal, Prosopis juriflora, LeucaenaND JQ013385 DQ450931 FJ745283 Y14158
 M. septentrionale HAMBI2582TChina Astragalus adsurgens ND JX392804 DQ450932 DQ450941 AF508207
 M. tarimense CCBAU83306TChinaGlycyrrhiza uralensis, Lotus corniculatus, Oxytropis glabra and Robinia pseudoacaciaND JQ013387 EU252607 EF050786 EF035058
 M. tianshanense A-1BSTChinaGlycyrrhiza pallidflora, Swansonia, Glycine, Caragana, SophoraND JQ013388 DQ450934 DQ450943 AF041447
 M. thiogangeticum SJTTIndia Clitoria ternatea ND JQ013389 NANA AJ864462
 M. 101-ÉvoraPortugal Cicer arietinum ND JQ013399 NANA DQ787138
 M. 6b-BejaPortugal Cicer arietinum ND JQ013398 DQ431732 DQ431753 AY225381
Mesorhizobium isolate
 M. BR-8-BragançaPortugal Cicer arietinum ND JQ013390 JQ033936 JQ033958 EU652123
 M. C-1-CoimbraPortugal Cicer arietinum ND JQ013400 NANA EF504313
 M. C-14-CoimbraPortugal Cicer arietinum ND JQ013401 NANA EU652110
 M. CV-18-ElvasPortugal Cicer arietinum ND JQ013402 DQ431741 DQ431762 AY225390
 M. EE-7-ElvasPortugal Cicer arietinum ND JQ013391 DQ431743 DQ431764 AY225397
 M. EE-14-ElvasPortugal Cicer arietinum ND JQ013403 DQ431745 DQ431766 AY225399
 M. EE-29-ElvasPortugal Cicer arietinum ND JQ013404 DQ431746 DQ431767 AY225400
 M. G-10-GuardaPortugal Cicer arietinum ND JQ013392 JQ033940 JQ033946 EU652147
 M. G-55-GuardaPortugal Cicer arietinum ND JQ013393 JQ033931 JQ033947 EU652149
 M. L-19-LeiriaPortugal Cicer arietinum ND JQ013394 NANA EU652111
 M. LMS-1Portugal Cicer arietinum ND JQ013395 JQ033935 JQ033957 JQ033929
 M. PII-1-PortoPortugal Cicer arietinum ND JQ013405 NANA EU652133
 M. PM-1-PortimãoPortugal Cicer arietinum ND JQ013396 NANA EU652175
 M. S-15-SintraPortugal Cicer arietinum ND JQ013397 NANA EU652170
 M. STR-16-SantarémPortugal Cicer arietinum ND JQ013406 NANA EU652174
 M. V5b-ViseuPortugal Cicer arietinum ND JQ013407 NANA EU652112

Detection of acdS genes by PCR

Mesorhizobium strains and isolates were grown in 5 mL of TY medium at 28 °C for 2–4 days. The bacterial cultures were centrifuged at 16 000 g for 1 min and used for genomic DNA isolation using the E.Z.N.A bacterial DNA kit (Omega Bio-tek) following the manufacturer's suggested protocol.

To amplify the acdS gene in Mesorhizobium type strains and chickpea Mesorhizobium isolates, PCR primers were designed based on the Mesorhizobium sp. MAFF303099 and M. ciceri bv. biserrulae WSM1271 acdS gene sequences, resulting in primers F2 (5′-CAAGCTGCGCAAGCTCGAATA-3′) and R6 (5′-CATCCCTTGC ATCGATTTGC-3′).

The acdS gene was amplified in a ‘T Personal Cycler’ (Biometra) thermocycler using the following program: 3 min of initial denaturation at 95 °C, 35 cycles of 1 min denaturation at 94 °C, followed by 1 min and 30 s of primer annealing at 49 °C and 1 min of elongation at 72 °C, and a final elongation step of 5 min at 72 °C. The amplification product is a 760-bp fragment. After amplification, the PCR product was purified using the GFX DNA purification Kit (GE Healthcare, UK) and sequenced by Macrogen Inc. (Seoul, Korea). The obtained acdS gene sequences are presented in Table 1.

Detection of acdS genes by Southern hybridization

Mesorhizobium type strains and chickpea Mesorhizobium isolates were studied for the presence of an acdS gene by Southern hybridization. The partial acdS gene (approximately 810 bp) from Mesorhizobium sp. MAFF303099 was amplified by PCR using primers F (5′-GGCAAGGTCGACATCTATGC-3′) (Duan et al., 2009) and R2 (5′-GCATCGATTTGCCCTCATAG-3′). The amplified sequence was used as a DNA hybridization probe that was constructed using the DIG-High Prime DNA Labeling Kit (Roche Applied Science, Germany), according to the manufacturer's instructions.

About 2 μg of total DNA was digested with the restriction enzyme BamHI and used for Southern hybridization as described by Sambrook & Russell (2001). The hybridization process was carried using Dig Easy Hyb hybridization buffer (Roche Applied Science) at 42 °C, followed by washes at 25 and 68 °C. After membrane treatment with anti-Dig Fab fragments (Roche Applied Science) and posterior washing, the hybridization signals were detected using the CDP-star chemiluminescent method (Roche Applied Science). Membranes were then sealed in folders and exposed to X-ray film (Kodak).

RT-PCR from root nodules

To assess Mesorhizobium ciceri UPM-Ca7T acdS gene expression in root nodules, an RT-PCR analysis amplification was conducted using the chickpea Mesorhizobium symbiosis system. This strain was used because it would give useful information about ACC deaminase expression and regulation in chickpea mesorhizobia, giving further insights into the ecology of these agronomical important strains.

Three chickpea plants were grown and inoculated with M. ciceri UPM-Ca7T as described by Nascimento et al. (2012a). Briefly, chickpea seeds were surface sterilized and sown in pots containing sterilized vermiculite. The plants were grown in a growth chamber (Walk-in fitoclima, Aralab, Portugal) and irrigated with nitrogen-free nutrient solution (Broughton & Dilworth, 1971). After 3 weeks of plant growth, nodules were collected and treated for posterior RNA extraction as described by Cabanes et al. (2000). Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Germany) according to the manufacturer's suggested protocol. After extraction, about 800 ng of total RNA was treated with 3U DNase I (Roche Applied Science) according to the enzyme manufacturer's protocol. The conversion of total RNA to cDNA was conducted using RevertAid H Minus Reverse Transcriptase (Fermentas) as suggested by the manufacturer. Amplification of the acdS gene by PCR from the cDNA product was performed using primers F2 and R6 with the conditions previously described.

ACC deaminase activity assay

ACC deaminase activity was assessed in Mesorhizobium type strains and in chickpea Mesorhizobium isolates growing in free-living conditions (Table 1). Determination of ACC deaminase activity in cells was performed following the method described by Duan et al. (2009). Strains were grown in TY at 28 °C for 2 days, and cells were collected by centrifugation and washed twice with 0.1 M Tris-HCl (pH 7.5). Cells were then re-suspended in M9 minimal medium containing 5 mM ACC. The bacterial suspension was incubated with shaking (150 r.p.m.) at 28 °C for 2 days.

Enzyme activity was measured based on the determination of α-ketobuty rate resulting from ACC cleavage by ACC deaminase (Penrose & Glick, 2003).

Pseudomonas putida UW4 and Mesorhizobium sp. MAFF303099 were used as a positive and negative control, respectively.

Analysis of genome regions containing the acdS gene

The region of the genomes from M. loti R7A, Mesorhizobium sp. MAFF303099, M. ciceri bv. biserrulae WSM1271, M. australicum WSM2073T, and M. opportunistum WSM2075T that contain the acdS gene were analyzed to determine the acdS gene ‘neighborhood’. The intergenic regions upstream of the acdS gene in Mloti R7A, Mesorhizobium sp. MAFF303099, Mciceri bv. biserrulae WSM1271, M. australicum WSM2073T, and Mopportunistum WSM2075T were examined for putative upstream activator sequences (UAS). Putative NifAUAS (5′-TGT-N9–11-ACA-3′) (Alvarez-Morales et al., 1986; Buck et al., 1986; Morett & Buck, 1988) were searched in the immediate upstream region of the acdS genes using FUZZNUC (http://mobyle.pasteur.fr/cgi-bin/portal.py#forms::fuzznuc), a Web-based program of the European Molecular Biology Open Software Suite (EMBOSS) (Rice et al., 2000).

Phylogenetic analysis of acdS, nifH, nodC, and 16S rRNA genes

The acdS, nifH, nodC, and 16S rRNA gene sequences (Table 1) were analyzed using bioedit v.7.0.5.3 (Hall, 1999) and aligned with muscle (Edgar, 2004).

To obtain the best substitution model for the construction of the phylogenetic trees, the resulting acdS, nifH, nodC, and 16S rRNA gene alignments were analyzed with jModeltest (Posada, 2008). The best substitution model for each phylogenetic analysis was chosen based on the lowest Bayesian Information Criteria and Akaike Information Criteria values.

All phylogenetic trees were constructed with mega v.5.05 (Tamura et al., 2011) using the maximum likelihood method and the corresponding best substitution model selected. A bootstrap analysis of 1000 replicates was conducted for every phylogenetic analysis.

Results

Detection of acdS genes and ACC deaminase activity

Genes encoding putative ACC deaminase were detected in 10 of 12 Mesorhizobium type strains, as well as in all 18 chickpea Mesorhizobium isolates studied in this work (Table 1).

In Mesorhizobium huakuii CCBAU2609T and Mesorhizobium amorphae ACCC19665T, the ACC deaminase gene was not detected by either PCR or Southern hybridization.

Southern hybridization showed that only one copy of the acdS gene is present in most of the acdS+ Mesorhizobium type strains (Supporting Information, Fig. S1). All Portuguese chickpea mesorhizobia showed one copy of the acdS gene (data not shown). In these isolates, the acdS gene is present in a fragment of about 8 kb, similar to the fragment obtained from M. ciceri UPM-Ca7T after total DNA digestion with BamHI.

Most Mesorhizobium strains used in this study possess an acdS gene; however, ACC deaminase activity under free-living conditions was not detected in any of these strains (Table 1).

The acdS gene sequences here obtained share high identity (84–99%) with the previously described acdS gene of Mesorhizobium sp. MAFF303099.

RT-PCR from root nodules

Using Mesorhizobium acdS specific primers, a product with the expected size (about 760 bp) was obtained after PCR amplification of the cDNA from root nodule extracted RNA (Fig. S2). No fragment was amplified when using RNA from root nodules (treated with DNase I), demonstrating that possible DNA contaminants were not present.

Analysis of the genomic regions containing the acdS gene

In strains M. loti R7A, Mesorhizobium sp. MAFF303099, M. ciceri bv. biserrulae WSM1271, M. australicum WSM2073T, and M. opportunistum WSM2075T, the acdS gene is located on a symbiosis island within the chromosome (Fig. S3). Interestingly, the acdS neighborhood genes show a similar organization in the different symbiosis islands belonging to organisms that nodulate different hosts (Fig. S3). In the upstream region of the acdS gene, an fdxB gene (encoding a ferredoxin 2[4Fe-4S] III) is present in the five genomes, followed by the nif genes cluster and the nifA gene.

In the region of the genome that is immediately upstream of the acdS gene, there is a putative NifA UAS in the five abovementioned mesorhizobia genomes (data not shown).

Phylogenetic analysis of acdS, nodC, nifH, and 16S rRNA genes in Mesorhizobium

Phylogenetic analysis of the acdS gene in Mesorhizobium indicates that strains able to nodulate the same plant host have a similar acdS gene. The phylogenetic tree-based onacdS gene sequences (Fig. 1) shows three main clusters. Strains that nodulate Cicer arietinum, namely M. ciceri UPM-Ca7T, M. mediterraneum UPM-Ca36T, and all Portuguese Mesorhizobium isolates, form one group (A). The strains nodulating Biserrula pelecinus, that is, M. ciceri bv. biserrulae WSM1271, M. australicum WSM2073T, and M. opportunistum WSM2075T, form another group (B). Strains M. loti LMG6025T, M. loti R7A, Mesorhizobium sp. MAFF303099, and Mesorhizobium tarimense CCBAU 83306, all able to nodulate Lotus corniculatus, form a third group (C).

Figure 1.

acdS gene phylogram. The evolutionary history was inferred using the maximum likelihood method based on the Hasegawa-Kishino-Yano model (Hasegawa et al., 1985). A discrete gamma distribution was used to model evolutionary rate differences among sites (four categories (+G, parameter = 1.2907)). Branch lengths are measured based on the number of substitutions per site. There were a total of 531 positions in the final dataset. All positions containing gaps and missing data were eliminated. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. (A) Strains nodulating Cicer arietinum, (B) strains nodulating Biserrula pelecinus, and (C) strains nodulating Lotus corniculatus. Mesorhizobium type strains are highlighted in bold.

The same grouping is observed in the phylogenetic trees constructed using nodC (Fig. 2) and nifH (Fig. 3) gene sequences. Strains within the same groups mentioned above do not necessarily belong to the same species. This is clear upon comparison of the phylogenetic trees for acdS, nodC, and nifH genes with the 16SrRNA gene phylogenetic tree of these bacterial strains (Fig. 4).

Figure 2.

nodC gene phylogram. The evolutionary history was inferred using the maximum likelihood method based on the Tamura model (Tamura, 1992). A discrete gamma distribution was used to model evolutionary rate differences among sites [four categories (+G, parameter = 1.4998)]. The rate variation model allowed for some sites to be evolutionarily invariable ([+I], 33.8172% sites). Branch lengths are measured based on the number of substitutions per site. There were a total of 468 positions in the final. All positions containing gaps and missing data were eliminated dataset. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. (A) Strains nodulating Cicer arietinum, (B) strains nodulating Biserrula pelecinus, and (C) strains nodulating Lotus corniculatus. Mesorhizobium type strains are highlighted in bold.

Figure 3.

nifH gene phylogram. The evolutionary history was inferred using the maximum likelihood method based on the Tamura model (Tamura, 1992). A discrete gamma distribution was used to model evolutionary rate differences among sites [four categories (+G, parameter = 0.7486)]. Branch lengths are measured based on the number of substitutions per site. All positions containing gaps and missing data were eliminated. There were a total of 472 positions in the final dataset. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. (A) Strains nodulating Cicer arietinum, (B) strains nodulating Biserrula pelecinus, and (C) strains nodulating Lotus corniculatus. Mesorhizobium type strains are highlighted in bold.

Figure 4.

16S rRNA gene phylogram. The evolutionary history was inferred using the maximum likelihood method based on the Kimura 2-parameter model (Kimura, 1980). A discrete gamma distribution was used to model evolutionary rate differences among sites [four categories (+G, parameter = 0.1812)]. The rate variation model allowed for some sites to be evolutionarily invariable ([+I], 65.7025% sites). Branch lengths are measured based on the number of substitutions per site. There were a total of 512 positions in the final dataset. All positions containing gaps and missing data were eliminated. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. Mesorhizobium type strains are highlighted in bold.

Discussion

The production of ACC deaminase by rhizobia has been shown to play an important role in their symbiotic performance (Ma et al., 2003a, 2004; Conforte et al., 2010; Nascimento et al., 2012ab). ACC deaminase genes are naturally present in many strains of Rhizobium spp. and are prevalent in isolates from different geographical locations (Ma et al., 2003b; Duan et al., 2009).

In this work, we report the presence of acdS genes in 10 of 12 Mesorhizobium type strains, obtained from different geographical locations and nodulating different leguminous plants, suggesting that ACC deaminase is a common feature in most Mesorhizobium spp.

In the study conducted by Ma et al. (2003b), two Mesorhizobium strains (Mesorhizobium sp. MAFF303099 and M. ciceri UPM-Ca7T) were tested for the presence of an acdS gene. The gene was detected in Mesorhizobium sp. MAFF303099 but not in M. ciceri UPM-Ca7T. Here, we report the presence of an acdS gene in M. ciceri UPM-Ca7T as well as in Mesorhizobium sp. MAFF303099. This result may be due to the fact that a hybridization probe based on the acdS gene of Mesorhizobium sp. MAFF303099 was used in the present study, while in the study performed by Ma et al. (2003b), the probe was based on the P. putida UW4 acdS gene. This notwithstanding, similar Southern hybridization results were obtained with the Mesorhizobium sp. MAFF303099 strain, where the acdS gene is present on a ~ 6-kb fragment, as previously described by Ma et al. (2003b).

Using the acdS gene of Mesorhizobium sp. MAFF303099 as a hybridization probe, acdS genes were detected in the 18 chickpea mesorhizobia isolates tested here. These isolates belong to a collection that includes soil isolates from all over Portugal (Alexandre et al., 2009), indicating that many of the Portuguese chickpea Mesorhizobium possess an acdS gene and suggesting that ACC deaminase genes are prevalent in these chickpea-nodulating mesorhizobia.

However, similar to the results obtained by Ma et al. (2003b) with M. ciceri UPM-Ca7T and Mesorhizobium sp. MAFF303099, ACC deaminase activity was not detected, under free-living conditions, in any of the Mesorhizobium strains tested. On the other hand, Uchiumi et al. (2004) demonstrated that Mesorhizobium sp. MAFF303099, despite showing no ACC deaminase under free-living conditions, produces ACC deaminase in the bacteroid state, indicating that ACC deaminase is only produced under symbiotic conditions. Subsequent studies by Nukui et al. (2006) showed that ACC deaminase production by Mesorhizobium sp. MAFF303099 is under transcriptional control of the NifA2 protein. In the work reported here, RNA was extracted from M. ciceri UPM-Ca7T nodules, and after RT-PCR amplification, it was possible to detect the acdS transcript using Mesorhizobium acdS specific primers. This indicates that M. ciceri UPM-Ca7T also expresses its acdS gene under symbiotic conditions. In addition to the data of Uchiumi et al. (2004) and Nukui et al. (2006), this result suggests that ACC deaminase production under symbiotic conditions may occur in many Mesorhizobium strains. Moreover, analysis of the upstream regions of the acdS gene in M. loti R7A, Mesorhizobium sp. MAFF303099, M. ciceri bv. biserrulae WSM1271, M. australicum WSM2073T, and M. opportunistum WSM2075T indicate a putative NifA UAS, suggesting that NifA regulation of acdS expression may be common within the Mesorhizobium genus.

The acdS phylogenetic tree shows a topology similar to the symbiosis (nodC and nifH) genes-based trees (Figs 2 and 3; Laranjo et al., 2008), grouping isolates that nodulate the same host, rather than grouping by species as in the 16S rRNA gene-based phylogeny. Several studies show that many Mesorhizobium strains have acquired the ability to nodulate a specific host by acquiring a symbiosis island carrying specific symbiosis genes (Sullivan et al., 1995; Sullivan & Ronson, 1998; Nandasena et al., 2006, 2007). Therefore, we suggest that the acdS gene is likely to be horizontally transferred between Mesorhizobium species by exchange of the symbiosis island. This hypothesis is supported by the presence of the acdS gene in the symbiosis island of M. loti R7A, Mesorhizobium sp. MAFF303099, M. ciceri bv. biserrulae WSM1271, M. australicum WSM2073T, and M. opportunistum WSM2075T, close to the nitrogen fixation genes cluster.

Curiously, in strains M. amorphae ACCC19665T and M. huakuii CCBAU2609T, the acdS gene was not detected. These strains have their symbiosis genes in plasmids (Wang et al., 1999; Zhang et al., 2000) and not in the chromosome on a symbiosis island, as in other Mesorhizobium strains (Kaneko et al., 2000; Sullivan et al., 2002).

Analysis of the symbiosis islands of strains M. loti R7A, Mesorhizobium sp. MAFF303099, M. ciceri bv. biserrulae WSM1271, M. australicum WSM2073T, and M. opportunistum WSM2075T shows a similar gene organization, suggesting that symbiosis islands may have evolved from a single common ancestor and that the acdS gene was already present in the symbiosis island at that time.

Following extensive gene transfer analysis, Slater et al. (2009) suggested that Mesorhizobium strains may have evolved by plasmid gene integration into the ancestral chromosome. In other members of α-Proteobacteria and in other rhizobial strains, acdS genes are often found on plasmids (Young et al., 2006; Kuhn et al., 2008; Kaneko et al., 2010). Interestingly, in Rhizobium leguminosarum bv. viciae 3841, the acdS gene is located on the pRL10 plasmid near the nitrogen fixation genes cluster (Young et al., 2006). This arrangement is also observed in Sinorhizobium meliloti BL225C on the plasmid pSINMEB01 (Lucas et al., 2011d).

All together these data suggest that the presence of the acdS gene in Mesorhizobium spp. dates to a common ancestor possessing this gene in a symbiosis island. Therefore, the acdS gene appears to be horizontally transferred between different Mesorhizobium species by exchange of the symbiosis island, keeping its regulatory system intact, so that this gene is only expressed in symbiotic conditions under the control of the NifA protein.

Acknowledgements

This work has received funding from Fundação para a Ciência e a Tecnologia (FCT), co-financed by EU-FEDER (PTDC/BIO/80932/2006) and from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 247669. C. Brígido acknowledges a PhD fellowship (SFRH/BD/30680/2006) from FCT. We thank G. Mariano for technical assistance.

Ancillary