Key roles of microsymbiont amino acid metabolism in ...

98 downloads 0 Views 2MB Size Report
Mar 7, 2014 - in this organism (Dessaux et al., 1986; Cho & Winans. 1996 ..... Bobik C, Meilhoc E, Batut J. (2006). .... Chen W-M, Prell J, James EK, et al.
Critical Reviews in Microbiology

ISSN: 1040-841X (Print) 1549-7828 (Online) Journal homepage: http://www.tandfonline.com/loi/imby20

Key roles of microsymbiont amino acid metabolism in rhizobia-legume interactions Michael Frederick Dunn To cite this article: Michael Frederick Dunn (2015) Key roles of microsymbiont amino acid metabolism in rhizobia-legume interactions, Critical Reviews in Microbiology, 41:4, 411-451, DOI: 10.3109/1040841X.2013.856854 To link to this article: http://dx.doi.org/10.3109/1040841X.2013.856854

Published online: 07 Mar 2014.

Submit your article to this journal

Article views: 234

View related articles

View Crossmark data

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=imby20 Download by: [UNAM Ciudad Universitaria]

Date: 18 January 2016, At: 07:26

http://informahealthcare.com/mby ISSN: 1040-841X (print), 1549-7828 (electronic) Crit Rev Microbiol, 2014; 41(4): 411–451 ! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/1040841X.2013.856854

REVIEW ARTICLE

Key roles of microsymbiont amino acid metabolism in rhizobia-legume interactions Michael Frederick Dunn

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Centro de Ciencias Geno´micas, UNAM, Cuernavaca, Mexico

Abstract

Keywords

Rhizobia are bacteria in the a-proteobacterial genera Rhizobium, Sinorhizobium, Mesorhizobium, Azorhizobium and Bradyrhizobium that reduce (fix) atmospheric nitrogen in symbiotic association with a compatible host plant. In free-living and/or symbiotically associated rhizobia, amino acids may, in addition to their incorporation into proteins, serve as carbon, nitrogen or sulfur sources, signals of cellular nitrogen status and precursors of important metabolites. Depending on the rhizobia-host plant combination, microsymbiont amino acid metabolism (biosynthesis, transport and/or degradation) is often crucial to the establishment and maintenance of an effective nitrogen-fixing symbiosis and is intimately interconnected with the metabolism of the plant. This review summarizes past findings and current research directions in rhizobial amino acid metabolism and evaluates the genetic, biochemical and genome expression studies from which these are derived. Specific sections deal with the regulation of rhizobial amino acid metabolism, amino acid transport, and finally the symbiotic roles of individual amino acids in different plant-rhizobia combinations.

Amino acid catabolism, amino acid synthesis, amino acid transport, nitrogen fixation, rhizobia-legume symbiosis

Introduction The biologically useful nitrogen present in organic compounds is derived from the reduction of atmospheric nitrogen. A major source of reduced nitrogen is biological nitrogen fixation, which is performed strictly by Archaebacteria and Eubacteria having the nitrogenase and accessory enzyme systems necessary for the reduction of atmospheric nitrogen to ammonia. While many free-living bacteria are diazotrophs, the greatest quantity of fixed nitrogen in the biosphere is produced by symbiotic biological nitrogen fixation. Reduced nitrogen is most frequently the limiting nutrient for plants, including those in agricultural systems. Besides being a fascinating system for studying the molecular basis of an ecologically essential process, a practical aspect of research on symbiotic nitrogen fixation is the potential for the development of improved rhizobal inoculants that could reduce the need for chemical fertilizers and contribute to a more sustainable agriculture (Olivares et al., 2013). Several genera of gram-negative a-proteobacteria, including species of Azorhizobium, Bradyrhizobium, Mesorhizobium, Rhizobium and Sinorhizobium (collectively called rhizobia), are able to establish nitrogen-fixing symbioses with leguminous host plants (for recent reviews, see Haag et al., 2013; Udvardi & Poole, 2013; Terpolilli et al., 2012). Rhizobia induce the formation of specialized

Address for correspondence: Michael Frederick Dunn, Centro de Ciencias Geno´micas, UNAM, Av. Universidad s/n, Col. Chamilpa, Cuernavaca 62210, Mexico. E-mail: [email protected]

History Received 22 July 2013 Revised 3 October 2013 Accepted 15 October 2013 Published online 6 March 2014

structures, called nodules, on the roots of compatible legumes. Within the plant cells of mature nodules intracellular rhizobia, called bacteroids in their nitrogen-fixing form, are present in a very low oxygen environment maintained by the nodule structure and by plant and bacteroid metabolism. This low-oxygen environment allows the oxygen-sensitive nitrogenase complex to reduce atmospheric nitrogen to ammonia, which then travels (by simple diffusion or perhaps a specific transport process) to the plant cytosol where it is incorporated into nitrogenous compounds for use by the plant. In turn, the bacteroids receive reduced carbon sources from the plant. In bacteroids, the production and export of fixed nitrogen occurs concomitantly with a severe reduction in ammonia assimilation into amino acids, allowing the export of the vast majority of ammonia to the plant. Consistent with the nondividing state of bacteroids, proteomic and transcriptomic studies show that many processes needed for the growth of rhizobia in cultures, including the synthesis of many amino acids, are downregulated but still present in bacteroids, which continue to produce proteins required for microaerobic metabolism and nitrogen fixation. In bacteroids, some amino acids appear to be synthesized de novo or by protein/ amino acid turnover, while others are obtained largely from the plant (Barnett et al., 2004; Becker et al., 2004; Djordjevic et al., 2003, 2004; Karunakaran et al., 2009; Mulley et al., 2011; Randhawa & Hassani, 2002; Vercruysse et al., 2011). From studying rhizobial amino acid metabolic mutants we know that the ability to metabolize (synthesize, transport or degrade) specific amino acids is required during specific

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

412

M. F. Dunn

stages of the interaction (Yadav, 2007). As shown in this review, these requirements differ markedly in different rhizobia-legume combinations. Rhizobial amino acid metabolism in symbiosis can be placed in four non-mutually-exclusive categories: (1) Symbioses where amino acid exchange occurs between the microsymbiont and the plant: Such exchange appears to be uncommon but is not completely discounted in playing a non-essential auxiliary role in nitrogen fixation, as discussed for the malate-aspartate shuttle and amino acid cycling hypotheses (Sections ‘‘Aspartate biosynthesis’’ and ‘‘Glutamate catabolism’’). (2) Microsymbiont uptake of plant-produced amino acids or derivatives: During the differentiation of rhizobia into bacteroids, there is a large requirement for plant-derived amino acids for protein synthesis (Mulley et al., 2011). The symbiotic auxotrophy hypothesis and microsymbiont uptake of -aminolevulinic acid are discussed in Sections ‘‘Leucine, isoleucine and valine biosynthesis and symbiotic auxotrophy’’ and ‘‘Glycine catabolism’’. (3) Export of amino acids or amino acid precursors to the plant: This category includes fixed nitrogen exported as ammonium and the alanine export hypothesis (Section ‘‘Alanine biosynthesis’’). (4) De novo synthesis, or degradation of amino acids or their derivatives by the microsymbiont: Rhizobial amino acid metabolism in this category has been elucidated largely by work showing that specific amino acid biosynthetic or catabolic mutants are affected in symbiosis. These studies are covered in Sections ‘‘Acidic amino acids: aspartate and glutamate’’ to ‘‘Neutral-nonpolar amino acids: alanine, leucine, isoleucine, valine, phenylalanine, proline, tryptophan, glycine and methionine’’. Some omissions in the amino acids discussed occur because either no relevant information was found on their metabolism in the microsymbionts (threonine), or information was found only for their biosynthesis but not their catabolism (lysine, cysteine and methionine). The synthesis of metabolites like polyamines and auxin from amino acid precursors is also discussed. The complete genome sequences available for rhizobia are an essential resource for exploring their metabolism using global methodologies (‘‘-omics’’) and in designing experiments aimed at affecting the function of a specific metabolic pathway. The major internet resources used for computational metabolic analyses in this review were Rhizobase (genome.kazusa.or.jp/rhizobase/), which includes the annotated genome sequences of rhizobia and other diazotrophs, and the KEGG database (www.genome.jp/kegg/pathway.html) containing metabolic pathway predictions based on the annotated genome sequences. In this review, the amino acids discussed in Sections ‘‘Acidic amino acids: aspartate and glutamate’’ to ‘‘Neutralnonpolar amino acids: alanine, leucine, isoleucine, valine, phenylalanine, proline, tryptophan, glycine and methionine’’ are grouped based on chemical similarities, which gives some group members similar biological properties (Watson et al., 2008). A general scheme of the relation between central metabolism and amino acid biosynthesis is shown in Figure 1.

Crit Rev Microbiol, 2014; 41(4): 411–451

Figure 1. Schematic representation of L-amino acid biosynthesis from precursors generated by central carbon metabolic pathways. Amino acids are color-coded in the following groups: acidic, red (Asp, Glu); basic, blue (Arg, His, Lys); neutral-polar, green (Asn, Cys, Gln, Ser, Thr, Tyr); neutral-nonpolar, purple (Ala, Gly, Ile, Leu, Met, Phe, Pro, Trp, Val). In almost all cases arrows represent multiple enzymatic reactions and intermediates (Kim & Gadd 2008; Voet & Voet 1995). Abbreviations: PRPP, 5-phospho-D-ribose a-1-pyrophosphate; TCA, tricarboxylic acid.

Establishment of the nitrogen-fixing symbiosis The formation of nitrogen-fixing root nodules on legumes can be conceptualized as occurring in three stages that all involve significant metabolic communication between the symbionts (Oldroyd et al., 2011). In the preinfection stage, rhizobia in the host plant rhizosphere multiply using mostly carbon and nitrogen sources produced by the root (Berg & Smalla, 2009; Prell & Poole, 2006). In response to rootexcreted flavanoids, rhizobial genes encoding enzymes for the synthesis of nodulation (Nod) factors are induced. Nod factors are chitin lipooligosaccharides that initiate the process of nodule organogenesis by affecting phytohormone balances and signalling processes in the root. The nodulation stage involves the formation of infection threads through which rhizobia enter the cytoplasm of host cells. This stage requires the multiplication of the microsymbiont within the infection threads using, at least in part, energy sources provided by the plant (Oldroyd et al., 2011; Prell & Poole, 2006). As discussed later, rhizobia within the infection threads are subjected to oxidative and other stresses and have developed mechanisms to resist these, some of which involve products generated by amino acid metabolism. The final stage of nodule development is the morphologic and metabolic transformation of rhizobia into their bacteroid forms, which are enveloped in a host-derived membrane that controls metabolite exchange between the symbionts via specific transport systems. The bacteroid membrane also governs metabolite exchange with its own transport and export systems. The structural unit containing the bacteroids and the surrounding plant-derived membrane is termed the symbiosome (Day et al., 2001; Ferraioli et al., 2002; Tate´ et al., 1999;). Nodule morphogenesis in legumes is divided into two types (Table 1). Indeterminate nodules have a persistent meristem with new nodule cells at the tip being continually

Microsymbiont amino acid metabolism

DOI: 10.3109/1040841X.2013.856854

413

Table 1. Rhizobia-legume combinations often used in studies of symbiotic physiology and the type of nodule formed. Host plant(s) Glycine max (soybean), Vigna unguiculata (cowpea) Leucaena leucocephala Lotus japonicum Cicer arietinum (chickpea) Medicago sativa (alfalfa), Medicago truncatula (barrel medic), Melilotus albus (bokhara clover) Phaseolus vulgaris (common bean)

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Vicia sativa (common vetch), Pisum sativum (pea) Trifolium alexandrinum (berseem clover) Sesbania rostrata

Microsymbiont(s)

Type of nodule formed

Bradyrhizobium japonicum, Sinorhizobium fredii Mesorhizobium loti Mesorhizobium loti Mesorhizobium ciceri Sinorhizobium meliloti

Determinate Indeterminate Determinate Determinate Indeterminate

Rhizobium etli, Rhizobium leguminosarum bv. phaseoli, Rhizobium tropici Rhizobium leguminosarum bv. viciae Rhizobium leguminosarum bv. trifolii Azorhizobium caulinodans

Determinate

infected by the microsymbiont, forming an elongated structure containing a young meristem at the tip and older, senescent tissue near the root. Symbiosomes in indeterminate nodules often contain single, morphologically swollen bacteroids whose formation probably requires the degradation of the reserve polymer poly-b-hydroxybutyrate (PHB) accumulated by the bacteria in the infection threads (Gage, 2004; Terpolilli et al., 2012). The swollen appearance of these bacteroids results from their endoreplication (repeated genome replication without cell division), a process that is triggered in response to plant-produced nodule-specific cysteine-rich peptides (NCRs). NCRs are produced by a variety of indeterminate nodule-forming hosts (e.g. pea, alfalfa; Table 1) and are necessary for the development of an effective symbiosis in these species. Indeterminate nodules generally form on temperate legumes, where ammonia derived from nitrogen fixation is incorporated into glutamine and asparagine by the host plant (Haag et al., 2013; Terpolilli et al., 2012). Determinate nodules usually form on tropical legumes, do not contain a persistant meristem, have no developmental gradient and are generally round in shape. Symbiosomes in determinate nodules typically contain several bacteroids that continue to accumulate PHB. Determinate nodule-forming legumes do not produce NCRs and the bacteroids do not endoreplicate. Fixed nitrogen in these nodules is usually incorporated into ureides (Gage, 2004; Prell & Poole, 2006; Terpolilli et al., 2012). Symbiotic phenotypes between rhizobia and their hosts are commonly described by the ability or inability of the microsymbiont to induce nodules (Nodþ and Nod, respectively). The ability or inability of bacteroids to fix nitrogen is phenotypically described as Fixþ and Fix, respectively, and this is often determined by assaying acetylene reduction activity (ARA), a generally accepted indirect measure of nitrogenase activity (Vessey, 1994). Some rhizobial mutants form Fix pseudonodules in which the plant cells are not infected by the microsymbiont. Rhizobial amino acid transporters Legume seed and root exudates contain some quantitity of virtually all of the protein amino acids. However, significant qualitative and quantitative differences in amino acids occur in exudates from different legumes or even different cultivars

Indeterminate Indeterminate Determinate and intermediate depending on conditions

of the same species (Barbour et al., 1991; Boulter et al., 1966; Kuo et al., 1982; Richter et al., 1968; Yaryura et al., 2008). Rhizobial genomes encode many transport systems for sugars, amino acids and other compounds likely to be present in the soil, rhizosphere or host root tissue (Hosie et al., 2001; Ramachandran et al., 2011; Urdvardi & Poole, 2013). Along with transport systems specific for the import of one or a few amino acids, many rhizobia encode two broad-specificity ABC transporters for amino acids: the general amino acid permease (Aap, composed of AapJQMP) and the branchedchain amino acid permease (Bra, composed of BraDEFGC). AapP, BraF and BraG are ATP-binding proteins and AapM, AapQ, BraD and BraE are integral membrane proteins. The solute binding proteins AapJ and BraC bind amino acids in the periplasm and transfer them to the membrane bound AapQMP and BraDEFG complexes, respectively, which catalyze the ATP-dependent transport of the amino acids (Djordjevic, 2004; Hosie et al., 2002; Oldroyd et al., 2011; Tateˆ et al., 2004; Walshaw & Poole, 1996). The uptake of some amino acids is dependent entirely on the Aap and/or Bra systems. For example, Rhizobium leguminosarum bv. viciae 3841 mutants lacking Aap or both uptake systems are unable to grow on glutamate as sole carbon and nitrogen source, while mutants lacking only Bra have significantly reduced growth on glutamate. The Aap permease is also a major route of glutamate excretion in strain 3841 (Hosie et al., 2002; Walshaw & Poole, 1996; Walshaw et al., 1997). In Sinorhizobium meliloti 1021, Aap and Bra (the components of the latter system are designated LivHMGFK in S. meliloti) transport system components and those of several other amino acid transporters were found in bacteroid proteomes but were absent from cells grown in culture (Djordjevic, 2004). In B. japonicum CPAC 15, AapJ was significantly more highly expressed in the proteome of genistein-induced cells in comparison to uninduced controls (Stefaˆnia da Silva Batista & Hungria, 2012). These results contrast the observed transcriptional downregulation of genes for the Aap transporter components and LivH in bacteroids as compared to cells grown in rich medium (Barnett et al., 2004). Key functions of the Aap and Bra broad-range transport systems are discussed in Sections ‘‘Glutamate catabolism’’ and ‘‘Leucine, isoleucine and valine biosynthesis and symbiotic auxotrophy’’, while more specific transporters are covered in the sections dealing with the amino acids that they transport.

414

M. F. Dunn

Global regulation of rhizobial amino acid metabolism Systems for regulating nitrogen fixation genes in free-living diazotrophs respond largely to nitrogen status while those of symbiotic rhizobia generally perceive changes in oxygen concentration. A brief survey of key regulatory systems of nitrogen metabolism that are mentioned in later sections is presented here.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

The FixLJ system Proteins encoded by rhizobial nif and fix genes include NifHDK, which form the nitrogenase complex, FixNOPQ, comprising the high O2-affinity cytochrome oxidase, and a variety of regulatory proteins that control the transcription of these and other genes for nitrogen fixation. At the low O2 concentrations required for nitrogen fixation, nif and fix genes are activated by NifA (a member of the enhancer-binding protein family of transcriptional regulators) in combination with 54 (RpoN), an alternative sigma factor. The activation of nifA transcription in response to oxygen depends on a regulatory cascade that varies in key components in different rhizobia. In S. meliloti, for example, the membrane-bound FixL protein contains an oxygen-sensing heme moiety that when not bound to oxygen triggers the autophosphorylation of its C-terminal transmitter domain. The phosphoryl group is then transferred to the receiver domain of the cytoplasmic FixJ protein, and phosphorylated FixJ activates transcription of nifA and fixK. NifA and FixK (a Crp-Fnr family regulator) then induce the transcription of other nif and fix genes (Dixon & Kahn, 2004). In addition to genes directly involved in nitrogen fixation, the S. meliloti FixJ regulon includes genes for polyamine, arginine, proline and alanine metabolism (Barnett et al., 2004; Bobik et al., 2006). The general nitrogen regulatory (Ntr) system The intracellular nitrogen status in the non-symbiotic diazotroph Klebsiella pneumoniae is sensed and transduced by the general nitrogen regulatory (Ntr) system. NtrB is a sensor kinase that phosphorylates and activates the NtrC response regulator when glutamine concentrations are low. Phosphorylated NtrC transcriptionally activates the glutamine synthetase (glnA) and nifLA operons, allowing glutamine synthesis and the activation of genes for nitrogen fixation, respectively. At high glutamine concentration, the NtrB phosphatase activity dephosphorylates NtrC, repressing glnA and nifLA transcription. Thus, the relative kinase/phosphatase activities of NtrB determine the response to glutamine levels. NtrC phosphorylation is ultimately controlled by the PII regulatory protein encoded by glnB. With low glutamine levels, PII is uridylylated by the uridylyl transferase GlnD: uridylylated PII does not bind to the NtrB sensor kinase, allowing the free NtrB to autophosphorylate and then transphosphorylate NtrC, promoting glnA and nifLA transcription. At high glutamine concentration, the GlnD uridylylremoving activity is activated, converting uridylylated PII into its free form. Unmodified PII activates NtrB phosphatase activity, dephosphorylating NtrC and thus reducing glnA and nifLA transcription. K. pneumoniae ntrC mutants do not fix nitrogen, are glutamine auxotrophs and are unable to

Crit Rev Microbiol, 2014; 41(4): 411–451

use nitrate or amino acids as nitrogen sources (Dixon & Kahn, 2004; Fischer, 1994). ntrBC homologs occur in rhizobia, where their products play a less significant role in nitrogen metabolism than in K. pneumoniae. ntrC mutants of S. meliloti and Bradyrhizobium japonicum are unaffected in nitrogen fixation since nifA transcription in these species is not NtrC dependent, although the S. meliloti NtrC can activate nifA under freeliving, nitrogen limited conditions. However, activation of the Azorhizobium caulinodans nifA gene by NtrC is significant in symbiosis, and ntrC mutants of this species are impaired in nitrogen fixation. The growth of ntrC mutants of S. meliloti and B. japonicum on amino acids is unaffected, while that of A. caulinodans is impaired (Fischer, 1994). Transcriptome studies in S. meliloti show that mutation of ntrC or glnB (the latter encoding the PII protein) significantly affects the expression of other genes related to nitrogen metabolism, including components of the Aap and Liv amino acid transporters (Davalos et al., 2004; Yurgel et al., 2013). In rhizobia, the Ntr system also integrates carbon and nitrogen metabolism by regulating the assimilation of ammonia by glutamine synthetase isoenzymes (GSI, GSII and GSIII; EC 6.3.1.2; Section ‘‘Glutamine biosynthesis’’) in response to the relative sizes of the glutamine and 2-oxoglutarate (2-OG) pools (Espı´n et al., 1994; Reitzer, 2003). The gene encoding GSI (glnA) forms an operon with glnB (glnBA), where glnB encodes the PII regulatory protein that plays a key role in the adenylylation/deadenylylation cascade regulating GSI activity (Arconde´guy et al., 1997; Espı´n et al., 1994; Schlu¨ter et al., 2000). In S. meliloti and Rhiobium etli, GSI is deadenylylated (activated) under conditions of ammonium limitation (Arconde´guy et al., 1996, 1997; Bravo & Mora, 1988). Although post-translational adenylylation control is the principal factor regulating GSI activity in response to nitrogen, glnA transcription is controlled either by initiation at its own promoter or from one or more Ntr-dependent promoters upstream of glnB, resulting in significant changes in glnA expression in response to nitrogen status (Arconde´guy et al., 1996; Carlson et al., 1985, 1987; Chiurazzi & Iaccarino, 1990; Davalos et al., 2004; de Bruijn et al., 1989; Martin et al., 1989). Two PII proteins, encoded by glnB and glnK, are present in rhizobia and their products in combination with the GlnD sensor protein have discrete functions in the Ntr regulatory circuit (Yurgel et al., 2010, 2012). A. caulinodans ORS571, which encodes only GSI (GlnA), can use molecular nitrogen for growth ex planta and single glnB or glnK mutants retain this ability, while a glnB glnK double mutant cannot grow on N2. In ORS571, both PII proteins are involved in GSI deadenylylation, as well as the repression of nitrogen fixation gene expression in the presence of ammonia. glnBA mutants lacking GSI activity are able to fix N2 in culture but excrete rather than assimilate ammonium (Michel-Reydellet & Kaminski, 1999). GSII (glnII) gene transcription is highly regulated in response to nitrogen availability but the enzyme is not regulated biochemically (Davalos et al., 2004; de Bruijn et al., 1989; Espı´n et al., 1994). In cultures, glnII is transcriptionally down-regulated by the Ntr system in response to ammonium excess (Davalos et al., 2004; de Bruijn et al., 1989;

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

Martin et al., 1989; Patriarca et al., 1992; Schlu¨ter et al., 2000). In B. japonicum, transcripts for both glnA and glnII are present in bacteroids formed either by the wild type or an ntrC mutant: this NtrC-independent expression of glnII results from its control by an overriding, O2-dependent system that probably involves NifA (Martin et al., 1988). The presence of glnII mRNA, but not GSII protein or activity in B. japonicum bacteroids is likely due to the action of the glutamine synthetase translational inhibitor protein (GstI), which is present in the B. japonicum bacteroid proteome but has been functionally characterized only in R. legumiosarum (Tate´ et al., 2001; Sarma & Emerich, 2006). The third rhizobial GS isoenzyme, GSIII (product of glnT), is encoded in several rhizobial genomes and has been studied in R. etli and S. meliloti (Section ‘‘Glutamine biosynthesis’’). Expression of the S. meliloti 1021 glnT is controlled by the Ntr system in a manner similar to that of glnII (de Bruijn et al., 1989; Espı´n et al., 1990, 1994). The stringent response Guanosine tetraphosphate (ppGpp) and guanosine pentaphosphate (pppGpp) (collectively referred to as (p)ppGpp) are the key components of the global regulatory system called the stringent response. When cells are in a non-growing state caused by entrance into stationary phase, amino acid starvation, or various stress conditions, the (p)ppGpp alarmone downregulates genes required for growth and activates those needed for survival. In rhizobia, (p)ppGpp levels are controlled by Rsh, a bifunctional (p)ppGpp synthetase/ hydrolase (EC 3.1.7.2; Braeken et al., 2006). Rhizobial rsh mutants unable to produce (p)ppGpp are severely affected in symbiosis. R. etli rsh mutants form nodules on bean, but fix nitrogen at 25% the level of the wild type. This phenotype might result from the mutant’s increased sensitivity to salt, heat and oxidative stress (Braeken et al., 2008a; Moris et al., 2005). S. meliloti rsh mutants overproduce the exopolysaccharide succinoglycan and are unable to nodulate alfalfa (Wells & Long, 2002). The expression of many genes involved in regulatory or metabolic processes were affected in stationary phase cells of a R. etli rsh mutant. Nitrogen metabolism genes with increased expression in the mutant included the polyamine periplasmic substrate binding protein encoded by potF, genes for components of the Aap and Bra amino acid transporters, and genes for cysteine and tryptophan biosynthesis (Vercruysse et al., 2011a). Other regulatory systems In Escherichia coli, Lrp (leucine-responsive regulatory protein), ArgR (arginine biosynthesis repressor protein) and TrpR (tryptophan biosynthesis repressor protein) transcriptionally regulate virtually all amino acid biosynthetic pathways, in addition to other metabolic processes. Homologs of argR and trpR are absent from the genomes of symbiotic rhizobia, but multiple lrp homologs are present. E. coli contains a master lrp gene whose product functions in global regulation. Up- or downregulation of a target gene by Lrp can be increased, decreased or unchanged in the presence of leucine, which acts as a molecular sensor of intracellular amino acid availability. E. coli and many other bacteria have

Microsymbiont amino acid metabolism

415

from one to many lrp homologs whose products have specific rather than global regulatory funcions. These Lrps often respond to amino acids other than leucine to regulate the expression of neighboring genes involved in amino acid metabolism (Brinkman et al., 2003; Cho et al., 2012). Rhizobia encode a large number of lrp homologs, for example 15 and 20 copies in S. meliloti and Mesorhizobium loti, respectively (Brinkman et al., 2003). The function of only three rhizobial Lrp paralogs has been demonstrated. The S. meliloti Lrp designated HpdR regulates the neighboring and divergently transcribed hpdA, encoding 4-hydroxyphenylpyruvate dioxygenase of the tyrosine catabolism pathway (King & O’Brian, 1997; Section ‘‘Tyrosine catabolism’’). The B. japonicum Lrp bll7260 regulates symbiotically-essential genes for -aminolevulinic acid uptake (Loprasert et al., 2007; Section ‘‘Glycine catabolism’’). In R. leguminosarum bv. viciae the transcription of aldA, encoding alanine dehydrogenase, is mediated by the product of the divergently transcribed lrp designated aldR (Lodwig et al., 2004; Section ‘‘Alanine biosynthesis’’). Many rhizobial Lrps remain to be characterized and it will be interesting to see whether any of them are global rather than specific-function regulators. Genes encoding the regulator NtrR, which does not belong to the Ntr regulatory system, occur only in the genomes of M. loti and S. meliloti. The status of NtrR as a global regulator comes from a transcriptomic analysis of a S. meliloti 1021 ntrR mutant that showed it directly or indirectly regulates a wide range of metabolic and other functions. Genes involved in nitrogen metabolism that were upregulated in the ntrR mutant include one for polyamine transport and several involved in amino acid biosynthesis or degradation (Puska´s et al., 2004). Riboregulation is a general term for gene regulation by RNA molecules, which act to change the elongation rate of mRNA transcripts, mRNA translation efficiency and/or mRNA stability. In bacteria, Hfq is a global regulatory RNA chaperone that interacts with regulatory RNAs to affect post-transcriptional regulation of gene expression (Sobrero & Valverde, 2012). In A. caulinodans, R. leguminosarum bv. viciae and S. meliloti, Hfq contributes to the regulation of nifA, and hfq mutants in these species are negatively affected in symbiosis. In S. meliloti, the Hfq regulon contains genes for amino acid transport (including the Aap and Liv systems) and metabolism (Barra-Bily et al., 2010, 2010a; Gao et al., 2010; Jime´nez-Zurdo et al., 2013; Torres-Quesada et al., 2010). A S. meliloti protein named YbeY was recently shown to riboregulate several of the same genes affected by Hfq (Pandey et al., 2011).

Acidic amino acids: aspartate and glutamate Aspartic acid and glutamic acid have negatively charged R-groups at physiological pH. In their charged state they are referred to as aspartate and glutamate, respectively, and both are frequently involved in transamination reactions (Voet & Voet, 1995). The biosynthetic requirement for glutamate in bacteria is very high since it is used to make glutamine, arginine, proline, peptidoglycan and polyamines. Aspartate is required for the synthesis of metabolites like purines and

M. F. Dunn

Rastogi & Watson (1991)

Alfano & Kahn (1993) Castillo et al. (2000) Karunakaran et al. (2009); Mulley et al. (2011)

Lewis et al. (1990) Ferraioli et al. (2002)

Nodþ Fix. AAT activity lowered 60%, unable to grow with Asp as C source. Nodþ Fixþ Normal growth with Asp as N source. Nodþ Fixþ with significantly higher ARA. Nodþ Fix. Formed few mature bacteroids.

NodþFixþ overall but many nodules white and Fix. Nodþ, but significantly reduced ARA.

aatA (aspartate aminotransferase). Asp biosynthesis/catabolism.

aatA (aspartate aminotransferase). Asp biosynthesis/catabolism.

gltB (glutamate synthase large subunit). Glu biosynthesis. gltD (glutamate synthase small subunit). Glu biosynthesis.

R. leguminosarum bv. viciae A34/P. sativum S. meliloti JJ1c10/M. sativa

S. meliloti 104A14/M. sativa R. etli CFN42/P. vulgaris R. leguminosarum bv. viciae 3841/P. sativum S. meliloti 102f34/M. sativa R. etli CE3/P. vulgaris

References Mutant characteristics versus wild type

þ

Two major reactant pairs for aspartate biosynthesis in rhizobia are L-glutamate þ 2-oxoglutarate (2-OG) and L-asparagine þ L-glutamate, participating in the reversible reactions catalyzed by aspartate aminotransferase (Aat; EC 2.6.1.1) and asparagine synthase (EC 6.3.5.4), respectively. A third alternative, aspartate dehydrogenase (EC 1.4.1.21), catalyzes the NAD(P)H-dependent conversion of oxaloacetate and NH3 to aspartate. A high level of aspartate dehydrogenase activity was detected in Rhizobium lupini 359a bacteroid extracts (Kretovich et al., 1981), although the gene is absent from the genomes of sequenced rhizobia (including R. lupini strain HPC(L); Agarwal & Purohit, 2013), except for a putative homolog in A. caulinodans (Lee et al., 2008), which nodulates the tropical legume Sesbania (Table 1) on both stems and roots. Aspartate has been proposed as a key amino acid linking micro- and macrosymbiont carbon and nitrogen metabolism. Suggested roles for aspartate in symbiotic metabolism include its participation in a malate-aspartate shuttle, described below, and in amino acid cycling (Section ‘‘Glutamate catabolism’’). The bacteroid malate-aspartate shuttle proposed by Kahn et al. (1985) is premised on the shuttle system operating in mitochondria. In mitochondria, glutamate and malate are taken in and malate is oxidized by malate dehydrogenase (EC 1.1.1.37) to form oxaloacetate. Aat transfers an amino group from glutamate to oxaloacetate, forming 2-OG and aspartate, which are then exported out of the mitochondria. The shuttle serves to introduce reducing equivalents, but not carbon, into the mitochondria. Legume nodule cytosol and bacteroid fractions contain the requisite Aat and malate dehydrogenase activities for a malate-aspartate shuttle between the symbiosome and the plant (Apppels & Haaker, 1991; Dunn, 1998; Walshaw, 1995). While some biochemical, metabolite labelling and transport studies support the shuttle hypothesis (Appels & Haaker, 1991; Batista et al., 2009), others do not. For example, in some rhizobia-legume combinations, transport capabilities across the symbiosome and/or bacteroid membrane appear to be too low to support significant exchange of the required metabolites (Prell & Poole, 2006; Walshaw, 1995). Although by general consensus the operation of a malate-aspartate shuttle seems doubtful, its operation cannot be discarded in all rhizobia-legume combinations since (i) plant-rhizobia metabolite exchange capabilities differ significantly between different symbiotic combinations (Terpolilli et al., 2012), (ii) labelling studies can be difficult to intrepret due to utilization of metabolites by other pathways (Appels & Haaker, 1991) and (iii) transport studies with isolated bacteroids or symbiosomes are prone to errors caused by damage either to the membranes or specific transporters within them (Prell & Poole, 2006). The Aat encoded by aatA has a symbiotically essential role in S. meliloti and R. leguminosarum bv. viciae, both of which

Table 2. Rhizobial acidic amino acid metabolism mutants and their symbiotic phenotypes.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Aspartate biosynthesis



pyrimidines (Reitzer, 2003). A summary of the symbiotic phenotypes of acidic amino acid metabolism mutants is presented in Table 2.

aatB (aspartate aminotransferase). Asp biosynthesis/catabolism. gltB (glutamate synthase large subunit). Glu biosynthesis. gltB (glutamate synthase large subunit). Glu biosynthesis.

Lodwig et al. (2003) Nod Fix .

Gene mutated and/or pathway affected

Crit Rev Microbiol, 2014; 41(4): 411–451

Rhizobia parent strain/host

416

Microsymbiont amino acid metabolism

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

appear to lack a malate-aspartate shuttle based on metabolite labelling and/or transport studies (Walshaw, 1995). The enzyme is highly active in bacteroids of various rhizobia but, as discussed in the next section, its principal symbiotic role is probably in aspartate catabolism rather than synthesis. In R. lupini bacteroid extracts, aspartase (EC 4.3.1.1) activity in the direction of aspartate synthesis was significantly higher than that for its deamination (Kretovich et al., 1981), in contrast to its role in aspartate catabolism in the closely related R. etli, described in the next section. Another potential source of aspartate in rhizobia is from the degradation of the compatible solute ectoine, which affords protection against salt stress in many bacteria (Schwibbert et al., 2011). Rhizobial genomes encode genes for the tranport and catabolism, but not biosynthesis, of ectoine. The Doe (Degradation Of Ectoine) gene products act to convert ectoine to L-aspartate (Jebbar et al., 2005; Schwibbert et al., 2011; Section ‘‘Alanine biosynthesis’’). Aspartate catabolism As mentioned, aspartate is a precursor of several metabolites including symbiotically essential purines and pyrimidines (George & Robert, 1991; Kerppola & Kahn, 1988; Kim et al., 1988), and the high levels present in nodules of some rhizobia-legume combinations also make it an attractive candidate as an alternative carbon and/or nitrogen source for bacteroids (Dunn, 1998). In R. leguminosarum bv. viciae and S. meliloti, aspartate is transported with low affinity by the Dct (dicarboxylic acid transport) system in addition to the quantitatively more important Aap system (Section ‘‘Rhizobial amino acid transporters’’). In S. meliloti a highaffinity aspartate/glutamate transporter has also been identified (Prell et al., 2010; Reid et al., 1996; Watson et al., 1993). In addition to the potential for aspartate utilization by asparagine synthetase (forming asparagine; Section ‘‘Asparagine biosynthesis’’) and Aat (forming oxaloacetate), aspartate can be converted to fumarate by three routes, followed by catabolism of fumarate in the tricarboxylic acid (TCA) cycle (Dunn, 1998). In the first route, aspartate is converted to arginosuccinate and then to fumarate and arginine by the sequential action of the arginine biosynthetic enzymes arginosuccinate synthase (ArgG; EC 6.3.4.5) and arginosuccinate lyase (ArgH; EC 4.3.2.1) (Section ‘‘Arginine biosynthesis’’). Transcriptome analysis shows that the R. etli argG is upregulated in bacteroids (Resendis-Antonio et al., 2011), while it and the majority of other arg genes are downregulated in S. meliloti bacteroids (Barnett et al., 2004). The second route occurs as a part of purine biosynthesis, where PurA (adenylosuccinate synthase; EC 6.3.4.4) converts aspartate to adenylosuccinate, which is converted to fumarate by adenylosuccinate lyase (PurB; EC 4.3.2.2). Rhizobial purine auxotrophs from different genera usually form pseudonodules on their hosts (Buendı´a-Claverı´a et al., 2003; Newman et al., 1992; Okazaki et al., 2007; Swamynathan & Singh, 1992). Finally, aspartate may be deaminated to fumarate and ammonium by aspartate ammonia lyase (aspartase, AnsB): L-aspartate

, fumarate þ NH3

417

As described in Section ‘‘Asparagine catabolism’’, the major function of aspartase is in asparagine degradation. Aspartate aminotransferase (Aat) catalyzes the pyridoxal phosphate-dependent reaction: L-aspartate

þ 2-OG , oxaloacetate þ L-glutamate

Rhizobia often contain two AAT-encoding genes, aatA and aatB. In rhizobia grown in culture, AatA functions in aspartate catabolism since S. meliloti aatA mutants do not grow with aspartate as a carbon source (Rastogi & Watson, 1991). A role for Aat in bacteroid aspartate catabolism is suggested by the fact that a aatA mutant of S. meliloti JJ1c10 formed nodules that were essentially Fix on alfalfa. The residual Aat activity in the mutant is attributable to an aromatic aminotransferase able to use aspartate as substrate (Rastogi & Watson, 1991) and to AatB (Alfano & Kahn, 1993). It is important to note that the aatA mutant was not deficient in aspartate transport, and that overexpression of the aromatic aminotransferase in the aatA mutant allowed partial restoration of growth on aspartate and nitrogen fixation (Rastogi & Watson, 1991). Consistent with a symbiotic role for AatA, proteome analysis showed that AatA was found only in S. meliloti 1021 bacteroids while AatB was present only in cells grown in culture (Djordjevic, 2004). Transcriptomic analysis of strain 1021 showed that both aatA and aatB were downregulated in bacteroids as compared to cells grown in tryptone/yeast extract medium (Barnett et al., 2004). This relative downregulation in bacteroids is perhaps not surprizing given the amino acid-rich nature of the medium used to grow the control cells used for transcriptomic comparison. Proteome analysis of B. japonicum CPAC 15 cells grown in culture showed significant induction of AatA by the flavanoid genistein (Stefaˆnia da Silva Batista & Hungria, 2012). On balance, these results support the possible use of aspartate as a carbon source during nodulation and by bacteroids (Dunn, 1998; Rastogi & Watson, 1991). Glutamate biosynthesis Under free-living conditions, rhizobia assimilate ammonium mainly by the glutamine synthetase (GS)/glutamate synthase (GOGAT) pathway. Following ammonium assimilation into glutamine by GS (EC 6.3.1.2; Section ‘‘Glutamine biosynthesis’’), GOGAT produces two molecules of glutamate by catalyzing the reductive transfer of the amide nitrogen of glutamine to 2-OG in a NADH (EC 1.4.1.14) or NADPHdependent (EC 1.4.1.13) reaction. The symbiotic phenotypes of GOGAT mutants of different rhizobia vary, even for mutants constructed in the same species. In general, inactivation of GOGAT results in a defective symbiosis. In R. leguminosarum bv. viciae 3841, inactivation of GOGAT (gltB) results in a Fix phenotype on pea because the mutant strain, aside from presumably lacking GOGAT activity, was unable to transport amino acids via the Aap and Bra systems. Overexpression of plasmid-borne copies of the Aap transport system or gltBD in the mutant restored amino acid transport and a normal symbiotic phenotype. Inhibition of amino acid transport in the gltB mutant occurred at the post-translational level. Interestingly, inactivation of the gene encoding the RNA chaperone Hfq

418

M. F. Dunn

(Section ‘‘Other regulatory systems’’) increased mRNA levels of Aap and Bra system components in the gltB mutant, restored its transport capabilities and allowed it to form a normal symbiosis. This shows that in this gltB mutant amino acid transport, rather than GOGAT activity, was required for an effective symbiosis (Mulley et al., 2011). Hfq is downregulated in R. leguminosaurm bv. viciae bacteroids (Karunakaran et al., 2009) and this could be important in promoting expression of the Aap and Bra systems. A second potential route of glutamate biosynthesis is catalyzed by glutamate dehydrogenase (Gdh; EC 1.4.1.3): þ H2 O þ NADðPÞþ , 2-OG þ NH3 þ NADðPÞH þ Hþ

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

L-glutamate

Consistent with the relatively rare occurence of gdh homologs in rhizobial genomes – only S. meliloti 1021, Sinorhizobium sp. NGR234 and Bradyrhizobium spp. strains BTAi1 and ORS278 have annotated gdhs – biosynthetic (NADPHdependent; EC 1.4.1.4) Gdh activity is undetectable in most rhizobia and, when present, probably has only a minor role in glutamate biosynthesis owing to its very low affinity for ammonium (Ali et al., 1981; Bravo & Mora, 1988; Brown & Dilworth, 1975; Donald & Ludwig, 1984; Kondorosi et al., 1977; Kouchi et al., 1991; O’Gara et al., 1984; Rossi et al., 1989). Thus, GS/GOGAT is regarded as the main pathway for ammonium assimilation in free-living rhizobia (Bravo & Mora, 1988; Kondorosi et al., 1977). Under acid stress in culture, peanut-nodulating Bradyrhizobium sp. SEMIA 6144 shows a concomitant increase in GOGAT specific activity and intracellular glutamate content, probably as a general consequence of acid stress and not as part of a resistance mechanism (Natera et al., 2006; Roccillo et al., 2000). Rhizobia also accumulate glutamate in response to salt stress, where it may function in the cytoplasm to neutralize the elevated Kþ levels that occur as an initial osmotic shock response in some bacteria (Wood et al., 2001). GS/GOGAT activity and intracellular Kþ concentration increase in salt-stressed Rhizobium sp. UMKL 20, indicating a role for increased glutamate synthesis in osmotic stress protection (Yap & Lim, 1983). Glutamate synthesis from 2-OG by Aat (Section ‘‘Aspartate biosynthesis’’) using aspartate as amino group donor may serve as an overflow pathway to dispose of excess carbon and reducing power in free-living R. leguminosarum bv. viciae 3841. This hypothesis stems from the observation that strain 3841 mutants in which succinyl-CoA synthetase (EC 6.2.1.4) or 2-OG dehydrogenase (EC 1.2.4.2) were not active did not accumulate intracellular 2-OG but accumulated and excreted (via the Aap permease) large quantities of glutamate. Thus, 2-OG produced in the TCA cycle-blocked mutants is converted to glutamate (Walshaw et al., 1997). In bacteroids from various symbiotic combinations, TCA cycle activity has been postulated to be redox inhibited at the level of 2-OG dehydrogenase because microaerobic conditions limit the oxidization of reduced nucleotides in the respiratory chain (Dunn, 1998). Glutamate accumulates to high levels in pea nodule bacteroids of strain 3841 and is

Crit Rev Microbiol, 2014; 41(4): 411–451

probably synthesized from 2-OG by Aat (Prell et al., 2009a). Glutamate synthesis and export via the Aap transporter, coupled with the synthesis of PHB, might serve to consume excess carbon and reductant and help relieve bacteroid TCA cycle inhibition in planta (Walshaw et al., 1997). Glutamate catabolism As mentioned, the Aat encoded by aatA is symbiotically essential in S. meliloti and R. leguminosarum bv. viciae. Aat catalyzes the interconversion of aspartate and 2-OG to glutamate and oxaloacetate and is highly active in bacteroids from various rhizobia, where its principal symbiotic role was proposed to be in aspartate degradation (see Section ‘‘Aspartate catabolism’’). However, the synthesis of aspartate from glutamate by Aat was a key reaction in the amino acid cycling model proposed by Poole and co-workers (Lodwig et al., 2003). This model was superseded by the symbiotic auxotrophy hypothesis (Section ‘‘Leucine, isoleucine and valine biosynthesis and symbiotic auxotrophy’’) and is described here for its historic significance (to rhizobiologists) and because it is still sometimes accepted as a fact in the literature. For example, the predictions of an in silico metabolic model of the S. meliloti-alfalfa symbiosis must be carefully evaluated since it incorporates amino acid cycling as a major component of nodule metabolism (Zhao et al., 2012). The original amino acid cycling hypothesis was derived from experimental results obtained with pea plants inoculated with R. leguminosarum bv. viciae wild type strain A34 or a aap bra double mutant. Plant phenotypes, nitrogen fixing activity (ARA or 15N labeling) and xylem amide content were analyzed. Inactivation of the Aap and Bra amino acid transporters caused a significant decrease in host xylem glutamine and asparagine content despite the fact that the mutant fixed nitrogen like wild type on a per bacteroid basis (Lodwig et al., 2003). It was hypothesized that glutamate imported from the nodule cytosol by the Aap/Bra transporters was converted by Aat to aspartate (Figure 2A). Aspartate exported from the nodule was presumed to be converted to asparagine by asparagine synthetase and, along with the glutamine formed in this reaction, exported to the host xylem. In addition, alanine could serve as an export product and be interconverted with glutamate and aspartate in the plant (Lodwig et al., 2003; Prell & Poole, 2006). A related model involves the import of g-aminobutyrate (GABA), instead of glutamate, from the nodule cytosol into the bacteroid using the Bra transporter (Figure 2B; Lodwig et al., 2003). This model was developed from 15N2 labeling studies with intact pea nodules formed by R. leguminosarum bv. viciae 3841. The GABA found in both the plant cytosol and bacteroid fractions was highly enriched in 15N2 and probably originated from the decarboxylation of glutamate in the plant cytosol by glutamate decarboxylase (Gdc; EC 4.1.1.15). Assays of enzymatic activities in the bacteroids were consistent with the hypothesis that the imported GABA was transaminated to form succinate semialdehyde, glutamate, and alanine (Figure 2B). Succinate semialdehyde would be oxidized by the TCA cycle following its conversion to succinate by succinate semialdehyde dehydrogenase(s) (EC 1.2.1.16), providing an additional energy input for

DOI: 10.3109/1040841X.2013.856854

Microsymbiont amino acid metabolism

419

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Figure 2. Bacteroid amino acid cycling models based on glutamate (Panel A) and g-aminobutyrate (GABA) (Panel B). As described in Section ‘‘Leucine, isoleucine and valine biosynthesis and symbiotic auxotrophy’’, these models are superceeded by the symbiotic auxotrophy hypothesis. Abbreviations: Ald, alanine dehydrogenase; AatA, aspartate aminotransferase A; Aap/ Bra; general and branched-chain amino acid transport systems; DctA, dicarboxylic acid transporter; Opa, omega amino acid pyruvate aminotransferase; GabT, GABA oxoglutarate aminotransferase; SSA, succinate semialdehyde; GabD, SSA dehydrogenase; GS/ GOGAT, glutamine synthetase/glutamate synthase cycle; GDC, glutamate decarboxylase; Bra, branched-chain amino acid transporter.

nitrogen fixation. The alanine formed from the transamination of GABA by GABA aminotransferase (EC 2.6.1.19) could participate in amino acid cycling, although it is clear from the labeling patterns that this amino acid was not a significant export product of fixed nitrogen (Lodwig et al., 2003; Prell & Poole, 2006). Several additional results argued against a major participation of GABA cycling in bacteroid metabolism. First, single and multiple mutants in which the GABA transaminases or succinate semialdehyde dehydrogenases were inactivated did not result in symbiotic defects. This finding precludes an essential role for GABA catabolism in pea nodules, but does not rule out the possible importance of GABA metabolism under some conditions (Prell et al., 2009a). Second, additional characterization of the Bra transport system has shown that only the transport of isoleucine, leucine and valine (but not GABA) are important for the development of an effective symbiosis

(Section ‘‘Leucine, isoleucine and valine biosynthesis and symbiotic auxotrophy’’). Enteric bacteria decarboxylate glutamate to form GABA using Gdc, a reaction that consumes protons and functions as part of their acid stress response (Bearson et al., 2009). The presence of Gdc seems to be rare in rhizobia: a low level of activity was detected in S. meliloti 102F34 cells grown on glutamate (Fitzmaurice & O’Gara, 1993), while B. japonicum strains grown in culture or isolated as bacteroids lacked the activity (Salminen & Streeter, 1990; Serraj et al., 1998). The genomes of sequenced rhizobia do not contain evident genes for this enzyme, probably precluding a role for glutamate decarboxylation in acid stress resistance. Glutamate is a precursor, along with cysteine and glycine, of the tripeptide glutathione (L-g-glutamyl-L-cystenylglycine [GSH]), a powerful antioxidant with an essential role in protecting cells from a variety of environmental stresses.

420

M. F. Dunn

Crit Rev Microbiol, 2014; 41(4): 411–451

Rhizobia infecting host plants experience oxidative stress both during infection and as nitrogen-fixing bacteroids, and oxidative stress is implicated in nodule senescence (Pauly et al., 2006; Mandon et al., 2009; Ramachandran et al., 2011). GSH biosynthesis involves two reactions catalyzed by g-glutamylcysteine synthetase (GshA) and GSH synthetase (GshB): L-glutamate

þ L-cysteine þ ATP

! -L-glutamyl-L-cysteine þ ADP þ PiðGshA; EC 6:3:2:2Þ

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

-L-glutamyl-L-cysteine þ L-glycine þ ATP ! GSH þ ADP þ PiðGshB; EC 6:3:2:3Þ In vitro, rhizobia require GSH for normal growth under osmotic, acid or oxidative stress conditions, and in some strains under non-stress conditions. Sinorhizobium meliloti 1021 gshA mutants accumulated neither g-L-glutamyl-Lcysteine or GSH, had severe growth defects even under non-stress conditions in vitro, and were Nod on alfalfa. Mutants in gshB were able to synthesize g-L-glutamyl-Lcysteine but not GSH and had less severe growth defects in cultures because g-L-glutamyl-L-cysteine can partially compensate for a lack of GSH. The gshB mutant had a delayed nodulation phenotype on alfalfa, and ultrastructural studies showed that the nodules underwent early senescence, leading to a drastic decrease in nitrogen fixation (Harrison et al., 2005). Similar results were obtained with a gshB mutant of R. etli CE3 (Ta´te et al., 2012). In contrast to these severe symbiotic phenotypes, a Rhizobium tropici CIAT899 gshB mutant and a Bradyrhizobium sp. SEMIA 6144 gshA mutant were affected only in their nodulation competitiveness on bean and peanut, respectively (Muglia et al., 2008; Sobrevals et al., 2006). In summary, the disruption of aspartate catabolism by inactivation of aatA, but not aatB, leads to a defective symbiosis. The importance of Aat in symbiosis might be due to its participation in overflow metabolism or in using aspartate as an alternate carbon source. The results with glutamate biosynthesis mutants are mixed, with most gltB or gltD mutants having serious symbiotic defects that, in at least one case, result from the GOGAT mutation decreasing amino acid transport activity.

Basic amino acids: arginine, lysine and histidine The arginine and lysine side chains contain strongly basic groups that give the molecules a net positive charge at physiological pH, while the side chain of histidine is nonionizable but still very polar (Rawn, 1989). Arginine and its metabolites are notable as precursors of a variety of biologically important compounds. Table 3 summarizes the symbiotic phenotypes of basic amino acid metabolism mutants. Arginine biosynthesis Arginine biosynthesis starts from glutamate and involves several N-acetylated intermediates (Figure 3). The final acetylated intermediate, N-acetylornithine (NAO), is deacetylated to form ornithine. Ornithine, a precursor of several

important metabolites in bacteria (Section ‘‘Ornithine utilization pathways’’), is converted to arginine in three additional steps. Two main classes of bacterial arginine synthesis pathways are recognized. A few genera use a linear pathway where NAO is hydrolyzed to ornithine by acetylornithine deacetylase (ArgE; EC 3.5.1.16), with the loss of the acetyl group to the environment. The first step of the linear pathway is the acetylation of glutamate by acetyl-CoA to form N-acetylglutamate (NAG), catalyzed by NAG synthase (ArgA; EC 2.3.1.1). Most other bacteria use a cyclic pathway in which NAO is deacetylated by ornithine N-acetyltransferase (ArgJ; EC 2.3.1.35) with transfer of the acetyl group to glutamate to form NAG. In some cyclic pathways NAG is formed from acetyl-CoA and glutamate by a bifunctional ArgJ (EC 2.3.1.1, EC 2.3.1.35) instead of or in addition to NAG synthase. Other bacteria have monofunctional ArgJs capable only of NAO deacetylation with concomitant glutamate transacetylation, and thus require ArgA to initiate carbon flow through the pathway (reviewed in Cunin et al., 1986; Xu et al., 2007). In bacteria with the cyclic pathway, arginine allosterically inhibits the second enzyme of the pathway, NAG kinase (ArgB, a hexamer; EC 2.7.2.8), and ArgJ activity is often inhibited by ornithine (Marc et al., 2000; Sankaranarayanan et al., 2010). In bacteria with the linear pathway, ArgA is arginine inhibited and ArgB (a homodimer) is not (Ramo´n-Maiques et al., 2002). The genome sequences of rhizobia have a number of interesting features with regard to arginine biosynthesis. They lack a gene encoding ArgA but contain candidates for a truncated form of the enzyme, Arg(A), that has so far only been described in Mycobacterium tuberculosis (Errey & Blanchard, 2005). Carbon flow through the rhizobial pathway could begin with a yet unidentified Arg(A) or a bifunctional ArgJ. In S. meliloti 1021 biochemical and genetic evidence suggests that the ArgJ is monofuncional, making an Arg(A) activity obligatory (unpublished results). Most interestingly, S. meliloti and other rhizobia contain argE (Figure 3), and thus are one of the relatively few bacteria having both argE and argJ. In rhizobia containing plasmids, argE is the sole plasmid-borne arg gene, suggesting a possible ancillary role for its product (Barnett et al., 2001; Capela et al., 2001). The non-essentiality of ArgE is shown by the fact that R. etli and S. meliloti argE mutants are arginine protrotrophs (Villasen˜or et al., 2011; M. Dunn, unpublished results). In S. meliloti 1021, both the argJ and argE gene products have been purified and biochemically confirmed to be a NAO deacetylases (unpublished results). An untested hypothesis is that ArgE might function under certain conditions to increase the production of ornithine, which feedback inhibits the S. meliloti ArgJ, for polyamine biosynthesis (Section ‘‘Polyamine biosynthesis and utilization’’). Biochemically, the S. meliloti ArgB and ArgJ are similar to those characterized in other bacteria using the cyclic pathway, and the purified proteins are significantly inhibited by arginine and ornithine, respectively. Neither ArgB or ArgJ production in S. meliloti is affected by exogenous arginine or ornithine (Dunn et al., 2010). The genome sequences of strain 1021 and other rhizobia lack an argR (arginine repressor) homolog, so whether arginine synthesis is subject to genetic

argF (anabolic ornithine carbamoyltransferase). Arg biosynthesis. Transposon-generated ornithine auxotrophs. Arg biosynthesis. Transposon-generated arginine auxotrophs. Arg biosynthesis. carA or carB (carbamoylphosphate synthase), chemically induced. Arg and pyrimidine biosynthesis. Chemically induced ornithine auxotrophs. Arg biosynthesis. Presumptive argF (anabolic ornithine carbamoyltransferase), chemically induced. Arg biosynthesis. Presumptive argG (arginosuccinate synthase), chemically induced. Arg biosynthesis. arcA (arginine deiminase). Arg catabolism. arcA (arginine deiminase). Arg catabolism. arcB (catabolic ornithine carbamoyltransferase). Arg catabolism. ocd (ornithine cyclodeaminase). Arg catabolism. hisA (phosphoribosylformimino-5-aminoimidazole carboxamide ribotide isomerase). His biosynthesis.

Chemically generated His auxotroph. His biosynthesis. Probable hisB (histidinol phosphate phosphatase), chemically generated. His biosynthesis. Probable hisD (histidinol dehydrogenase), chemically generated. His biosynthesis. Tn5 induced His auxotrophs. His biosynthesis.

Tn5 induced His auxotrophs. His biosynthesis. Tn5 induced His auxotroph (histidinol phosphate phosphatase (hisB)?). His biosynthesis. Tn5-induced His auxotrophs. His biosynthesis

hutHe2 (histidase). His catabolism. hutH2 (histidase). His catabolism. lysA (diaminopimelate decarboxylase). Lys biosynthesis.

lysA (diaminopimelate decarboxylase). Lys biosynthesis.

M. loti MAF303099/L. japonicus S. meliloti Rmd201/M. sativa S. meliloti Rmd201/M. sativa S. meliloti 104A14/M. sativa

R. etli CE3/P. vulgaris R. leguminosarum bv. trifolii/ T. alexandrinum R. leguminosarum bv. trifolii/ T. alexandrinum B. japonicum USDA I-110/G. max

B. japonicum USDA 110/G. max B. japonicum USDA122/G. soja and G. max Bradyrhizobium spp. IRC256/ V. unguiculata R. etli CNPAF512/P. vulgaris S. meliloti 1021/M. sativa R. etli CE3/P. vulgaris

M. ciceri TAL620/C. arietinum

R. etli CNPAF512/P. vulgaris S. meliloti 1021/M. sativa S. meliloti 1021/M. sativa S. meliloti GR4/M. sativa R. etli CE3/P. vulgaris

S. meliloti 104A14/M. sativa

S. meliloti 104A14/M. sativa S. meliloti 104A14/M. sativa

argC (acetyl-g-glutamyl phosphate reductase). Arg biosynthesis.

Gene mutated and/or pathway affected

R. etli CE3/P. vulgaris

Rhizobia parent strain/host

Table 3. Rhizobial basic amino acid metabolism mutants and their symbiotic phenotypes.

Kummer & Kuykendall (1989)

Significantly lower ARA. Strains reisolated from nodules were His prototrophs. Nod. Nod on both hosts. Nodulation restored with exogenous His. Nodþ Fixþ but greatly reduced ARA. Nod . Nodþ Fixþ. Elicited few, tumorlike structures (Fix). Nodulation partially restored with exogenous Lys. Nodþ Fix. ARA restored with exogenous Lys.



Yadav et al. (1998)

Nodþ Fix.

Das et al. (2010)

Braeken et al. (2008) Boncompagni et al. (2000) Ferraioli et al. (2002)

McLaughlin et al. (1987)

So et al. (1987) Sadowsky et al. (1986)

Newman et al. (1995) Yadav et al. (1998).

Nod Fix , but lower ARA. Nodþ Fixþ. Nodþ Fixþ. Delayed nodulation. Nodule number, fresh weight and ARA significantly reduced. Developed nodules showed necrosis of infected and uninfected plant cells. Nodþ Fixþ. Nod.

D’Hooghe et al. (1997) Bobik et al. (2006) Bobik et al. (2006) Soto et al. (1994a) Ferraioli et al. (2002)

Kerppola & Kahn (1988)

þ

Nodþ Fixþ. þ

Kerppola & Kahn (1988) Kerppola & Kahn (1988)

Nodþ Fix. Nodþ Fixþ.

Ferraioli et al. (2001)

References

Mishima et al. (2008) Kumar et al. (2003) Kumar et al. (2003) Kerppola & Kahn (1988, 1988a)

Mutant characteristics versus wild type Nod . Nod factors not made in culture without supplemental Arg. Nodþ Fix. Deficient in bacteroid proliferation. Nodþ Fix. Deficient in bacteroid proliferation. Nodþ Fixþ. Nodþ Fix.



Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

Microsymbiont amino acid metabolism 421

422

M. F. Dunn

Crit Rev Microbiol, 2014; 41(4): 411–451

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Figure 3. Rhizobial arginine biosynthesis based on the genome annotation of S. meliloti 1021. Abbreviations: Arg(A), short-form N-acetylglutamate synthase; ArgB, N-acetylglutamate kinase; ArgC, N-acetylglutamyl phosphate reductase; ArgD, acetylornithine/succinyldiaminopimelate aminotransferase; ArgJ, mono- or bifunctional ornithine N-acetyltransferase; b-ArgJ, bifunctional N-acertylglutamate synthase/ ornithine N-acetyltransferase; ArgE, N-acetylornithine deacetylase; ArgF, anabolic ornithine carbamoyltransferase; ArgG, arginosuccinate synthase; ArgH, arginosuccinate lyase. The ArgJ catalyzed reaction(s) (in red) are specific to the cyclic pathway.

control in rhizobia is an open question. Significant variations in arg gene transcript levels have been observed between S. meliloti bacteroids and cells from culture (Barnett et al., 2004), suggesting some sort of genetic regulation. Inactivation of argC, encoding acetyl-g-glutamyl phosphate reductase (EC 1.2.1.38) (Figure 3) in R. etli CE3 and S. meliloti 1021 results in arginine auxotrophy (Dı´az et al., 2011; Ferraioli et al., 2001). On bean plants, the R. etli mutant was Nod and unable to induce host early nodulation responses. This phenotype coincides with the inability of the mutant to synthesize Nod factors in response to the nod gene-inducer naringenin. Supplementation of cultures with arginine restored growth and Nod factor synthesis by the argC mutant, but arginine addition in the plant nodulation tests (50 mg/ml, a level 2.5-fold higher than that restoring growth on minimal medium) allowed the mutant to cause a limited early nodulation response in the plant, resulting only in the formation of small, nodule-like structures (Ferraioli et al., 2001). Bean roots probably produce insufficient arginine for argC mutant invasion and multiplication. The lack of inducible Nod factor synthesis in the R. etli argC mutant is probably a general feature of rhizobial auxotrophs. This is because nod gene induction occurs only in dividing rhizobia during the pre-infection and infection stages, and not in non-dividing bacteroids or in minimal medium cultures in which auxotrophs cannot grow. Nod factor synthesis also requires fatty acids, whose production is positively correlated with growth rate (Ferrailoi et al., 2002). Thus, rhizobial amino acid auxotrophs capable of nodulation must obtain the lacking amino acid from the plant host. The anabolic ornithine carbamoyltransferase (ArgF; EC 2.1.3.3) catalyzes the carbamoyl phosphate-dependent conversion of ornithine to citrulline (Figure 3). In M. loti MAFF303099, a argF insertion mutant did not form colonies on minimal medium except when supplemented with

citrulline or arginine, but not with ornithine. The mutant complemented with the cloned argF grew normally without supplements. The mutant was Nodþ Fix on Lotus japonicus, but fixed nitrogen normally in nodulation assays with exogenous citrulline or arginine, or if the mutant was complemented with argF (Mishima et al., 2008). The Nodþ Fix phenotype of the mutant indicates that it produced Nod factors using plant derived arginine or citrulline and that insufficiency of these compounds occurred at a later stage of symbiosis. Citrulline and arginine have been detected in L. japonicus roots and nodules (Desbrosses et al., 2005). In S. meliloti 104A14, two chemically induced mutants with phenotypes indicating that argF and argG (arginosuccinate synthase; EC 6.3.4.5), respectively, had been mutated formed effective nodules on alfalfa. In contrast, a 104A14 mutant growing on acetylornithine but not on NAG was classified as an ‘‘ornithine biosynthesis’’ mutant and was presumably blocked at a step between ArgB and ArgD (acetylornithine/succinyldiaminopimelate aminotransferase; EC 2.6.1.11, EC 2.6.1.17) (Figure 3). The Fix phenotype of these mutants led to the conclusion that S. meliloti arginine mutants affected in the early part of the pathway are ineffective while those affected in later steps are effective (Kerppola & Kahn, 1988). This contention is supported by the work of Kumar et al. (2003), who isolated arginine auxotrophs of S. meliloti strain Rmd201 by Tn5 mutagenesis. On alfalfa, the mutants that were also ornithine auxotrophs formed ineffective nodules, while the arginine auxotrophs made fully effective nodules. It appears that the ornithine requirement for symbiosis extends beyond its use as an intermediate in arginine biosynthesis (Section ‘‘Ornithine utilization pathways’’). In E. coli, ornithine transaminase (DapC; EC 2.6.1.17) has N-succinyl-1,1-diaminopimelate: 2-OG aminotransferase activity and is part of the lysine pathway (Charlier & Glansdorff, 2004). The lack of dapC in the S. meliloti

DOI: 10.3109/1040841X.2013.856854

Microsymbiont amino acid metabolism

423

to glutamate by the P5C dehydrogenase activity of PutA. Proline degradation is discussed in Section ‘‘Proline catabolism’’. As discussed in detail below, arginine degradation in rhizobia is required for its use as a carbon and nitrogen source, or for its conversion to metabolites important in stress resistance and other processes.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

The arginase pathway

Figure 4. Arginine degradative pathways in selected rhizobia as inferred from their genome sequences. Reactions of the arginase (ArgI, Ocd) and Arc (ArcA, ArcB, ArcC) pathways are highlighted in red and green, respectively. For the enzymes catalyzing the reactions, the presence of one or more of the encoding genes in an organism is numerically indicated as follows: 1, A. caulinodans ORS571; 2, B. japonicum USDA110; 3, M. loti MAFF303099; 4, R. etli CFN42; 5, R. leguminosarum bv. viciae 3841; 6, S. meliloti 1021. Abbreviations: ArgI, arginase; ArgF, anabolic ornithine carbamoyltransferase; UreABC, urease; ArcA, arginine deiminase; ArcB, catabolic ornithine carbamoyltransferase; ArcC, carbamate kinase; RocD, ornithine oxo-acid transaminase; Ocd, Ornithine cyclodeaminase; PutA, proline dehydrogenase; spon., spontaneous (non-enzymatic). The unlabled arrows denote reactions in the latter part of arginine biosynthesis (Figure 3).

genome might mean that ArgD, which can catalyze the succinyldiaminopimelate aminotransferase reaction, also functions in lysine biosynthesis (Section ‘‘Lysine biosynthesis’’). It would be interesting to generate an argD mutant to test this hypothesis. Arginine catabolism Multiple pathways for arginine degradation occur in bacteria (reviewed in Lu, 2006) and those found in rhizobia are described here and in Section ‘‘Polyamine biosynthesis and utilization’’. Genome analysis of selected rhizobia indicates that they have the pathways shown in Figure 4. These predicted pathways range from those of S. meliloti, which has complete arginine deiminase (Arc) and arginase pathways, to that of A. caulinodans, containing only ornithine cyclodeaminase (Ocd) and urease. The other root-nodulating genera lack a complete Arc pathway but have arginase pathways where either Ocd and/or ornithine oxo-acid transaminase (RocD) participate in converting ornithine into proline or glutamate-g-semialdehyde (GSA), respectively. In R. etli and R. leguminosarum bv. viciae, but not M. loti and B. japonicum, RocD is encoded next to arginase (Gonza´lez et al., 2006; Kaneko et al., 2000; Watanabe et al., 2002; Young et al., 2006). In the two latter species, arginine deiminase (ArcA) is encoded next to a putative arginine/ ornithine antiporter (ArcD), indicating a functional link between these gene products in polyamine metabolism (Sections ‘‘The arginine deiminase pathway’’ and ‘‘Polyamine biosynthesis and utilization’’). In all but Azorhizobium, which lacks proline dehydrogenase (PutA), GSA formed by arginine degradation would be converted

In Agrobacterium tumefaciens, a plant pathogen closely related to the nitrogen-fixing rhizobia, arginine is degraded by the arginase pathway, ultimately producing proline and glutamate (Dessaux, et al., 1986; Schrell et al., 1989; Figure 4). Arginase (EC 3.5.3.1) catalyzes the hydrolysis of arginine to ornithine and urea. S. meliloti 1021 has two arginase genes, argI1 and argI2, encoded on the chromosome and pSyma, respectively. Neither gene is located near other genes for arginine metabolism and their deduced amino acid sequences are only 28% identical. ArgI1 has nearly 50% amino acid identity to the A. tumefaciens arginase and its transcript is more abundant in cells grown on argininecontaining minimal medium. Arginase enzyme activity in S. meliloti is increased several fold by exogenous arginine but less than two-fold by exogenous ornithine (A. Arteaga and M. Dunn, unpublished). This indicates that argI1 encodes an inducible arginase that functions in arginine degradation under aerobic conditions. The urea produced by the arginase reaction can be hydrolyzed to ammonia and CO2 by the nickel metalloenzyme urease (EC 3.5.1.5; Figure 4). Urease in bacteria has two major functions: production of ammonia (a preferred nitrogen source) and alkalinization of the acidic environments inhabited by some urease producers, thus allowing their growth (Collins & D’Orazio, 1993). In the human pathogen Helicobacter pylori, arginase and urease occur as part of a urea cycle that regulates nitrogen balance inside the cell and affords protection against acid pH (Mendz & Hazell, 1996). A urease cycle was also proposed to operate in A. tumefaciens (Vissers et al., 1986), and if it also exists in rhizobia it could help protect bacteroids against the acidic environment in planta and free-living cells in acidic soils (Choudhury et al., 2010; Perez-Galdonat & Kahn, 1994). However, in S. meliloti 1021 neither urease or arginase is upregulated in response to acid stress (Hellweg et al., 2009). Most bacterial ureases have three subunits (a, b and g, encoded by ureC, ureB and ureA, respectively) that are usually clustered with accessory proteins needed for the assembly of the holo-urease. In some taxons urease gene expression is induced by the transcriptional regulator UreR in response to exogenous urea while in others its production is controlled by NtrA (Collins & D’Orazio, 1993; Section ‘‘The general nitrogen regulatory (Ntr) system’’). In rhizobial genomes ureA, ureB and ureC are clustered near urease accessory proteins but UreR-type regulators are not present. In S. meliloti AK631 the activity of a ureC::lacZ transcriptional fusion was not affected by exogenous urea in the growth medium, but was significantly decreased in minimal medium containing ammonium, or in rich medium. This suggests that ureC is regulated by NtrC and functions not in catabolizing

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

424

M. F. Dunn

exogenous urea but in degrading arginine (the effect of exogenous arginine on ureC transcription was not tested) (Miksch & Eberhardt, 1994). Urease is also downregulated by ammonium in free-living cowpea bradyrhizobia (Sen et al., 2008), so one might expect little urease expression in the ammonium-rich environment of the nodule. Consistent with this, urease specific activity in Bradyrhizobium sp. bacteroid extracts from green gram nodules was only a fraction of that found in free-living cells (Sen et al., 2008). In S. meliloti 1021, urease subunit or accessory proteins were absent from bacteroid proteomes (Djordjevic, 2004) and their gene transcripts downregulated (Barnett et al., 2004). Urease activity in pea bacteroids formed by R. leguminosarum bv. viciae UPM791 were about 40% that found in cells grown in rich medium, and mutants devoid of urease activity induced normal, Fixþ nodules (Toffanin et al., 2002). Ornithine cyclodeaminase (Ocd; EC 4.3.1.12) converts ornithine to proline and ammonia. In the genera considered here (Figure 4), the number of possible Ocds encoded in their genomes range from none in B. japonicum to five each in M. loti and R. leguminosarum bv. viciae. S. meliloti strains GR4 and 2011 grow with ornithine as sole carbon and nitrogen source and convert this substrate to proline, consistent with Ocd activity (Soto et al., 1994b). In S. meliloti GRM8 Ocd activity was induced by ornithine, but not by proline (Jime´nez-Zurdo et al., 1995). Inactivation of an apparent ocd located on a S. meliloti GR4 cryptic plasmid lowered the nodulation efficiency of the mutant (Soto et al., 1994a). Ocd activity was still present in a cryptic plasmid-cured strain, indicating the presence of functional Ocds elsewhere in the genome (Soto et al., 1994b). The GR4 Ocd is atypical, having about 25% amino acid similarity to the annotated Ocds or EutCs (see below) in other rhizobia and somewhat higher identity to putative Ocds from some anaerobic bacteria. Perhaps not unexpectedly, the GR4 ocd did neither hybridize with S. meliloti 2011 DNA nor did expression of the cloned ocd in E. coli confer Ocd activity (Soto et al., 1994a, b). In S. meliloti 1021 the deduced product of the Ocd encoded by smb21494 has high amino acid identity to the biochemically confirmed Ocd from A. tumefaciens (Schindler et al., 1989; Soto et al., 1994a) and is probably the Ocd involved in converting ornithine to proline. By proteome analysis, both smb21494 and arginase argI1 are upregulated in a S. meliloti 2011 hfq insertion mutant grown in culture (Torres-Quesada et al., 2010; Section ‘‘Other regulatory systems’’). The significant downregulation of both hfq and smb21494 in S. meliloti 1021 bacteroids (Barnett et al., 2004; Becker et al., 2004) suggest that regulatory controls besides repression by Hfq control ocd expression in planta. The existence of other ocd homologs in S. meliloti is described in Section ‘‘Alanine biosynthesis’’. The arginine deiminase pathway This pathway comprises the enzymes arginine deiminase (ArcA; EC 3.5.3.6), catabolic ornithine transcarbamoylase (ArcB; EC 2.1.3.3) and carbamate kinase (ArcC; EC 2.7.2.2) (Figure 4). Arginine degradation by the deiminase pathway in Pseuodmonas species is induced by low oxygen and is required for growth on arginine under anaerobic conditions.

Crit Rev Microbiol, 2014; 41(4): 411–451

In rhizobia, gene fusion and transcriptomic studies show that the Arc pathway is upregulated in microaerobiosis with the participation of the FixJ and NtrR regulators (Bobik et al., 2006; D’Hooghe et al., 1997; Puska´s et al., 2004). The arginine deiminase pathway can function in acid tolerance, where the NHþ 3 generated by ArcC is combined with a proton, yielding ammonium and increasing the intracellular pH. ATP produced by ArcC can be used to pump out protons via the F0F1-ATPase and give further pH homeostasis (Chen et al., 2011). In S. meliloti 1021, gene fusion and transcriptome analyses showed that the gene encoding the periplasmic solute binding component of a putative arginine ABC transporter (SMc03124) was induced by arginine or by salt stress, but not in response to acidity (Hellweg et al., 2009; Mauchline et al., 2006; Ru¨berg et al., 2003). Soil bacteria encounter elevated osmolarities in the rhizosphere (Miller & Wood, 1996), and arginine imported from the alfalfa rhizosphere (Richter et al., 1968) by S. meliloti could be catabolized by the arginine deiminase or arginase pathways to produce proline, a compatible solute (Miller & Wood, 1996). The arcABC genes in R. etli CNPAF512 are present close to fixLJ, and arcB transcription was highly induced under low oxygen conditions but was not dependent on FixLJ (Section ‘‘The FixLJ system’’). Interestingly, the ArcA activity in bacteroid extracts was inhibited by spermidine, indicating its participation in the biosynthesis of this polyamine (Section ‘‘Polyamine biosynthesis and utilization’’). On bean, the symbiotic phenotypes of two arcA mutants (one having a very reduced ArcB activity due to arcA insertion polarity) were normal except for a slight reduction in ARA (D’Hooghe et al., 1997). In S. meliloti 1021, arcA1BC are pSyma-encoded and are induced by microaerobiosis but not in symbiosis, where Ntr system control (Section ‘‘The general nitrogen regulatory (Ntr) system’’) probably overrides microaerobic induction (Becker et al., 2004; Puska´s et al., 2004). Unlike R. etli (discussed above), the low oxygen upregulation of arcA1BC in S. meliloti is dependent on FixJ and FixK in the case of arcA1 and arcB, and on FixJ in the case of arcC. The expression of arcA1 and arcB was decreased in bacteroids formed by a fixJ mutant, consistent with the presence of FixK boxes preceeding arcA1. Mutations in either arcA1 or arcB had no effect on symbiosis with M. sativa based on plant dry weights (Bobik et al., 2006). Genes located near arcA1BC that probably encode products involved in polyamine biosynthesis and transport are discussed in Section ‘‘Polyamine biosynthesis and utilization’’. Thus, the rhizobial arcABC operons encode the enzymes of the arginine deiminase pathway, which appears to operate under low oxygen conditions. Why do some rhizobial genomes contain a second arcA, or a single arcA in the absence of arcBC? In S. meliloti 1021, a second arcA occurs next to arcD1 and arcD2, encoding possible arginine/ ornithine antiporters (Figure 5). Mesorhizobium loti and B. japonicum do not contain an arcABC operon (Figure 4), but encode single arcAs in an arrangement similar to that of the S. meliloti arcA2 (Figure 5). In rhizobia containing only a arcA, could ArgF function bidirectionally to convert the citrulline produced by the sole ArcA to ornithine, or is citrulline the end-product (Figure 4; Sections ‘‘Arginine biosynthesis’’ and ‘‘Arginine catabolism’’)?

DOI: 10.3109/1040841X.2013.856854

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

In summary, a diversity of arginine catabolic genes are present in rhizobia and sometimes differ within the same species (e.g. R. etli CE3 versus CNPAF512 arc genes). It is notable that the degradative pathways depicted in Figure 4 generate a large quantity of ammonium and, ultimately, glutamate that could be used for the biosynthesis of other amino acids (Section ‘‘Glutamate catabolism’’). Little biochemical characterization of the enzymes or dedicated regulatory studies of genes for rhizobial arginine degradation have been done, but global expression studies indicate that the arginase and Arc pathways operate under aerobic and microaerobic conditions, respectively. The Arc pathway appears to be active in bacteroids and is controlled by symbiotically relevant regulatory systems.

Figure 5. The arginine deiminase (arcA)-arginine/ornithine antiporter (arcD) gene regions in three rhizobial genera. Arrows representing the genes are labelled with their Rhizobase accession number and genetic designation, if any. Arrows are color coded as follows: arcA, black (sma1670, mll6733, bll7310); arcD, green (sma1667, sma1668, mll6735, mll6736, bll3711); genes encoding components of a multidrug efflux system, red (mll6730, mll6731, bll7312); genes encoding hypothetical proteins, light gray.

Figure 6. Abbreviated scheme for the utilization of ornithine in the biosynthesis of polyamines, ornithine lipids, and the siderophore vicibactin. A portion of the arginine biosynthetic pathway from the intermediate N-acetylornithine (NAO) is shown (see also Figure 3). Abbreviations: Polyamine biosynthesis (Section ‘‘Polyamine biosynthesis and utilization’’): Adc, arginine decarboxylase; Hss, homospermidine synthase; Odc, ornithine decarboxylase; SpeB, agmatinase. Ornithine lipid biosynthesis (Section ‘‘Ornithine lipid biosynthesis and function’’): OlsA, lyso-ornithine lipid acyltransferase; OlsB, ornithine acyltransferase; OlsC and OlsE, ornithine lipid hydroxylases. S1, S2, P1 and P2 refer to structurally different species of ornithine lipids. The pathway shown is that of R. tropici CIAT899 (Vences-Guzma´n et al., 2011) but ornithine lipids are also made by S. meliloti by a similar pathway (Gao et al., 2004). Vicibactin biosynthesis (Section ‘‘Vicibactin biosynthesis and function’’): VbsA, condensation of D-3-hydroxybutyrate with ornithine; VbsC, acetylase; VbsL, epimerase; VbsO, ornithine hydroxylase; VbsS, non-ribosomal peptide synthase. This pathway is specific for R. leguminosarum bv. viciae (Carter et al., 2002). Other abbreviations: L-Arg, L-arginine; L-Orn, L-ornithine.

Microsymbiont amino acid metabolism

425

Ornithine utilization pathways The L-ornithine generated during arginine biosynthesis or degradation is a major branch-point metabolite in bacteria (Xu et al., 2007). In the following sections, the biosynthesis and function of three classes of rhizobial metabolites made from ornithine are discussed. Polyamine biosynthesis and utilization Polyamines are ubiquitous, aliphatic hydrocarbons containing quaternary nitrogen groups. In both eucaryotes and procaryotes, polyamines influence or are essential for DNA replication, transcription, translation and growth under normal or stress conditions. The detailed mechanisms by which polyamines affect these processes are not well understood, and polyamine metabolism and its regulation are complex (Schneider & Wendisch, 2011; Shah & Swiatlo, 2008; Shaw et al., 2010). Rhizobia grown in minimal medium produce a variety of intracellular polyamines including the diamines putrescine and cadaverine and the triamines spermidine and homospermidine (Fujihara & Harada, 1989; Hamana et al., 1990; Shaw et al., 2010). Polyamines are synthesized by the decarboxylation of basic amino acids by amino acid decarboxylases that are usually specific for one or sometimes two amino acid substrates (Lee et al., 2007). The decarboxylation products of ornithine, arginine and lysine are putrescine, agmatine and cadaverine, respectively (Figure 6). A given rhizobial genome will often encode several amino acid decarboxylases, usually annotated as ornithine/arginine/ lysine decarboxylases. In S. meliloti 1021, for example, two decarboxylases (SMa0680 and SMa0682) are encoded near the arcA1BC operon (Section ‘‘The arginine deiminase pathway’’). Based on their overall sequence similarity to decarboxylases present

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

426

M. F. Dunn

in other organisms, SMa0680 may be an ornithine decarboxylase (Odc; EC 4.1.1.17) and SMa0862 a lysine decarboxylase (Ldc; EC 4.1.1.18). The transcription of sma0680 is increased over wild type levels in a ntrR mutant of 1021 grown under low oxygen conditions (Puska´s et al., 2004; Section ‘‘Other regulatory systems’’). Transcription of both sma0680 and sma0682 increased 2–4-fold in the wild type grown with low oxygen and were significantly reduced in a fixJ or fixK mutant background (Bobik et al., 2006). The membrane component of an amino acid transporter (SMa0678, which has significant amino acid sequence similarity to the S. meliloti ArcD1 and ArcD2 putative arginine/ornithine antiporters) that occurs next to SMa0680 is regulated in the same manner as the decarboxylases. A consensus FixK box occurs in front of the genes for both the SMa0682 putative Ldc and the amino acid transporter, consistent with their regulation by FixK (Bobik et al., 2006). Without biochemical characterization, it is not possible to know if the putative Ldc is specific for lysine, or if it could perhaps decarboxylate arginine taken in by the neighboring transporter. In addition to the plasmid-encoded decarboxylases, strain 1021 has a chromosomal gene annotated as a ornithine/diaminopimelate/arginine decarboxylase that by sequence similarity could encode a dual specificity Ldc/Odc (Lee et al., 2007). Genes putatively encoding enzymes that produce other polyamines from the products of amino acid decarboxylation are also encoded in rhizobial genomes. These include agmatinase (SpeB), which hydrolyzes agmatine to putrescine and urea (Figure 6). In S. meliloti 1021, for example, speB is contiguous with but independently transcribed from the arginine biosynthetic gene argC (Dı´az et al., 2011), and speB2 (smc01967) occurs immediately downstream of a putative spermidine/putrescine ABC transport system encoded by genes smc01963!smc01966, which are homologs of potC, B, A, and D, respectively. The PotABCD transporter of enteric bacteria imports both spermidine and putrescine, but has a higher affinity for and is feedback inhibited by spermidine (Shah & Swiatlo, 2008). Without a functional characterization of these gene products, we do not know if they function in the uptake of putrescine and its conversion to agmatine. According to the Microbes Online Operon Prediction for S. meliloti 1021 (http:// meta.microbesonline.org/operons/gnc266834.html), the transporter genes form an operon, but speB2 is not predicted to be part of it. Spermidine is produced by the transfer of the propylamine group from S-adenosylmethionine to putrescine by the enzyme spermidine synthase. Spermidine can be monoacetylated by spermidine acetyltransferase, a modification that makes spermidine physiologically inactive in E. coli (Limsunwun & Jones, 2000). Genes possibly encoding spermidine synthase (sometimes annotated as hypothetical proteins) are present in some rhizobial genomes while spermidine acetyltransferase is annotated only in M. loti. Homospermidine is formed by the condensation of two putrescine molecules by homospermidine synthase (Hss), and hss genes are present in many a-proteobacterial genomes (Shaw et al., 2010). The functionality of the Hss from B. japonicum N2P5549 was biochemically confirmed

Crit Rev Microbiol, 2014; 41(4): 411–451

(Shaw, 2011). A R. etli hss mutant was unable to swarm, had a slight growth defect in rich medium but fixed nitrogen like wild type (Braeken et al., 2008). A R. leguminosarum bv. viciae 3841 with a transposon insertion in the promoter region of its hss gene produced about 35% the wild type level of homospermidine in culture but was unaffected in its freeliving growth or in symbiosis with pea (Shaw, 2011). Legume root nodules contain relatively high levels of polyamines that may influence nodule formation and transport processes across the symbiosome membrane (Shah & Swiatlo, 2008; Terakado-Tonooka & Fujihara, 2008; Whitehead et al., 2001), and could be used as carbon and nitrogen sources by rhizobia. Rhizobia contain multiple genes encoding polyamine transporter components and some of these are upregulated in bacteroids (Resendis-Antonio et al., 2011; Vercruysse et al., 2011). In S. meliloti 1021 transcription of the gene encoding the ATP-binding subunit of the putative spermidine/putrescine transporter (SMc01965) is repressed by phosphate stress in a PhoB-independent manner, probably as part of a general downregulation of amino acid metabolism that occurs with phosphate starvation (Krol & Becker, 2004). Transcriptome analysis shows that components of some putative polyamine transporters in S. meliloti are upregulated during microaerobic growth and that their expression is affected by mutations in ntrR or hfq (Becker et al., 2004; Puska´s et al., 2004; Torres-Quesada et al., 2010; Section ‘‘Other regulatory systems’’). However, experimental confirmation of the participation of these regulators in polyamine transport is lacking. Little is known about the biological functions of polyamines in rhizobia. Bacteroids of B. japonicum USDA110 had homospermidine levels that were three-fold higher than those of cells grown in culture (Vauclare et al., 2013). In Sinorhizobium fredii P220, a soybean nodulating strain, large changes in intracellular homospermidine levels occurred in response to acid and osmotic shock and were hypothesized to be involved in the high intrinsic stress resistance of this strain. In contrast, little change in homospermidine content occurred in the stress-sensitive B. japonicum strain A1017 under these conditions (Fujihara & Yoneyama, 1993). Difluoromethylornithine is a specific, irreversible inhibitor of ornithine decarboxylase (Odc) (Tabor & Tabor, 1985), and the growth of R. leguminosarum bv. viciae 3841 was progressively and severely reduced with increasing concentrations of this inhibitor in the medium. Odc specific activity and the putrescine and homospermidine content of the treated cells was also significantly reduced. Their growth in the presence of the inhibitor was restored with exogenous putrescine or homospermidine, indicating a specific requirement for these polyamines (Shaw et al., 2010). Based on biochemical and genetic data from E. coli (Schneider & Reitzer, 2012), a pathway for the degradation of putrescine to GABA can be inferred from the S. meliloti 1021 genome sequence (Figure 7). On the chromosome, four genes comprising a putative spermidine/putrescine ABC-type transporter lie close to contiguous genes potentially encoding a g-glutamyl-g-aminobutyraldehyde dehydrogenase (PuuC; EC 1.2.1._), g-glutamylputrescine oxidoreductase (PuuB; EC 1.4.3._), and g-glutamylputrescine synthetase (PuuA; EC

DOI: 10.3109/1040841X.2013.856854

Microsymbiont amino acid metabolism

427

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Figure 7. Possible putrescine degradation pathway in S. meliloti involving chromosomal, clustered spermidine/putrescine transporter and puu gene products, and pSymB-encoded, contiguous gene products. Succinate would be further metabolized by the TCA cycle. Abbreviations are as listed in Figure 6.

6.3.1.11). These gene products could serve to import putrescine from the environment and catalyze its transformation to the intermediate N-g-glutamylaminobutyrate. Additionally, agmatine produced by arginine decarboxylation could be converted to putrescine by the product of the agmatinase (speB2) encoded in the same region. A homolog encoding g-glutamyl-g-aminobutyrate hydrolase (PuuD; EC 3.5.1.94) is absent from this region but SMc03943, a conserved hypothetical protein with about 40% amino acid identity to the E. coli K12 PuuD could possibly serve this function. Further metabolism of GABA to succinate would involve GabT (GABA transaminase) and GabD2 (succinic semialdehyde dehydrogenase), encoded together on pSymb. In R. leguminosarum bv. viciae, gabT and gabD2 are highly upregulated in bacteroids as compared to succinate-grown cells (Karunakaran et al., 2009). In S. meliloti 1021, transcript levels for both genes are significantly reduced in bacteroids versus control cells grown in rich medium (Barnett et al., 2004). As the rich culture media used for the controls probably contained polyamines (Shaw et al., 2010), gabT and gabD2 transcripts in the controls might be expected to be high if their products function in polyamine degradation, thus making the transcript levels seen in bacteroids appear comparatively low. No direct experimental evidence for putrescine degradation in symbiosis exists. However, a S. meliloti 102F34 mutant partially impared in GABA catabolism due to reduced succinate semialdehyde dehydrogenase activity had a 30–40% reduction in ARA activity and plant biomass in combination with alfalfa. As mentioned in Section ‘‘Glutamate catabolism’’, strain 102F34 has glutamate decarboxylase activity and can convert glutamate to GABA. Because the GABA catabolism-impaired mutant was also affected in its ability to metabolize glutamate (Fitzmaurice & O’Gara, 1993), its symbiotic phenotype can not be attributed

to an inability to metabolize GABA produced by putrescine degradation. In summary, polyamines have vital roles in bacterial metabolism and transcriptomic data suggest that in rhizobia polyamine biosynthesis and transport are differentially regulated in symbiosis versus free-life. Additional studies with gene-directed rhizobial mutants affected in polyamine synthesis, transport or catabolism should lead to interesting findings on the specific roles of these metabolites in the symbiotic and free-living states. Ornithine lipid biosynthesis and function When grown under phosphate limitation, S. meliloti replaces membrane phospholipids with a variety of phosphate-free lipids, including ornithine lipids (OLs) and hydroxylated-OLs (Lo´pez-Lara et al., 2005). A S. meliloti 1021 olsA (encoding lyso-ornithine lipid acyltransferase; EC 2.3.1.; Figure 6) mutant lacked ornithine lipids, but had no growth defect in phosphate-limited media and formed symbiotically effective nodules on alfalfa (Lo´pez-Lara et al., 2005; Weissenmayer et al., 2002). In Rhizobium tropici CIAT899, however, mutants deficient in hydroxylated-OL synthesis were more sensitive to acid and temperature stress, and on bean were less competitive and formed poorly developed nodules that fixed significantly less nitrogen than the wild type. Although the physiological function of OLs and hydroxylated-OLs is not known, they probably alter membrane structure to increase stress resistance (Rojas-Jime´nez et al., 2005; Vences-Guzma´n et al., 2011). Vicibactin biosynthesis and function Because leghaemoglobin and nitrogenase contain iron, this element is essential for symbiosis and the synthesis of iron transport systems is induced in bacteroids

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

428

M. F. Dunn

(Barnett et al., 2004; Djordjevic, 2004; Tuskada et al., 2009; Vercruysse et al., 2011). In R. leguminosarum bv. viciae genes for the synthesis of the ornithine-derived siderophore vicibactin (Figure 6) are induced under low iron conditions. However, inactivation of various genes in the vicibactin pathway did not result in a defective symbiosis with pea or vetch inoculated with the mutants under either iron-rich or iron-deficient conditions. This result may derive from the ability of R. leguminosarum to synthesize other siderophores and/or iron transport systems (Todd et al., 2002). It would be of interest to determine how the vicibactin mutants are affected in nodulation competitiveness (Carter et al., 2002; Dilworth et al., 1998). Exogenously supplied vicibactin also reduces alumnium toxicity in R. leguminosarum cultures, possibly because the Al-vicibactin complex is not transported into the cell or, if taken up, remains in a complexed state (Rogers et al., 2001). Aluminum toxicity in soils is an important factor affecting the growth of rhizobia and other soil bacteria in some agricultural regions (Wood, 1995), but a possible role for vicibactin in symbiosis has not been explored from this angle. Histidine biosynthesis Based on genome annotation, rhizobia have a canonical histidine biosynthesis pathway but lack a gene encoding histidinol phosphatase (EC 3.1.3.15, denoted by ? in Figure 8), nor do they appear to have a bifunctional HisB that can catalyze both the histidinol phosphatase and imidazoleglycerol-phosphate dehydratase (EC 4.2.1.19) steps of the pathway. Lacking these alternatives, the missing reaction in rhizobia could be catalyzed by a member of the inositol monophosphatase family, as occurs in some bacteria (Mormann et al., 2006). Histidinol dehydrogense (HisD; EC 1.1.1.23) catalyzes the last two steps of the pathway and the

Crit Rev Microbiol, 2014; 41(4): 411–451

single gene encoding it is upregulated in R. etli bacteroids (Resendis-Antonio et al., 2011), while in S. meliloti 1021 two hisD homologs are downregulated in bacteroids (Barnett et al., 2004). HisA (phosphoribosylformimino-5 aminoimidazole carboxamide ribotide isomerase; EC 5.3.1.16) catalyzes a middle step in the pathway, prior to the generation of the purine precursor 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR; Figure 8). A R. etli CE3 hisA mutant induced a significantly reduced number of small nodules with decreased ARA on bean plants in comparison to the wild type. Plant inoculation with the mutant in the presence of exogenous histidine allowed the formation of significantly more nodules than in its absence (Ferraioli et al., 2002). A genetically uncharacterized R. etli CE3 histidine auxotroph unable to synthesize AICAR was symbiotically proficient on bean (Newman et al., 1995). The reason for the different symbiotic phenotypes of the two R. etli histidine auxotrophs is unknown, but might be due to the different P. vulgaris cultivars used in the nodulation assays. In Rhizobium leguminosarum bv. trifolii RTH48, two independent histidine auxotrophs having the phenotypic and enzymatic characteristics of a hisB and a hisD mutant were generated by chemical mutagenesis. Although HisB and HisD participate in histidine biosynthesis after the AICAR branchpoint (Figure 8), the symbiotic phenotypes of the mutants on berseem clover were Nod and NodþFix, respectively. Symbiotic competence was not restored by exogenous histidine (Yadav et al., 1998), but it is not known whether the supplied histidine was able to reach the infecting bacteria. The severe symbiotic defect of these late-step mutants indicates a requirement for histidine, rather than a pathway intermediate, for the formation of an effective symbiosis with clover. The symbiotic phenotypes of several genetically uncharacterized Tn5-induced histidine auxotrophs of B. japonicum

Figure 8. Histidine biosynthesis in R. etli CFN42 as inferred from the genome sequence. Abbreviations: HisA, phosphoribosylformimino5-aminoimidazole carboxamide ribotide isomerase; HisB, imidazoleglycerol-phosphate dehydratase; HisD, histidinol dehydrogenase; HisE, phosphoribosyl-ATP pyrophosphatase; HisF and HisH, imidzole glycerol phosphate synthase subunits; HisG and HisZ, ATP-phosphoribosyltransferase catalytic and regulatory subunits, respectively; HisI, phosphoribosyl-AMP cyclohydrolase;?, unidentified enzyme catalyzing the histidinolphosphatase reaction.

DOI: 10.3109/1040841X.2013.856854

USDA110 and USDA122 inoculated on G. max and/or G. soja ranged from Nod to Nodþ Fixþ (Kummer & Kuykendall, 1989; Sadowsky et al., 1986; So et al., 1987). Two histidine auxotrophs generated by Tn5 mutagenesis of the slow-growing cowpea Bradyrhizobium sp. IRC256 were Nodþ Fixþ, but their ARA was less than half that of nodules formed by the wild type (McLaughlin et al., 1987). Explanations for the different symbiotic outcomes seen with these mutants include quantitatively different histidine requirements of the microsymbionts or their abilities to transport plant-produced histidine, different quantities of histidine produced by the roots or nodules of the three hosts, or the point at which the pathway was blocked in the different mutants (e.g. before or after the purine branchpoint).

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Histidine catabolism The probable histidine catabolic pathway in rhizobia, using R. etli CFN42 as an example, is shown in Figure 9. The first step is catalyzed by histidase (HutH; EC 4.3.1.3), which deaminates histidine to urocanate and NH3. A R. etli CNPAF512 hutH transposon mutant was unable to nodulate bean plants (Braeken et al., 2008), but its ability to catabolize histidine was not determined. The R. etli CFN42 gene corresponding to the CNPAF512 hutH is designated hutHe2 and is plasmid encoded near another hutH homolog annotated hutHe1. hutHe1 occurs in an apparent histidine catabolism gene cluster, the products of which are highlighted in red in Figure 9. The cluster also contains hutC, encoding a putative histidine utilization transcriptional repressor protein (Allison & Phillips, 1990). Both the R. etli hutC and hutHe2 genes are upregulated in R. etli bacteroids (Resendis-Antonio et al., 2011), suggesting that perhaps hutHe2 transcription is not

Microsymbiont amino acid metabolism

429

completely repressed by HutC. In R. etli, another copy of hutH is encoded on the chromosome but not near genes for histidine catabolism. In S. meliloti the chromosomal histidase hutH2 is encoded contiguously with hutXWV, encoding a histidine uptake system. The transcription of hutX, encoding the periplasmic solute binding protein, is specifically and significantly upregulated by histidine (Boncompagni et al., 2000; Mauchline et al., 2006), and growth of a hutH2 or hutX mutant was reduced about 30% in medium with histidine as a nitrogen source. Histidine transport was reduced by about 50% in hutX or hutV mutants. The presence of other histidine catabolic and uptake systems in S. meliloti could explain the normal symbiotic phenotypes of these hut mutants on alfalfa (Boncompagni et al., 2000). S. meliloti also has a pSymbencoded hutH1 clustered with other genes involved in histidine catabolism, similar to the plasmid-encoded hutH homologs described for R. etli, above. hutH1 transcription is downregulated in S. meliloti bacteroids (Barnett et al., 2004). The number of potential hutH genes and their genomic context in rhizobia is variable: S. meliloti 1021 and R. leguminosarum bv. viciae 3841 encode three and five copies, respectively, where the hutH copies clustered with other hut genes are plasmid-borne. M. loti MAFF303099 encodes four copies (one in a hut cluster), and B. japonicum USDA110 a single copy localized with other hut genes. A. caulinodans ORS571 appears to lack all of the hut genes yet can grow with histidine as a nitrogen source (MichelReydellet & Kaminski, 1999), indicating that histidine deamination does occur in this strain. In Rhizobium phaseoli Ch24-10 genes encoding histidine catabolic enzymes HutH, HutG (formiminoglutamase; EC 3.5.3.8) and HutU (urocanate hydratase; EC 4.2.1.49) (Figure 9) are significantly upregulated in the rhizosphere of maize and bean plants (Lo´pez-Guerrero et al., 2012). These R. phaseoli genes are homologous to those clustered with the plasmid-encoded hut genes in R. leguminosarum bv. viciae 3841, which are not significantly upregulated in strain 3841 in the rhizospheres of host or non-host plants (Ramachandran et al., 2011). In summary, rhizobia contain more than one homolog of some of the histidine catabolic genes that are sometimes upregulated in rhizospheres or in bacteroids. The complete degradation of histidine generates NH3, 2-OG and glutamate that could be used in the synthesis of other amino acids. The contrasting symbiotic phenotypes of the R. etli and S. meliloti hutH gene mutants might stem from differences in the functionality of their HutH paralogs. Lysine biosynthesis

Figure 9. Histidine catabolism in R. etli CFN42 as inferred from the genome sequence. Hut enzymes encoded by genes present in the plasmidic histidine utilization gene cluster are shown in red (HutHe1, HutUe, HutI, HutG). Abbreviations: HutH; histidine ammonia-lyase; HutI, imidazolonepropionase; HutG, formiminoglutamase; HutU, urocanate hydratase.

Lysine is synthesized in a nine-step pathway beginning with aspartate (Figure 1). LysA (diaminopimelate decarboxylase; EC 4.1.1.20) catalyses the final step of the pathway by decarboxylating diaminopimelate to lysine. A R. etli CE3 lysA::Tn5 mutant grew at a slower rate than wild type on minimal medium and required lysine supplementation for normal growth. On bean plants, the mutant was Nod despite its ability to synthesize Nod factors in unsupplemented minimal medium. Inclusion of exogenous lysine in the

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

430

M. F. Dunn

nodulation assays allowed the mutant to form some nodulelike structures (Ferraioli et al., 2002). The partial auxotrophy of the R. etli lysA mutant is somewhat puzzling since strain CFN42 (parent strain of CE3) contains only one annotated lysA gene and no apparent alternatives to circumvent disruption of this step. A lysA transposon mutant of Mesorhizobium ciceri was a complete lysine auxotroph and formed only small, Fix nodules when inoculated onto chickpea. Undifferentiated bacteroids and bacteroids undergoing early senescence were present in plant cells of these nodules. Inoculation of the mutant in the presence of supplemental lysine allowed normal nodulation and nitrogen fixation (Das et al., 2010). In R. leguminosarum bv. viciae 3841, the second intermediate of the pathway, L-aspartate 4-semialdehyde, can be produced from aspartate by the sequential reactions of aspartate kinase (EC 2.7.2.4) and aspartate-semialdehyde dehydrogenase (EC 1.2.1.11), or directly from homoserine by the action of homoserine dehydrogenase (Hom: EC 1.1.1.3). The genes encoding Hom and the mid-pathway enzyme succinyl-diaminopimelate desuccinylase (DapE: EC 3.5.1.18) are highly upregulated in pea (hom) or host and non-host rhizospheres (dapE) (Ramachandran et al., 2011). Upregulation of these genes could be due to the presence of homoserine in root mucilage, which has been hypothesized to confer a selective growth advantage to R. leguminosaurm (Ramachandran et al., 2011; van Egeraat, 1975). Transcription of several lysine biosynthetic genes is downregulated in mature S. meliloti nodules, as are genes for the synthesis of most other amino acids (Barnett et al., 2004). A review of the experimental evidence for the roles of the basic amino acids in symbiosis shows that the synthesis of arginine or intermediates of the arginine pathway is essential in a majority of cases. In contrast, preventing arginine catabolism by the Arc pathway or Ocd does not greatly disrupt symbiosis, although compensating pathways or enzyme paralogs may be operational in these mutants. Histidine and lysine biosynthesis are in almost all cases important for a functional symbiosis, while histidine catabolism has not been extensively studied in symbiosis but is important in R. etli.

Neutral-polar amino acids: glutamine, asparagine, tyrosine, cysteine and serine Of the amino acids considered in this section, glutamine is notable as a major product of ammonium assimilation in free-living rhizobia and an important precursor of other essential nitrogen-containing compounds. Serine, along with aspartate, provides a large amount of the carbon skeletons needed for the synthesis of other metabolites (Reitzer, 2003). Cysteine in bacteroids may serve as a source of sulfur. The symbiotic phenotypes of rhizobial mutants affected in the metabolism of the amino acids considered in this section are shown in Table 4. Glutamine biosynthesis Rhizobia assimilate ammonium to produce glutamine using the glutamine synthetase (GS)/glutamate synthase (GOGAT)

Crit Rev Microbiol, 2014; 41(4): 411–451

pathway (Section ‘‘Glutamate biosynthesis’’) and alanine dehydrogenase (Section ‘‘Alanine catabolism’’). GS catalyzes the reaction: L-glutamate

þ ATP þ NH3 ) L-glutamine þ ADP þ Pi

As mentioned (Section ‘‘The general nitrogen regulatory (Ntr) system’’), most rhizobia have three GS isoenzymes designated GSI, GSII and GSIII, encoded by the glnA, glnII and glnT genes, respectively. Rhizobial GSI enzymes are structurally similar to enterobacterial GSIs and are also biochemically regulated by reversible adenylylation via the Ntr system (Arconde´guy et al., 1996, 1997; Espı´n et al., 1994; Section ‘‘The general nitrogen regulatory (Ntr) system’’). In S. meliloti and B. japonicum glnA transcript and GSI protein or activity are present at low levels in mature bacteroids while the activities of the other isoforms are absent (Arconde´guy et al., 1996; Carlson et al., 1985; Djordjevic, 2004; Espı´n et al., 1994). Mutants lacking GSI are glutamine prototrophs and symbiotically effective. S. meliloti GMI708 glnB mutants making no PII protein or a PII protein that is non-uridylylatable (Section ‘‘The general nitrogen regulatory (Ntr) system’’) both produce little inactive (adenylylated) GSI and have drastically lowered glnII expression. On alfalfa, both PII mutant types formed defective symbioses that were deficient in infection thread progression, but Fixþ nodules were eventually formed. The plants inoculated with either PII mutant developed symptoms of nitrogen starvation despite their Fixþ phenotypes, suggesting a role for PII in coupling bacteroid nitrogen fixation to plant nitrogen assimilation. This phenotype cannot be unambiguously attributed to changes in GSI/GSII activity/expression in the mutants, since the mutant also had reduced nod gene operon inducibility by luteolin (Arconde´guy et al., 1997). Similarly, S. meliloti mutants producing modified forms of GlnD, the uridylyltransferase/uridylyl removing enzyme interacting with PII, formed nitrogen-fixing nodules that failed to supply the plant with assimilable nitrogen (Yurgel & Kahn, 2008). Proteome analysis of S. meliloti 1021 wild type bacteroids isolated from M. trunculata or Melilotus alba detected no PII protein, in contrast to cells grown in culture (Djordjevic, 2004). Assuming PII is also absent in strain GMI708 PII mutant bacteroids, the symbiotic phenotypes described above might originate from the action of PII during earlier stages of nodule formation rather than in mature bacteroids. In rhizobia, GSII (glnII) gene transcription is regulated by the Ntr system as described in Section ‘‘The general nitrogen regulatory (Ntr) system’’. The third rhizobial GS isoenzyme, GSIII (GlnT) has been studied in R. etli CE3 and S. meliloti 1021 (de Bruijn et al., 1989; Espı´n et al., 1990, 1994). The glnT genes of these species were isolated by functional complementation of glutamine auxotrophic K. pneumoniae and E. coli glnA mutants, respectively (Espı´n et al., 1990; Shatters et al., 1993). In the E. coli glnA mutant complemented to glutamine prototrophy with the S. meliloti 1021 glnT, no GS activity was detected (de Bruijn et al., 1989). Purified GSIII from S. meliloti 104A14 was enzymatically active, had a very high Km (over 30 mM) for ammonium and was not expressed in the presence of the GSI or GSII isoenzymes (Shatters et al., 1993). glnT mutants of both

meliloti meliloti meliloti meliloti

S. S. S. S.

S. meliloti 1021/M. sativa R. etli CE3/P. vulgaris M. ciceri TAL620/C. arietinum

S. meliloti 104A14/M. sativa

R. etli CE3/P. vulgaris S. meliloti 1021/M. sativa R. etli CE3/P. vulgaris R. etli CE3/P. vulgaris

S. meliloti 104A14/M. sativa

1021/M. sativa 1021/M. sativa 104A14/M. sativa 1021/M. sativa

etli CE3/P. vulgaris japonicum Bj110/G. max japonicum Bj110/G. max japonicum Bj110/G. max

R. B. B. B.

de Bruijn et al. (1989) de Bruijn et al. (1989) Somerville et al. (1989) de Bruijn et al. (1989) Somerville et al. (1989)

Nodþ Fixþ. Partial Gln auxotroph.

Possible tyrC (cyclohexadienyl dehydrogenase). Tyr biosynthesis. hmgA (homogentisate dioxygenase). Tyr catabolism. melA (tyrosinase). Tyr catabolism. serA (3-phosphoglycerate dehydrogenase). Ser biosynthesis.

cysG (siroheme synthetase). Cys biosynthesis. glnA (glutamine synthetase I). Gln biosynthesis. glnII (glutamine synthetase II). Gln biosynthesis. glnA glnII (glutamine synthetases I and II). Gln biosynthesis. glnA (glutamine synthetase I). Gln biosynthesis. glnII (glutamine synthetase II). Gln biosynthesis. glnII (glutamine synthetase II). Gln biosynthesis. glnA glnII (glutamine synthetases I and II). Gln biosynthesis. glnA glnII (glutamine synthetases I and II). Gln biosynthesis. glnT (glutamine synthetase III). Gln biosynthesis. glnT (glutamine synthetase III). Gln biosynthesis. gdhB (glutamate dehydrogenase). Gln degradation. glsA (glutaminase A). Gln degradation.

Espı´n et al., 1994 de Bruijn et al. (1989) Tate´ et al. (2004) Dura´n et al. (1995); HuertaSaquero et al. (2004) Kerppola & Kahn (1988) Milcamps & de Bruijn (1999) Pin˜ero et al. (2007) Das et al. (2006)

Nodþ Fixþ. Nodþ (slight delay) Fixþ but lower ARA. Formed empty Fixswellings on root. Nodþ Fixþ with exogenous Ser.

þ

Nod Fix . Gln prototroph. Nodþ Fixþ. Gln prototroph. Nodþ Fixþ. Unable to use Gln as C and N source. Nodþ Fixþ. Significantly less GlsA activity, grew poorly on Gln as C and N source. Nodþ Fix.

þ

Tate´ et al. (1997) Carlson et al. (1987) Carlson et al. (1987) Carlson et al. (1987)

Pobigaylo et al. (2008)

Nodþ Fixþ. CysE activity reduced 95% but not a Cys auxotroph. Nod-delayed, Fixþ. Greatly reduced competititivity. Nodulation defects not prevented by introduction of the cloned cysG. Nodþ Fixþ. Cys auxotroph, poor nodulation competitor. Nodþ Fixþ. Increased ARA, Gln prototroph. Nodþ Fixþ. Increased ARA, Gln prototroph. Nod. Gln auxotroph. Gln supplementation allowed nodulation but not nitrogen fixation. Nodþ Fixþ. Gln prototroph. Nodþ Fixþ. Gln prototroph. Nodþ Fixþ. Gln prototroph Nodþ Fixþ. Gln auxotroph.

cysE (serine acetyltransferase). Cys biosynthesis. cysG (siroheme synthetase). Cys biosynthesis.

Parker et al. (2001)

Nodþ Fix.

Nodþ Fixþ. Unable to grow with Asn as C and N source.

Cys auxotroph, Tn5-generated. Cys biosynthesis.

Bradyrhizobium sp. CP283/M. atropurpureum R. leguminosarum bv. viciae RL3841/P. sativum S. meliloti Rm2011/M. sativa

Kerppola & Kahn (1988)

References

Huerta-Zepeda et al. (1997); Ortun˜o-Olea & Dura´n-Vargas (2000) Huerta-Zepeda et al. (1997); Ortun˜o-Olea & Dura´n-Vargas (2000) Cen et al. (1982)

ansB (aspartase), Tn5-generated. Asn degradation.

þ

Nod Fix . Asn auxotroph. þ

Mutant characteristics versus wild type 

þ

Nod Fix . Unable to grow with Asn as C and N source.

R. etli CFN42/P. vulgaris

R. etli CFN42/P. vulgaris

Presumptive asnB (asparagine synthetase), chemicallyinduced. Asn biosynthesis. ansA (asparaginase), Tn5-generated. Asn degradation.

Gene(s) mutated and/or pathway affected

S. meliloti 104A14/M. sativa

Rhizobia parent strain/host

Table 4. Rhizobial neutral-polar amino acid metabolism mutants and their symbiotic phenotypes.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

Microsymbiont amino acid metabolism 431

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

432

M. F. Dunn

species form effective, nitrogen fixing nodules on their respective hosts (de Bruijn et al., 1989; Espı´n et al., 1994). The free-living and symbiotic phenotypes of mutants lacking one or more GS isoenzyme depends on the microsymbiont species and the specific isoenzyme gene inactivated (Table 4). In the B. japonicum-soybean interaction, for example, strains singly mutated in either glnA or glnII retained glutamine prototrophy and formed nitrogenfixing nodules with higher ARA than the wild type, indicating that reducing the ammonium assimilation capacity of the microsymbiont enhanced nitrogenase activity. In contrast, a B. japonicum DglnA DglnII mutant displayed a free-living glutamine auxotrophy and was unable to form nodules, evidence that glutamine biosynthesis is important early in the nodulation process. Addition of exogenous glutamine to the rooting medium of the soybean plants inoculated with the double mutant allowed nodulation but not nitrogen fixation (Carlson et al., 1987), showing that GSI and GSII are also involved in later stages of the symbiosis. The rescue of the double mutant by exogenous glutamine indicates that the specific cause of the Nod phenotype was glutamine auxotrophy. In contrast to B. japonicum, S. meliloti glnA glnII double mutants constructed in strains 1021 and 104A14 formed effective nodules on their host. Interestingly, the growth defects of S. meliloti glnA glnII double mutants are strain dependent. A double mutant derived from 1021 had a strict glutamine auxotrophy (de Bruijn et al., 1989) while one derived from 104A14 was only partially auxotrophic for glutamine (Somerville et al., 1989) due to the expression of GlnT activity in this genetic background (Shatters et al., 1993). It would be interesting to construct and determine the symbiotic phenotype of an S. meliloti glnA glnII glnT triple mutant. Glutamine catabolism In R. etli CFN42, glutamine catabolism proceeds via the glutamine transaminase/!-amidase pathways (reactions 1 and 2) or glutaminase (reaction 3; Dura´n & Caldero´n, 1995): (1) L-glutamine þ glyoxylate , L-glycine þ 2-oxoglutaramate (EC 2.6.1.15) (2) 2-oxoglutaramate þ H2O ) 2-OG þ NHþ 3 (EC 3.5.1.3) (3) L-glutamine þ H2O ) L-glutamate þ NHþ 3 (EC 3.5.1.2) These reactions function as part of a glutamine cycle in R. etli that links nitrogen and carbon metabolism and promotes the more efficient utilization of succinate as a carbon source in vitro (Bravo & Mora, 1988; Dura´n & Caldero´n, 1995; Encarnacio´n et al., 1998). R. etli encodes two glutaminases, with the majority of the total glutaminase activity belonging to GlsA, a thermolabile, glutamine-induced enzyme. A R. etli glsA mutant had greatly reduced glutaminase activity and grew poorly on glutamine as sole carbon and nitrogen source but was unaffected in symbiosis with bean (Dura´n et al., 1995, 1996). In R. etli and A. caulinodans, NADH-dependent glutamate dehydrogenase (Gdh; EC 1.4.1.2) functions in concert with glutaminase in glutamine catabolism (Bravo & Mora, 1988; Donald & Ludwig, 1984; Encarnacio´n et al., 1998; Tate´ et al., 2004, 2012). Inactivation of gdhB in R. etli resulted in a mutant lacking Gdh activity and unable to utilize glutamine

Crit Rev Microbiol, 2014; 41(4): 411–451

or glutamate as a source of carbon and nitrogen. The gdhB mutant was also significantly impared in its ability to transport glutamine, but was symbiotically proficient on bean (Tate´ et al., 2004, 2012). The ability of R. etli to synthesize the tripeptide glutathione (GSH; Section ‘‘Glutamate catabolism’’) is essential for growth of strain CE3 on glutamine or glutamate. Mutants unable to synthesize GSH transported glutamine at a greatly reduced rate but expressed Gdh at near wild type levels. This indicates that the GSH mutants fail to grow on glutamine or glutamate due to a defect in transport rather than catabolism. The mechanism by which GSH affects glutamine transport, which occurs primarily via the Aap/Bra systems, remains to be discovered but parallels with GSHdependent amino acid transport in yeast and mammalian cells were postulated (Tate´ et al., 2012). In bacteria, glutamine is a precursor for purines, pyrimidines, amino sugars and carbamoyl phosphate. In S. meliloti 1021, the asnO gene encodes a asparagine synthetase-like protein that participates not in asparagine synthesis but in transferring the amide nitrogen from glutamine to N-acetylglutaminylglutamine, forming N-acetylglutaminylglutamine amide (NAGGN) (Berge´s et al., 2001; Sagot et al., 2010). In S. meliloti the NAGGN dipeptide serves as an osmoprotectant whose synthesis, along with asnO transcription, is induced under osmotic stress (Domı´nguezFerreras et al., 2006; Ru¨berg et al., 2003; Sagot et al., 2010; Virezen et al., 2013). Consistent with this, asnO null mutants grow significantly slower than the wild type in a high-salt medium but are symbiotically proficient with alfalfa inoculated under non-stress conditions (Sagot et al., 2010; Virezen et al., 2013). In S. meliloti GM1211, AsnO represses the anti-kinase activity of the regulator FixT. FixT inhibits phosphorylation (activation) of the FixL sensor kinase, causing less phosphorylation (activation) of FixJ and, ultimately, reduced expression of genes for nitrogen fixation (Section ‘‘The FixLJ system’’). This asnO mutant was not a asparagine auxotroph and formed Fixþ nodules on alfalfa (Berge´s et al., 2001). Asparagine biosynthesis In both the plant cytosol and bacteroids of nodules formed by R. leguminosarum bv. viciae 3841 on pea, asparagine is by far the most abundant amino acid on a molar basis. However, in the plant cytosol relatively little fixed nitrogen is incorporated into asparagine in comparison to that incorporated into amino acids that are directly involved in plant ammonium assimilation, like glutamate (Prell et al., 2009). In rhizobia, asparagine synthetase (AsnB; EC 6.3.5.4) catalyzes the interconversion of asparagine and glutamate to aspartate and glutamine: ATP þ L-aspartate þ L-glutamine þ H2 O , AMP þ diphosphate þ L-asparagine þ L-glutamate Most rhizobial genomes encode one or more asnB genes and AsnB activity is present in R. lupini bacteroids and lupin nodule cytosol (Kretovich et al., 1981). The growth requirements of a chemically-induced presumptive asnB mutant of S. meliloti 104A14 suggest that asparagine synthetase in this

DOI: 10.3109/1040841X.2013.856854

species acts in the biosynthetic direction. The mutant formed ineffective nodules on alfalfa (Kerppola & Kahn, 1988). Asparagine catabolism In R. etli, asparagine is catabolized as a carbon and nitrogen source by the successive actions of asparaginase (AnsA) and aspartase (AnsB; Section ‘‘Aspartate catabolism’’): L-asparagine

þ H2 O , L-aspartate þ NH3

ðAnsA; EC 3:5:1:1Þ , fumarate þ NH3 ðAnsB; EC 4:3:1:1Þ

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

L-aspartate

In R. etli the genes encoding aspartase and asparaginase occur in a ansRPAB operon: ansR and ansP encode a regulator of the operon and an asparagine permease, respectively (Ortun˜o-Olea & Dura´n-Vargas, 2000). A R. etli asnA::Tn5 mutant had greatly reduced asparaginase activity (R. etli encodes a second, constitutive asparaginase) and almost null aspartase activity, indicating a polar effect of the insertion in asnA on asnB. The asnA mutant was unable to grow on asparagine but was symbiotically proficient on bean (Huerta-Zepeda et al., 1997; Ortun˜o-Olea & Dura´n-Vargas, 2000). Both asparaginase and aspartase activities are significantly increased in cells grown with asparagine as a carbon and nitrogen source (Huerta-Zepeda et al., 1997) and asnA transcription is upregulated in bacteroids (Resendis-Antonio et al., 2011). The ansRPAB operon structure explains the co-regulation of asparaginase and aspartase activities by asparagine, which is transported by AnsP and acts as a co-activator, with AnsR, of the downstream ansA and ansB

Microsymbiont amino acid metabolism

433

genes (Huerta-Zepeda et al., 1997; Ortun˜o-Olea & Dura´nVargas, 2000). R. etli CFN42 grew well on asparagine and had very high induced aspartase activity when cultured in media containing asparagine as nitrogen or carbon and nitrogen sources. Aspartase activity was not induced in media containing aspartate and, in fact, the wild type did not appreciably utilize aspartate as sole carbon and nitrogen source but did grow with aspartate as nitrogen source (Huerta-Zepeda et al., 1996, 1997). A asnB::Tn5 mutant with normal asparaginase activity but lacking significant aspartase activity grew poorly relative to the wild type on asparagine, very poorly (like the wild type) on aspartate but formed wild-type nodules on bean plants (Huerta-Zepeda et al., 1997; Ortun˜o-Olea & Dura´n-Vargas, 2000). Tyrosine biosynthesis In rhizobia, tyrosine is synthesized as a branch of the shikimate pathway (Figure 10) in two reaction sequences starting from prephenate. In the first, prephenate is converted by cyclohexadieny/prephenate dehydrogenase (TyrC; EC 1.3.1.43) to 4-hydroxyphenylpyruvate, from which tyrosine is produced by histidinol-phosphate aminotransferase (HisC; EC 2.6.1.9) or, possibly, aromatic amino acid or aspartate aminotransferases (EC 2.6.1.57 and EC 2.6.1.1, respectively). In the second route, prephenate is converted to L-arogenate by the above-named aminotransferases, and TyrC converts this product to tyrosine. Two independent, chemically induced tyrosine auxotrophs of S. meliloti 104A14 were able to grow on minimal medium

Figure 10. Biosynthesis of the aromatic amino acids L-phenylalanine, L-tyrosine and L-tryptophan in S. meliloti 1021 based on the genome sequence. Abbreviations: AatA and AatB, aspartate aminotransferases; AroA, 3-phosphoshikimate 1-carboxyvinyltransferase; AroB, 3-dehydroquinate synthase; AroC, chorismate synthase; AroE shikimate 5-dehydrogenase; AroF, 3-deoxy-7-phosphoheptulonate synthase; AroK, shikimate kinase; AroQ, 3-dehydroquinate dehydratase; HisC, histidinol-phosphate aminotransferase; PheAa, chorismate mutase; PheA, prephenate dehydratase; TatA, aromatic amino acid aminotransferase; TrpA and TrpB, tryptophan synthase alpha and beta subunits, respectively; TrpC, indole-3-glycerol phosphate synthase; TrpD, anthranilate phosphoribosyltransferase; TrpE, anthranilate synthase; TrpF, N-(5’-phosphoribosyl)anthranilate isomerase; TyrC, indole-3-glycerol phosphate synthase.

434

M. F. Dunn

with tyrosine but not with shikimate or phenylalanine. Most sequenced S. meliloti strains lack a gene encoding phenylalanine 4-monooxygenase (EC 1.14.16.1), which converts phenylalanine to tyrosine and explains why the mutant was unable to utilize phenylalanine. Since shikimate is a middle intermediate of the shikimate pathway, failure to grow with it could be due to inactivation of any one of several steps occurring before tyrosine synthesis. The 104A14 auxotroph formed ineffective nodules on alfalfa, indicating a requirement for tyrosine biosynthesis by the microsymbiont mainly in the later stages of symbiosis (Kerppola & Kahn, 1988). Tyrosine is found in exudates of sand-grown alfalfa (Richter et al., 1968), but its concentration in nodules is not known.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Tyrosine catabolism During tyrosine degradation, nitrogen is derived by deamination (first step of the pathway shown in Figure 11), while utilization of tyrosine as a carbon source requires the subsequent steps of the pathway. S. meliloti 1021 is able to utilize tyrosine as a sole nitrogen or carbon source, and a hmgA (encoding homogentisate dioxygenase; EC 1.13.11.5) mutant loses the ability to metabolize this amino acid as a carbon source (Milcamps & de Bruijn, 1999). During nitrogen-limited growth hmgA expression is positively regulated by the ArsR-family transcriptional regulator NirR (Milcamps et al., 2001). Although hmgA mutants form a normal symbiosis with alfalfa, their survival during prolonged incubation in minimal medium is significantly reduced in comparison to the wild type. This might indicate an important role for intracellular tyrosine degradation (turnover) in the ex planta ecology of S. meliloti (Milcamps & de Bruijn, 1999). In response to nutrient deprivation or the presence of phenylalanine or tyrosine, S. meliloti 1021 4-hydroxyphenylpyruvate dioxygenase (HpdA; EC 1.13.11.27) gene expression is markedly increased by the Lrp-family transcriptional regulator HpdR (Loprasert et al., 2007; Section ‘‘Other regulatory systems’’). Tyrosine degradation triggered by HpdR is part of a nutrient deprivation response that may

Figure 11. Tyrosine catabolism in S. meliloti (Loprasert et al., 2007; Milcamps & de Bruijn, 1999).

Crit Rev Microbiol, 2014; 41(4): 411–451

enhance S. meliloti survival in the soil or growth in the rhizosphere (Milcamps et al., 2001). In B. japonicum 110spc4, HpdA protein was abundant in bacteroids obtained from cowpea nodules but absent from those obtained from siratro or soybean, suggesting a host-specific function (Koch et al., 2010). To test these ideas, hpdA mutants of S. meliloti and B. japonicum will be required. Tyrosinase (MelA; EC 1.14.18.1) is the first enzyme of the melanin biosynthesis pathway. The redox properties of this pigment might protect cells against reactive oxygen species, UV light and toxic compounds like phenolics. In Rhizobium and Sinorhizobium species, melA is plasmid encoded and regulated by the NifA-RpoN system (Section ‘‘The FixLJ system’’), indicating a possible symbiotic role, although rhizobial mutants unable to produce melanin are little affected in symbiosis (Borthakur et al., 1987; Pin˜ero et al., 2007). For example, a melA mutant of R. etli CE3 had a slightly delayed onset of nodule formation and made 25% fewer nodules on bean plants in comparison to the wild type or a melAcomplemented mutant. Nodules formed by the mutant had wild type ARA. The melA mutant was slightly more sensitive to hydrogen peroxide challenge than the wild type or melAcomplemented mutant (Pin˜ero et al., 2007). The relatively mild symbiotic defect and only slightly higher peroxide sensitivity of the mutant are consistent with additional oxidative stress resistance systems present in R. etli (Vargas et al., 2003). Cysteine biosynthesis Although not yet experimentally demonstrated, cysteine is probably a major source of sulfur for the synthesis of the metal sulfur clusters present in nitrogenase and other proteins (Evans et al., 1991; Zheng et al., 1993). In rhizobia, cysteine is synthesized from L-serine by the sequential action of serine acetyltransferase (CysE; EC 2.3.1.30) and cysteine synthase (CysK; EC 2.5.1.47), or possibly from pyruvate by cystathionine b-lyase (MetC; EC 4.4.1.8; only some MetCs are able to catalyze this reaction). Often, two or more homologs of CysE and/or CysK genes exist in a given species, while MetC is usually encoded by a single locus. In S. meliloti 1021, cysE is highly upregulated in bacteroids, contrasting the general downregulation of most other genes (including metC) involved in amino acid biosynthesis (Barnett et al., 2004). In B. japonicum CPAC 15, CysK is induced in the proteome of genistein-treated cells grown in culture and thus may also be induced in the early stages of nodulation (Stefaˆnia da Silva Batista & Hungria, 2012). In R. leguminosarum bv. viciae 3841, Tn5-insertion mutagenesis of a cysE gene (RL2209) reduced CysE activity to 5 % that of the wild type but did not cause cysteine auxotrophy or a symbiotic defect on pea (Parker et al., 2001). Consistent with the speculation of Parker et al. (2001) and the results of later genome sequencing (Young et al., 2006), the mutant’s cysteine prototrophy and residual CysE activity are probably due to the second cysE present in R. leguminosarum. To test the requirement for cysteine biosynthesis in strain 3841 a double cysE mutant will be required. In ‘‘Rhizobium parasponia’’ CP283, genetically uncharacterized Tn5-generated cysteine auxotrophs formed

DOI: 10.3109/1040841X.2013.856854

atypically small, Fix nodules on Siratro (Cen et al., 1982). Although closely related to bradyrhizobia (Lafay et al., 2006), which encode more than one cysE and cysK gene, the cysteine-requiring phenotype of the CP283 mutant suggests a lack of redundancy in its cysteine pathway.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Serine biosynthesis Rhizobial genome annotations indicate several routes for serine biosynthesis from different precursors, including the pathway converting 3-phosphoglycerate to serine using 3-phosphoglycerate dehydrogenase (SerA; EC 1.1.1.95), phosphoserine aminotransferase (SerC; EC 2.6.1.52) and phosphoserine phosphatase (SerB; EC 3.1.3.3). Most rhizobia encode several serA paralogs but contain single loci for serC and serB. A possible riboswitch regulatory sequence occurs upstream of the serC of some rhizobia, but its function in regulation has not been demonstrated (Corbino et al., 2005). In M. ciceri TAL620, which has not been sequenced, inactivation of a serA gene resulted in serine auxotrophy, indicating a lack of functional serA paralogs (note that 3 serA genes are annotated in the sequenced M. ciceri bv bisurrulae strain WSM1271, although this biovar does not nodulate chickpea as does TAL620; Nandasena et al., 2007). Inoculation of the mutant onto chickpea resulted only in the formation of swellings devoid of bacteria. Symbiotic proficiency was restored to the mutant in plant inoculations performed with exogenous serine, suggesting that chickpea roots supply insufficient serine to the mutant for symbiotic development (Das et al., 2006). A R. etli CE3 serA mutant had a less drastic phenotype on bean, inducing only a slightly reduced number of nodules and having nearly 50 % of the wild type ARA. Serine is relatively abundant in bean root exudates, presumably in quantities sufficient to allow microsymbiont multiplication, Nod factor synthesis and nodulation, but apparently not abundant enough in the nodule to promote efficient symbiosis (Ferraioli et al., 2002). In contrast to the results with M. ciceri and R. etli, a chemically-induced serine- or glycine-requiring auxotroph of S. meliloti 104A14 was symbiotically proficient on alfalfa (Kerppola & Kahn, 1988). Of the amino acids present in alfalfa root exudates, both glycine and serine, which are enzymatically interconvertible via serine hydroxymethyltransferase (EC 2.1.2.1), are present in relatively high concentrations (Richter et al., 1968). As described in Section ‘‘Glycine biosynthesis’’, serine is also derived from the degradation of glycine betaine. Serine catabolism Rhizobial genomes encode several enzymes potentially catalyzing the conversion of serine to pyruvate, hydroxypyruvate, L-tryptophane (a tryptophan precursor), or glycine. Serine dehydratase (Sda; EC 4.3.1.17) reversibly converts serine to pyruvate and ammonia. In S. meliloti 1021, the transcript for sda is notably increased in bacteroids versus cells grown in rich medium (Barnett et al., 2004). If operating in the direction of pyruvate formation, this product could be catabolized in the TCA cycle. With the exception of glutamine, relatively little data exists with which to evaluate the importance of neutral-polar amino

Microsymbiont amino acid metabolism

435

acid metabolism in symbiosis. In the great majority of cases, glutamine biosynthetic or catabolic mutants are not symbiotically defective. Cysteine synthesis mutants were symbiotically proficient in the three different rhizobia-legume combinations studied, while the available data for asparagine, tyrosine and serine metabolism are too limited to make a generalization.

Neutral-nonpolar amino acids: alanine, leucine, isoleucine, valine, phenylalanine, proline, tryptophan, glycine and methionine The branched-chain amino acids isoleucine, valine and leucine are often the most abundant amino acids in bacterial proteomes and leucine serves as one indicator of amino acid availability in enteric bacteria (Prell et al., 2009; Reitzer, 2003). The biosynthetic requirement for glycine in the synthesis of other essential metabolite is high in bacteria (Reitzer, 2003). In some symbiotic combinations, alanine may serve as an additional fixed nitrogen export product. Table 5 summarizes the symbiotic phenotypes of rhizobia affected in the metabolism of the amino acids discussed in this section. Alanine biosynthesis Alanine is synthesized from pyruvate (Figure 1) in the NADH-dependent reaction catalyzed by alanine dehydrogenase (Ald; EC 1.4.1.1). In R. leguminosarum bv. viciae aldA expression is strongly upregulated during growth on dicarboxylic acids or alanine, or in the rhizosphere of host and non-host plants (Karunakaran et al., 2009; Lodwig et al., 2004; Ramachandran et al., 2011). Induction of aldA by alanine is mediated by the product of the divergently transcribed aldR gene (Lodwig et al., 2004; Section ‘‘Global regulation of rhizobial amino acid metabolism’’). Although not symbiotically essential in R. leguminosarum bv. viciae, AldA is important for full symbiotic efficiency on pea plants (Allaway et al., 2000). Alanine synthesized by AldA was proposed to be a major fixed nitrogen excretion product from bacteroids. Waters et al. (1998) showed that isolated B. japonicum bacteroids incubated with 15N exported highly labelled alanine, and not ammonium, to the medium. Alanine excretion was postulated not to have been observed in earlier labelling experiments due to differences in bacteroid isolation protocols and assay oxygen concentrations. However, results from Bergersen’s laboratory using B. japonicum bacteroids isolated by methods similar to those used by Waters et al. (1998) detected only the significant excretion of labelled ammonium, and not alanine, in flow chamber 15N labelling experiments (Li et al., 2002). Poole and co-workers showed a contributing but not major role for alanine biosynthesis and secretion during nitrogen fixation by R. leguminosarum on pea (Allaway et al., 2000). Further work in R. leguminosarum and M. loti showed essentially wild-type symbiotic phenotypes of legume hosts inoculated with aldA mutants of their corresponding microsymbionts, supporting the contention that alanine is not a major fixed nitrogen secretion product (Prell & Poole, 2006; White et al., 2007). In S. meliloti AldA and enzymes responsible for ectoine catabolism (DoeA, B, C, D) are induced in cultures grown

Nod.

putA (proline dehydrogenase). Pro catabolism. putA (proline dehydrogenase). Pro catabolism. putA (proline dehydrogenase). Stachydrine and Pro catabolism. Tn5-generated mutants lacking PutA activity. Pro catabolism. Tn5-generated mutants lacking PutA activity. Pro catabolism.

de las Nieves Peltzer (2008) Prell et al. (2009)

Nod-delayed, Fix. Nodþ Fixþ with exogenous Leu. Nod. Nodulating ability partially restored by exogenous Leu. Nodþ, but bacteria not released from infection threads. Nod. Met auxotroph unable to produce Nod factors in culture without exogenous Met. With exogenous Met formed essentially Fix nodules. Nodþ (reduced number) Fix. Nodþ Fix; nodules small and white with few viable bacteria. Mutant was a Pro auxotroph. Exogenous Pro did not rescue phenotype. Nodþ Fixþ. Nodþ Fixþ, but less competitive for nodulation. Nodþ Fixþ, but less competitive for nodulation.

leuD (2-isopropylmalate isomerase). Leu biosynthesis. leuD (isopropylmalate isomerase small subunit). Leu biosynthesis Tn5-generated Leu auxotrophs. Leu biosynthesis. metZ (O-succinyl-homoserinesulfhydrylase). Met biosynthesis.

Symbiotic efficiency reduced by ca. 50 % based on plant phenotypes. Nod.

(2002) (2008) (2008) (2008)

Sharma & Yadov (2012)

Chaudhary et al. (1999)

Straub et al. (1996) Jime´nez-Zurdo et al. (1995, 1997) Phillips et al. (1998)

Ferraioli et al. (2002) King et al. (2000)

Nichik et al. (1995) Tate´ et al. (1999)

Chen et al. (2012)

Pobigaylo et al. (2008)

Sanjua´n-Pinilla et al. de las Nieves Peltzer de las Nieves Peltzer de las Nieves Peltzer Chen et al. (2012)

Formed pseudonodules lacking infection threads.

Nod. Nodþ Fixþ with exogenous Leu or Leu precursors. Nod-delayed, Fix. Nodþ Fixþ with exogenous Leu. Nod-delayed, Fix. Nodþ Fixþ with exogenous Leu. Nod-delayed, Fix. Nodþ Fixþ with exogenous Leu. Leu auxotroph, formed pseudonodules.

de las Nieves Peltzer (2008)

Nod-delayed, Fix. Nodþ Fixþ with exogenous Ile and Val.

leuA (2-isopropylmalate synthase). Leu biosynthesis. leuA1(2-isopropylmalate synthase). Leu biosynthesis. leuB (b-isopropylmalate dehydrogenase). Leu biosynthesis leuC (2-isopropylmalate isomerase). Leu biosynthesis. leuC (2-isopropylmalate isomerase). Leu biosynthesis.

Prell et al. (2009)

Auxotrophic for Ile and Val, essentially Nod . Nod genes not inducible. Nod, even with exogenous Ile, Val and Leu.

de las Nieves Peltzer (2008) Pobigaylo et al. (2008) Steele et al. (2003)

B. japonicum JH/G. max S. meliloti GR4/M. sativa S. meliloti 1021/M. sativa

Rhizobium sp. Cajanus Rspc-4/C. cajan

Lo´pez et al. (2001)

Nodþ (reduced number) Fix. Nod-delayed, Fix. Nodþ Fixþ with exogenous Ile and Val.

Aguilar & Grasso (1991)

Allaway et al. (2000); Mulley et al. (2011) Ferraioli et al. (2002) de las Nieves Peltzer (2008)

Nodþ Fixþ, but plants showed signs of N deficiency.



Kumar et al. (2005)

References

Prototrophic for branched-chain amino acids, formed pseudonodules. Nod-delayed, Fix. Nodþ Fixþ with exogenous Ile and Val. Induced a few pseudonodules lacking infection threads. Nod on both hosts. Nodþ Fixþ with exogenous Leu.

pheA (prephenate dehydrogenase). Phe biosynthesis. proC (P5C reductase). Pro biosynthesis.

Rhizobium spp. Mo6/V. radiata

Mutant characteristics versus wild type þ

Nod Fix judged by plant dry weights.

þ

ilvD2 (dihydroxy-acid dehydratase). Ile and Val biosynthesis. ilvD2 (dihydroxy-acid dehydratase). Ile and Val biosynthesis. ilvE (branched-chain amino acid aminotransferase). Branched-chain amino acid biosynthesis. ilvI (acetohydroxy acid synthase). Ile and Val biosynthesis. ilvI (acetohydroxy acid synthase). Ile and Val biosynthesis. leuA (2-isopropylmalate synthase). Leu biosynthesis.

aroK (shikimate kinase). Aromatic amino acid biosynthesis. ilvC (acetohydroxyacid isomeroreductase). Ile and Val biosynthesis. ilvC (acetohydroxyacid isomeroreductase). Ile and Val biosynthesis. ilvC (acetohydroxyacid isomeroreductase). Ile and Val biosynthesis. ilvD (dihydroxy-acid dehydratase). Ile and Val biosynthesis.

aldAc aldAp (chromosomal and plasmidic alanine dehydrogenases). Ala biosynthesis. aldA (alanine dehydrogenase). Ala biosynthesis.

Gene(s) mutated and/or pathway affected

M. F. Dunn

R. etli CE3/P. vulgaris B. japonicum I110/G. max

S. meliloti 1021/M. sativa S. meliloti 1021/M. sativa R. tropici CIAT899/L. leucocephala and P. vulgaris S. meliloti GR4/M. sativa S. meliloti 1021/M. sativa S. meliloti 1021/M. sativa S. meliloti 1021/M. sativa C. tiawanensis LMG19424/ M. pudica S. meliloti 1021/M. sativa R. leguminosarum bv. viciae 3841/ P. sativum S. meliloti SAM1/M. sativa R. etli CE3/P. vulgaris

B. phymatum STM815/M. pudica

S. meliloti 2011/M. sativa

R. leguminosarum bv. viciae 3841/ P. sativum S. meliloti 1021/M. sativa

S. meliloti 1021/M. sativa

S. meliloti 1021/M. sativa

M. loti MAFF303099/ L. corniculatus R. leguminosarum bv. viciae 3841/ P. sativum R. etli CE3/P. vulgaris S. meliloti 1021/M. sativa

Rhizobia parent strain/host

Table 5. Rhizobial neutral nonpolar amino acid metabolism mutants and their symbiotic phenotypes.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

436 Crit Rev Microbiol, 2014; 41(4): 411–451

Microsymbiont amino acid metabolism

Barsomian et al. (1992) Nodþ, mix of Fixþ and Fix nodules. Isolates from both Fixþ and Fix nodules retained Trp auxotrophy.

R. leguminosarum bv. viciae VF39SM/P. sativum and V. cracca S. meliloti 1021/M. sativa

B. japonicum I 110/G. max

trpE(G) (anthranilate synthase). Trp biosynthesis.

Noel (1998) Nodþ on both hosts, but significantly fewer plant cells infected.

Kuykendall & Hunter (1995)

Tate´ et al. (1999a)

Nodule number and ARA greatly reduced. Bacteria recovered from nodules retained Trp auxotrophy. Nod (formed only cortical proliferations on roots). Trp auxotrophs.

trpCD deletion mutants (indole glycerol phosphate synthase and phosphoribosyl anthranilate transferase). Trp biosynthesis. trpE(G) (anthranilate synthase). Trp biosynthesis.

Noel (1998) Nod on both hosts, but nodules nearly devoid of bacteroids.

R. leguminosarum bv. viciae VF39SM/P. sativum and V. cracca R. etli CE3/P. vulgaris

trpB (tryptophan synthase b chain). Trp biosynthesis.

Barsomian et al. (1992)

þ

Noel (1998) Nodþ on both hosts, but significantly fewer plant cells infected.

Nodþ Fixþ. All were Trp auxotrophs.

trpA, trpB, trpC, trpD or trpF single mutants. Trp biosynthesis. trpB (tryptophan synthase b chain). Trp biosynthesis.

Chien et al. (1991) Nodþ Fixþ.

Tn5-generated mutants lacking PutA activity. Pro catabolism. trpA (tryptophan synthase a chain). Trp biosynthesis.

R. leguminosarum bv. viciae C1204b/P. sativum R. leguminosarum bv. viciae VF39SM/P. sativum and V. cracca S. meliloti 1021/M. sativa

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

437

with this compatible solute. Ectoine is an osmoprotectant in S. meliloti and its protective effect requires the intracellular catabolism of ectoine taken up from the environment (Jebbar et al., 2005). Ectoine catabolism generates L-aspartate-4semialdehyde that could feed into the biosynthesis of lysine, threonine and aspartate. Why AldA levels are increased in the proteome of ectoine-induced cultures, along with the Doe enzymes, is not apparent. Of the four possible ocds in S. meliloti 1021, three (sma0486, sma1871, and smb20433) are clustered with genes whose products are probably involved in degrading ectoine. Although ocd genes are part of many ectoine degradation gene clusters, they have no known role in the catabolism of this compound (Schwibbert et al., 2011). In S. meliloti 102F34, a smb20433 homolog is induced by ectoine (Jebbar et al., 2005). Alanine catabolism As mentioned, AldA catalyzes the reversible conversion of ammonium plus pyruvate to alanine. While overexpression of AldA promotes alanine catabolism in R. leguminosarum bv. viciae, its principal role is in alanine synthesis (Allaway et al., 2000; Lodwig et al., 2004; Section ‘‘Alanine biosynthesis’’). L -alanine is catabolized after its conversion to D-alanine by alanine racemase (DadX; EC 5.1.1.1), followed by oxidation of D-alanine to pyruvate and ammonia by D-alanine dehydrogenase (DadA; EC 1.4.5.1). The expression of dadA and dadX in R. leguminoarum is dependent on the regulator DadR, which is required for alanine catabolism (Allaway et al., 2000). Somewhat unexpectedly, transcriptomic analysis of glucose- versus succinate-grown cells of R. leguminosarum bv viciae showed that growth on succinate significantly induced not only the expression of genes encoding the degradative enzymes DadA and DadX, but also that encoding the biosynthetic AldA (Karunakaran et al., 2009; Section ‘‘Alanine biosynthesis’’). Leucine, isoleucine and valine biosynthesis and symbiotic auxotrophy Bacterial biosynthesis of isoleucine and valine begins from the precursors threonine and pyruvate, respectively, utilizing the same set of ilv gene-encoded enzymes. Leucine is synthesized in reactions branching from valine synthesis and uses enzymes encoded by the leu genes (Kim & Gadd, 2008; Figure 12). Genetic and biochemical experiments performed by Poole and co-workers in R. leguminosarum bv. viciae demonstrate the symbiotic essentiality of leucine, isoleucine and valine (LIV) transport by bacteroids. As described in Section ‘‘Rhizobial amino acid transporters’’, the R. leguminosarum Aap and Bra transporters have broad specificity, making it difficult to determine which non-transported amino acid(s) are responsible for the defective symbioses formed by aap bra double mutants. This limitation was addressed by inactivating, in a R. leguminosarum bv. viciae 3841 aapJ transport mutant, the gene encoding the BraC broad-specificity solute binding protein that forms part of the BraCDEFG complex. This allowed the Bra complex to interact only with the orphan solute binding protein BraC3. The BraC3DEFG complex

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

438

M. F. Dunn

Figure 12. Abbreviated scheme of isoleucine, leucine and valine biosynthesis showing the gene products catalyzing key steps and some of the pathway intermediates. Abbreviations: IlvA, threonine dehydratase; IlvC, keto acid reductoisomerase; IlvD, dihydroxy-acid dehydratase; IlvE, branched chain amino acid aminotransferase; IlvG/H/ I, components of acetohydroxyacid synthase; LeuA, 2-isopropylmalate synthase; LeuB, 3-isopropylmalate dehydrogenase; LeuC/D, subunits of isopropylmalate isomerase.

present in the aapJ mutant background allows transport of only LIV and alanine (Hosie et al., 2002; Prell et al., 2009). Pea plants inoculated with a aapJ braC braC3 triple mutant showed severe nitrogen stress symptoms (significantly reduced plant dry weight and chlorosis) in comparison to the wild type. In contrast, plants inoculated with a aapJ braC double mutant, capable only of LIV and alanine transport, were symbiotically indistinguishable from plants inoculated with the wild type. The importance of alanine transport in the restoration of the symbiotic phenotype of the double mutant was discarded by heterologous complementation of a bv. viciae strain A34 aap bra double mutant (whose symbiotic defects on pea are identical to those of the strain 3841 double mutant) with the E. coli livKHMGF transporter, which is specific for LIV. The resulting strain transported alanine at a significantly reduced rate in comparison to the wild type, while LIV were transported at wild type rates. Pea plants inoculated with this strain were symbiotically proficient (Prell et al., 2009). A similar symbiotic requirement for LIV (but not alanine) transport was demonstrated for bacteroids of R. leguminosarum bv. phaseoli in combination with bean (Prell et al., 2010). Thus, R. leguminosarum must transport LIV to establish an effective symbiosis with pea or bean plants and, interestingly, selected ilv and leu genes are downregulated in bacteroids (Prell et al., 2009). The phenomenon by which LIV synthesis is specifically repressed in symbiosis, requiring the uptake of plant-produced LIV by bacteroids, was termed symbiotic auxotrophy (Prell et al., 2009, 2010; Terpolilli et al., 2012).

Crit Rev Microbiol, 2014; 41(4): 411–451

The hypotheses offered by Prell et al. (2009) to explain why bacteroids become dependent on the plant for LIV include a possible bacteroid LIV pseudoauxotrophy caused by plant-synthesized leucine-containing peptides that inhibit the acetohydroxyacid synthase step of LIV synthesis (Figure 12). Pseudoauxotrophy has been described in E. coli but not rhizobia. Symbiotic auxotrophy could also result from a shortage of LIV precursors like pyruvate in bacteroids. Additional data obtained by Poole and co-workers are consistent with leucine playing the key role in symbiotic auxotrophy, and the speculation that this could be linked to the function of leucine-responsive regulatory proteins is intriguing (Section ‘‘Other regulatory systems’’). Unlike R. leguminosarum, S. meliloti does not display symbiotic auxotrophy on alfalfa when its Aap and LIV transport systems are inactivated, perhaps because the mutants retain slightly higher rates of LIV transport than the corresponding R. leguminosarum mutants (Prell et al., 2010; Terpolilli et al., 2012). The significant upregulation of a LIV periplasmic binding protein (SMa0476) of an ABC transporter in S meliloti 1021 bacteroids (Barnett et al., 2004) is consistent with this notion. In S. meliloti 1021 ilvI and ilvD2 encode a subunit of acetolactate synthase (EC 2.2.1.6) and dihydroxy-acid dehydratase (EC 4.2.1.9), respectively, which are needed for making both valine and isoleucine (Figure 12). Independent null mutants in these genes resulted in auxotrophy for both amino acids, greatly reduced alfalfa nodulation competitiveness in comparison to the wild type and the induction of pseudonodules on the roots. Complementation of these mutants with their corresponding cloned ilv genes or inclusion of exogenous isoleucine and valine in plant nodulation assays restored their wild type symbiotic phenotypes (de las Nieves Peltzer et al., 2008; Pobigaylo et al., 2008). Similarly, ilvC (ketol-acid reductoisomerase (EC 1.1.1.86; Figure 12)) mutants of three different S. meliloti strains (including 1021) that required both isoleucine and valine for growth on minimal medium were unable to form nodules on the majority of the alfalfa plants inoculated, or formed small pseudonodules on a minority of the plants. Complementation of the ilvC mutants with the corresponding gene cloned from E. coli and/or S. meliloti restored prototrophy and symbiotic competence to the mutants (Aguilar & Grasso, 1991; Lo´pez et al., 2001). Both Nod factor production (analyzed by thin-layer chromotrogaphy) and nod gene promoter activity were diminished in the ilvC mutants (Lo´pez et al., 2001). Experiments with S. meliloti 1021 ilvC, ilvI, and ilvD2 mutants showed that their apparent decrease in nod gene transcription was caused by neomycin present in the assay cultures. Without neomycin, the S. meliloti ilv mutants had normal luteolin-inducible nodC expression (de las Nieves Peltzer et al., 2008). Together, the ability to synthesize Nod factors, the complemention of the mutant phenotypes with cloned ilv genes and the restoration of a normal symbiotic phenotype by exogenous isoleucine and valine suggest that the symbiotic phenotypes of the different S. meliloti mutants were due to isoleucine/valine auxotrophy (Aguilar & Grasso, 1991; de las Nieves Peltzer et al., 2008). Similarly, genetically undefined Tn5-generated leucine auxotrophs of S. meliloti SAM1 formed nodules on alfalfa, but

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

the bacteria were not released from the infection threads. Leucine supplementation did not rescue the defective symbiotic phenotype (Nichik et al., 1995). In S. meliloti 1021 independent mutants with lesions in leuA1, leuB, leuC or leuD (encoding 2-isopropylmalate synthase (EC 2.3.3.13), 3-isopropylmalate dehydrogenase (EC 1.1.1.85) and the large and small subunits of isopropylmalate isomerase (EC 4.2.1.33), respectively; Figure 12) were created. These mutants were leucine auxotrophs and, like the ilv mutants described above, were Nod-delayed and Fix on alfalfa. nodC-lacZ inducibility was not affected in these strains and a Nodþ Fixþ phenotype for all of the mutants was restored with exogenous leucine (de las Nieves Peltzer et al., 2008). Similarly, a S. meliloti GR4 leuA mutant devoid of 2isopropylmalate synthase activity (Figure 12) was auxotrophic for leucine and unable to induce nodules on alfalfa unless complemented with the cloned leuA gene (allowing full symbiotic restoration) or when nodulation assays contained exogenous leucine (allowing partial symbiotic restoration) (Sanjua´n-Pinilla et al., 2002). Similar to the situation described for the 1021 ilvC mutant characterized by Aguilar & Grasso (1991), the GR4 leuA mutant showed no induction of a nodC::lacZ fusion in the presence of luteolin in culture media with or without exogenous leucine. Interestingly, viable cell counts of the GR4 wild type and leuA mutant following inoculation onto alfalfa showed that the mutant was unable to multiply in the rhizosphere, unlike the wild type whose numbers increased 10-fold shortly after inoculation. The leuA mutant was thus symbiotically hampered mostly by its lack of multiplication in the rhizosphere and not by an inability to induce the nod genes (Sanjua´n-Pinilla et al., 2002). With regard to the symbiotic defects of the S. meliloti LIV mutants, it is relevant that alfalfa root exudates contain relatively low levels of these amino acids (Richter et al., 1968). A S. meliloti 1021 insertion mutant in livM (a homolog of the R. leguminosarum braE), encoding the branched-chain amino acid transporter permease, was competition-defective but otherwise symbiotically normal (Pobigaylo et al., 2008). It appears that LIV biosynthesis, possibly coupled with some uptake of the plant-produced amino acids, is required for an effective symbiosis by S. meliloti and is consistent with the lack of symbiotic auxotrophy in this species. A leucine auxotroph of R. leguminosarum bv. phaseoli TAL182 isolated by Tn5 mutagenesis had drastically reduced nodule formation on bean and was Fix (George & Robert, 1991). Gene-directed mutation of ilvD in R. leguminosarum bv. viciae 3841 blocks biosynthesis of LIV (Figure 12) and results in an inability to nodulate pea with or without the inclusion of LIV (1 mM each) in the nodulation assays. A strain 3841 leuD mutant was blocked specifically in leucine biosynthesis (Figure 12) and able to nodulate only with LIV supplementation. Addition of leucine alone (1 mM) to the leuD mutant nodulation assays also allowed nodulation, but resulted in root growth abnormalities (Prell et al., 2009). Thus, although LIV biosynthesis is downregulated in R. leguminosarum, biosynthesis of these amino acids is still required for an effective symbiosis. The importance of LIV biosynthesis by bacteroids extends to symbiotic nitrogen-fixing bacteria outside of the alpha-

Microsymbiont amino acid metabolism

439

proteobacterial clade. The beta-proteobacteria Burkholderia phymatum and Cupriavidus tiawanensis both nodulate and fix nitrogen in combination with Mimosa pudica. Inactivation of ilvE (encoding a branched-chain amino acid aminotransferase; EC 2.6.1.42) in B. phymatum and leuC in C. taiwanensis both cause Fix pseudonodules to be formed on their host. The ilvE mutant retained some branched-chain amino acid amino aminotransferase activity and was not auxotrophic for branched-chain amino acids while the leuC mutant was auxotrophic for leucine. Introduction of the relevant cloned genes into each of the mutants restored their symbiotic efficiency (Chen et al., 2012). Tn5-induced but genetically uncharacterized isoleucine/ valine auxotrophs of S. fredii HH303 formed a mixture of pseudonodules and mature nodules on soybean that, on a per plant basis, fixed nitrogen poorly in comparison to the wild type. However, bacteria reisolated from the nodules were all kanamycin sensitive, isoleucine/valine prototrophs, suggesting an inability of the original auxotrophic strains to successfully invade the root (Kim et al., 1988). In summary, all of the LIV biosynthetic mutants so far generated in Sinorhizobium and Rhizobium have severe symbiotic phenotypes, indicating the essentiality of the biosynthesis (and transport, in R. leguminosarum) of these amino acids for an effective symbiosis. Phenylalanine biosynthesis A genetically uncharacterized R. leguminosarum bv. phaseoli TAL182 Tn5-generated phenylalanine auxotroph had a Nodþ (with greatly reduced nodule number) Fix phenotype on bean (George & Robert, 1991). While bv. phaseoli strains have not been sequenced, phenylalanine biosynthesis in the related R. leguminosarum bv. viciae and other rhizobia begins with the AroF-catalyzed condensation of erythrose-4phosphate and phosphoenolpyruvate to form 3-deoxy-Darabinoheptulosonic acid-7-phosphate (DAHP) and proceeds through the shikimate pathway to form prephenate (Figure 10). This intermediate is then converted to phenylpyruvate by prephenate dehydratase (PheA; EC 4.2.1.51), and finally to phenylalanine in a reaction catalyzed by histidinol-phosphate aminotransferase (HisC; EC 2.6.1.9). Presumably, the R. leguminosarum bv. phaseoli phenylalanine mutant described above was affected in one of the later enzymes of the pathway, after prephenate, or it should have also been auxotrophic for tyrosine, which is synthesized from this precursor. Genetically uncharacterized Tn5-generated S. meliloti 1021 mutants with greatly decreased AroF (DAHP synthetase; EC 2.5.1.54) activity were Nodþ but essentially Fix on alfalfa (Jelesko et al., 1993). It is not known if aroF was the gene disrupted in the mutants, and the residual DAHP synthetase activity in the mutants is surprizing since S. meliloti 1021 encodes only one aroF. R. etli CE3 mutants in pheA, which is specific to the phenylalanine pathway, or aroK (encoding shikimate kinase; EC 2.7.1.71; Figure 10), required for the biosynthesis of all aromatic amino acids, have the expected growth requirements for phenylalanine and phenylalanine plus tryptophan plus tyrosine, respectively. On bean both mutants formed a

440

M. F. Dunn

drastically reduced number of nodules in comparison to the wild type, and these were Fix and had developmental abnormalities (Ferraioli et al., 2002).

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

Phenylalanine catabolism In rhizobia, the pathway for phenylalanine catabolism has not been established. In S. meliloti 1021, this amino acid is used as a nitrogen but as not a carbon source, and it appears that under nitrogen-limiting conditions phenylalanine is converted to tyrosine (Milcamps & de Bruijn, 1999), which is degraded by the pathway shown in Figure 11. In R. leguminosarum bv. viciae 3841, genes for the entire phenylalanine degradation pathway are induced in host and non-host rhizospheres (Ramachandran et al., 2011). In R. etli CFN42 bacteroids, a gene encoding a chromosomal 4-hydroxyphenylpyruvate dioxygenase (EC 1.13.11.27), part of the phenylalanine and tyrosine degradation pathways, is upregulated relative to its transcription level in cultured, exponential phase cells (Vercruysse et al., 2011). Interestingly, the presence of the dioxygenase protein in B. japonicum bacteorids is host-plant specific, being present in strain 110spc4 bacteroids isolated from cowpea, but not in those isolated from Siratro or soybean (Koch et al., 2010). When participating in tyrosine catabolism, the dioxygenase converts 4-hydroxyphenylpyruvate to homogenistate, which can apparently be further metabolized to 4-fumarylacetoacetate. However, in R. etli and most other rhizobia, genes encoding the enzyme fumarylacetoacetase, required for the conversion of 4-fumarylacetoacetate into fumarate and acetoacetate, appear to be absent. Proline biosynthesis In rhizobia, proline is synthesized by the action of g-glutamyl phosphate reductase (ProA; EC 1.2.1.41), g-glutamyl kinase (ProB; EC 2.7.2.11) and pyrroline-5-carboxylate (P5C) reductase (ProC; EC 1.5.1.2; Figure 13). In rhizobial genomes, the proline biosynthetic genes proA and proB are contiguous, while proC is located elsewhere in the genome. S. meliloti 1021 is unique in having two proB genes, proB1 and proB2: proB1 is in operon with proA, has a higher deduced amino acid sequence identity with the ProB of E. coli and likely encodes the ProB enzyme participating in the housekeeping synthesis of proline. Mutation of proB2

Figure 13. Proline biosynthesis and degradation. Although all of the enzymatic reactions are reversible, the arrows show the apparent in vivo reaction direction based on mutagenesis studies in rhizobia. Abbreviations: ProA, g-glutamyl phosphate reductase; ProB, g- glutamyl phosphate kinase; ProC, pyrroline-5-carboxylate reductase; PutA, bifunctional proline dehydrogenase/pyrroline-5-carboxylate dehydrogenase.

Crit Rev Microbiol, 2014; 41(4): 411–451

does not cause proline auxotrophy (mutation of proB1 was not done), and mutation of both proB genes might be necessary to cause this. proB2 is regulated by the FixLJ system (Section ‘‘The FixLJ system’’) and is highly induced under microaerobic conditions (Bobik et al., 2006; Ferrie´res et al., 2004), especially in a ntrR mutant background (Puska´s et al., 2004), as well as in bacteroids (Barnett et al., 2004). proB2 expression in strain 1021 bacteroids was significantly decreased in a fixJ mutant, but increased in a nifA or nifH mutant relative to the wild type (Bobik et al., 2006). Since fixLJ does not appear to be transcriptionally regulated by NtrR (Puska´s et al., 2004), the effect of these two nitrogen metabolic regulatory systems on proB2 appear to be separate and each could perhaps act independently to finetune proB2 expression under different conditions. In B. japonicum I110, a proC mutant generated by genedirected mutagenesis was a strict proline auxotroph unless complemented with the cloned proC borne on a plasmid or integrated at another location in the genome. On soybean, the mutant formed small nodules containing few bacteroids and was Fix. This phenotype could not be rescued with exogenous proline (King et al., 2000). In contrast, Tn5generated proline auxotrophs of R. leguminosarum bv. viciae C1204b (which were not genetically defined but could not be functionally complemented with the cloned proBA or proC genes of E. coli) formed normal nodules on pea. The mutants were also able to grow in cultures supplemented with glutamate (Chien et al., 1991), indicating that proline dehydrogenase (PutA, described below) in R. leguminosarum may also act in the biosynthetic direction, converting glutamate to glutamate-5-semialdehyde, thus by-passing ProBA. Proline catabolism A bifunctional proline dehydrogenase/P5C dehydrogenase (PutA; EC 1.5.99.8) is responsible for the catabolism of proline to glutamate in many bacteria, including rhizobia. In the first step of the reaction, the proline dehydrogenase domain of the enzyme uses FAD as a cofactor to oxidize proline, forming P5C (Figure 13). Following the spontaneous hydrolysis of this intermediate into GSA, the P5C dehydrogenase domain of the enzyme catalyzes a NAD(P)-dependent oxidation of GSA into glutamate (Krishnan & Becker, 2005). putA mutants of A. tumefaciens are unable to grow on proline, arginine or ornithine as sole carbon source, since arginine is degraded by the arginase and Arc pathways in this organism (Dessaux et al., 1986; Cho & Winans 1996; Sections ‘‘The arginase pathway’’ and ‘‘The arginine deiminase pathway’’). In A. tumefaciens, the putA promoter is positively regulated by the PutR regulator in the presence of proline or valine (a gratuitous inducer) and is not autorepressed (Cho & Winans, 1996). In S. meliloti GR4, putA is not linked to other genes of proline metabolism such as putR or a proline permease, as occurs in some bacteria. In S. meliloti, putA::lacZ fusions are induced by proline but not by valine. When the S. meliloti putA::lacZ construct was expressed in a putA mutant, no proline induction occurred, suggesting

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

PutA is an autogenous repressor of its encoding gene (Soto et al., 2000). A putA mutant of S. meliloti GRM8 was devoid of PutA activity and unable to grow on ornithine or proline. PutA activity in the parent strain was detectable only after induction by proline or ornithine. The putA mutant had reduced nodulation efficiency, suggesting a role for proline catabolism early in nodule formation (Jime´nez-Zurdo et al., 1995). The overexpression of PutA activity in S. meliloti 2011 strain resulted in a slight but transient increase in the nodulation competitiveness of the modified strain, especially in drought-stressed alfalfa plants, which probably secrete more proline into the rhizosphere (van Dillewijn et al., 2001). These results indicate a role for proline (or possibly arginine, which is metabolized to proline) catabolism in the early stages of the symbiotic interaction. In later stages, bacteroid putA transcription (in strain 1021) is downregulated (Barnett et al., 2004). Widely varying symbiotic phenotypes have been documented for PutA mutants of other rhizobia. A Rhizobium sp. Cajanus transposon-generated mutant lacking detectable PutA activity was unable to form nodules on Pigeonpea (Sharma & Yadav, 2012). Presumptive putA mutants of mungbean-nodulating rhizobia, genetically uncharacterized but lacking detectable PutA activity, were severely defective in symbiosis as quantified by nodule number and fresh weight per plant, and shoot nitrogen content and dry weight (Chaudhary et al., 1999). In contrast, presumptive putA mutants of R. leguminosarum bv. viciae C1204b lacking PutA activity were symbiotically unaffected in combination with pea (Chien et al., 1991). Stachydrine (N,N-dimethylproline, or proline betaine) is a proline derivative produced by Medicago species that has a role in drought stress protection. The compound is found in developing seeds and in the nodules of salt-stressed alfalfa. Stachydrine is efficiently transported by S. meliloti bacteroids (Fouge`re & Le Rudulier, 1990) and can serve as a sole source of carbon and nitrogen and as an osmoprotectant in this species (Alloing et al., 2006; Goldmann et al., 1994). Stachydrine catabolism involves successive demethylations catalzyed by products of the stc genes (Burnet et al., 2000) to form proline, which is further metabolized by PutA (Phillips et al., 1998) A S. meliloti putA::Tn5lux mutant showed induction by micromolar amounts of stachydrine or proline and failed to grown in media containing stachydrine, N-methylproline or proline as carbon and nitrogen sources. Proline uptake in the putA mutant was impaired by nearly 80% versus the wild type. Independent putA and stcD (stachydrine utilization protein; EC 1.-.-.-) mutants of S. meliloti were somewhat competition defective for infecting alfalfa (Goldman et al., 1994; Phillips et al., 1998). Three transport systems for stachydrine uptake in S. meliloti are known. The ATP-binding cassette histidine transporter (Hut) is a high-affinity uptake system for stachydrine and proline as well as histidine. Because this system is induced by exogenous histidine (but not by high osmolarity, proline or stachydrine), it appears to function in the uptake of these substrates for degradation rather than osmotic stress protection. As mentioned in Section

Microsymbiont amino acid metabolism

441

‘‘Histidine catabolism’’, the symbiotic properties of S. meliloti Rm5000 hutV and hutX interposon mutants, interrupted in the genes encoding the ATPase and periplasmic binding protein, respectively, were not significantly different from those of the wild type on alfalfa (Boncompagni et al., 2000). BetS, encoding a BCCT (betaine/carnitine/choline)family transporter with a high affinity for glycine betaine and stachydrine, is constitutively expressed in S. meliloti but is posttranslationally activated under salt stress. Stachydrine uptake was reduced by about 60% in a betS mutant, which had significantly slower growth in high-salt medium compared to the wild type (Boscari et al., 2002). BetS is expressed in S. meliloti bacteroids, especially in salt-stressed alfalfa plants. While a betS mutant had unimpaired nodulation and nitrogenfixing activity, overexpression of betS from a pTrp promotor in bacteroids led to a transient protection of nitrogenase activity from the effects of salt stress in alfalfa. The net accumulation of stachydrine in the BetS overproducing bacteroids would of course be determined not only by its increased transport but by its decreased rate of degradation (Boscari et al., 2006; Goldmann et al., 1994). The most recently described S. meliloti stachydrine transport system is Prb, belonging to the oligopeptide permease transporter family. Stachydrine and NaCl exert a strong, positive, synergistic effect on the expression of the Prb transporter. When grown in the presence of stachydrine, both prb single and prb betS double mutant strains grew significantly slower than wild type under osmotic stress conditions, but their symbiotic phenotypes were not determined (Alloing et al., 2006). The proline derivative trans-4-hydroxy-L-proline (hydroxyproline) is present in legume root nodules as well as in the rhizosphere and could be a rich source of reduced carbon and nitrogen for rhizobia. In S. meliloti 1021, the hydroxyproline transport and catabolic pathways and their regulation have been described in some detail, with the final product of the degradative pathway being the TCA cycle intermediate 2-OG. However, genes for neither transport or catabolism were highly expressed in bacteroids, in contrast to the clear and significant induction that occurred in cells cultured in the presence of hydroxyproline (MacLean et al., 2009; White et al., 2012). Mutants in hydroxyproline uptake are not affected in symbiosis, although it is reasonable to speculate that rhizobia might take advantage of hydroxyproline during rhizosphere colonization (MacLean et al., 2009). In summary some, but not all, proline biosynthetic and catabolic/utilization pathways are important for normal symbiotic function. Those that are important for nodulation efficiency or nitrogen fixation may be linked to their involvement in stress resistance, which is important during nodulation (Section ‘‘Establishment of the nitrogen-fixing symbiosis’’). Tryptophan biosynthesis Like the other aromatic amino acids, tryptophan synthesis initiates from the condensation of erythrose-4-phosphate and phosphoenolpyruvate (Figure 10), with chorismate being formed following several reactions catalyzed by aro gene

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

442

M. F. Dunn

products. This intermediate is converted to (3-indoyl)glycerolphosphate via the action of several trp gene products, and tryptophan is formed from this by the action of tryptophan synthase (EC 4.2.1.20: trpBA encode the b and a subunits of the enzyme, respectively). R. leguminosarum bv. viciae trpA and trpB mutants formed nodules on pea or vetch, but had a significantly reduced ability to infect plant cells. In a trpA::lacZ fusion strain, trpA expression was severely repressed by exogenous tryptophan but not by the tryptophan biosynthetic intermediate anthranilate (Noel, 1998). A trpB::Tn5 mutant of R. etli CE3 was a tryptophan auxotroph, except when complemented with the cloned trpB gene, and lacked tryptophan synthase activity. When inoculated onto bean, the trpB mutant formed significantly fewer nodules than the wild type, and these were small, white or pale pink, and essentially Fix. Bacteria isolated from the nodules formed by the mutant were all kanamycin resistant, tryptophan auxotrophs. Histological examination of nodules formed by the mutant showed that bacteroid maturation was greatly delayed or lacking. The effect of exogenous tryptophan in the root nodulation assays could not be tested since this amino acid severely affected bean root and root hair growth (Tate´ et al., 1999a). Assays in cultures induced with naringenin showed that the R. etli trpB mutant did not produce Nod factors in the absence of supplemental tryptophan. Since even a growth rate-limiting amount of exogenous tryptophan allowed some nod factor production by the mutant, it was concluded that the plant root probably supplied the mutant with sufficient tryptophan to allow infection but not normal multiplication in infection threads or just after release into plant cells (Tate´ et al., 1999a). In contrast to the trpB mutant of R. etli, genetically defined S. meliloti 1021 mutants inactivated in trpB or any other gene encoding a later step in tryptophan biosynthesis (trpA, trpC, trpD, or trpF; Figure 10), were symbiotically proficient on alfalfa (Barsomian et al., 1992). However, S. meliloti 2011 trpF (encoding N-5’-phosphoribosylanthranilate isomerase; EC 5.3.1.24) transposon mutants were completely infection-deficient and could not invade alfalfa in competition experiments with the wild type (Pobigaylo et al., 2008). The first committed step in tryptophan biosynthesis is catalyzed by anthranilate synthase (TrpE(G); EC 4.1.3.27; Barsomian et al., 1992). A R. leguminosarum bv. viciae VF39SM trpE(G) transposon mutant nodulated both pea and vetch but was deficient in its ability to invade plant cells of both hosts (Noel, 1998). A S. meliloti 1021 trpE(G) deletion mutant formed two morphologically distinct nodule types on alfalfa. The first type were large, elongated and white in color, except for a pink zone at the base of the nodule. These nodules were Fixþ. The second type of nodules were uniformly white in color and did not fix nitrogen. Histological examination of both nodule types showed that the trpE(G) deletion had caused an arrest of nodule development at an intermediate stage. Bacteria isolated from the nodules retained the expected antibiotic resistances and tryptophan auxotrophy. The intrepretation of these results was that S. meliloti requires the synthesis of anthranilate, but not tryptophan, for normal nodulation and nitrogen fixation.

Crit Rev Microbiol, 2014; 41(4): 411–451

One idea proposed is that the importance of anthranilate might relate to its role as a siderophore (Barsomian et al., 1992). A trpCD operon deletion mutant of B. japonicum USDA110 lacked indole glycerol phosphate synthase (EC 4.1.1.48) and phosphoribosyl anthranilate transferase (EC 2.4.2.18) activities, which are the products of trpC and trpD respectively, but contained greatly elevated levels of anthranilate synthase activity. The mutant formed only cortical proliferations, rather than nodules, on soybean (Kuykendall & Hunter, 1995). In contrast, a genetically uncharacterized mutant lacking tryptophan synthase was symbiotically proficient in the same series of experiments. These results were interpreted as indicating that the ability to synthesize a tryptophan biosynthetic intermediate (e.g. indole), but not tryptophan itself, were required for nodulation and nitrogen fixation by B. japonicum (Kuykendall & Hunter, 1995; Wells & Kuykendall, 1983). In summary, the symbiotic importance of the ability of rhizobia to synthesize tryptophan, as opposed to metabolites derived from the tryptophan biosynthetic pathway, is highly dependent on the symbiosis in question. This may relate in part to the widely varying tryptophan content of plant root exudates (Spaepen et al., 2007). Tryptophan catabolism Tryptophan is the precursor of indole acetic acid (IAA) (Patten et al., 2013), and synthesis and redistribution of this phytohormone in legumes plays an essential role in nodule organogenesis. Much interest in rhizobial tryptophan degradation has centered on determining whether IAA synthesized by the microsymbiont also contributes to nodulation efficiency (Spaepen et al., 2007). For example, a genetically engineered IAA overproducer of S. meliloti 1021 formed more nodules on Medicago hosts, while on bean nodulation was unaffected when it was inoculated with R. leguminosarum bv. phaseoli bearing the same IAA overproduction construct (Pii et al., 2007). Genetically uncharacterized IAAdeficient mutants of Bradyrhizobium elkanii USDA31 formed less nodules on soybean relative to the wild type, with the decreased number of nodules correlating roughly with the decreased IAA production seen in cultures (Fukuhara et al., 1994). Although these results suggest the possible importance of IAA synthesis from tryptophan in nodulation efficiency, null mutants devoid of IAA synthesis are still lacking in large part due to the multiple IAA synthesis pathways from tryptophan that exist in rhizobia (Spaepen et al., 2007). Glycine biosynthesis In rhizobia, glycine can be made from serine (using glycine hydroxymethyltansferase; EC 2.1.2.1) or threonine (using L -threonine aldolase; EC 4.1.2.5). Three independent chemically induced glycine auxotrophs of S. meliloti 2011 were symbiotically effective on lucerne judging by plant dry weights. Although the genetic lesions in these mutants were not identified, they were presumably affected in one of the enzymes allowing glycine synthesis from serine or threonine (Scherrer & De´narie´, 1971). Similarly, serine auxotrophs of

DOI: 10.3109/1040841X.2013.856854

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

S. meliloti 104A14 whose auxotrophy could be satisified by serine or by glycine were symbiotically proficient on alfalfa (Kerppola & Kahn, 1988), perhaps owing to the relatively high abundance of serine and glycine in alfalfa root exudates (Richter et al., 1968). In S. meliloti, glycine is also produced as an intermediate in the degradation of glycine betaine, which functions as an osmoprotectant (under high osmolarity conditions) or as a carbon and nitrogen source (under low osmolarity conditions) (Smith et al., 1988). Glycine betaine is present in micromolar amounts in nodules formed on some legumes (Ashraf & Iram, 2005) and is transported with high affinity by isolated bacteroids of S. meliloti (Fouge`re & Le Rudulier, 1990). The normal symbiotic phenotype of the S. meliloti 2011 glycine auxotrophs could be due to plant-supplied glycine, endogenous biosynthesis from serine or threonine, and/or uptake and catabolism of glycine betaine. Glycine catabolism In rhizobia and legumes, the heme precursor -aminolevulinic acid (ALA) is formed from glycine and succinyl-CoA by ALA synthase (HemA; EC 2.3.1.37). In wild type B. japonicum I110 grown in culture or isolated from soybean or cowpea bacteroids, HemA activity is present, while a hemA mutant strain lacked detectable activity in culture or in bacteroids. In nodules formed by either the wild type or hemA mutant, very high HemA activity was present in the plant fraction. Interestingly, the hemA mutant formed fully effective nodules on soybean or cowpea because it has a high affinity transport system for ALA that is active in bacteroids. In contrast, hemA mutants of several other rhizobia are symbiotically ineffective and lack ALA transport activity in bacteroids. Thus, the B. japonicum hemA mutant is symbiotically rescued by its ability to take up plant-produced ALA (McGinnis & O’Brian, 1995). ALA uptake in B. japonicum is negatively regulated by the product of a lrp homolog (King & O’Brian, 1997; Section ‘‘Other regulatory systems’’). Methionine biosynthesis Methionine is made from aspartate with aspartate kinase (LysC, or Ask; EC 2.7.2.4) catalyzing the first committed step of the pathway (Figure 14). Rhizobial genome sequences contain a single lysC or ask, although the inactivation of the sole LysC-encoding gene present in B. japonicum USDA110 did not cause methionine auxotrophy (King & O’Brian, 2001) despite the apparent lack of a by-pass for this step in the genome sequence. In Rhizobium and Sinorhizobium homoserine, a middle intermediate in the pathway is O-succinylated by homoserine O-succinyltransferase (MetA; EC 2.3.1.46), the product O-succcinyl-L-homoserine is converted in two steps catalyzed by cystathionine gamma-synthase (MetB; EC 2.5.1.48) and MetC into L-homocysteine, with L-cystathionine as intermediate, or directly to L-homocysteine by MetB and/or MetZ (O-succinylhomoserine sulfhydrylase; EC 2.5.1.-) (Tate´ et al., 1999). This penultimate product of the pathway is converted to methionine by methionine synthase (MetH; EC 2.1.1.13), the gene for which is induced by the flavanoid

Microsymbiont amino acid metabolism

443

Figure 14. Methionine biosynthesis in rhizobia. Numbers in parentheses next to gene product abbreviations correspond to the rhizobial strains encoding putative genes for these enzymes: 1, A. caulinodans ORS571; 2, B. japonicum USDA110; 3, M. loti MAFF303099; 4, R. etli CFN42; 5, R. leguminosarum bv. viciae 3841; 6, S. meliloti 1021. Abbreviations: Asd, aspartate semialdehyde dehydrogenase; CysD, O-acetylhomoserine sulfhydrylase; LysC, aspartate kinase; MecB, cystathionine-gamma-lyase; MetA, homoserine O-succinyltransferase; MetB, cystathionine gamma-synthase; MetC, cystathionine beta-lyase; MetH, methionine synthase; MetX, homoserine O-acetyltransferase.

naringenin in R. leguminosarum bv. viciae 3841 (Tolin et al., 2013). In S. meliloti the MetH reaction can also be catalyzed by betaine-homocysteine methyltransferase, part of the glycine betaine degradation pathway (Barra et al., 2006; Section ‘‘Glycine biosynthesis’’). In Bradyrhizobium and Azorhizobium, homoserine is O-acetylated by homoserine O-acetyltransferase (MetX; EC 2.3.1.31), O-acetyl-L-homoserine is converted to L-homocysteine by O-acetylhomoserine sulfhydrylase (CysD; EC 2.5.1.49), and methionine produced via MetH. In B. japonicum, M. loti and S. meliloti, regulation of methionine synthesis may depend on a riboswitch sequence present in the 50 untranslated leaders of the mRNAs for metA (S. meliloti only), metX and metZ. In A. tumefaciens this riboswitch specifically binds S-adenosylmethionine and probably functions as a genetic OFF switch of methionine synthesis in response to this metabolite (Corbino et al., 2005). Rhizobium leguminosarum bv. viciae 3841 contains all of the met genes required for the conversion of homoserine to methionine by two distinct routes (Figure 14). Because metX expression is highly upregulated in bacteroids versus succinate-grown cells, it was inactivated by Karunakaran et al. (2009) and the symbiotic phenotype of the resulting mutant determined. The mutant formed nodules with virtually wild type ARA on pea (Karunakaran et al., 2009; P. Poole, pers. comm.) perhaps because MetA and MetZ bypass the missing MetX reaction. Screening of a library of S. meliloti 1021 signature-tagged transposon mutants for defects in nodulating alfalfa led to the isolation of a Fixþ metA mutant with a delayed nodulation phenotype and greatly impaired nodulation competitiveness (Pobigaylo et al., 2008). The methionine requirement of the metA mutant was not determined but it should have been a methionine auxotroph, since S. meliloti lacks metX (Figure 14). Similarly, chemically generated

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

444

M. F. Dunn

methionine auxotrophs of S. meliloti 2011 were effective on alfalfa (Scherrer & De´narie´, 1971). The Fix phenotype of a chemically induced S. meliloti 104A14 methionine auxotroph probably lacking a functional metB (Kerppola & Kahn, 1988) indicates that the MetZ pathway is absent or functions inefficiently in this strain. A metZ null mutant of R. etli CE3 devoid of O-succinylhomoserine sulfhydrylase activity failed to grow on minimal medium in the absence of methionine. This indicates that MetZ, rather than MetC and/or MetB (also encoded in the genome) are responsible for the formation of L-homocysteine in the penultimate step of the R. etli pathway (Figure 14). When inoculated onto bean the mutant was Nod, but inoculation in the presence of methionine gave a Nodþ Fix phenotype. In culture, the metZ strain produced Nod factors only with supplemental methionine, explaining its Nod phenotype without added methionine. The Fix nodules formed in the presence of methionine contained a low number of infected cells with bacteroids, indicating that methionine biosynthesis is probably necessary for bacterial multiplication during infection (Tate´ et al., 1999). Chemically induced methionine, or cysteine þ methionine, auxotrophs of S. meliloti 2011 were symbiotically indistinguishable from the wild type based on alfalfa plant dry weight and phenotype (Scherrer & De´narie´, 1971). In contrast, S. meliloti strain 41 cysteine and/or methionine auxotrophs (obtained by chemical or Tn5 mutagenesis) were reported to be ‘‘symbiotically defective", though no further details were given (Forrai et al., 1983). In comparison to most other amino acids present in alfalfa root exudates, methionine is present at low levels (cysteine was not assayed for) (Richter et al., 1968). Tn5-induced but genetically uncharacterized S. fredii HH303 mutants requiring cysteine or methionine for growth formed Fixþ nodules on soybean. Bacteria reisolated from the nodules retained the auxotrophy and so were not revertants to wild type (Kim et al., 1988). The data presented in this section show an invarient symbiotic requirement for leucine, isoleucine and valine (LIV) biosynthesis in a variety of rhizobia-host combinations. In symbiotic auxotrophs like R. leguminosarum, bacteroid transport of host-produced LIV is also necessary for an effective symbiosis. Whether the importance of LIV availability in symbiosis is due to the large amount needed for protein synthesis and/or to its potential role in regulating nitrogen metabolism is an open question. For methionine, phenylalanine and proline, the limited data available suggests a symbiotic requirement for their synthesis by the microsymbiont. A relatively large number of tryptophan synthesis mutants have been characterized and generally have slight to moderate symbiotic or nodulation competitiveness defects. Whether this relates to a disruption of their ability to synthesize IAA remains to be determined. Consistent with the notion that alanine is not a major product of fixed nitrogen export to the plant, alanine biosynthesis mutants show little or no symbiotic defect. Proline catabolism mutants are sometimes less competitive for nodulation, suggesting a possible role for this amino acid as a nutrient in the early stages of symbiosis. The contrasting symbiotic phenotypes of the

Crit Rev Microbiol, 2014; 41(4): 411–451

rhizobial methionine auxotrophs might be due to different strain or species-specific requirements for methionine biosynthesis in symbiosis, differences in the quantity of methionine or pathway intermediates produced by different hosts (Barbour et al., 1991; Richter et al., 1968; Yaryura et al., 2008), or symbiotic induction of alternative Met enzymes.

Concluding remarks Much has been learned about rhizobial amino acid metabolism in recent decades and summaries of the symbiotic importance of specific amino acids are included at the ends of the relevant sections in this review. For rhizobia, relatively few generalizations can be made regarding the symbiotic roles of specific amino acids in symbiosis. This has to do with the host plant involved (widely differing amino acid contents in exudates or root tissue), the nature of the specific amino acid metabolism gene affected, the presence of gene redundancies and the great metabolic diversity of the microsymbionts. In addition, the evaluation of symbiotic effectiveness is greatly affected by experimental conditions of plant growth and cultivar, inoculation protocol, and how symbiotic proficiency is quantitated. Clearly, rhizobial amino acid metabolic mutants frequently have symbiotic defects. While the symbiotic characterization of amino acid auxotrophs and those deficient in catabolism has been instructive, the random mutagenic approaches used in many of these experiments often yields mutants with collateral metabolic effects such as the inability to make Nod factors or synthesize important precursor metabolites. With the genomic sequences now available, amino acid metabolic mutants can be constructed to either minimize or selectively exacerbate these effects. In the studies dealt with in this review, the great majority involve amino acid biosynthesis mutants. Much more remains to be done in generating and characterizing amino acid transport and catabolism mutants. For instance, would increasing the catabolic or transport capacity of a rhizobial strain for an amino acid abundant in the host rhizosphere or root lead to greater nodulation competitiveness? Global methodologies are providing a lot of information on which amino acid metabolic genes are expressed or proteins present in rhizobia grown in culture or in symbiosis. Like much of the work discussed here, future studies will continue to use this data to select metabolic genes for inactivation or enhanced expression and thus give more precise information on the roles of amino acid metabolism in symbiosis.

Acknowledgements Many thanks to those who sent reprints of their work and to Drs Lourdes Girard and Ismael Herna´ndez-Lucas for discussions or reading parts of the manuscript, and to Dr Jaime Mora for continued support.

Declaration of interest The author has no declarations of interest. Partial financial support for work in the author’s laboratory was provided by

DOI: 10.3109/1040841X.2013.856854

grants IN212008 and IN208811-3 from DGAPA-PAPIIT, Mexico.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

References Agarwal L, Purohit HJ. (2013). Genome sequence of Rhizobium lupini HPC(L) isolated from saline soil, Kutch (Gujarat). Genome Announcements 1:e00071–12. Aguilar OM, Grasso DH. (1991). The product of the Rhizobium meliloti ilvC gene is required for isoleucine and valine synthesis and nodulation of alfalfa. J Bacteriol 173:7756–64. Alfano JR, Kahn ML. (1993). Isolation and characterization of a gene coding for a novel aspartate aminotransferase from Rhizobium meliloti. J Bacteriol 175:4186–96. Ali H, Niel C, Guillaume JB. (1981). The pathways of ammonium assimilation in Rhizobium meliloti. Arch Microbiol 129:391–4. Allaway D, Lodwig EM, Crompton LA, et al. (2000). Identification of alanine dehydrogenase and its role in mixed secretion of ammonium and alanine by pea bacteroids. Mol Microbiol 36:508–15. Allison SL, Phillips AT. (1990). Nucleotide sequence of the gene encoding the repressor for the histidine utilization genes of Pseudomonas putida. J Bacteriol 172:5470–6. Alloing G, Travers I, Sagot B, et al. (2006). Proline betaine uptake in Sinorhizobium meliloti: characterization of Prb, an Opp-like ABC transporter regulated by both proline betaine and salinity stress. J Bacteriol 188:6308–17. Appels MA, Haaker H. (1991). Glutamate oxaloacetate transaminase in pea root nodules. Participation in a malate/aspartate shuttle between plant and bacteroid. Plant Physiol 95:740–7. Arconde´guy T, Huez I, Fourment J, Kahn D. (1996). Symbiotic nitrogen fixation does not require adenylylation of glutamine synthetase I in Rhizobium meliloti. FEMS Microbiol Lett 145:33–40. Arconde´guy T, Huez I, Tillard P, et al. (1997). The Rhizobium meliloti PII protein, which controls bacterial nitrogen metabolism, affects alfalfa nodule development. Genes Develop 11:1194–206. Ashraf M, Iram A. (2005). Drought stress induced changes in some organic substances in nodules and other plant parts of two potential legumes differing in salt tolerance. Flora 200:535–46. Barbour WM, Hattermann DR, Stacey G. (1991). Chemotaxis of Bradyrhizobium japonicum to soybean exudates. Appl Environ Microbiol 57:2635–9. Barnett MJ, Fisher RF, Jones T, et al. (2001). Nucleotide sequence and predicted functions of the entire Sinorhizobium meliloti pSymA megaplasmid. Proc Natl Acad Sci USA 98:9883–8. Barnett, MJ, Toman, CJ, Fisher, RF, Long, SR. (2004). A dual-genome Symbiosis Chip for coordinate study of signal exchange and development in a prokaryote-host interaction. Proc Natl Acad Sci USA 101:16636–41. Barra L, Fontenelle C, Ermel G, et al. (2006). Interrelations between glycine betaine catabolism and methionine biosynthesis in Sinorhizobium meliloti strain 102F34. J Bacteriol 188:7195–204. Barra-Bily L, Pandey SP, Trautwetter A, et al. (2010). The Sinorhizobium meliloti RNA chaperone Hfq mediates symbiosis of S. meliloti and alfalfa. J Bacteriol 192:1710–18. Barra-Bily L, Fontenelle C, Jan G, et al. (2010a). Proteomic alterations explain phenotypic changes in Sinorhizobium meliloti lacking the RNA chaperone Hfq. J Bacteriol 192:1719–29. Barsomian GD, Urzainqui A, Lohman K, Walker GC. (1992). Rhizobium meliloti mutants unable to synthesize anthranilate display a novel symbiotic phenotype. J Bacteriol 174:4416–26. Batista S, Patriarca EJ, Tate´ R, et al. (2009). An alternative succinate (2- oxoglutarate) transport system in Rhizobium tropici is induced in nodules of Phaseolus vulgaris. J Bacteriol 191:5057–67. Bearson BL, Lee IS, Casey TA. (2009). Escherichia coli O157:H7 glutamate- and arginine- dependent acid-resistance systems protect against oxidative stress during extreme acid challenge. Microbiol 155: 805–12. Becker A, Berge´s H, Krol E, et al. (2004). Global changes in gene expression in Sinorhizobium meliloti 1021 under microoxic and symbiotic conditions. Mol Plant-Microbe Interact 17:292–303. Berg G, Smalla K. (2009). Plant species and soil type cooperatively shape the structure and function of microbial communities in the rhizosphere. FEMS Microbiol Ecol 68:1–13.

Microsymbiont amino acid metabolism

445

Berge`s H, Checroun C, Guiral S, et al. (2001). A glutamineamidotransferase-like protein modulates FixT anti-kinase activity in Sinorhizobium meliloti. BMC Microbiol 1:6. Bobik C, Meilhoc E, Batut J. (2006). FixJ: a major regulator of the oxygen limitation response and late symbiotic functions of Sinorhizobium meliloti. J Bacteriol 188:4890–902. Boncompagni E, Dupont L, Mignot T, et al. (2000). Characterization of a Sinorhizobium meliloti ATP-binding cassette histidine transporter also involved in betaine and proline uptake. J Bacteriol 182:3717–25. Borthakur D, Lamb JW, Johnston AWB. (1987). Identification of two classes of Rhizobium phaseoli genes required for melanin synthesis, one of which is required for nitrogen fixation and activates the transcription of the other. Mol Gen Genet 207:155–60. Boscari A, Mandon K, Dupont L, et al. (2002). BetS is a major glycine betaine/proline betaine transporter required for early osmotic adjustment in Sinorhizobium meliloti. J Bacteriol 184:2654–63. Boscari A, Van de Sype G, Le Rudulier D, Mandon, K. (2006). Overexpression of BetS, a Sinorhizobium meliloti high-affinity betaine transporter, in bacteroids from Medicago sativa nodules sustains nitrogen fixation during early salt stress adaptation. Mol Plant- Microbe Interact 19:896–903. Boulter D, Jeremy JJ, Wilding M. (1966). Amino acids liberated into the culture medium by pea seedling roots. Plant Soil 24:121–7. Braeken K, Moris M, Daniels R, et al. (2006). New horizons for (p)ppGpp in bacterial and plant physiology. Trends Microbiol 14: 45–54. Braeken K, Daniels R, Vos K, et al. (2008). Genetic determinants of swarming in Rhizobium etli. Microbiol Ecol 55:54–64. Bracken K, Fauvart M, Vercruysse M, et al. (2008a). Pleiotropic effects of a rel mutation on stress survival in Rhizobium etli CNPAF512. BMC Microbiol 8:219. Available from: http://www.biomedcentral. com/1471-2180/8/219. Bravo A, Mora J. (1988). Ammonium assimilation in Rhizobium phaseoli by the glutamine synthetase-glutamate synthase pathway. J Bacteriol 170:980–4. Brinkman AB, Ettema TJG, de Vos WM, van der Oost J. (2003). The Lrp family of transcriptional regulators. Mol Microbiol 48:287–94. Brown CM, Dilworth MJ. (1975). Ammonium assimilation by Rhizobium cultures and bacteroids. J Gen Microbiol 86:39–48. Buendı´a-Claverı´a AM, Moussaid A, Ollero FJ, et al. (2003). A purL mutant of Sinorhizobium fredii HH103 is symbiotically defective and altered in its lipopolysaccharide. Microbiol 149:1807–18. Burnet MW, Goldman A, Message B, et al. (2000). The stachydrine catabolism region in Sinorhizobium meliloti encodes a multi-enzyme complex similar to the xenobiotic degrading systems in other bacteria. Gene 244:151–61. Capela D, Barloy-Hubler F, Gouzy J, et al. (2001). Analysis of the chromosome sequence of the legume symbiont Sinorhizobium meliloti strain 1021. Proc Natl Acad Sci USA 98:9877–82. Carlson TA, Guerinot ML, Chelm BK. (1985). Characterization of the gene encoding glutamine synthetase I (glnI) from Bradyrhizobium japonicum. J Bacteriol 162:698–703. Carlson TA, Martin GB, Chelm BK. (1987). Differential transcription of the two glutamine synthetase genes of Bradyrhizobium japonicum. J Bacteriol 169:5861–6. Carter RA, Worsley PS, Sawers G, et al. (2002). The vbs genes that direct synthesis of the siderophore vicibactin in Rhizobium leguminosarum: their expression in other genera requires ECF  factor RpoI. Mol Microbiol 44:1153–66. Castillo A, Taboada H, Mendoza A, et al. (2000). Role of GOGAT in carbon and nitrogen partitioning in Rhizobium etli. Microbiol 146: 1627–37. Cen Y, Bender GL, Trinick MJ, et al. (1982). Transposon mutagenesis in rhizobia which can nodulate both legumes and the nonlegume Parasponia. Appl Environ Microbiol 43:233–6. Charlier D, Glansdorff N. (2004). Biosynthesis of arginine and polyamines. In: Cohen G. editor. Escherichia coli and Salmonella: cellular and molecular biology. Module 3.6.1.10, EcoSal http:// www.ecosal.org. Washington, DC: ASM Press. Chaudhary S, Dudeja SS, Sharma HR, et al. (1999). Proline dehydrogenase activity of mungbean rhizobia and their proline prototrophs in relation to their efficiency in symbiotic association. Ind J Exp Biol 37: 1234–40. Chen J, Cheng C, Xia Y, et al. (2011). Lmo0036, an ornithine and putrescine carbamoyltransferase in Listeria monocytogenes,

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

446

M. F. Dunn

participates in arginine deiminase and agmatine deiminase pathways and mediates acid tolerance. Microbiol 157:3150–61. Chen W-M, Prell J, James EK, et al. (2012). Biosynthesis of branchedchain amino acids is essential for effective symbiosis between betarhizobia and Mimosa pudica. Microbiol 158:1758–66. Chien C-T, Rupp R, Beck S, Orser CS. (1991). Proline auxotrophic and catabolic mutants of Rhizobium leguminosarum biovar viciae strain C1204b are unaffected in nitrogen fixation. FEMS Microbiol Lett 77: 299–302. Chiurazzi M, Iaccarino M. (1990). Transcriptional analysis of the glnB-glnA region of Rhizobium leguminosarum biovar viciae. Mol Microbiol 4:1727–35. Cho K, Winans SC. (1996). The putA gene of Agrobacterium tumefaciens is transcriptionally activated in response to proline by an Lrp-like protein and is not autoregulated. Mol Microbiol 22:1025–33. Cho B-K, Federowicz S, Part Y-S, et al. (2012). Deciphering the transcriptional regulatory logic of amino acid metabolism. Nature Chem Biol 8:65–71. Choudhury B, Azad P, Kalita MC. (2010). Variability in symbiotic effectiveness of native rhizobia in acid stress. Curr Microbiol 61: 85–91. Collins CM, D’Orazio SEF. (1993). Bacterial ureases: structure, regulation of expression and role in pathogenesis. Mol Microbiol 9:907–13. Corbino KA, Barrick JE, Lim J, et al. (2005). Evidence for a second class of S-adenosylmethionine riboswitches and other regulatory RNA motifs in alpha-proteobacteria. Genome Biol 6:R70. Available from: http://genomebiology.com/2005/6/8/R70. Cunin R, Glansdorff N, Pie´rard A, Stalon V. (1986). Biosynthesis and metabolism of arginine in bacteria. Microbiol Rev 50:314–52. Das SK, Gautam US, Chakrabartty PK, Singh A. (2006). Characterization of a symbiotically defective serine auxotroph of Mesorhizobium ciceri. FEMS Microbiol Lett 263:244–51. Das SK, Gautam US, Sandhu KV, et al. (2010). Mutation in the lysA gene impairs the symbiotic properties of Mesorhizobium ciceri. Arch Microbiol 192:69–77. Davalos M, Fourment J, Lucas A, et al. (2004). Nitrogen regulation in Sinorhizobium meliloti probed with whole genome arrays. FEMS Microbiol Lett 241:33–40. Day DA, Poole PS, Tyerman SD, Rosendahl L. (2001). Ammonia and amino acid transport across symbiotic membranes in nitrogen-fixing legume nodules. Cell Mol Life Sci 58:61–71. de Bruijn FJ, Rossbach S, Schneider M, et al. (1989). Rhizobium meliloti 1021 has three differentially regulated loci involved in glutamine biosynthesis, none of which is essential for symbiotic nitrogen fixation. J Bacteriol 171:1673–82. de las Nieves Peltzer M, Roques N, Poinsot V, et al. (2008). Auxotrophy accounts for nodulation defect of most Sinorhizobium meliloti mutants in the branched-chain amino acid biosynthesis pathway. Mol Plant-Microbe Interact 21:1232–41. Desbrosses GG, Kopka J, Udvardi MK. (2005). Lotus japonicus metabolic profiling. Development of gas chromatography-mass spectrometry resources for the study of plant- microbe interactions. Plant Physiol 137:1302–18. Dessaux Y, Petit A, Tempe´ J, et al. (1986). Arginine catabolism in Agrobacterium strains: role of the Ti plasmid. J Bacteriol 166:44–50. D’Hooghe I, Vander Wauven C, Michiels J, et al. (1997). The arginine deiminase pathway in Rhizobium etli: DNA sequence analysis and functional study of the arcABC genes. J Bacteriol 179:7403–9. Dı´az R, Vargas-Lagunas C, Villalobos MA, et al. (2011). argC orthologs from Rhizobales show diverse profiles of transcriptional efficiency and functionality in Sinorhizobium meliloti. J Bacteriol 193:460–72. Dilworth MJ, Carson KC, Giles RGF, et al. (1998). Rhizobium leguminosarum bv. viciae produces a novel cyclic trihydroxamate siderophore, vicibactin. Microbiol 144:781–91. Dixon R, Kahn D. (2004). Genetic regulation of biological nitrogen fixation. Nat Rev Microbiol 2:621–31. Djordjevic MA. (2004). Sinorhizobium meliloti metabolism in the root nodule: A proteomic perspective. Proteomics 4:1859–72. Djordjevic MA, Chen HC, Natera S, et al. (2003). A global analysis of protein expression profiles in Sinorhizobium meliloti: discovery of new genes for nodule occupancy and stress adaptation. Mol PlantMicrobe Interact 16:508–24. Domı´nguez-Ferreas A, Pe´rez-Arnedo R, Becker A, et al. (2006). Transcriptome profiling reveals the importance of plasmid pSymB

Crit Rev Microbiol, 2014; 41(4): 411–451

for osmoadaptation of Sinorhizobium meliloti. J Bacteriol 188: 7617–25. Donald RGK, Ludwig RA. (1984). Rhizobium sp. strain ORS571 ammonium assimilation and nitrogen fixation. J Bacteriol 158: 1144–51. Dunn MF. (1998). Tricarboxylic acid cycle and anaplerotic enzymes in rhizobia. FEMS Microbiol Rev 22:105–23. Dunn MF, Cruz A, Girard L, Mora J. (2010). Regulacio´n de enzimas para la sı´ntesis de arginina en Sinorhizobium meliloti. In: Resumenes del XXVIII Congreso Nacional de Bioquı´mica (abstracts on CDROM). Dura´n S, Caldero´n J. (1995). Role of glutamine transaminase-!-amidase pathway and glutaminase in glutamine degradation in Rhizobium etli. Microbiol 141:589–95. Dura´n S, Du Pont G, Huerta-Zepeda A, Caldero´n J. (1995). The role of glutaminase in Rhizobium etli: studies with a new mutant. Microbiol 141:2883–9. Dura´n S, Sa´nchez-Lineares L, Huerta-Saquero A, et al. (1996). Identification of two glutaminases in Rhizobium etli. Biochem Genet 34:453–65. Encarnacio´n S, Caldero´n J, Gelbard AS, et al. (1998). Glutamine biosynthesis and the utilization of succinate and glutamine by Rhizobium etli and Sinorhizobium meliloti. Microbiol 144: 2629–38. Errey JC, Blanchard JS. (2005). Functional characterization of a novel ArgA from Mycobacterium tuberculosis. J Bacteriol 187:3039–44. Espı´n G, Moreno S, Wild M, et al. (1990). A previously unrecognized glutamine synthetase expressed in Klebsiella pneumoniae from the glnT locus of Rhizobium leguminosarum. Mol Gen Genet 223: 513–16. Espı´n G, Moreno S, Guzman J. (1994). Molecular genetics of the glutamine synthetases in Rhizobium species. Crit Rev Microbiol 20: 117–23. Evans DJ, Jones R, Woodley PR, et al. (1991). Nucleotide sequence and genetic analysis of the Azotobacter chroococcum nifUSVWZM cluster, including a new gene (nifP) which encodes a serine acetyl transferase. J Bacteriol 173:5457–69. Ferrie´res L, Francez-Charlot A, Gouzy J, et al. (2004). FixJ-regulated genes evolved through promoter duplication in Sinorhizobium meliloti. Microbiol 150:2335–45. Ferraioli S, Tate´ R, Caputo E, et al. (2001). The Rhizobium etli argC gene is essential for arginine biosynthesis and nodulation in Phaseolus vulgaris. Mol Plant-Microbe Interact 14:250–4. Ferraioli S, Tate´ R, Cermola M, et al. (2002). Auxotrophic mutant strains of Rhizobium etli reveal new nodule development phenotypes. Mol Plant-Microbe Interact 15:501–10. Fischer H-M. (1994). Genetic regulation of nitrogen fixation in rhizobia. Microbiol Rev 58:352–86. Fitzmaurice AM, O’Gara F. (1993). A Rhizobium meliloti mutant, lacking a functional g- aminobutyrate (GABA) bypass, is defective in glutamate catabolism and symbiotic nitrogen fixation. FEMS Microbiol Lett 10:195–202. Forrai T, Vinsze E, Ba´nfalvi Z, et al. (1983). Localization of symbiotic mutations in Rhizobium meliloti. J Bacteriol 153:635–43. Fouge`re F, Le Rudulier D. (1990). Uptake of glycine betaine and its analogues by bacteroids of Rhizobium meliloti. J Gen Microbiol 136: 157–63. Fujihara S, Harada Y. (1989). Fast-growing root nodule bacteria produce a novel polyamine, aminobutylhomospermidine. Biochem Biophys Res Comm 165:659–66. Fujihara S, Yoneyama T. (1993). Effects of pH and osmotic stress on cellular polyamine contents in the soybean rhizobia Rhizobium fredii P220 and Bradyrhizobium japonicum A1017. Appl Environ Microbiol 59:1104–9. Fukuhara H, Minakawa Y, Akao S, Minamisawa K. (1994). The involvement of indole-3-acetic acid produced by Bradyrhizobium elkanii in nodule formation. Plant Cell Physiol 35: 1261–5. Gage DJ. (2004). Infection and invasion of roots by symbiotic, nitrogenfixing rhizobia during nodulation of temperate legumes. Microbiol Mol Biol Rev 68:280–300. Gao J-L, Weissenmayer B, Taylor AM, et al. (2004). Identification of a gene required for the formation of lyso-ornithine lipid, an intermediate in the biosynthesis of ornithine-containing lipids. Mol Microbiol 53: 1757–70.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

Gao, M, Barnett MJ, Long SR, Tiplitski M. (2010). Role of Sinorhizobium meliloti global regulator Hfq in gene regulation and symbiosis. Mol Plant-Microbe Interact 23:355–65. George MLC, Robert FM. (1991). Autoregulatory response of Phaseolus vulgaris L. to symbiotic mutants of Rhizobium leguminosarum bv. phaseoli. Appl Environ Microbiol 57:2687–92. Goldmann A, Lecoeur L, Message B, et al. (1994). Symbiotic plasmid genes essential to the catabolism of proline betaine, or stachydrine, are also required for efficient nodulation by Rhizobium meliloti. FEMS Microbiol Lett 115:305–12. Gonza´lez V, Santamarı´a RI, Bustos P, et al. (2006). The partitioned Rhizobium etli genome: genetic and metabolic redundancy in seven interacting replicons. Proc Natl Acad Sci USA 103:3834–9. Haag AF, Arnold MFF, Myka KK, et al. (2013). Molecular insights into bacteroid development during Rhizobium-legume symbiosis. FEMS Microbiol Rev 37:364–83. Hamana K, Minamisawa K, Matsuzaki S. (1990). Polyamines in Rhizobium, Bradyrhizobium, Azorhizobium and Agrobacterium. FEMS Microbiol Lett 71:71–6. Harrison J, Jamet A, Muglia CI, et al. (2005). Glutathione plays a fundamental role in growth and symbiotic capacity of Sinorhizobium meliloti. J Bacteriol 187:168–74. Hellweg C, Pu¨hler A, Weidner S. (2009). The time course of the transcriptomic response of Sinorhizobium meliloti 1021 following a shift to acidic pH. BMC Microbiol 9:37. Hosie AHF, Allaway D, Jones MA, et al. (2001). Solute- binding proteindependent ABC transporters are responsible for solute efflux in addition to solute uptake. Mol Microbiol 40:1449–59. Hosie AHF, Allaway D, Galloway CS, et al. (2002). Rhizobium leguminosarum has a second general amino acid permease with unusually broad substrate specificity and high similarity to branchedchain amino acid transporters (Bra/LIV) of the ABC family. J Bacteriol 184:4071–80. Huerta-Saquero A, Caldero´n-Flores A, Dı´az-Villasen˜or A, et al. (2004). Regulation of transcription and activity of Rhizobium etli glutaminase A. Biochim Biophys Acta 1673:201–7. Huerta-Zepeda A, Dura´n S, Du Pont G, Calderon J. (1996). Asparagine degradation in Rhizobium etli. Microbiol 142:1071–6. Huerta-Zepeda A, Ortun˜o L, Du Pont G, et al. (1997). Isolation and characterization of Rhizobium etli mutants altered in degradation of asparagine. J Bacteriol 179:2068–72. Jebbar M, Sohn-Bo¨sser L, Bremer E, et al. (2005). Ectoine-induced proteins in Sinorhizobium meliloti include an ectoine ABC-type transporter involved in osmoprotection and ectoine catabolism. J Bacteriol 187:1293–304. Jelesko JG, Lara JC, Leigh JA. (1993). Rhizobium meliloti mutants with decreased DAHP synthase activity are sensitive to exogenous tryptophan and phenylalanine and form ineffective nodules. Mol Plant-Microbe Interact 6:135–43. Jime´nez-Zurdo JI, van Dillewijn P, Soto MJ, et al. (1995). Characterization of a Rhizobium meliloti proline dehydrogenase mutant altered in nodulation efficiency and competitiveness on alfalfa roots. Mol Plant-Microbe Interact 6:492–8. Jime´nez-Zurdo JI, Garcı´a-Rodrı´guez FM, Toro N. (1997). The Rhizobium meliloti putA gene: its role in the establishment of the symbiotic interaction with alfalfa. Mol Microbiol 23:85–93. Jime´nez-Zurdo JI, Valverde C, Becker A. (2013). Insights into the noncoding RNome of nitrogen-fixing endosymbiotic a-proteobacteria. Mol Plant-Microbe Interact 26:160–7. Kahn ML, Kraus J, Sommerville JE. (1985). A model of nutrient exchange in the Rhizobium- legume symbiosis. In: Evans H, Bottemley P, Newton WE editors. Nitrogen fixation research progress. New York: M. J. Nijhoff. pp. 193–9. Kaneko T, Nakamura Y, Sato S, et al. (2000). Complete genome structure of the nitrogen-fixing symbiotic bacterium Mesorhizobium loti. DNA Res 7:331–8. Karunakaran R, Ramachandran VK, Seaman JC, et al. (2009). Transcriptomic analysis of Rhizobium leguminosarum biovar viciae in symbiosis with host plants Pisum sativum and Vicia cracca. J Bacteriol 191:4002–14. Kerppola TK, Kahn ML. (1988). Symbiotic phenotypes of auxotrophic mutants of Rhizobium meliloti 104A14. J Gen Microbiol 134:913–19. Kerppola TK, Kahn ML. (1988a). Genetic analysis of carbamoylphosphate synthesis in Rhizobium meliloti 104A14. J Gen Microbiol 134: 921–9.

Microsymbiont amino acid metabolism

447

Kim BH, Gadd GM. (2008). Bacterial physiology and metabolism. Cambridge, UK: Cambridge University Press. Kim C-H, Kuykendall LD, Shah KS, Keister DL. (1988). Induction of symbiotically defective auxotrophic mutants of Rhizobium fredii HH303 by transposon mutagenesis. Appl Environ Microbiol 54:423–7. King ND, O’Brian MR. (1997). Identification of the lrp gene in Bradyrhizobium japonicum and its role in regulation of -aminolevulinic acid uptake. J Bacteriol 179:1828–31. King ND, O’Brian MR. (2001). Evidence for direct interaction between Enzyme INtr and aspartokinase to regulate bacterial oligopeptide transport. J Biol Chem 276:21311–16. King ND, Hojnacki D, O’Brian MR. (2000). The Bradyrhizobium japonicum proline biosynthesis gene proC is essential for symbiosis. Appl Environ Microbiol 66:5469–71. Koch M, Delmotte N, Rehrauer H, et al. (2010). Rhizobial adaptation to hosts, a new facet in the legume root-nodule symbiosis. Mol PlantMicrobe Interact 23:784–90. Kondorosi A, Sva´b Z, Kiss GB, Dixon RA. (1977). Ammonia assimilation and nitrogen fixation in Rhizobium meliloti. Mol Gen Genet 151:221–6. Kouchi H, Fukai K, Kihara A. (1991). Metabolism of glutamate and aspartate in bacteroids isolated from soybean root nodules. J Gen Microbiol 137:2901–10. Kretovich WL, Karikina TI, Weinova MK, et al. (1981). The synthesis of aspartic acid in Rhizobium lipini bacteroids. Plant Soil 61: 145–56. Krishnan N, Becker DF. (2005). Characterization of a bifunctional PutA homologue from Bradyrhizobium japonicum and identification of an active site residue that modulates proline reduction of the flavin adenine dinucleotide cofactor. Biochem 44:9130–9. Krol E, Becker A. (2004). Global transcriptional analysis of the phosphate starvation response in Sinorhizobium meliloti strains 1021 and 2011. Mol Genet Genomics 272:1–17. Kumar A, Vij N, Randhawa GS. (2003). Isolation and symbiotic characterization of transposon Tn5-induced arginine auxotrophs of Sinorhizobium meliloti. Indian J Exp Biol 41:1198–204. Kumar S, Bourde`s A, Poole P. (2005). De novo alanine synthesis by bacteroids of Mesorhizobium loti is not required for nitrogen transfer in determinate nodules of Lotus corniculatus. J Bacteriol 187:5493–5. Kummer RM, Kuykendall LD. (1989). Symbiotic properties of amino acid auxotrophs of Bradyrhizobium japonicum. Soil Biol Biochem 21: 779–82. Kuo Y-H, Lambien F, Ikegami F, Van Parijs R. (1982). Isoxazolin-5-ones and amino acids in root exudates of pea and sweet pea seedlings. Plant Physiol 70:1283–9. Kuykendall JD, Hunter WJ. (1995). Symbiotic ineffectiveness of trpCD deletion mutants of Bradyrhizobium japonicum. Soil Biol Biochem 27:1035–9. Lafay B, Bullier E, Burdon JJ. (2006). Bradyrhizobia isolated from nodules of Parasponia (Ulmaceae) do not constitute a separate coherent lineage. Int J Syst Evol Microbiol 56: 1013–18. Lee J, Michael AJ, Martynowski D, et al. (2007). Phylogenetic diversity and the structural basis of substrate specificity in the b/a-barrel fold basic amino acid decarboxylases. J Biol Chem 282:27115–25. Lee KB, De Backer P, Aono T, et al. (2008). The genome of the versatile nitrogen fixer Azorhizobium caulinodans ORS571. BMC Genomics 9: 271. Lewis TA, Gonzalez R, Botsford JL. (1990). Rhizobium meliloti glutamate synthase: cloning and initial characterization of the glt locus. J Bacteriol 172:2413–20. Li Y, Parsons R, Day DA, Bergersen FJ. (2002). Reassessment of major products of N2 fixation by bacteroids from soybean root nodules. Microbiol 148:1959–66. Limsunwun K, Jones PG. (2000). Spermidine acetyltransferase is required to prevent spermidine toxicity at low temperatures in Escherichia coli. J Bacteriol 182:5373–80. Lodwig EM, Hosie AHF, Bourde`s A, et al. (2003). Amino-acid cycling drives nitrogen fixation in the legume-Rhizobium symbiosis. Nature 422:722–6. Lodwig E, Kumar S, Allaway D, et al. (2004). Regulation of L-alanine dehydrogenase in Rhizobium leguminosarum bv. viciae and its role in pea nodules. J Bacteriol 186:842–9. Lo´pez JC, Grasso DH, Frugier F, et al. (2001). Early symbiotic responses induced by Sinorhizobium meliloti ilvC mutants in alfalfa. Mol PlantMicrobe Interact 14:55–62.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

448

M. F. Dunn

Lo´pez-Guerrero MG, Ormen˜o-Orrillo E, Acosta JL, et al. (2012). Rhizobial extrachromosomal replicon variability, stability and expression in natural niches. Plasmid 68:149–58. Lo´pez-Lara IM, Gao J-L, Soto MJ, et al. (2005). Phosphorus-free membrane lipids of Sinorhizobium meliloti are not required for symbiosis with alfalfa but contribute to inreases cell yields under phosphorus-limiting conditions of growth. Mol Plant-Microbe Interact 18:973–82. Loprasert S, Whangsuk W, Dubbs JM, et al. (2007). HpdR is a transcriptional activator of Sinorhizobium meliloti hpdA, which encodes a herbicide-tagreted 4-hydroxyphenylpyruvate dioxygenase. J Bacteriol 189:3660–4. Lu C-D. (2006). Pathways and regulation of bacterial arginine metabolism and perspectives for obtaining arginine overproducing strains. Appl Microbiol Biotechnol, 70:261–72. MacLean AM, White CE, Fowler JE, Finan TM. (2009). Identification of a hydroxyproline transport system in the legume endosymbiont Sinorhizobium meliloti. Mol Plant-Microbe Interact 22:1116–27. Mandon K, Pauly N, Boscari A, et al. (2009). ROS in the legumeRhizobium symbiosis. In: del Rı´o LA, Puppo A eds. Reactive oxygen species in plant signaling. Berlin Heidelberg: Springer-Verlag. pp. 135–47. Marc F, Weigel P, Legrain C, et al. (2000). Characterization and kinetic mechanisms of mono- and bifunctional ornithine acetyltransferases from thermophilic microorganisms. Eur J Biochem 267: 5217–26. Martin GB, Chapman KA, Chelm BK. (1988). Role of the Bradyrhizobium japonicum ntrC gene product in differential regulation of the glutamine synthetase II gene (glnII). J Bacteriol 170: 5452–9. Martin GB, Thomashow MF, Chelm BK. (1989). Bradyrhizobium japonicum glnB, a putative nitrogen-regulatory gene, is regulated by NtrC at tandem promoters. J Bacteriol 171:5638–45. Mauchline TH, Fowler JE, East AK, et al. (2006). Mapping the Sinorhizobium meliloti 1021 solute-binding protein-dependent transportome. Proc Natl Acad Sci USA 103:17933–8. McGinnis SD, O’Brian MR. (1995). The rhizobial hemA gene is required for symbiosis in species with deficient -aminolevulinic acid uptake activity. Plant Physiol 108:1547–52. McLaughlin W, Singh I, Ahmad MH. (1987). Characterization of Tn5-induced symbiotically defective mutants of cowpea rhizobia. FEMS Microbiol Lett 41:331–6. Mendz GL, Hazell SL. (1996). The urea cycle of Helicobacter pylori. Microbiol 142:2959–67. Michel-Reydellet N, Kaminski PA. (1999). Azorhizobium caulinodans PII and GlnK proteins control nitrogen fixation and ammonia assimilation. J Bacteriol 181:2655–8. Miksch G, Eberhardt U. (1994). Regulation of urease activity in Rhizobium meliloti. FEMS Microbiol Lett 120:149–54. Milcamps A, de Bruijn FJ. (1999). Identification of a novel nutrientdeprivation-induced Sinorhizobium meliloti gene (hmgA) involved in the degradation of tyrosine. Microbiol 145:935–47. Milcamps A, Struffi P, de Bruijn FJ. (2001). The Sinorhizobium meliloti nutrient-deprivation- induced tyrosine degradation gene hmgA is conrolled by a novel member of the arsR family of regulatory genes. Appl Environ Microbiol 67:2641–8. Miller KJ, Wood JM. (1996). Osmoadaptation by rhizosphere bacteria. Annu Rev Microbiol 50:101–36. Mishima E, Hosokawa A, Imaizumi-Anraku H, et al. (2008). Requirement for Mesorhizobium loti ornithine transcarbamoylase for successful symbiosis with Lotus japonicus as revealed by an unexpected long-range genome deletion. Plant Cell Physiol 49: 301–13. Mormann S, Lo¨mker A, Ru¨ckert C, et al. (2006). Random mutagenesis in Corynebacterium glutamicum ATCC 13032 using IS6100-based transposon vector identified the last unknown gene in the histidine biosynthesis pathway. BMC Genomics 7:205. Moris M, Braeken K, Schoeters E, et al. (2005). Effective symbiosis between Rhizobium etli and Phaseolus vulgaris requires the alarmone ppGpp. J Bacteriol 187:5460–9. Muglia C, Comai G, Spegazzini E, et al. (2008). Glutathione produced by Rhizobium tropici is important to prevent early senescence in common bean nodules. FEMS Microbiol Lett 286:191–8. Mulley G, White JP, Karunakaran R, et al. (2011). Mutation of GOGAT prevents pea bacteroid formation and N2 fixation by globally

Crit Rev Microbiol, 2014; 41(4): 411–451

downregulating transport of organic nitrogen sources. Mol Microbiol 80:149–67. Nandasena KG, O’Hara GW, Tiwari RP, et al. (2007). Mesorhizobium ciceri biovar biserrulae, a novel biovar nodulating the pasture legume Biserrula pelecinus L. Int J Syst Evol Microbiol 57:1041–5. Natera V, Sobrevals L, Fabra A, Castro S. (2006). Glutamate is involved in acid stress response in Bradyrhizobium sp. SEMIA 6144 (Arachis hypogaea L.) microsymbiont. Curr Microbiol 53:479–82. Newman JD, Schultz BW, Noel KD. (1992). Dissection of nodule development by supplementation of Rhizobium leguminosarum biovar phaseoli purine auxotrophs with 4- aminoimidazole-5-carboxamide riboside. Plant Physiol 99:401–8. Newman JD, Rosovitz MJ, Noel KD. (1995). Requirement for rhizobial production of 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR) for infection of bean. Mol Plant-Microbe Interact 8: 407–14. Nichik MM, Yerko VN, Klts SY, et al. (1995). Rhizobium mutants with changed nitrogen and cold resistance and leucine auxotrophy. In: Tikhonovich IA, Provorov NA, Romanov, VI, Newton WE eds. Nitrogen fixation: fundamentals and applications. Dordrecht: Kluwer Academic Publishers. pp. 412. Noel TC. (1998). Interaction of Rhizobium leguminosarum tryptophan and adenosine auxotrophs with host plants and non-legumes. PhD Thesis, University of Calgary. O’Gara F, Manian S, Meade J. (1984). Isolation of an Asm-mutant of Rhizobium japonicum defective in symbiotic N2-fixation. FEMS Microbiol Lett 24:241–5. Okazaki S, Hattori Y, Saeki K. (2007). The Mesorhizobium loti purB gene is involved in infection thread formation and nodule development in Lotus japonicus. J Bacteriol 189:8347–52. Oldroyd GED, Murray JD, Poole PS, Downie JA. (2011). The rules of engagement in the legume-rhizobial symbiosis. Annu Rev Genet 45: 119–44. Olivares J, Bedmar EJ, Sanjua´n J. (2013). Biological nitrogen fixation in the context of global change. Mol Plant-Microbe Interact 26:486–94. Ortun˜o-Olea L, Dura´n-Vargas S. (2000). The L-asparagine operon of Rhizobium etli contains a gene encoding an atypical asparaginase. FEMS Microbiol Lett 189:177–82. Pandey SP, Minesinger BK, Kumar J, Walker GC. (2011). A highly conserved protein of unknown function in Sinorhizobium meliloti affects sRNA regulation similar to Hfq. Nuc Acids Res 39:4691–708. Parker G, Walshaw D, O’Rourke K, et al. (2001). Evidence for redundancy in cysteine biosynthesis in Rhizobium leguminosaum RL3841: analysis of a cysE gene encoding serine acetyltransferase. Microbiol 147:2553–60. Patriarca EJ, Chiurazzi M, Manco G, et al. (1992). Activation of the Rhizobium leguminosarum glnII gene by NtrC is dependent on upstream DNA sequences. Mol Gen Genet 234:337–45. Patten CL, Blakney AJC, Coulson TJD. (2013). Activity, distribution and function of indole-3- acetic acid biosynthetic pathways in bacteria. Crit Rev Microbiol 39:395–415. Pauly N, Pucciariello C, Mandon K, et al. (2006). Reactive oxygen and nitrogen species and glutathione: key players in the legume-Rhizobium symbiosis. J Exp Bot 57:1769–76. Perez-Galdonat R, Kahn, ML. (1994). Effects of organic acids and low pH on Rhizobium meliloti 104A14. Microbiol 140:1231–5. Phillips DA, Sande ES, Vriezen JAC, et al. (1998). A new genetic locus in Sinorhizobium meliloti is involved in stachydrine utilization. Appl Environ Microbiol 64:3954–60. Pii Y, Crimi M, Cremonese G, et al. (2007). Auxin and nitric oxide control indeterminate nodule formation. BMC Plant Biol 7:21. Pin˜ero S, Rivera J, Romero D, et al. (2007). Tyrosinase from Rhizobium etli is involved in nodulation efficiency and symbiosis- associated stress resistance. J Mol Microbiol Biotechnol 13:35–44. Pobigaylo N, Szymczak S, Nattkemper TW, Becker A. (2008). Identification of genes relevant to symbiosis and competitiveness in Sinorhizobium meliloti using signature-tagged mutants. Mol PlantMicrobe Interact 21:219–31. Prell J, Poole P. (2006). Metabolic changes of rhizobia in legume nodules. Trends Microbiol 14:161–8. Prell J, White JP, Bourdes A, et al. (2009). Legumes regulate Rhizobium bacteroid development and persistence by the supply of branched-chain amino acids. Proc Natl Acad Sci USA 106: 12477–82.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

DOI: 10.3109/1040841X.2013.856854

Prell J, Bourde`s A, Karumakaran R, et al. (2009a). Pathway of g- aminobutyrate metabolism in Rhizobium leguminosarum 3841 and its role in symbiosis. J Bacteriol 191:2177–86. Prell J, Bourde`s A, Kumar S, et al. (2010). Role of symbiotic auxotrophy in the Rhizobium-legume symbiosis. Plos One 5:11. Puska´s IG, Nagy ZB, Keleman JZ, et al. (2004). Wide-range transcriptional modulating effect of ntrR under microaerobosis in Sinorhizobium meliloti. Mol Gen Genomics 272:275–89. Ramachandran VK, East AK, Karunakaran R, et al. (2011). Adaptation of Rhizobium leguminosarum to pea, alfalfa and sugar beet rhizospheres investigated by comparative transcriptomics. Genome Biol 12: R106. Available from: http://genomebiology.com/2011/12/10/R106. Ramo´n-Maiques S, Marina A, Gil-Ortiz F, et al. (2002). Structure of acetylglutamate kinase, a key enzyme for arginine biosynthesis and a prototype for the amino acid kinase family, during catalysis. Structure 10:329–42. Randhawa GS, Hassani R. (2002). Role of rhizobial biosynthetic pathways of amino acids, nucleotide bases and vitamins in symbiosis. Ind J Exp Biol 40:755–64. Rastogi VK, Watson RJ. (1991). Aspartate aminotransferase activity is required for aspartate catabolism and symbiotic nitrogen fixation in Rhizobium meliloti. J Bacteriol 173:2879–87. Rawn JD. (1989). Biochemistry. Burlington North Carolina: Neil Patterson Publishers. Reid CJ, Walshaw DL, Poole PS. (1996). Aspartate transport by the Dct system in Rhizobium leguminosarum negatively affects nitrogenregulated operons. Microbiol 142:2603–12. Reitzer L. (2003). Nitrogen assimilation and global regulation in Escherichia coli. Annu Rev Microbiol 57:155–76. Resendis-Antonio O, Herna´ndez M, Salazar E, et al. (2011). Systems biology of bacterial nitrogen fixation: high-throughput technology and its integrative description with constraint-based modeling. BMC Syst Biol 5:120. Richter M, Wilms W, Scheffer F. (1968). Determination of root exudates in a sterile continuous flow culture. II. Short-term and long-term variations of exudation intensity. Plant Physiol 43: 1747–54. Roccillo PM, Muglia CI, de Bruijn FJ, et al. (2000). Glutathione is involved in environmental stress responses in Rhizobium tropici, including acid tolereance. J Bacteriol 182:1748–53. Rogers NJ, Carson KC, Glenn AR, et al. (2001). Alleviation of aluminum toxicity to Rhizobium leguminosarum bv. viciae by the hydroxamate siderophore vicibactin. BioMetals 14:59–66. Rossi M, Defez R, Chiurazzi M, et al. (1989). Regulation of glutamine synthetase isoenzymes in Rhizobium leguminosarum biovar viciae. J Gen Microbiol 135:629–37. Rojas-Jime´nez K, Sohlenkamp C, Geiger O, et al. (2005). A CIC chloride channel homolog and ornithine-containing membrane lipids of Rhizobium tropici CIAT899 are involved in symbiotic efficiency and acid tolerance. Mol Plant-Microbe Interact 18: 1175–85. Ru¨berg S, Tian Z-X, Krol E, et al. (2003). Construction and validation of a Sinorhizobium meliloti whole genome DNA microarray: genomewide profiling of osmoadaptive gene expression. J Biotechnol 106: 255–68. Sadowsky MJ, Rostas K, Sista PW, et al. (1986). Symbiotically defective histidine auxotrophs of Bradyrhizobium japonicum. Arch Microbiol 144:334–9. Sagot B, Gaysinski M, Mehiri M, et al. (2010). Osmotically induced synthesis of the dipeptide N-acetylglutaminylglutamine amide is mediated by a new pathway conserved among bacteria. Proc Natl Acad Sci USA 107:12652–7. Salminen SO, Streeter JG. (1990). Factors contributing to the accumulation of glutamate in Bradyrhizobium japonicum bacteroids under microaerobic conditions. J Gen Microbiol 136:2119–26. Sanjua´n-Pinilla JM, Mun˜oz S, Nogales J, et al. (2002). Involvement of the Sinorhizobium meliloti leuA gene in activation of nodulation genes by NodD1 and luteolin. Arch Microbiol 178:36–44. Sankaranarayanan R, Cherney MM, Garen C, et al. (2010). The molecular structure of ornithine acetyltransferase from Mycobacterium tuberculosis bound to ornithine, a competitive inhibitor. J Mol Biol 397:679–90. Sarma AD, Emerich DW. (2006). A comparative proteomic evaluation of culture grown vs nodule isolated Bradyrhizobium japonicum. Proteomics 6:3008–28.

Microsymbiont amino acid metabolism

449

Scherrer A, De´narie´ J. (1971). Symbiotic properties of some auxotrophic mutants of Rhizobium meliloti and or their prototrophic revertants. Plant Soil 1971:39–45. Schindler U, Sans N, Schro¨der J. (1989). Ornithine cyclodeaminase from octopine Ti plasmid Ach5: identification, DNA sequence, enzyme properties, and comparison with gene and enzyme from nopoline Ti plasmid C58. J Bacteriol 171:847–54. Schlu¨ter A, Nohlen M, Kra¨mer M, et al. (2000). The Rhizobium leguminosarum bv. viciae glnD gene, encoding a uridylyltransferase/ uridylyl-removing enzyme, is expressed in the root nodule but is not essential for nitrogen fixation. Microbiol 146:2987–96. Schneider BL, Reitzer L. (2012). Pathway and enzyme redundancy in putrescine catabolism in Escherichia coli. J Bacteriol 194:4080–8. Schneider J, Wendisch VF. (2011). Biotechnological production of polyamines in bacteria: recent achievements and future perspectives. Appl Microbiol Biotechnol 91:17–30. Schrell A, Alt-Moerbe J, Lanz T, Schroeder J. (1989). Arginase of Agrobacterium Ti plasmid C58. DNA sequence, properties, and comparison with eucaryotic enzymes. Eur J Biochem 184:635–41. Schwibbert K, Marin-Sanguino A, Bagyan I, et al. (2011). A blueprint of ectoine metabolism from the genome of the industrial producer Halomonas elongata DSM 2581T. Environ Microbiol 13:1973–94. Sen D, Appunu C, Singh RK. (2008). Regulation of urease in Bradyrhizobium colonizing green gram (Vigna radiata (L.) Wilczek). Ind J Exp Biol 46:846–51. Serraj R, Shelp BJ, Sinclair TR. (1998). Accumulation of g-aminobutyric acid in nodulated soybean in response to drought stress. Physiol Plantarum 102:79–86. Shah P, Swiatlo E. (2008). A multifaceted role for polyamines in bacterial pathogens. Mol Microbiol 68:4–16. Sharma P, Yadav AS. (2012). Symbiotic characterization of mutants defective in proline dehydrogenase in Rhizobium sp. cajanus under drought stress condition. Eur J Exp Biol 2:206–16. Shatters RG, Liu Y, Kahn ML. (1993). Isolation and characterization of a novel glutamine synthetase from Rhizobium meliloti. J Biol Chem 268:469–75. Shaw FL. (2011). From prediction to function: Polyamine biosynthesis and formate metabolism in the a- and "-Proteobacteria. PhD Thesis, University of East Anglia. Shaw FL, Elliott KA, Kinch LN, et al. (2010). Evolution and multifarious horizontal transfer of an alternative biosynthetic pathway for the alternative polyamine sym-homospermidine. J Biol Chem 285: 14711–23. Smith LT, Pocard J-A, Bernard T, Le Rudulier, D. (1988). Osmotic control of glycine betaine biosynthesis and degradation in Rhizobium meliloti. J Bacteriol 170:3142–9. So J.-S, Hodgson ALM, Haugland R, et al. (1987). Transposon-induced symbiotic mutants of Bradyrhizobium japonicum: isolation of two gene regions essential for nodulation. Mol Gen Genet 207:15–23. Sobrero P, Valverde C. (2012). The bacterial protein Hfq: much more than a mere RNA-binding factor. Crit Rev Microbiol 38:276–99. Sobrevals L, Mu¨ller P, Fabra A, Castro S. (2006). Role of glutathione in the growth of Bradyrhizobium sp. (peanut microsymbiont) under different environmental stresses and in symbiosis with the host plant. Can J Microbiol 52:609–16. Somerville JE, Shatters RG, Kahn ML. (1989). Isolation, characterization and complementation of Rhizobium meliloti 104A14 mutants that lack glutamine synthetase II activity. J Bacteriol 171:5079–86. Soto MJ, Zorzano A, Garcı´a-Rodriguez FM, et al. (1994a). Identification of a novel Rhizobium meliloti nodulation efficiency nfe gene homolog of Agrobacterium orithine cyclodeaminase. Mol Plant-Microbe Interact 7:703–7. Soto MJ, van Dillewijn P, Olivares J, Toro N. (1994b). Ornithine cyclodeaminase activity in Rhizobium meliloti. FEMS Microbiol Lett 119:209–14. Soto MJ, Jime´nez-Zurdo JI, van Dillewijn P, Toro N. (2000). Sinorhizobium meliloti putA gene regulation: a new model within the family Rhizobiaceae. J Bacteriol 182:1935–41. Spaepen S, Vanderleyden J, Remans R. (2007). Indole-3-acetic acid in microbial and microorganism-plant signaling. FEMS Microbiol Rev 31:1–24. Stefaˆnia da Silva Batista J, Hungria M. (2012). Proteomics reveals differential expression of proteins related to a variety of metabolic pathways by genistein-induced Bradyrhizobium japonicum strains. J Proteomics 75:1211–19.

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

450

M. F. Dunn

Steele HL, Vinuesa P, Warner D. (2003). A leucine biosynthesis mutant of Rhizobium tropici CIAT899 which survives at pH 3.5. Biol Fertil Soils 38:84–8. Straub PF, Reynolds PHS, Althomsons S, et al. (1996). Isolation, DNA sequence analysis, and mutagenesis of a proline dehydrogenase gene (putA) from Bradyrhizobium japonicum. Appl Environ Microbiol 62: 221–9. Swamynathan SK, Singh A. (1992). Rhizobium meliloti purine auxotrophs are nodþ but defective in nitrogen fixation. J Genet 71:11–21. Tabor CW, Tabor H. (1985). Polyamines in microorganism. Microbiol Rev 49:81–99. Tate´ R, Riccio A, Iaccarino M, Patriarca EJ. (1997). A cysG mutant strain of Rhizobium etli pleiotropically defective in sulfate and nitrate assimilation. J Bacteriol 179:7343–50. Tate´ R, Riccio A, Caputo R, et al. (1999). The Rhizobium etli metZ gene is essential for methionine biosynthesis and nodulation of Phaseolus vulgaris. Mol Plant- Microbe Interact 12:24–34. Tate´ R, Riccio A, Caputo E, et al. (1999a). The Rhizobium etli trpB gene is essential for an effective symbiotic interaction with Phaseolus vulgaris. Mol Plant-Microbe Interact 12:926–33. Tate´ R, Mandrich L, Spinosa MR, et al. (2001). The Rhizobium GstI protein reduces the NHþ4 assimilation capacity of Rhizobium leguminosarum. Mol Plant-Microbe Interact 14:823–31. Tate´ R, Ferraioli S, Filosa S, et al. (2004). Glutamine utilization by Rhizobium etli. Mol Plant-Microbe Interact 17:720–8. Tate´ R, Cermola M, Riccio A, et al. (2012). Glutathione is required by Rhizobium etli for glutamine utilization and symbiotic effectiveness. Mol Plant-Microbe Interact 25:331–40. Terakado-Tonooka J, Fujihara S. (2008). Involvement of polyamines in the root nodule regulation of soybeans (Glycine max). Plant Root 2: 46–53. Terpolilli JJ, Hood GA, Poole PS. (2012). What determines the efficiency of N2-fixing Rhizobium-legume symbioses? Adv Microbiol Physiol 60:325–89. Todd JD, Wexler M, Sawers G, et al. (2002). RirA, an iron-responsive regulator in the symbiotic bacterium Rhizobium leguminosarum. Microbiol 148:4059–71. Toffanin A, Cadahia E, Imperial J, et al. (2002). Characterization of the urease gene cluster from Rhizobium leguminosarum bv. viciae. Arch Microbiol 177:290–8. Tolin S, Arrigoni G, Moscatiello R, et al. (2013). Quantitative analysis of the naringenin-inducible proteome in Rhizobium leguminosarum by isobaric tagging and mass spectrometry. Proteomics 13:1961–72. Torres-Quesada O, Oruezabal RI, et al. (2010). The Sinorhizobium meliloti RNA chaperone Hfq influences central carbon metabolism and the symbiotic interaction with alfalfa. BMC Microbiol 10:71. Available from: http://www.biomedcentral.com/1471-2180/10/71. Tuskada S, Aono T, Akiba N, et al. (2009). Comparative genome-wide transciptional profiling of Azorhizobium caulinodans ORS571 grown under free-living and symbiotic conditions. Appl Environ Microbiol 75:5037–46. Udvardi M, Poole PS. (2013). Transport and metabolism in legumerhizobia symbioses. Annu Rev Plant Biol 64:781–805. van Dillewijn P, Soto MJ, Villadas PJ, Toro N. (2001). Construction and environmental release of a Sinorhizobium meliloti strain genetically modified to be more competitive for alfalfa nodulation. Appl Environ Microbiol 67:3860–5. van Egeraat AWSM. (1975). The possible role of homoserine in the development of Rhizobium leguminoarum in the rhizosphere of pea seedlings. Plant Soil 42:381–6. Vargas MC, Encarnacio´n S, Da´valos A, et al. (2003). Only one catalase, KatG, is detectable in Rhizobium etli, and is encoded along with OxyR on a plasmid replicon. Microbiol 149:1163–76. Vauclare P, Bligny R, Gout E, Widmer F. (2013). An overview of the metabolic differences between Bradyrhizobium japonicum 110 bacteria and differentiated bacteroids from soybean (Glycine max) root nodules: an in vitro 13C- and 31P-nuclear magnetic resonance spectroscopy study. FEMS Microbiol Lett 343:49–56. Vences-Guzma´n MA, Guan Z, Ormen˜o-Orillo E, et al. (2011). Hydroxylated ornithine lipids increase stress tolerance in Rhizobium tropici CIAT899. Mol Microbiol 79:1496–514. Vercruysse M, Fauvart M, Beullens S, et al. (2011). A comparative transcriptome analysis of Rhizobium etli bacteroids: specific gene expression during symbiotic growth. Mol Plant-Microbe Interact 24: 1553–61.

Crit Rev Microbiol, 2014; 41(4): 411–451

Vercruysse M, Fauvart M, Jans A, et al. (2011a). Stress response regulators identified through genome-wide transcriptome analysis of the (p)ppGpp-dependent response in Rhizobium etli. Genome Biol 12: R17. Available from: http://genomebiology.com/2011/12/2/R17. Vessey JK. (1994). Measurement of nitrogenase activity in legume root nodules: in defense of the acetylene reduction assay. Plant Soil 158: 151–62. Villasen˜or T, Brom S, Da´valos A, et al. (2011). Housekeeping genes essential for pantothenate biosynthesis are plasmid-encoded in Rhizobium etli and Rhizobium leguminosarum. BMC Microbiol 11: 66. Vissers S, Legrain C, Wiame J-M. (1986). Control of a futile urea cycle by arginine feedback inhibition of ornithine carbamoyltransferase in Agrobacterium tumefaciens and Rhizobia. Eur J Biochem 159: 507–11. Voet D, Voet JG. (1995). Biochemistry. 2nd ed. New York: John Wiley & Sons. Vriezen JAC, de Bruijn FJ, Nu¨sslein K. (2013). Identification and characterization of a NaCl- responsive genetic locus involved in survival during desiccation in Sinorhizobium meliloti. Appl Environ Microbiol 79:5693–700. Walshaw DL. (1995). The general amino acid permease of Rhizobium leguminosarum biovar viciae. PhD Thesis, University of Reading. Walshaw DL, Poole PS. (1996). The general L-amino acid permease of Rhizobium leguminosarum is an ABC uptake system that also influences efflux of solutes. Mol Microbiol 21:1239–52. Walshaw DL, Wilkinson A, Mundy M, et al. (1997). Regulation of the TCA cycle and the general amino acid permease by overflow metabolism in Rhizobium leguminosarum. Microbiol 143: 2209–21. Watanabe A, Ideawa K, Iriguchi M, et al. (2002). Complete genomic sequence of nitrogen-fixing symbiotic bacterium Bradyrhizobium japonicum USDA110. DNA Res 9:189–97. Waters JK, Hughes BL, Purcell LC, et al. (1998). Alanine, not ammonia, is excreted from N2-fixing soybean nodule bacteroids. Proc Natl Acad Sci USA 95:12038–42. Watson JD, Baker TA, Bell SP, et al. (2008). Molecular biology of the gene. 6th ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Watson RJ, Rastogi VK, Chan Y-K. (1993). Aspartate transport in Rhizobium meliloti. J Gen Microbiol 139:1315–23. Weissenmayer B, Gao J-L, Lo´pez-Lara IM, Geiger O. (2002). Identification of a gene required for the biosynthesis of ornithinederived lipids. Mol Microbiol 45:721–33. Wells SE, Kuykendall LD. (1983). Tryptophan auxotrophs of Rhizobium japonicum. J Bacteriol 156:1356–8. Wells DH, Long SR. (2002). The Sinorhizobium meliloti stringent response affects multiple aspects of symbiosis. Mol Microbiol 43: 1115–27. White CE, Gavina JMA, Morton R, et al. (2012). Control of hydroxyproline catabolism in Sinorhizobium meliloti. Mol Microbiol 85:1133–47. White J, Prell J, James EK, Poole P. (2007). Nutrient sharing between symbionts. Plant Physiol 144:604–14. Whitehead LF, Tyerman SD, Day DA. (2001). Polyamines as potential regulators of nutrient exchange across the peribacteroid membrane in soybean root nodules. Aust J Plant Physiol 28:675–81. Wood JM, Bremer E, Csonka LN, et al. (2001). Osmosensing and osmoregulatory compatible solute accumulation by bacteria. Comp Biochem Physiol Part A 130:437–60. Wood M. (1995). A mechanism of aluminum toxicity to soil bacteria and possible ecological implications. Plant Soil 171:63–9. Xu Y, Lebedan B, Glansdorff N. (2007). Surprising arginine biosynthesis: a reapprasial of the enzymology and evolution of the pathway in microorganisms. Microbiol Mol Biol Rev 71:36–47. Yadav AS. (2007). Auxotrophy in rhizobia revisited. Ind J Microbiol 47: 279–88. Yadav AS, Vashishat RK, Kuykendall LD, Hashem FM. (1998). Biochemical and symbiotic properties of histidine-requiring mutants of Rhizobium leguminosarum biovar trifolii. Lett Appl Microbiol 26: 22–6. Yap SF, Lim ST. (1983). Response of Rhizobium sp. UMKL 20 to sodium chloride stress. Arch Microbiol 135:224–8. Yaryura PM, Leo´n M, Correa OS, et al. (2008). Assessment of the role of chemotaxis and biofilm formation as requirements for colonization of

DOI: 10.3109/1040841X.2013.856854

Downloaded by [UNAM Ciudad Universitaria] at 07:26 18 January 2016

roots and seeds of soybean plants by Bacillus amyloliquefaciens BNM339. Curr Microbiol 56:625–32. Young JP, Crossman LC, Johnston AWB, et al. (2006). The genome of Rhizobium leguminosarum has recognizable core and accessory elements. Genome Biol 7:R34. Yurgel SN, Kahn ML. (2008). A mutant GlnD nitrogen sensor protein leads to a nitrogen-fixing but ineffective Sinorhizobium meliloti symbiosis with alfalfa. Proc Natl Acad Sci USA 105: 18958–63. Yurgel SN, Rice J, Mulder M, Kahn ML. (2010). GlnB/GlnK PII proteins and regulation of the Sinorhizobium meliloti Rm1021 nitrogen stress response and symbiotic function. J Bacteriol 192:2473–81.

Microsymbiont amino acid metabolism

451

Yurgel SN, Rice J, Kahn ML. (2012). Nitrogen metabolism in Sinorhizobium meliloti-alfalfa symbiosis: dissecting the role of GlnD and PII proteins. Mol Plant-Microbe Interact 25:355–62. Yurgel SN, Rice J, Kahn ML. (2013). Transcriptome analysis of the role of GlnD/GlnBK in nitrogen stress adaptation by Sinorhizobium meliloti Rm1021. Plos One 8:e58028. Zhao H, Li M, Fang K, et al. (2012). In silico insights into the symbiotic nitrogen fixation in Sinorhizobium meliloti via metabolic reconstruction. Plos One 7:e31287. Zheng L, White RH, Cash VL, et al. (1993). Cysteine desulfurase activity indicates a role for NIFS in metallocluster biosynthesis. Proc Natl Acad Sci USA 90:2754–8.