Author's personal copy REGULAR ARTICLE

REGULAR ARTICLE

The hisC1 gene, encoding aromatic amino acidaminotransferase-1 in Azospirillum brasilense Sp7,expressed in wheat

Julio Castro-Guerrero & Angelica Romero & José J. Aguilar & Ma. Luisa Xiqui &Jesús O. Sandoval & Beatriz E. Baca

Received: 9 February 2011 /Accepted: 25 September 2011 /Published online: 30 November 2011# Springer Science+Business Media B.V. 2011

AbstractBackground and aims Production of indole-3-aceticacid (IAA) by Azospirillum brasilense is one of themost important mechanisms underlying the beneficialeffects observed in plants after inoculation with thisbacterium. This study determined the contribution ofthe hisC1 gene, which encodes aromatic amino acidaminotransferase-1 (AAT1), to IAA production, andanalyzed its expression in the free-living state and inassociation with the roots of wheat.Methods We determined production of IAA and AATactivity in the mutant hisC::gusA-smR. To study theexpression of hisC1, a chromosomal gene fusion wasanalyzed by following β-glucuronidase (GUS) activityin vitro, in the presence of root exudates, and inassociation with roots.Results IAA production in the hisC mutant was notreduced significantly compared to the activity of thewild-type strain. AAT1 activity was reduced by 50%when tyrosine was used as the amino acid donor,

whereas there was a 30% reduction when tryptophanwas used, compared to the activity of the wild-typestrain. Expression of the fusion protein was up-regulated in both logarithmic and stationary phasesby several compounds, including IAA, tryptophan,tyrosine, and phenyl acetic acid. We observed theexpression of hisC1 in bacteria associated with wheatroots. Root exudates of wheat and maize were able tostimulate hisC1 expression.Conclusions The expression data indicate that hisC1is under a positive feedback control in the presence ofroot exudates and on plants, suggesting that AAT1activity plays a role in Azospirillum–plant interactions.

Keywords Azospirillum brasilense . Aromatic aminoacid aminotransferase-1 . Indole acetic production .

Radical exudates regulation

AbbreviationsIAA Indole-3-acetic acidAAT1 Aromatic amino acid aminotransferase-1IPyA Indole-3-pyruvic acidIAdhl Indole-3-acetaldehydeTrp TryptophanTyr TyrosinePhe PhenylalaninepHPP p-HydroxyphenylpyruvatePAA Phenyl acetic acidHPLC High-performance liquid chromatography

Plant Soil (2012) 356:139–150DOI 10.1007/s11104-011-1009-2

Responsible Editor: Euan K. James.

J. Castro-Guerrero :A. Romero : J. J. Aguilar :M. L. Xiqui : J. O. Sandoval :B. E. Baca (*)Centro de Investigaciones en Ciencias Microbiológicas,Instituto de Ciencias,Benemérita Universidad Autónoma de Puebla,Av. San Claudio s/n, Ciudad Universitaria,72570 Puebla, Mexicoe-mail: [email protected]

Author's personal copy

Introduction

The amino acid aminotransferase-1 (AAT1) proteinfrom Azospirillum brasilense Sp7 is encoded byhisC1. AAT1 belongs to the α subfamily of FamilyI aminotransferases (Jensen and Gu 1996). Theseaminotransferases participate in several chemicalreactions of cellular metabolism, aromatic amino acidbiosynthesis, and function in the catabolism of aminoacids to provide nitrogen and carbon sources forgrowth (Sergeeva et al. 2007). Individual subfamilymembers may be relatively specific for aromaticamino acids, but exhibit broad specificity. In additionto using aromatic amino acids as substrates, AAT1from A. brasilense is involved in the catabolism ofhistidine, and produces indol-3-pyruvic acid (IPyA)from tryptophan (Trp) and phenyl pyruvic acid fromphenylalanine. These compounds are precursors ofindole acetic acid (IAA) and phenyl acetic acid(PAA), respectively (Baca et al. 1994; Soto-Urzúa etal. 1996; Somers et al. 2005).

Azospirillum belongs to a group of soil bacteriaoften referred to as plant-growth promoting rhizobac-teria (PGPR; Bashan et al. 2004). Several studies havedescribed their beneficial effects on several plants:cereals; Compositae, such as sunflower; Cruciferae,such as mustard, vegetable crops; legumes, such aschick pea; clover, and alfalfa (Okon and Labandera-González 1994; Bashan et al. 2004). Several mecha-nisms of plant–microbe interactions can affect plantgrowth, including N2 fixation, hormonal effects,production of nitric oxide (NO), and enhancedmineral uptake (Bashan et al. 1990, 2004; Creus etal. 2005), all of which contribute to the beneficialeffects observed in plants (Bashan and de-Bashan2010).

Azospirillum can synthesize four classes of phyto-hormones; auxins, gibberellins, cytokinins, and absci-sic acid (Tien et al. 1979; Carreño-López et al. 2000;Cohen et al. 2008; Croizier et al. 1988; Martínez-Morales et al. 2003; Piccoli et al. 1996). Production ofauxins is considered the most important (Spaepen etal. 2008). The biosynthesis of IAA involves distinctpathways: three pathways requiring Trp have beenproposed. The main pathway in Azospirillum involvesthe intermediary (IPyA), [Trp IPyA IAdhl

IAA] (Costacurta et al. 1994; Carreño-Lópezet al. 2000; Spaepen et al. 2008). The first reactionof this pathway is transamination of Trp to IPyA. The

enzymatic activity of two AATs is found in several A.brasilense strains (Pedraza et al. 2004) and four A.lipoferum strains (Ruckdäschel et al. 1988). Thesecond step is catalysis by the key enzyme phenylpyruvate decarboxylase (PPDC; Spaepen et al.2007a, b) to produce IAdlh, which is further oxidizedto IAA.

The present study identified the hisC1 gene fromA. brasilense Sp7 that encodes AAT1, and examinedits contribution to biosynthesis of IAA, as well as itsexpression in both the free-living state and inassociation with the plant.

Materials and methods

Bacterial strains and growth conditions

Escherichia coli strains were grown overnight at 37°Cin Luria broth (LB, Sambrook et al. 1989) containingantibiotics: 25 μg/mL kanamycin, 100 μg/mL ampi-cillin, and 75 μg/mL streptomycin (Sm). Azospirillumstrains were grown in minimal K-lactate medium(Carreño-López et al. 2000), overnight at 30°C. Themating and isolation of the transcriptional fusion wasperformed as described by Carreño-López et al.(2000). E. coli S17.1 (pAB144-4A) or E. coli S17.1(pAB144-4B) strains (Table 1), with the cassette smR-gusA or A. brasilense Sp7 were used as the donor andacceptor strain, respectively. For selection of trans-conjugants, plates (minimal K-lactate medium with150 μg/mL Sm) were incubated at 30°C for 2 days.

Construction of plasmid clones and DNA sequencing

Plasmid DNA isolation, restriction analysis, transforma-tion and molecular cloning were performed by methodsdescribed in Sambrook et al. (1989). For Southernhybridization experiments, genomic, plasmid, and cos-mid DNA were digested with BglII, EcoRI, KpnI,EcoRV, or SalI and were blotted on Hybond N+ nylonmembranes (Amersham Biosciences, Piscataway, NJ)with a vacuum blotter (Fisher Scientific, Waltham,MA). DNA was fixed by exposure to a 312-nmtransilluminator for 4 min. Prehybridization, labeling,and purification of probes, as well as hybridization,were performed as described by Carreño-López et al.(2000). The physical maps of the inserts for this work

140 Plant Soil (2012) 356:139–150


are shown in Fig. 1. Both DNA strands weresequenced by Macrogen (Rockville, MD).

To identify the hisC1 gene, we used primers Faat1(5′ GCTGACGGTCAATCCGCTGG) and Raat1 (5′CCGCAGGACCATCTCCGACGG) designed fromthe sequence of peptides LSLTVNPLGP, and YIP-SEMVLR, which were obtained from AAT1 protein

purified to homogeneity as described by Soto-Urzúaet al. (1996).The amino acid sequences of the twopeptides were determined at the Stanford UniversityProtein and Nucleic Acid Facility (Stanford, PaloAlto, CA). PCR amplification was performed accord-ing to directions of the manufacturer of Taq polymer-ase (Invitrogen, Carlsbad, CA), in a final volume of

Table 1 Bacterial strains and plasmids used in this study

Strains/plasmids Description Reference or source

Strain

Escherichiacoli DH5α

endA1 hsdR17 supE44 thi-1 λ- recA1 gyrA96 relA1 ΔlacU169 Ø80(ΔlacZΔM15)

Invitrogen, Carlsbad, CA

E. coli S17.1 pro, thi hsd, recA, :: RP4-2 Tc:: Mu- Km Tn7 Simon et al. 1983

Azospirillumbrasilense Sp7

Wild-type strain ATCC29729

A. brasilense7445 Derivative of Sp7 hisC1:: gusA- smR This study

A. brasilense7446 Derivative of Sp7 hisC1::smR-gusA This study

Plasmids

pSUP202 Suicide vector ApR, CmR, TcR Simon et al. 1983

SuperCos 1 Cloning vector ApR, KmR, Stratagene, La Jolla, CA

pCR 2.1 Cloning vector Invitrogen

pBSL197 Vector pBSL carrying the cassette smR-gusA

pABaat1 Derivative of pCR2.1 carrying a fragment of 432 bp of hisC1 gene This study

pAB144 Cosmid from A. brasilense Sp7 library, carrying a fragment of20 kb harboring the hisC1 gene

This study

pAB144-1 Derivative of pAB144 carrying a 13 kb EcoRI fragment This study

pAB144-2 Derivative of pSUP202 carrying a 4.2 kb BglII fragment of pAB1444-2 This study

pAB144-3 Derivative of pAB144-1 carrying a 4.2 kb BglII fragment, cloned in pCR2.1 This study

pAB144-4A Derivative of pAB144-2 carrying the cassette hisC1:: gusA- smRinserted in the 5’ 3’ direction

This study

pAB144-4B Derivative of pAB144-2 carrying the cassette hisC1::smR-gusA This study

E C Sm Bg S K Bg K E

probe hisC1 pABaat1

pAB144-1

pAB144-3

V EV Bg S K Bg S V

ORF1 ORF2 AAT1 ORF31kb

smR gusA

gusA smR

S

S

S

SS

SpAB144-4A; 7445

pAB144-4B; 7446

Fig. 1 Schematic map of the sequenced hisC1 region in the plasmid, strains and constructs. The insertion of the smR-gusA cassette isshown, and arrows indicate the direction of transcription. E EcoRI, C ClaI , Sm SmaI, Bg BglI, S SalI, K KpnI, EV EcoRV, V Vector

Plant Soil (2012) 356:139–150 141


25 μl, using 50 ng genomic A. brasilense Sp7 DNA.The amplification cycle consisted of 10 min at 96°C,30 cycles of 1 min at 95°C, 1.5 min at annealingtemperature of 67°C, 1 min at 72°C, and a final 10-min extension at 72°C. The amplicon of 473 bp wascloned into the pCR2.1 vector (Invitrogen) as de-scribed by the manufacturer to obtain the plasmidpABaat1 (Table 1). This amplicon was used as aprobe to screen an A. brasilense Sp7 library. Thecosmid pAB144 was identified by DNA/DNA hy-bridization. pAB144-1 was obtained by digestion ofpAB144 with EcoRI, hybridization with the probeaat1, and then subcloning of a 13-kb fragment into theSupercos vector (Invitrogen). The cosmid pAB144-1was digested with enzymes Cla I, Kpn I, Sma I, orBglI after hybridization to probe aat1. A 4.7-kb BglIIfragment was subcloned into pCR2.1, which wasdigested with the enzyme BamHI and transformedinto E. coli DH5α. This plasmid was designatedpAB144-2. Both strands were sequenced at Macrogento obtain the full sequence of the hisC1 gene. TheGenBank accession number is JF509704.1.

Construction of trancriptional fusion and mutationof the hisC1 gene

The same 4.7 bp BglII fragment from pAB144-1 wasalso subcloned into the suicide vector pSUP202digested with BamHI. The ligated fragments weretransformed into E. coli S17.1 to generate the plasmidpAB144-3. This plasmid was digested with KpnI andthe gusA-smR cassette was obtained from plasmidpSBL197 by digestion with KpnI (Alexeyev et al.1995). The fragments were ligated and transformedinto E. coli S17.1. Both cloned directions wereselected (see Table 1). E. coli strains S17.1(pAB144-4A) and S17.1 (pAB144-4B) were used asdonors and transferred to A. brasilense Sp7 to obtaintranscriptional fusions (Fig. 1). The correct orientation(sense strand of gusA) was determined by Southernblot analysis as described by Carreño-López et al.(2000), using aat1 and the gusA-smR cassette asprobes.

IAA production and AAT1 enzyme assays

IAA production was determined from bacteria grown in10 mL K-malate minimal medium supplemented with100 μg/mLTrp at 30°C for 48 h with shaking at 120 rpm.

The cells were then pelleted by centrifugation(10,000 g, 5 min, at 20°C). Aliquots of 20 μlsupernatant were taken, and the IAA concentra-tion was determined by HPLC as described byCarreño-López et al. (2000). Three replicates wereperformed, and the experiment was repeated twice.Total AAT enzyme activity was determined from the cellpellets, which was used to produce a cell-free extractdescribed by Pedraza et al. (2004). Specific activity wasdefined as micromoles of product [(IPyA and p-hydroxyphenyl pyruvic acid (pHPP)] per minute permilligram of protein described by Diamondstone(1966).

Preparation of wheat and maize root exudates

To prepare root exudates, seed from wheat (Triticumaestivum var Zacatecas) and maize (Zea mays var.Criollo) were surface-sterilized by gentle shaking for1 min in 0.2% HgCl2 with 2% Tween 20 (P1379,Sigma). The sterilized seeds were soaked six times for30 min in sterile, demineralized water. The seeds wereallowed to germinate on Petri plates containing agar-water under dark conditions at 20°C for 2 days.Seedlings were transferred aseptically to Magentavessels (Sigma-Aldrich, St. Louis, MO) equippedwith a stainless steel perforated tray and filled with150 mL plant nutrient solution (Hoffland 1992).Germinated wheat seedlings (65) and germinatedmaize seedlings (25), (corresponding to 30 g) wereplaced on the tray with the root in the solution, andthen placed in a cycle of 16-h light at 24°C and 8-h ofdark at 16°C for 7 and 10 days.

Root exudates were recovered aseptically andchecked for bacterial contamination using LB mediumbefore sterile filtration through a 0.45-μm disposablecellulose-nitrate filter (EMD Millipore, Bedford, MA)and then stored at −20°C until use. Only materials fromexudates in which no microbial growth could bedetected were used.

Reporter gene expression in minimal mediumand root exudates in the free living-state

The activity of β-glucuronidase (GUS) and OD600 nmof A. brasilense 7445 and A. brasilense 7446 strainswere determined in K-lactate minimal medium con-taining: Trp, Phe, pHPP, phenyl-pyruvic acid (PPA),IAA, or PAA, each at a concentration of 0.5 mM. The

142 Plant Soil (2012) 356:139–150


pH of all media was maintained at 6.8. The strainswere grown in K-lactate medium overnight at 30°Cwith shaking, then inoculated at 0.01 OD600nm in a125-mL Erlenmeyer flask containing 15 mL of eachtest medium in individual flasks and incubated at 30°Cwith shaking at 200 rpm. At various points during theexponential and stationary growth phases, 2 mLculture per flask were centrifuged and resuspendedin K-lactate medium (for OD600nm measurements)and frozen at −20°C for later GUS activity determi-nations. GUS activity was measured by the method ofMiller (1972) using PNPG (para-nitrophenyl-β-D-glucuronide, Sigma) as substrate. Enzyme activitieswere expressed as Ea=[A420(OD600nm× t)]×1,000,where Ea is GUS activity in Miller units and t isincubation time (in minutes). When required, the totalprotein was determined with a protein assay reagent(Bio-Rad, Hercules, CA) using serum bovine albuminas the standard (Sigma). Erlenmeyer flasks containing15 mL root exudates diluted to 50% with a solutiondescribed by Hoffland (1992) were inoculated asdescribed above and incubated at 30°C with shakinguntil reaching 0.3 or 0.4 OD600nm. Cells were thencentrifuged and GUS activity determined as indicatedabove.

Plant cultivation and in situ detection of bacteriacolonizing the root surface

Wheat seeds were germinated as described above, andgrown hydroponically under sterile conditions asdescribed by Zeman et al. (1992). One-week-oldwheat plantlets were inoculated with 1 mL A.brasilense Sp7 or A. brasilense strains 7445 or 7446at approximately 108 bacterial cells mL−1 in hydro-ponic culture as described by Arsène et al. (1994).Five days after inoculation, root material was weighedand homogenized in a mortar with 2 mL sterilephosphate buffer (20 mM at pH 7). Activity of thereporter gene was measure in cells from 1 mL of thehomogenized material. As soon as bacteria wereisolated from root wheat, GUS activity was deter-mined as described by Miller (1972). The coloniza-tion pattern of the root surfaces was estimated by thenumber of Azospirillum cells present in 1 mL ho-mogenized material from wheat roots by the most-probable number (MPN) technique. Cell counts for A.brasilense 7445 and 7446 were made in K lactatemedium supplemented with Sm (150 μg/mL). In situ

detection of GUS activity in 2-cm root segmentsinoculated with Azospirillum 7445 and 7446 strainswas performed as described by Arsène et al. (1994).Afterwards, the segments were stained (XGlc; 5-bromo-4-chloro-indolyl-D-glucopyranoside, Sigma).After staining, the segments were examined under alight microscope (Nikon-2000), and sections wereviewed and photographed using the Nikon system.

Results

Identification and mutagenesis of hisC1

We isolated the hisC1 gene using a PCR-generatedprobe with oligonucleotides designed from peptidesobtained from purified and digested AAT1 protein, asdescribed in Materials and methods. The Sp7strainlibrary was screened and a clone including the hisC1gene was obtained. Figure 1 describes this region.The deduced protein product of ORF1 shares homol-ogy with an acetyl transferase of the GNAT familyfrom Rhospirillum centenum SW (YP_002297377.1)with 55% similarity, Roseobacter SpAzwK-3b(ZP_01903442) with 60% similarity, and Azospiril-lum B510 (AZL012950), with 88% similarity. ORF2appears to be a hypothetical membrane proteinshowing homology to the deduced protein productsfrom Aurantimonas manganoxidans SI85-A1(ZP_01228200.1), with 61% similarity and Rhizobi-um etli CIAT with 56% similarity. The hisC1 gene, ina convergent orientation, displayed a deduced proteinsequence with high similarity to the histidol-aromaticamino acid aminotransferases of various bacteria, inparticular the gene hisC (AZL012940) coding for ahistidol-phosphate aminotransferase, putatively iden-tified in the recent genome sequence of Azospirillumspp. IB510 (http://genome.kazusa.or.jp/rhizobase).The predicted product from Azospirillum spp. IB510shares 85% similarity with hisC1 from A. brasilenseSp7. The sequence of AAT1 indicated a protein of360 amino acids with a predicted molecular mass of39,248 Da; the pyridoxal 5′ phosphate binding(FLANPNNPTGTY) was also identified. ORF3 cor-responded to a partial sequence of a histidine kinase,which is predicted to be a signaling protein. Theconservation of gene arrangement in the geneticregion identified in A. brasilense Sp7 matched thecorresponding region in Azospirillum spp B510; in

Plant Soil (2012) 356:139–150 143


http://genome.kazusa.or.jp/rhizobase

the latter strain, this region is located on the bacterialchromosome (Kaneko et al. 2010).

To analyze the contribution of AAT1 to IAAbiosynthesis, we constructed an AAT1 mutant byinserting a gusA-smrR cassette in the natural KpnI sitefound in the hisC1 sequence. The correct cassetteorientation relative to the promoter was created instrain 7445, and strain 7446 was generated byinserting the cassette in the opposite orientation.Growth of strains Sp7, 7445, and 7446 was similar(data not shown). IAA production in 7445 was lowercompared with wild type Sp7, but the difference was notsignificant, as expected: 11.3±1.2 IAA μg/mg protein−1

for the wild type strain, and 8.9±1.1 IAA μg/mgprotein−1 for A. brasilense 7445. However, total AATactivity was lower in strain 7445 than in the Sp7 strain(Fig. 2). When Trp was used as the amino acid donor,the reduction was 30% (0.023 μmol/mg protein−1)compared with the wild type (0.039 μmol/mgprotein−1). When Tyr was used as the amino aciddonor, the reduction was 50% (0.034 to 0.016 μmol/mg

protein−1). The results indicate a role for AAT1 inIPyA intermediate production.

Chromosomal gene fusion expression in the freeliving state

To measure the influence of individual compounds onexpression of the hisC1 gene, GUS activity of thereporter gene cassette inserted into A. brasilensestrains 7445 and 7446 was tested. The cassetteinsertion was confirmed by Southern blot analysison A. brasilense 7445 and 7446 genomic DNAprobed with the gusA-smR and aat1 DNA cassettes(data not shown). The resulting chromosomal fusionconferred the ability to produce blue colonies onmedium containing the GUS indicator X-Glc; thecorrect orientation for expression was present only inthe case of A. brasilense 7445. To check expression ofthe reporter fusion, GUS activity was determined inboth permeabilized A. brasilense 7445 and 7446 cells.As was expected, only A. brasilense 7445 displayedactivity. To further study gene expression, A. brasi-lense 7445 was grown in minimal medium (K-lactate)and K-lactate medium supplemented with Tyr, Trp,Phe, pHPP, IAA, or PAA. GUS activity was deter-mined as described in Materials and methods.Expression of hisC1 was constitutive and was seenin both exponential and stationary growth phases(Fig. 3). In medium supplemented with IAA, PPA,Trp or Tyr, expression was significantly up-regulatedduring both phases. Up-regulation was 1.8-fold forTrp, 2.6-fold for Tyr, 2.5-fold for PAA, and 4.6-fold for IAA in early stationary phase comparedwith expression observed in K-lactate medium.These data indicate that the compound that bestfunctioned as a regulator was IAA, followed byPAA, Tyr, and Trp.

In view of this, we tested the expression of strain7445 grown in radical exudates derived from wheat andmaize. We found that both exudates drive the transcrip-tion of the hisC1 promoter, with maize root exudatesshowing greater enhancement than wheat exudates(Fig. 4). Maximal bacteria growth was detected withmaize root exudates at 10 days (107 CFU/mL),compared with wheat root exudates (106 CFU/mL).Currently, we do not know the concentrations ofcarbon and nitrogen sources or the nature of theexogenous inducer(s). Additionally, Hoffland solutionwas inoculated with strain 7445, and wheat exudates

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

Spec

ific

Act

ivit

y A

AT

1

1 2

Fig. 2 Aminotransferase-1 (AAT) specific activity. 1 Trypto-phan (Trp) AAT activity was assayed by measuring indole-3-pyruvic acid (IPyA) at 330 nm (ε=10,000 cm2/mol, Grannerand Tomkins 1970), at pH 8.5 with pyridoxal 5′ phosphate(1 mM), 2-oxoglutarate (1 mM), and Trp (1 mM). 2 Tyrosine(Tyr) ATT was measuring as the concentration of p-hydrox-yphenylpyruvate (pHPP) at 331 nm (ε=19,900 cm2/mol,Granner and Tomkins 1970), at pH 8.5 with pyridoxal 5′phosphate (1 mM), 2-oxoglutarate (1 mM), and Tyr (1 mM).Each bar represents the mean of three independent replicates,error bars represent the standard error of the mean, black barsSp7 strain, grey bars 7445 strain

144 Plant Soil (2012) 356:139–150


were inoculated with strain 7446 strain: no GUSactivity was detected in either case (1.7±0.15 MU).

Chromosomal gene fusion expression in associationwith plants

Expression of the reporter gene in a hydroponic plantwas also determined using the GUS assay and MPN.The results shown in Fig. 5 indicate that activity wasmore significant in A. brasilense 7445 (155.3±43.5MU) than in A. brasilense 7446 (3.9±0.15 MU), asexpected from the direction of cassette insertion(Fig. 5a). The CFU of both strains was set at 107.8

to 108.01 CFU/g root, suggesting that growth of theintroduced bacteria occurred in both strains to thesame extent (Fig. 5b). To monitor establishment ofbacteria expressing the hisC gene on wheat roots,

GUS expression of bacteria on roots was determinedat 5 days after inoculation. When plants wereinoculated with A. brasilense 7445, hisC expressionwas detected on roots, whereas no GUS staining wasobserved in roots inoculated with A. brasilense 7446(Fig. 5c).

Discussion

Amino acids in the rhizosphere are of both plant andmicrobial origin (González-López et al. 2005). Therole of amino acids produced by rhizobacteria duringtheir interaction with plants is practically unknown,although it has been established that some aminoacids, such as Met and Trp, act in soil as the mainprecursors for synthesis of the plant hormones

0

5000

10000

15000

20000

25000

KL KL + Trp KL + Tyr KL +PAA KL +IAAM

iller

Uni

ts/m

g pr

otei

n-1

Fig. 3 β-glucuronidase(GUS) activity of the hisC1::gusA fusion in the free-living state. Bacteria weregrown to both exponentialand stationary phase in K-lactate (KL) minimal mediumsupplemented with 0.5 mMTrp, Tyr, IAA and PPA or intheir absence. GUS activitywas determined at 12 h(white bars), 24 h (greybars), and 36 h (black bars).The data shown are aver-ages of three independentreplicates; error bars stan-dard error of the mean

0

2000

4000

6000

8000

10000

12000

14000

0.3 0.4 0.3 0.4

7 Days 10 Days

Mill

erU

nits

/ mg

pro

tein

-1

OD600nm

Fig. 4 GUS activity of thehisC1::gusA fusion in thepresence of exudates fromwheat and maize in the free-living state. GUS activitywas determined in cells atOD600 = 0.3 or 0.4. Greybars Wheat exudates (har-vested at 7 and 10 days),black bars maize exudates(harvested at 7 and10 days). The data shownare averages of three inde-pendent replicates; errorbars standard error of themean

Plant Soil (2012) 356:139–150 145


ethylene and IAA, respectively (Frankenberger andArshad 1995; Murcia et al. 1997; Kamilova et al.2006). Biosynthesis of phytohormones by Azospirillumis documented widely, as is involvement of thesesubstances in plant responses observed after inoculationof plants with bacteria (Dobbelaere et al. 2001; Spaepenet al. 2008). Here, we identified and studied theregulation of the hisC1 gene from A. brasilense Sp7

strain coding for AAT1, which we had previouslypurified (Soto-Urzúa et al. 1996). We had alsodemonstrated its substrate specificity and found thatthe enzyme can convert Trp to IPyA, pHPP to Tyr, andphenyl pyruvate to Phe as reversible reactions of AATs(Carreño-López et al. 2000; Soto-Urzúa et al. 1996).

A knock-out mutant of the hisC1 gene showed asignificant decrease in specific AAT activity (∼50%).

0

20

40

60

80

100

120

140

160

180

A.brasilense

7445

A.brasilense

Mill

erun

its/

mg

prot

ein-1

/g R

oot

7446

A

B

A. brasilense 7445

A. brasilense 7446

C

i ii

iii iv

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

A.brasilense 7445 A.brasilense 7446

Log

MP

N/g

Roo

t

Fig. 5a–c Colonization and expression of A. brasilense strains7445 and 7446 in association with wheat. After 5 days of plantgrowth, root tips were isolated; bacteria were isolated from theroot tip and then plated for determination of MPN and GUSactivity as described in Materials and methods. a hisC1promoter expression: black bar 7445 strain, grey bar 7446strain. Values represent the average of 15 plants for each

experiment; error bars standard error of the mean. b Mostprobable number (MPN) determination as log CFU/mL: blackbar 7445 strain, grey bar 7446 strain. c Bacterial hisC::gusAexpression on wheat root. In situ detection of GUS activity ofwheat root segments, inoculated with A. brasilense strains 7445or 7446 strains. Bars: i, iii 100 μm; ii, iv 5 μm

146 Plant Soil (2012) 356:139–150


However, IAA production was not as low asexpected, which could happen if other AATs isoformsthat contribute to IAA production are present inAzospirillum (Pedraza et al. 2004). The lack of asignificant effect on IAA production of loss of hisC1also suggests that AAT does not catalyze the rate-limiting step in the IAA biosynthesis pathway. Ge etal. (2009) identified and mutated the atrC gene fromA. brasilense Y62, demonstrating that atrC wasinvolved in IAA production; moreover, a recombinantAtrC protein expressed and purified in E. colifunctioned as an aminotransferase with high substratespecificity for Trp. Trp aminotransferase activity wasalso found to be associated with IAA production inEnterobacter cloacae (Koga et al. 1994), and inArabidopsis thaliana (Tao et al. 2008). In Pantoeaagglomerans rhizosphere isolates, AAT activity wasrelated to the amount of IAA produced in thepresence of Trp, but not when Tyr or Phe was usedas a sole nitrogen source; all isolates accumulated Trpinternally and released IAA (Sergeeva et al. 2007).

GUS activity was compared in the presence of twoauxins, IAA and PAA, and amino acids Trp and Tyr,both of which are AAT1 substrates. hisC1, whichencodes AAT1, catalyzes the first step in theconversion of Trp to IPyA, Tyr to pHPP and Phe toPPA; it is involved in Tyr biosynthesis (by the reversereaction, pHPP to Tyr), and is up-regulated by PPAand Tyr in addition to IAA. PPA and Tyr arethemselves products of the pathway. The resultsindicate that IAA was the best inducer, followed byPAA, and suggest a mechanism of feedback control.Another positive feedback mechanism has beendescribed for A. brasilense Sp245 by Vande Broeket al. (1999). The expression of the key gene, ipdC,coding for the enzyme PPDC, showed increasedexpression in the presence of Trp at its maximumduring stationary phase, as this study also demon-strated for hisC1. Also, the ipdC gene from A.brasilense strains Sp7 and SM is upregulated bydifferent auxins such as IAA and indole-3-butyricacid (Rothballer et al. 2005; Malhotra and Srivastava2008). In addition, the ipdC gene from Enterobactercloacae strain UW5 showed upregulation by TyR, Trpand Phe (Ryu and Patten 2008). Despite the fact that,in A. brasilense, the addition of Trp to the culturemedium strongly increased IAA production, Trp didnot upregulate ipdC expression (Vande Broek et al.1999; Carreño-López et al. 2000). Unlike ipdC

expression, hisC1 is upregulated by Trp, albeit to alesser extent than with the other compounds tested.This upregulation could be due to the binding of thesecompounds to a transcriptional regulator at the hisC1promoter. Such transcriptional regulation of IAAbiosynthesis has been described in E. cloacae UW5acting on the ipdC promoter (Ryu and Patten 2008).Our data confirm the additional critical role of IAA ingene expression. IAA has a profound effect onbacterial metabolism, as a signaling molecule inbacteria acting during microorganism–plant interac-tions (Yang et al. 2007; Spaepen et al. 2007a, b).

The plant root is thought to be the major source ofnutrients for microorganisms living in the rhizosphere(Lynch and Whipps 1990). Therefore, bacterialgrowth on compounds present in the rhizosphere isassumed to be essential for efficient colonization andestablishment in the rhizosphere. In addition tomeasuring its expression in the free-living state, wealso determined regulation of hisC using maize andwheat root exudates as growth media. Both were ableto support growth, but the highest expression wasobserved in cells growing in root maize exudates. It iswell established that exudates from different plantspecies provide different compounds acting as carbonand nitrogen sources for bacteria associated withplants. Kamilova et al. (2006) found that the totalamount of organic acids is much higher than that oftotal sugar or Trp in the exudates of several plants.However, the concentration differs with plant speciesand plant age at exudate collection, as observed here.Reporter gene induction in vitro by maize and wheatexudates for genes involved in proline catabolism hasbeen described in Pseudomonas fluorescens andPseudomonas putida (van Overbeek and van Elsas1995). Induction of the put promoter by maizeexudates collected at 7 days was approximately 20-fold (Vilchez et al. 2000).

The extent of hisC1 expression was also deter-mined in association with wheat. Bacteria isolatedfrom plants after 5 days of growth showed asubstantial difference between the negative control,A. brasilense 7446, and A. brasilense 7445, whichcarried the reporter gene fusion in the correctorientation. Additionally we tested hisC::gusA ex-pression in situ in wheat roots. In agreement with dataobtained from bacteria isolated from roots, significantexpression of hisC was detected in plants after 5 days.The tip and emergence point of lateral roots were

Plant Soil (2012) 356:139–150 147


favored sites of bacterial microcolonies expressing thereporter gene, as also described by others (Arsène etal. 1994; Vande Broek et al. 1993).

Using global analysis of P. putida genes expressedduring their interaction with the maize rhizosphere,Matilla et al. (2007) found 90 rhizosphere-upregulated genes: genes coding for an aminotrans-ferase (PP3786), an aldhehyde dehydrogenase familyprotein (PP3281), and phenyl acetic degradationprotein (PP2694), as well as genes coding for aminoacid transport and metabolism of aromatic com-pounds. Some of the bacteria were analyzed incompetitive colonization, and mutants were foundsthat were hampered for survival in the rhizosphere incompetition with wild type. In our work, both hisC1strains (7445 and 7446) tested were indistinguishablefrom each other and from the wild type strain forcolonization of wheat under laboratory conditions(data not shown). Furthermore, these mutants alsobehaved like their parent in this study; A. brasilense7445 and 7446 were not auxotrophic, as shown bygrowth in K-lactate minimal medium with NH4Cl asthe sole source of nitrogen. On the other hand, bothstrains were able to use Tyr as sole source of nitrogen.Uncharacterized genes coding for AATs with roles indetermining the competitive ability of A. brasilense inthe rhizosphere may explain why no difference wasobserved in colonization, although no such geneconferring a host-specific effect has yet been identified.

Knowledge of bacterial growth conditions in therhizosphere is important for understanding rhizo-sphere colonization. The composition of root exudateswas not analyzed for the presence of carbon sourcesand amino acids. However, the presence of com-pounds such as dicarboxylic acids and Trp, Tyr andPhe have been described in the exudates of wheat andmaize, suggesting the availability of these particularnutrients (Phillips et al. 2004). Future work shouldaim to identify the regulatory protein that controls thehisC gene at the transcriptional level, as well asdetermine the amino acid, IAA, and PAA compositionof axenically collected wheat (var Zacatecas) andmaize (var. Criollo) root exudates.

Acknowledgments This work was supported by a ConsejoNacional de Ciencia y Tecnología (CONACyT) Grant 49227-Z,together with Vicerrectoría de investigación y estudios deposgrado-Secretaría de educación pública (VIEP-SEP), GrantNAT-IC-1. J.G.C. was the recipient of a scholarship fromCONACyT.

References

Alexeyev MF, Shokolenko IN, Croughan TP (1995) Improvedantibiotic-resistance gene cassettes and omega elementsfor Escherichia coli vector construction and in vitrodeletion/insertion mutagenesis. Gene 160:63–67

Arsène F, Katupitiya S, Kennedy IR, Elmerich C (1994) Use oflacZ fusions to study the expression of nif genes ofAzospirillum brasilense in association with plants. MolPlant Microbe Interact 7:748–757

Baca EB, Soto-Urzúa L, Xochihua YG, Cuervo A (1994)Characterization of aromatic amino acid aminotransferasesand production of indolacetic acid in Azospirillum strains.Soil Biol Biochem 26:57–63

Bashan Y, de-Bashan L (2010) How the plant growth-promotingbacterium Azospirillum promotes plant growth—a criticalassessment. Adv Agron 108:77–136

Bashan Y, Harrison SK, Whitmoyer RE (1990) Enhanced growthof wheat and soybean plants inoculated with Azospirillumbrasilense is not necessarily due to general enhancementof mineral uptake. Appl Environ Microbiol 56:769–775

Bashan Y, Holguin G, de-Bashan LE (2004) Azospirillum–plantrelationships: physiological, molecular, agricultural, andenvironmental advances (1997–2003). Can J Microbiol50:521–577

Carreño López R, Campos Reales N, Elmerich C, Baca BE(2000) Physiological evidence for differently regulatedtryptophan-dependent pathways for indole-3-acetic acidsynthesis in Azospirillum brasilense Sp7. Mol Gen Genet264:521–530

Cohen AC, Bottini R, Piccoli PN (2008) Azospirillum brasi-lense Sp 245 produces ABA in chemically-defined culturemedium and increases ABA content in arabidopsis plants.Plant Growth Regul 54:97–103

Costacurta A, Keijers V, Vanderleyden J (1994) Molecularcloning and sequence analysis of an Azospirillum brasi-lense indole-3-pyruvate decarboxylase gene. Mol GenGenet 243:463–472

Creus CM, Graziano M, Casanovas EM, Pereyra MA,Simontacchi M, Puntarulo S, Barassi CA, Lamattina L(2005) Nitric oxide is involved in the Azospirillumbrasilense-induced lateral root formation in tomato. Planta221:297–303

Crozier A, Arruda P, Jasmin JM, Monterio AM,Sandberg G(1988) Analysis of indole-3 acetic acidand related indolesin culture medium and Azospirillumbrasilense. ApplEnviron Microbiol 54:2833–2837

Diamondstone TI (1966) Assay of tyrosine transaminaseactivity by conversion of p-hydroxyphenylpyruvate tohydroxybenzaldehyde. Anal Biochem 16:395–401

Dobbelaere S, Croonenborghs A, Thys A, Ptacek D, VanderleydenJ, Dutto P et al (2001) Responses of agronomically importantcrops to inoculation with Azospirillum. Aust J Plant Physiol28:871–879

Frankenberger WT, Arshad M (1995) Phytohormones in soil:microbial production and function. Dekker, New York

Ge SM, Li T, Chen SF (2009) Expression and functionalanalysis of aminotransferase involved in indole-3-aceticacid biosynthesis in Azospirillum brasilense Y62. Biochem(Moscow) 74:81–84

148 Plant Soil (2012) 356:139–150


González-López J, Rodelas B, Pozo C, Salmerón-López V,Martínez-Toledo MV, Salmerón V (2005) Liberation ofamino acids by heterotrophic nitrogen fixing bacteria.Amino Acids 28:363–367

Granner DK, Tomkins GM (1970) Tyrosine aminotransferase(rat liver). Methods Enzymol 17:633–637

Hoffland E (1992) Quantitative evaluation of the role of organicacid exudation in the mobilization of rock phosphate byrape. Plant Soil 140:279–289

Jensen RA, Gu W (1996) Evolutionary recruitment of bio-chemically specialized subdivisions of Family I within theprotein superfamily of aminotransferases. J Bacteriol178:2161–2171

Kamilova F, Kravchenko LV, Shaposhnikov AI, Azarova T,Makarova N, Lugtenberg B (2006) Organic acids, sugarsand L-tryptophane in exudates and vegetables growing onstonewool and their effects on activities of rhozospherebacteria. Mol Plant Microbe Interact 19:250–256

Kaneko T et al (2010) Complete genomic structure of thecultivated rice endophyte Azospirillum sp B510. DNA Res17:37–50. doi:10.1093/dnares/dsp026

Koga J, Syono K, Ichikawa T, Adachi T (1994) Involvement ofL-tryptophan in indole-3-acetic acid biosynthesis in Enter-obacter cloacae. Biochem Biophys Acta 1209:241–247

Lynch JM, Whipps JM (1990) Substrate flow in the rhizo-sphere. Plant Soil 129:1–10

Malhotra M, Srivastava S (2008) organization of the ipdCregion regulates IAA levels in different Azospirillumbrasilense strains: molecular and functional analysis ofipdC in strain SM. Environ Microbiol 10:1365–1373

Martínez-Morales LJ, Soto-Urzúa L, Baca BE, Sánchez-AhédoJA (2003) Indole-3-butyric acid (IBA) production inculture medium by wild-type Azospirillum brasilense.FEMS Microbiol Lett 228:167–173

Matilla MA, Espinosa-Urgel M, Rodríguez-Herva JJ, RamosJL, Ramos-González MI (2007) Genomic analysis revealsthe major driving forces of bacterial life in the rhizosphere.Genome Biol 8:R1791–13

Miller JH (1972) Experiments in molecular genetics. ColdSpring Harbor Laboratory Press, Cold Spring Harbor,NewYork

Murcia R, Rodelas B, Salmerón V, Pozo C, González-López J(1997) Effect of simazine on the production of lisine andmethionine by Azotobacter chroococcum and Azotobactervinelandii. Amino Acids 12:249–255

Okon Y, Labandera-González CA (1994) Agronomic applica-tionsof Azospirillum: an evaluation of 20 years worldwidefieldinoculation. Soil Biol Biochem 26:1591–1601

Van Overbeek LS, Van Elsas JD (1995) Root exudate-inducedpromoter activity in Pseudomonas fluorescensmutants inthe wheat rhizosphere. Appl Environ Microbiol 61:889–898

Pedraza RO, Ramírez Mata A, Xiqui Vázquez ML, Baca BE(2004) Aromatic amino acid aminotransferase activity andíndole-3-acetic acid production by associate nitrogen-fixing bacteria. FEMS Microbiol Lett 233:15–21

Phillips DA, Fox TC, King MD, Bhuvaneswari TV, Teuber LR(2004) Microbial products trigger amino acid exudationfrom plant roots. Plant Physiol 136:2887–2894

Piccoli P, Masciarelli O, Bottini R(1996) Metabolism of 17,17-[2H2]-gibberellin A4, A9 and A20 by Azospirillum

lipoferum in chemically defined culture medium. Symbi-osis 21:263–274

Rothballer M, Schmid M, Fekete A, Hartmann A (2005)Comparative in situ analysis of ipdC-gfpmut3 promoterfusions of Azospirillum brasilense strains Sp7 and Sp245.Environ Microbiol 7:1839–1846

Ruckdäschel E, Lewis-Kittell B, Helinski DR, Klingmüller W(1988) Aromatic amino acid aminotransferase of Azospir-illum lipoferum and their possible involvement in IAAbiosynthesis. In: Klingmüller W (ed) Azospirillum IV,genetics, physiology, ecology. Springer, The Netherlands,pp 49–53

Ryu RJ, Patten CL (2008) Aromatic amino acid dependentexpression of índole-3-pyruvate decarboxylase is regulatedby TyrR in Enterobacter cloacae UW5. J Bacteriol190:7200–7208

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning:a laboratory manual. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, New York

Sergeeva E, Hirkala DLM, Nelson LM (2007) Production ofindole-3-acetic acid, aromatic amino aminotransferaseactivities and plant growth promotion by Pantoea agglom-erans rhizosphere isolates. Plant Soil 297:1–13

Simon R, Priefer U, Pühler A (1983) A broad host rangemobilization system for in vivo genetic engineering:transposon mutagenesis in gram negative bacteria. Bio-technology 1:784–790

Somers E, Ptacek D, Gysegom P, Srinivasan M, Vanderleyden J(2005) Azospirillum brasilense produces the auxin-likephenylacetic acid by using the key enzyme for indol-3-acetic acid biosynthesis. Appl Environ Microbiol71:1803–1810

Soto-Urzúa L, Xochihua-Corona YG, Flores-Encarnación M,Baca BE (1996) Purification and properties of aromaticamino acid aminotransferase from Azospirillum brasilenseUAP14 strain. Can J Microbiol 42:294–298

Spaepen S, Vanderleyden J, Remans R (2007a) Indole-3-aceticacid in microbial and microorganism-plant signaling.FEMS Microbiol Rev 31:425–448

Spaepen S, Versées W, Gocke D, Pohl M, Steyaert J,Vanderleyden J (2007b) Characterization of phenylpyru-vate decarboxylase, involved in auxin production ofAzospirillum brasilense. J Bacteriol 189:7626–7633

Spaepen S, Dobbelaere S, Croonenborghs A, Vanderleyden J(2008) Effects of Azospirillum brasilense of indol-e-3acetic-acid production on inoculated wheat plants. PlantSoil 312:15–23

Tao Y et al (2008) Rapid synthesis of auxin via a newtryptophan-dependent pathway is required for shadeavoidance in plants. Cell 133:164–176

Tien TM, Gaskins MH, Hubbell DH (1979) Plant growthsubstances23 produced by Azospirillum brasilense andtheir effect on the growth of pearl millet24 (Pennisetumamericanum L.). Appl Environ Microbiol 37:1016–1024

Vande Broek A, Michiels J, Van Gool A, Vanderleyden J(1993) Spatial-temporal colonization patterns of Azospir-illum brasilense on the wheat root surface and expressionof bacterial nifH gene during association. Mol PlantMicrobe Interact 6:592–600

Vande Broek A, Lambrecht M, Eggermont K, Vanderleyden J(1999) Auxins up-regulate expression of the indole-3-

Plant Soil (2012) 356:139–150 149


http://dx.doi.org/10.1093/dnares/dsp026

pyruvate decarboxylase gene in Azospirillum brasilense. JBacteriol 181:1338–1342

Vílchez S, Molina L, Ramos C, Ramos JL (2000) Prolinecatabolism by Pseudomonas putida: cloning, characteriza-tion, and expression of the put genes in the presence ofroot exudates. J Bacteriol 182:91–99

Yang S, Zhang Q, Guo J, Charkowski AM, Glick BR, IbekweAM, Cooksey DA, Yang CH (2007) Global effect of

indole-3-acetic acid biosynthesis on multiple factors ofErwinia chrysanthemi 3937. Appl Environ Microbiol73:1079–1088

Zeman AMM, Tchan YT, Elmerich C, Kennedy IR (1992)Nitrogenase activity in wheat seedlings bearing para-nodules induced by 2,4-dichlorophenoxyacetic acid (2,4-D) and inoculated with Azospirillum. Res Microbiol143:847–855

150 Plant Soil (2012) 356:139–150


Author's personal copy REGULAR ARTICLE

Documents

Transcript of Author's personal copy REGULAR ARTICLE