two potential indole-3-acetaldehyde dehydrogenases in the phytopathogenic fungus ustilago maydis

Eur. J . Biochem. 242, 648-656 (1996) 0 FEBS 1996

Two potential indole-3-acetaldehyde dehydrogenases in the phytopathogenic fungus Ustilago maydis Christoph W. BASSE', Friedrich LOTTSPEICH2, Wolfgang STEGLICH3 and Regine KAHMANN' ' Institut fur Cenetik und Mikrobiologie der Universitat Munchen, Germany * Max-Planck-Institut fur Biochemie, Martinsried, Germany

Institut fur Organische Chemie der Universitat Munchen, Germany

(Received 25 July 1996) - EJB 96 111 8/4

The phytopathogenic basidiomycete Ustilago maydis produces indole-3-acetic acid (IndCH,COOH) and indole-3-pyruvic acid (Ind-Prv) from tryptophan. Indole-3-acetaldehyde (IndCH,CH,O) is the coni- mon intermediate in the conversion of Ind-Prv and tryptamine to IndCH,COOH. We purified an enzyme (Iadl ) from 0. maydis that catalyzes the NAD+-dependent conversion of IndCH,CH,O to IndCH,COOH and isolated corresponding cDNA and genomic clones. The identity of the cDNA clone was confirmed by expression in Escherichiu coli and demonstration of enzymatic activity. In U. muydis, iudl-null mutants were generated by gene replacement. The ability to convert IndCH,CH,O to IndCH,COOH was at least 100-fold reduced in U. maydis iadl-null mutants grown in medium with glucose as carbon source. However, the iadl-null mutants were not diminished in their capacity to produce IndCH,COOH from tryptophan, indicating that IndCH,COOH formation from tryptophan apparently proceeds in the absence of IndCH,CH,O dehydrogenase activity under these conditions. ladl expression was strongly induced during growth on ethanol while under these conditions iudl-null mutants were unable to grow. This reveals that Iadl is primarily engaged in the conversion of ethanol to acetate. In iadl-null mutants we detected an additional NAD'-dependent IndCH,CH,O dehydrogenase activity that was induced during growth on L-arabinose but repressed in the presence of D-glucose. In arabinose-containing medium the conversion of tryptophan to IndCH,COOH was approximately 5-fold reduced in wild-type strains but 10- 15-fold reduced in iadl-null mutant strains compared to IndCH,COOH formation in glucose-containing medium. In addition, the formation of Ind-Prv from tryptophan was abolished in wild-type and iadl- null mutant strains. During growth on arabinose, the conversion of tryptamine to IndCH,COOH was strongly favored suggesting that the glucose-repressible IndCH,CH,O dehydrogenase is required to convert IndCH,CH,O derived from tryptamine to IndCH,COOH.

Keywords: Ustilago maydis ; indole-?acetic acid ; indole-3-acetaldehyde dehydrogenase.

The basidiomycete Ustilago maydis is the cause of smut disease in maize. Disease development is characterized by chloro- sis, stunting, and tumor development (see Banuett, 1995). Dur- ing its life-cycle, U. maydis exists in a haploid and a dikaryotic form. Haploid cells grow vegetatively by budding and can be propagated on artificial media. The dikaryon results from fusion of two compatible haploid cells, is filamentous and needs the host plant for proliferation. Hyphal cells induce tumors in which

Correspondence to R. Kdhmann, Institut fur Genetik und Mikrobio-

Fax: +49 89 178 56 33. Abbreviations. IndCH,COOH, indole-3-acetic acid ; Ind-Prv, indole-

3-pyruvic acid; IndCH,CH,O, indole-3-acetaldehyde; CM, complete medium (Holliday, 1961) ; Iadl and Iad2, indole-3-acetaldehyde dehydrogenases; incompletely specified bases are given in the single-letter code, where N represents any nucleotide, B represents G, T, or C, M represents A or C, R represents G or A, and Y represents T or C.

Eruyrnes. Aldehyde dehydrogenase (EC 1.2.1.3); aldehyde oxidase (EC 1.2.3.1); alcohol dehydrogenase (EC 1.1.1.1); indole-3-pymvate decarboxylase (EC 4.2.1.-); tryptophan decarboxylase (EC 4.1.1.28); nitrilase (EC 3.5.5.1): nitrile hydratase (EC 4.2.1.84); tryptophan monooxygenase (EC 1.13.12.3).

Nore. The novel nucleotide sequence published here has been sub- mitted to the GenBank sequence data bank and is available under the accession number U74468.

logie, Maria-Ward-Str. 1 a, D-80638 Munchen, Germany

they differentiate to form diploid teliospores (Banuett, 1995). Mating and pathogenic development are genetically controlled by two unlinked loci, termed a and b. The a locus, with two alleles, encodes a pheromone-based recognition system, required for fusion of two compatible, haploid cells. The multiallelic b locus encodes a pair of unrelated homeodomain proteins and controls filamentous growth and pathogenicity (Rowell and De- Vay, 1 954 ; Rowell, 1955 ; Holliday, 1961 ; Puhalla, 1968 ; Bolker et al., 1992; Spellig et al., 1994; Schulz et al., 1990; Gillissen et al., 1992; Kamper et al., 1995).

Whereas the central genes controlling pathogenic development have been identified and are studied extensively (see Kah- mann et al., 1995), very little is known about the interaction between U. maydis and its host. A particularly intriguing ques- tion is how the massive proliferation of plant tissue is induced after infection. Biochemical analyzes of fungal culture supernatants and of plant tumors have indicated that U. maydis is capa- ble of producing plant growth-promoting substances. U. maydis was shown to produce indole-3-acetic acid (IndCH,COOH) when grown in medium containing tryptophan (Wolf, 1952). In addition, in tumor tissue 5-20-fold elevated levels of auxins have been demonstrated (Wolf, 1952; Tuiian and Hamilton, 1960). IndCH,COOH has profound effects on biological pro- cesses such as cell expansion, cell division, induction of adventi-

B a s e et al. (Eur: J. Biochem. 242) 649

tious roots, control of apical dominance, and tropisms (see Klee and Estelle, 1991). Since tumor development in plants infected by U. maydis is characterized by increased mitotic divisions and an enlargement of cells (Christensen, 1963; Callow and Ling, 1973) it is conceivable that IndCH,COOH produced by U. maydis is involved in this process.

IndCH,COOH biosynthesis was extensively studied in plants, plant-associated and phytopathogenic bacteria. In plants, three different IndCH,COOH pathways from tryptophan have been described (see Marumo, 1986). These include IndCH,COOH biosynthesis via indole-3-pyruvic acid (Ind-Prv), tryptamine, and indole-3-acetaldoxime. In the Ind-Prv pathway the conversion of tryptophan to Jnd-Prv is catalyzed by a tryptophan transaminase. A decarboxylase converts Ind-Prv to indole- 3-acetaldehyde (IndCH,CH,O). The aldehyde is converted in the presence of NADH by an alcohol dehydrogenase to tryptophol (indole-3-ethanol). An oxidase or an aldehyde dehydrogenase in the presence of NAD' form IndCH,COOH from IndCH,CH,O (Wightman and Cohen, 1968). In the tryptamine pathway tryptophan is decarboxylated to tryptamine followed by oxidative de- amination to IndCH2CH,0 (Percival and Purves, 1974). Trypto- phan decarboxylase has been detected in numerous plant systems and the gene encoding tryptophan decarboxylase has been cloned from Cutharunthus roseus (De Luca et al., 1989). Com- mon to the Ind-Prv and tryptamine pathways is the conversion of IndCH,CH,O to IndCH,COOH. As a third possibility IndCH,COOH is synthesized from tryptophan via indole-3-acetaldoxime and indole-3-acetonitrile (Ludwig-Muller and Hil- genberg, 1988). Indole-3-acetonitrile can be hydrolyzed to IndCH,COOH by nitrilase. Genes for nitrilases have been isolated from Arabidopsis thaliana (Bartling et al., 1992, 1994; Bartel and Fink, 1994). There is also evidence that indole-3- acetonitrile is produced in A. thaliana by a non-tryptophan pathway (Normanly et al., 1993). As yet another alternative, tryptophan can be converted via indole-3-acetamide to IndCH,COOH by the successive action of a tryptophan monooxygenase and an indole-3-acetamide hydrolase. These pathways occur in Agro- bacterium tumefaciens and Pseudomanas savastanoi (Klee et al., 1984; Yamada et al., 1985). Furthermore, in A. tumefaciens and Rhizobium strains indole-3-acetonitrile is processed to IndCH,COOH via indole-3-acetamide by the action of nitrile hydratase (Kobayashi et al., 1995).

Although auxin production has been reported for many fungi (see Gruen, 1959) little is known about the biosynthetic pathways by which it is produced. In U. maydis tryptamine has not been observed as an intermediate in IndCH,COOH formation from tryptophan (Wolf, 1952) and neither tryptamine nor indole- 3-acetamide could serve as precursors for IndCH,COOH formation in U. rnaydis cultures (Navarre and Damann, 1990). Ind- Prv has therefore been suggested as an intermediate in lndCH,COOH biosynthesis in U. maydis.

In this study, we analyzed the involvement of IndCH,CH,O dehydrogenase activities in IndCH,COOH biosynthesis in U. maydis. We demonstrate the presence of two acetaldehyde dehydrogenases (ladl and Iad2) with the ability to convert IndCH,- CH,O to IndCH,COOH. The gene for one of these enzymes was cloned and analyzed molecularly.

MATERIALS AND METHODS

Strains and plasmids. The Escherichiu coli strain DHSa (BRL) was used for cloning. The haploid U. maydis strains FB1 and FB2 (see Banuett and Herskowitz, 1989) are non-pathogenic and grow like yeast on complete solid medium containing charcoal. For cloning and expression analysis in E. coli, the plasmids

pUC19, pTZ19R, and pTrc99A (Pharmacia) were used. As se- lectable marker for U. maydis a hygromycin resistance cassette was isolated as a 3.2-kb PvuII fragment from the U. rnaydis vector pCM54 (Tsukuda et al., 1988).

Growth conditions and media. U. maydis strains were grown in complete medium (CM) (Holliday, 1961) or minimal medium (Holliday, 1974) at 28°C. CM and minimal medium contained 1 % D-glucose. In CM with either arabinose or xylose, glucose was replaced by 1 % L-arabinose or 1 % D-XyloSe, respectively. Liquid cultures were shaken at 200 rpm. Absorbance (A) of cultures was determined at 600 nm with a LKB Novaspec spectrophotometer. Ind-Prv, IndCH,CH,O, tryptamine, and tryptophol were from Sigma, IndCH,COOH was from ICN Bio- chemicals, tryptophan was from Calbiochem. All chemicals used were of analytical grade.

Purification of culture supernatants for mass spectro- scopic analysis. FBI was grown in CM containing 2 mM tryptophan to a cell density of about 2X108/ml. Cell-free supernatant was acidified to pH 2.5 with HCl and extracted three times with ethyl acetate. The combined organic phases were evaporated to dryness, redissolved and subjected to HPLC (Beckman Gold) on a CIS ultrasphere column (5 pm, 4.6X250 mm; Beckman) eluting with 20% (by vol.) acetonitrile, 0.1 % trifluoroacetic acid in water for 10 min, followed by a linear gradient to 40% acetonitrile, 0.1% trifluoroacetic acid in 20 rnin at a flow rate of 1 ml min-' (gradient I). Absorbance was measured at 220nm. The detection limit for lndCH,COOH was in the range of 1-2 ng corresponding to 5 - 10 pmol/ml IndCH,COOH. The IndCH,COOH-containing fractions (IndCH,COOH eluted at 17 min) were combined. Since IndCH,COOH and tryptophol showed similar retentention times with the acidic solvent, a basic methanol solvent was used for separating these compounds. IndCH,COOH-containing fractions were further purified on a C18 ultrasphere column ( 5 pm, 4.6X250 mm; Beckman) eluting with 25% methanol in 20 mM ammonium acetate pH 6.5 for IOmin, followed by a linear gradient to 40% methanol in 20min at a flow rate of 1 ml min-' (gradient 11). Absorbance was measured at 220 nm. Using this gradient IndCH,COOH eluted at 7.1 min and tryptophol at 30.3 min. The chemical identity of the HPLC-purified IndCH,COOH was confirmed by electron-impact mass spectroscopy at 70 eV using a Finnigan MAT 95Q mass spectrometer.

Feeding experiments. U. maydis was cultivated with either tryptophan ( 5 mM final concentration unless otherwise indicated), IndCH,CH20 (0.2 mM), or tryptamine (0.5 mM). The cultures were grown to a density of about 2X108 ml- I . Cell-free supernatants (1 ml) were removed, acidified to pH 2.5 with HCI, and extracted with ethyl acetate. The organic phase was evaporated to dryness, redissolved and subjected to HPLC running with gradients I or I1 (see above). Fractions were collected, dried under vacuum, redissolved and tested with the Salkowsi reagent (Gordon and Weber, 1951).

Preparation of a crude enzyme preparation and enzyme assays. U. maydis was grown in CM to an A,, of about 1.0, the cells were harvested by centrifugation (3500 rpni for 10 rnin), washed with water and resuspended in 50 mM Tris/HCI pH 7.8 containing 5 mM dithiothreitol, 0.2 mM EDTA, 1 mM Pefabloc (Boehringer), 3 pg/ml pepstatin (Boehringer), and 1.5 pg/ml leupeptin (Boehringer). Cells were stored frozen (-80°C) before cell lysis. Cells were thawed and subsequently disrupted in a chilled (OOC) French pressure cell (SLM Instruments). Cell ly- sates were centrifuged (1 0 min, 31 000 g, 4°C) and the supernatant was desalted on a PD-10 column (Pharmacia) eluting with 20 mM Tris/HCI pH 7.0, 10 mM KCI, 1 mM dithiothreitol (buffer A). Protein concentrations were determined according to Bradford (1976). Protein preparations (1 50 pl) were incubated

650 Basse et al. ( E m J . Biochem. 242)

in the presence of IndCH2CH,0 (1 mM final concentration) in a reaction volume of 300 pl containing 70 mM Tris/HCI pH 7.0, 10 mM KCI, 0.5 mM dithiothreitol for 90 min at 30°C. Oxida- tion of IndCH,CH,O to IndCH,COOH was tested in the presence or absence of NAD' (1 mM final concentration). After incubation, the reaction mixtures were acidified, extracted with ethyl acetate and analyzed with Salkowski reagent for the determination of IndCH,COOH formation or subjected to HPLC (gradient 11) for the quantification of IndCH,COOH.

Purification of IndCH,CH,O dehydrogenase. A crude enzyme preparation obtained from 2 1 of a FBI culture grown in CM to an A,,,, , of 1.0 was desalted on a PD-I0 columii eluting with buffer A, pH 7.8, containing 0.5 mM Pefabloc, 1.5 pg/ml leupeptin and applied to a 5 ml Econo-Pac High Q cartridge (Bio-Rad) equilibrated with buffer A, pH 7.8. The column was washed with three column volumes of starting buffer at a flow rate of 1 ml min-' and then developed with a 0-0.6 M NaCl gradient in buffer A, pH 7.8, over six column volumes; 1-ml fractions were collected and analyzed for IndCH,CHZO dehydrogenase activity. The active fractions eluting between 150- 250 mM NaCl were combined, desalted on a PD-10 column eluting with buffer A containing 0.5 mM Pefabloc, 1.5 pg/ml leupeptin and applied to a 1-ml HiTrap Blue column (Pharmacia Biotech). The column was washed with 10 column volumes of this buffer. A step gradient of 3 mM NAD' was used to elute IndCH,CH,O dehydrogenase activity ; 1 -ml fractions were collected. The active fractions were combined and loaded onto a MonoQ HR5/5 anion-exchange column (Pharmacia). The column was equilibrated in buffer A, pH 7.4, and washed with 10 column volumes of starting buffer at 1 ml min-l after loading. IndCH,CH,O dehydrogenase activity was eluted with a gradient of 0-0.5 M NaCl over 30 column volumes. Fractions of 0.5 ml were collected and analyzed for IndCH,CH,O dehydrogenase activity. The activity eluted between 100- 125 mM NaCI. Purifi- cation of IndCH,CH,O dehydrogenase occurred at 4°C. The assay, described above, was used throughout the purification protocol. The assay mixture contained in addition 1 ing m-' BSA. An aliquot (0.1 ml) of the fractions was incubated in this mixture for 90 min at 30°C in a final volume of 0.2 ml. One unit of enzyme activity was defined as the quantity of enzyme required to catalyze the production of 1 nmol IndCH,COOH min I at 30°C. SDS/PAGE was performed as described by Laemmli (1970) employing a 4% stacking gel and a 10% gel for separation. Gels were run at 80-100 V and protein bands were visualized by silver staining. For N-terminal sequencing proteins were blotted onto polyvinylidenedifluoride membrane (Millipore) after SDS/PAGE and stained with Coomassie bril- liant blue G-250 (Serva). After destaining, the N-terminal sequence of the proteins was determined on a 479A pulsed liquid- phase sequencer (Applied Biosystems).

Hydroxyapatite chromatography. Hydroxyapatite chromatography was performed using a 1-ml hydroxyapatite cartridge (Bio-Rad). The column was eluted with 18 mM sodium phosphate pH 6.8 containing 1 mM dithiothreitol for 10 min followed by a linear gradient to 0.5 M sodium phosphate, pH 6.8, 1 mM dithiothreitol in 38 min at a flow rate of 0.75 ml min-'. The activity eluted between 250-325 mM sodium phospate. The IndCHZCH20 dehydrogenase activities of fractions 14- 19 were determined. The activities were 0.1, 0.1, 1.3, 3.6, 2.5, and 0.6 U for fractions 14- 19, respectively.

Molecular techniques. Standard molecular techniques followed Sambrook et al. (1989). U. rnaydis chromosomal DNA was isolated as described by Hoffman and Winston (1987). Southern analyzes followed the procedure of Church and Gilbert (1984). RNA isolation and northern analysis were as described by Schmitt et al. (1990). Filters were hybridized at 60°C. For

quantification of radioactive signals a Storm 840 PhosphorIm- ager system (Molecular Dynamics) was used. Radioactive labeling of DNA was performed with a random primer labeling kit (Amersham). Nucleotide sequences were determined by the di- deoxynucleotide chain-termination method (Sanger et al., 1977) using T7 DNA polymerase (T7 sequencing kit, Pharmacia) and automated sequencing using an ABl Prism 377 DNA sequencer. Both strands of the cDNA were sequenced. Nucleotide sequences were compared using BLAST (Altschul et al., 1990).

Isolation of genomic and cDNA clones of iadl. Degenerate oligonucleotide primers IADHl : 5'-CGCGGATCCACNYTBA- AYYTNGA-3' and IADH3 5'-CGCGGATCCGGNACRAAYT- TRTTRTT-3' containing BamHI sites were used for PCR with 140 ng genomic DNA of FBI as template in a volume of 70 PI. Reactions contained: 10 mM Tris/HCI pH 8.3, 1.5 mM MgCI?, SO mM KCI, 0.2 mM dNTPs, 400 pmol primer IADHl , 7 5 pmol primer IADH3 and 5 U Tuq polymerase (Boehringer). The following program was used for amplification (30 cycles): 1 rnin 94"C, 1 min 50°C and 0.75 min 72°C. A 96-bp fragment was isolated and cloned into the BarnHl site of pUC19. The BarnHI fragment was used as probe to isolate a full-length cDNA from a cDNA library of FBDl1 (Schauwecker et al., 1995). The 1.7-kb EcoRI cDNA fragment was introduced into the EcoRI site of pTZ19R to give pTZiadlc. From a cosmid library of FBD11 (Bolker et al., 1995) a 2.9-kb genomic PstI fragment hybridizing to the cDNA was identified and cloned in pUC19 to give pIadl g.

Constructs for gene replacement. To construct pdIad1 a 1.6-kb AgeI-BstBI fragment from the coding region of iadl was replaced i n pIad1g with a 3.2-kb hygromycinB cassette. To replace the resident iadl gene with the Aiadl allele, a 6.1-kb Sspl-AlwNI fragment was isolated from pAIadl and transformed into U. maydis strains FB1 and FB2 as described (Schulz et al., 1990). Gene replacement was verified by Southern analysis. Of 20 FBI and FB2 transformants recovered, 5 and 4, respectively, carried the Aiudl allele in place of the wild-type allele.

Expression of indl in DHSn. The gene was isolated from pTZiadlc as a SphI-EcoRI fragment and introduced into pTrc99A cleaved with NcoI and EcoRI. The SphI site containing the ATG start codon of iadl was treated with T4 polymerase and the NcoI site of pTrc99A was treated with Klenow polymerase prior to ligation. The resulting plasmid pladl was introduced into DH5a. Control strains contained pTrc99A. For overexpres- sion, cells were grown in Luria-Bertani medium containing ampicillin (100 pg/ml) at 37°C. At an A , , , of 0.5, 1 mM isopropyl P-o-thiogalactoside (Gerbu) was added and incubation continued for 5 h. After induction, cells were harvested, resuspended i n ice-cold enzyme extraction buffer (see above) and stored frozen at -80°C. Enzyme extracts were prepared and desalted on PD- 10 as described above. The enzyme extract was further purified by chromatography on High Q (see above) to eliminate NADH oxidase or dehydrogenase activities. IndCH,CH,O and acetaldehyde dehydrogenase activities were measured at 25 "C by following the reduction of NAD' at 340 nm in 1 .O-cm light-path cuvettes using a Perkin Elmer UVNIS spectrophotometer. The reaction mixture consisted of 60 mM Tris/HCI pH 8.0, 1 mM NAD', 10 niM KCI, and various concentrations of substrate in a volume of 0.2 ml.

Test for filamentous growth aiid pathogenicity. Yeast-like or mycelial phenotypes were distinguished on charcoal nutrient medium (Holliday, 1974). To test for mating, strains were co- spotted on charcoal-containing CM plates (Holliday, 1974) and incubated at room temperature for 48 h. To assay pathogenicity. 6 - 8-day-old corn seedlings (variety Early Golden Bantam)

B a s e et al. (ELK J. Biochem. 242) 65 1

Table 1. Summary of purification of U. muydis IndCH,CH,O dehydrogenase from 2 I FB1 suspension culture. 1 U is the amount of enzyme required to catalyze the production of 1 nmol lndCHzCOOH min ' at 30°C.

Purification step Volume Total protein Total activity Specific activity Purification Yield

ml mg U U/mg -fold 5% - - PD 10 eluent of crude extract 28.0 270.1 115.2 0.425

High Q 7.0 20.6 57.9 2.81 6.6 50.3 Blue 3G-A 2.5 0.308 27.8 90.31 212.4 24.1 MonoQ HR515 1.5 0.050 19.7 390.87 919.1 17.1

1.5 - 1.2 6 8 0.9 N

f :: 0

0 5 10 15 20 25 30 35 40 45

Time (min)

m/z

Fig. 1. Identification of IndCH,COOH. (A) Supernatant from strain FB1 grown in CM supplemented with 1 mM tryptophan was extracted with ethyl acetate and subjected to reverse-phase HPLC on a C18 column running with gradient I. (-) Continuous registration of the absorbance at 220 nm; (0) Salkowski test of the individual fractions. The arrows point to the peaks that elute like IndCH,COOH (IAA) and Ind- Prv (IPyA). (B) Mass spectrometric identification of the indole compound eluting like IndCH,COOH. A plot of massfcharge ( m k ) ratios obtained from this compound (B) is compared with that of IndCH,COOH standard (C). Mass spectra were obtained by electron impact at 70 eV.

were infected with mixtures of compatible haploid strains as described (Holliday, 1974).

RESULTS IndCH,COOH formation in U. muydis cultures and purification of Iadl. We have reinvestigated IndCH,COOH biosynthe-

sis in U. maydis by using HPLC for the detection of Ind.CH,COOH and indole derivatives in culture supernatants or in cell-free extracts. The formation of IndCH,COOH by U. maydis strain FBI in liquid medium was at least 100-fold elevated when CM was supplemented with 1 mM tryptophan (not shown). The HPLC profile of ethyl-acetate-extracted culture medium revealed single peaks with retention times identical to IndCH,COOH and Ind-Prv (Fig. 1). The corresponding peak fractions reacted with the Salkowski reagent which gives a red color with IndCH,COOH and other indole derivatives (Glick- mann and Dessaux, 1995). The peak fraction eluting like IndCH,COOH was further purified (see Materials and Methods) and analysed by mass spectroscopy. As shown in Fig. 1, the mass spectrum of IndCH,COOH isolated from U. maydis corre- sponded to the spectrum of authentic IndCH,COOH. It has been proposed that IndCH,COOH formation proceeds by the Ind-Prv pathway in U. maydis. Since Ind-Prv is processed via IndCH,- CH,O, we have attempted to purify IndCH,CH,O dehydrogenase, an activity which was demonstrated in extracts of strain FBI cultivated in liquid CM (Table 1). The reaction products were analyzed by reverse-phase HPLC monitoring their absorbance at 220 nm. A crude enzyme preparation, which was desalted on a PD-20 column, converted IndCH,CH,O to IndCH,COOH in the presence of either NAD' or NADH indicating the presence of NADH oxidase or dehydrogenase activity in these extracts. Further purification by anion-exchange chromatography indicated a strict NAD' dependence. The active fractions were applied to a Blue 3G-A affinity column and eluted with NAD' from this column. This step led to 30-fold enrichment of the specific activity. The activity was subsequently chromatographed by anion-exchange chromatography on a MonoQ column resulting in an approximate 1000-fold increase in the specific IndCH,CH,O dehydrogenase activity (Table 1). Fractions of the individual purification steps and from MonoQ chromatography were applied to SDS/PAGE (Fig. 2 A). The presence of a protein with an apparent molecular mass of 54 kDa correleted with the IndCH,CH,O dehydrogenase activity (Fig. 2A and B). We have designated this protein Iadl. Addi- tional proteins with molecular masses of about 41 kDa and 49 kDa that copurified with the IndCH,CH,O dehydrogenase activity were excluded as candidates upon chromatography on a hydroxyapatite column (Fig. 2 C).

Cloning and characterization of the iud gene. Iadl was sequenced by automated Edman degradation and the N-terminal sequence was determined to be PTLNLDLPNGIKSTIQADL- FINNKFVPALD. With the help of degenerate oligonucleotide primers representing amino acid residues 3-7 and 23-28, a fragment of 96 bp was amplifed by PCR from genomic DNA of FB1 as template. This fragment was used as probe to isolate a full-length cDNA from a cDNA library of FBDl1 (Schauwecker et al., 1995). The iadl cDNA was used as probe to isolate a 2.9- kb genomic PstI fragment encompassing iadl from a cosmid library of FBDll (Bolker et at., 1995). Sequencing of the iadl

652 Basse et al. ( E m J . Biochem. 242)

kDa

-91

- 66 4 - 49

- 29

B

C

lo 1

12 13 14 15 16 17 Fraction

91 66

49 4

Fig. 2. Purification of Iadl. (A) SDSPAGE analysis of the individual fractions from purification of Tadl. First lane, crude extract after desalt- ing on PD-10 (3.0 pg protein); second lane, pooled fractions from High Q anion-exchange chromatography (2.4 pg); third lane, pooled fractions from Blue 3G-A affinity chromatography (1.0 big); lanes 12-17, individual fractions from MonoQ chromatography containing 0.06 pg, 0.23 pg, 2.21 pg, 0.78 pg, 0.09 pg, and 0.02 pg, respectively. Protein bands were visualized by silver staining. The arrowhead indicates Tadl . (B) IndCH,CH,O dehydrogenase activity of the individual fractions froin MonoQ chromatography. (C) IndCH,CH,O dehydrogenase was purified by chroinatography on High Q and Blue 3C-A as described in Materials and Methods followed by chromatography on a hydroxyapatite column. Lanes H14, HI5 and H18, fractions 14 (0.48 pg), 15 (0.30 pg) and 18 (0.34 pg) from hydroxyapatite chromatography. Fractions 14 and 15 contained no IndCH,CH,O dehydrogenase activity whereas a major portion of the activity (see Materials and Methods) was present in fraction 18. Lanes MQ13 and MQ14, fractions 13 (0.38 pg) and 14 (3.69 pg) from MonoQ chromatography (see Fig. 2A). Lane M, molecular mass stan- dards (values on the right). Protein bands were visualized by silver staining. The arrowhead indicates lad1 .

cDNA indicated the presence of an OR€: encoding a protein of 497 amino acids with a calculated molecular mass of 53.7 kDa (Fig. 3 A). The N-terminal sequence determined by automated sequencing matched the deduced amino acid sequence of iadl. The sequence surrounding the presumed translation start codon conforms to the fungal consensus for initiation of translation CAMMATGNC (Ballance, 1990). Sequence analysis of 218 nu- cleotides upstream of the start codon did not reveal the presence of additional translation initiation sites (not shown). The N-ter-

A -54

8 2 2

33 9 7

5 8 1 7 2

83 2 4 7

1 0 8 322

1 3 3 397

1 5 8 4 7 2

1 8 3 547

2 0 8 6 2 2

233 697

2 5 8 7 7 2

2 8 3 847

3 0 8 9 2 2

333 9 9 7

3 5 8 1 0 7 2

383 1 1 4 7

4 3 8 1 2 2 2

4 J 3 1297

458 1 3 7 2

483 1 4 4 7

1 5 2 2

1 5 9 7

B ( I ) - ----- - - - P T L N L w I X S T - L Q A - D L F I " K F V P B L a

(2) --------SSSGTPaLSVLLTDLK~YTKI~E~DSVS

(3) - - - - - - - - S S S G T P o L P V L L T D L K F Q Y T K I ~ ~ D S V S

(4) - - - - - - - ~ S S P A Q P A V p A P L ~ L K ~ H T K I ~ E ~ N S L N

( 5 ) M T K L H F D T A E P V K I T W L T Y - E - Q P T G m M K & Q

Fig. 3. Nucleotide sequence of iudl and derived amino acid sequence of Zudl. (A) The cDNA sequence is shown with the ORF extending from nucleotide +1 to nucleotide +1491. The amino acid sequence is given in capital letters. Polyadenylation was observed at the most 3' G residue of the sequence shown. In the construct used for gene replacement the iadl ORF was deleted between the AgeI and BstBI site. The translation stop is indicated by an asterisk. (B) Alignment of the N- terminal amino acid sequences determined for U. may& Iadl (l), human cytosolic aldehyde dehydrogenase E, (2) (Hempel et al., 1984), cytosolic aldehyde dehydrogenases from horse (3) (von Bahr-Lindstroni et al., 1984), rat (4) (Dunn et al., 1989), and the N-terminal amino acid sequence from yeast cytosolic aldehydc dehydrogenase ( 5 ) (Weiner, Wang, and Bai, Genbank accession number US6604). Gaps have been inserted to maximize the number of identities to Iadl (underlined).

minal sequence of Iadl showed similarity to cytosolic aldehyde dehydrogenases (Fig. 3 B) and did not reveal N-terminal targeting sequences. Therefore, the purified Iadl activity is likely to be cytosolic.

653 Basse et al. ( E m J. Biochem. 242)

2.5

2

3

E l

$

f 1.5 ", B

0.5

a

0.8

'100 2 0.6

72 2

h

- 8 - 0.4

-1 -2

2 -In<,

A

pTrc99A

& 14.0 16.0 18.0

B

pIadl !L 14.0 16.0 18.0

Time (mm)

0.3 1 D

Fig.4. Expression of iadl in E. coli. E. coli strains DHSa harbouring pladl (B) or pTrc99A (A) were grown in Luria-Bertani medium containing ampicillin (100 pglml), supplemented with 0.4 mM IndCH,CH,O. Isopropylthio-~-~-thiogalactoside (1 mM) was added to both cultures at an A,,,,, of 0.5 and incubation was continued for an additional 4 h. An aliquot of the supernatant was acidified, extracted with ethyl acetate and subjected to reverse-phase HPLC on a C18 column running with gradient I. Fractions (1 ml) were collected and tested with the Salkowski reagent. Each data point represents the average of three measurements with cell cultures of independent E. coli transformants. IAA = IndCH,COOH. (C, D) Lineweaver-Burk plot for the determination of the K, value of Iddl for IndCH,CH,O (C) and acetaldehyde (D). Enzyme assays were performed with partially purified extracts from E. coli harbouring pIddl as described in Materials and Methods. Each data point represents the average of four or five individual measurements.

When compared with sequences in the database, Iadl (Fig. 3 A) revealed the highest similarity to aldehyde dehydrogenases from Aspergillus nidulans (Pickett et al., 1987), Aspergil- lus niger (O'Connell and Kelly, 1988), Cladosporium herberum (Achatz et al., 1995), Alternaria alternata (Achatz et al., 1995), and human (Braun et al., 1987) with 54-62% identity. The conserved regions encompass the highly conserved Cys302 residue which is part of the active site in rat liver mitochondria1 aldehyde dehydrogenase (FarrCs et al., 1995) and the conserved Glu268 residue required for activity in human liver mito- chondrial aldehyde dehydrogenase (Wang and Weiner, 1995). The corresponding amino acids are represented by Cys299 and Glu265 in U. maydis Tad1 (Fig. 3A and not shown).

To demonstrate enzymatic activity of Iadl, the iadl cDNA was ftised to the isopropyl-0-D-thiogalactoside-inducible trc pro- motor of the E. coli expression vector pTrc99A and the resulting plasmid pIadl was transformed into the E. coli strain DHSa. As shown in Fig. 4A and B, cells containing pIadl exhibited a 20-fold increased capacity to convert IndCH,CH,O to IndCH,COOH over control cultures harbouring pTrc99A. The apparent K, value of Iadl for IndCH,CH,O and acetaldehyde was determined with partially purified enzyme extracts obtained from E. coli expressing pIadl (see Materials and Methods). Data were analysed in a Lineweaver-Burk plot and from this Michae- lis constants for IndCH,CH,O and acetaldehyde of 0.19 mM and

1.7 mM, respectively, could be calculated (Fig. 4 C and D). This shows that Iadl exhibits a higher affinity for IndCH,CH,O compared to acetaldehyde.

IndCH,COOH synthesis in strains carrying a iadl-null allele. To investigate the role of Iadl in IndCH,COOH formation and pathogenic development in U. maydis, iadl-null mutants were generated in the compatible haploid strains FBI and FB2 by gene disruption (see Materials and Methods). In FBldiadl and FB2diadl the ORF of iadl is deleted from amino acids 44-497 (see Fig. 3A) and replaced by a hygromycin resistance cassette. Northern analysis revealed that the iudl mRNA was not detectable in FBldiadl (Fig. 5, lanes 3 and 4). In FBldiadl cultivated in CM with glucose as carbon source the capacity to convert IndCH,CH,O to IndCH,COOH was more than 100-fold reduced compared to the parental strain FBI (Table 2) and the IndCHFH,O dehydrogenase activity was 40-fold reduced (Ta- ble 3). Instead of producing IndCH,COOH, diadl mutants converted IndCH,CH,O efficiently to tryptophol, a reaction unfa- vored in wild-type strains (Table 2). However when subjected to different growth conditions we detected an IndCH,CH,O dehydrogenase activity in diadl mutant strains which is subject to glucose repression (not shown). Therefore, we have analyzed IndCH,CH,O dehydrogenase activity in diadl mutants grown in the presence of various carbon sources (Table 2). This revealed that in FBI h a d l the additional IndCH,CH,O dehydrogenase activity, designated Iad2, is strongly induced during growth on L- arabinose compared to growth on D-glUCOse or D-XylOSe (Table 2). I n vitro assays indicated that the Iad2 activity was also NAD'-dependent and markedly increased in diadl mutants cultivated in CM with arabinose compared to the activity of diadl mutants grown in the same medium in the presence of glucose (Table 3). In dIad1 mutants cultivated in CM with arabinose and glucose this IndCH,CH,O dehydrogenase activity could no longer be detected and the conversion of IndCH,CH,O to IndCH,COOH occurred as inefficiently as during growth on glucose (not shown). This illustrates that Iad2 is subject to glucose repression.

Northern analysis revealed that iadl was expressed at similar levels in FB1 during growth in CM medium with either glucose or arabinose (Fig. 5). This suggests that in arabinose-containing medium both IndCH,CH,O dehydrogenase activities are present. When FB1 was cultivated in CM with arabinose high amounts of IndCH,COOH were produced from IndCH,CH,O (Table 2) and these exceeded the amounts produced by FBI grown in CM with glucose by twofold.

Tryptamine represents another intermediate in IndCH,- COOH biosynthesis that is converted to IndCH,COOH via IndCH,CH,O. In previous investigations tryptamine was not found to be a precursor of IndCH,COOH formation in U. maydis (Wolf, 1952; Navarre and Damann, 1990). Indeed, tryptamine was poorly converted to IndCH,COOH in FB1 grown in glucose-containing medium while it was efficiently converted to tryptophol (Table 2). Interestingly, however, IndCH,COOH formation from tryptamine was substantial in FBI and FBldiadl cultivated in CM with arabinose (Table 2). This suggests that IndCH,CH,O derived from tryptamine is preferentially converted to IndCH,COOH by lad2 in vivo.

Role of iadl in IndCH,COOH formation. To gain insight into the possible function of Iadl in the conversion of tryptophan to IndCH,COOH, FBI and FBldiadl strains were cultivated in CM with glucose and tryptophan. As shown above, Iad2 activity is absent under these conditions. Therefore, an involvement of Iadl in IndCH,COOH formation should be reflected by lower amounts of IndCH,COOH in diadl mutant strains compared to

654 Basse et al. (EM J , Biochern. 242)

Table 2. In vivo conversion of IndCH,CH,O and tryptamine to IndCH,COOH and tryptophol. Strains were cultivated in CM with the added sugar in the presence of either 0.2 mM IndCH,CH,O or 0.5 mM tryptamine. The amounts of IndCH,COOH and tryptophol were determined as described in Materials and Methods. Values are given as means 2 SD of three individual measurements.

Strain Precursor Sugar IndCH,COOH Tryptophol

nmolhl

FB 1 FBldiadl FB 1 FB 1 diad 1 FB 1 FBldiadl

FB 1 FB ldiadl FB 1 FB 1 4 iadl

IndCH,CH,O glucose glucose arabinose ardbinose xylose xylose

try ptamine glucose glucose arabinose arabinose

17.70 i 2.29 0.05 5 0.02

42.87 2 0.81 21.57 2 11.07 16.47 t 3.30 0.26 i 0.18

0.151 +- 0.097 0.027 I 0.025

10.80 i: 2.14 6.88 t- 1.66

0.00+ 0.00 38.262 0.72 0 . 0 9 I 0.03

16.362 11.16 0.022 0.03

22.87 214.54

33.75 115.64 22.32 t 3.79 99.55 t 10.50 87.25 t 0.89

Table 3. In vitru conversion of IndCH,CH,O to IndCH,COOH. The enzyme extracts were prepared and the enzyme assayed as described in Materials and Methods. Strains were cultivated in CM with the added sugar. Data are presented as mean values of two individual measurements.

Strain Sugar Specific activity of IndCH2COOH formation _____~_____ +NAD' - NAD +

1 2 3 4 5 6 7

4

U/mg --__ ___

FB 1 glucose 0 6256 0 009 5 FB 1 did 1 glucose 0 0153 0 0025 FB 1 drabinose 2 0478 0 0144 FBl41adl arabinoae 0 1140 0 0068

Table 4. In vivo conversion of tryptophan to IndCH,COOH and Ind- Prv. U. inuydi.7 cells (FBI) were cultivated in the presence of 5 mM tryptophan. The amounts of IndCH,COOH, Ind-Prv were determined as described in Materials and Methods. Valucs with glucose are mean values 5 SD of four or five individual measurements; values with arabinose are mean values -+ SD of 10 individual measurements.

~ ~ ~

Strain Sugar IndCH,COOH Ind-Prv ~ ~ ~~~~~

nmol/ml ~-

FB 1 glucose 2.631 2 1.382 27.33 t 19.98 FB1 diadl glucose 3.7422 1.231 31.99 514.55 FB 1 arabinose 0.653 20.125 0.019 2 0.008 FB 1 d iad 1 arabinose 0.200 2 0.045 0.020 2 0.003

wild-type strains. Surprisingly, the IndCH,COOH levels were not significantly different in FBI and FBldiadl strains (Table 4). This demonstrates that Iadl does not contribute to lndCH,COOH formation under these conditions. However, the situation was different in arabinose-containing medium. The formation of IndCH,COOH was 2-3-fold more reduced in FB tdiadl strains than in wild-type strains (Table 4). In addition, we observed that Ind-Prv formation was abolished both in wild- type and diadl rnutant strains grown in CM with arabinose. These results reveal a participation of Iadl in IndCH,COOH biosynthesis. We suggest that, in the FBI wild-type strain grown

Fig.5. Expression of indl in U. rnuydis grown on different carbon sources. RNA was isolated fi-or11 FB1 and FBldiadl, grown for 16 h in CM with glucose or CM with arabinose, or in minimal medium for 19 h. Lane 1. FBI grown in CM with glucose; lane 2, FB1 grown in CM with arabinose; lane 3, FBldiadl grown in CM with glucose; lane 4, FBldiadl grown in CM with ardbinose; lanes 5 and 6, FBI grown in minimal medium containing glucose as carbon source; lane 7, FBI grown in minimal medium containing 1 % ethanol as carbon source. For northern analysis about equal amounts (5 pg) of RNA were loaded. After gel electrophoresis, RNA was hybridized to a "P-labelled 1.45-kb Agc.l-B.rtBI fragment of iadl. The arrowhead points to the iadl tran- script.

in glucose-containing medium, the high levels of IndCH,COOH originating from Ind-Prv in the absence of IndCH,CH,O dehydrogenase activity mask the detection of small amounts of IndCH,COOH arising from enzymatic formation. This is supported by the observation that Ind-Prv is a rather unstable compound and breaks down spontaneously in aqueous solutions to give a number of products, including IndCH,COOH (Kaper and Veldstra, 1958 ; Sheldrake, 1973). The nonenzymatic conversion of lnd-Prv to IndCH,COOH was also observed in this study (not shown).

Iadl is required for growth on ethanol as sole carbon source. FBI and FBlAiadl strains were tested for growth in minimal medium containing 1% ethanol as sole carbon source. While strain FBI could grow with a generation time of about 5 h, FBlAiadl was unable to grow (not shown). When acetate was used as sole carbon source FBI and FBlAiadl strains showed similar growth profiles. Nothern analysis revealed that iadl expression was more than SO-fold elevated i n minimal medium containing ethanol as carbon source (Fig. 5) illustrating that Iadl is primarily engaged in the conversion of ethanol to acetate.

655 Basse et al. (ELK J. Biochern. 242)

Compared with the K,,, value of yeast mitochondria1 acetaldehyde dehydrogenase for acetaldehyde, which was determined to be 5 1 M (Saigal et al., 1991), Iadl exhibited a relative high K,, value for acetaldehyde. This difference might be inherent to the enzymes with one being cytosolic and the other being mito- chondrial. Alternatively, the differences in K,, could be caused by the assay conditions (Saigal et al., 1991) which were not comparable in the two studies.

Pathogenicity and filament formation of iudl-null mutants. To analyze if iudl affects the mating reaction and subsequent pathogenic development in planta, the mutant strains FB 1 diadl and FB2diadl were spotted together on CM charcoal plates. Formation of dikaryotic filaments was comparable to that observed in a cross of FBI with FB2 (not shown). When mixtures of FBIdiadl and FB2Aiadl were injected into corn plants, of 24 infected plants all developed tumors. In time of appearance and morphology these tumors were indistinguishable from tumors obtained from infection with a mixture of FB1 and FB2 strains. These results indicate that iadl plays no crucial role in mating, filament formation, and pathogenicity.

DISCUSSION

We purified from U. maydis an acetaldehyde dehydrogenase which is able to convert IndCH,CH,O to IndCH,COOH and isolated the corresponding gene iadl. The identity of iadl was confirmed by expression in E. coli and by generating iadl-null mutant strains in U. maydis. In these mutants, IndCH,CH,O dehydrogenase activity was severely reduced both in vivo and in vitro. diadl mutant strains therefore offer a useful tool in the in- vestigation of IndCH,COOH biosynthesis. It has been proposed that Ind-Prv is an intermediate in IndCH,COOH biosynthesis in U. maydis (Wolf, 1952; Navarre and Damann, 1990). The conversion of Ind-Prv to IndCH,COOH involves the consecutive action of Ind-Prv decarboxylase and IndCH,CH,O dehydrogenase. Since IndCH,COOH formation from tryptophan was not reduced in h a d l mutants cultivated in CM with glucose, Iadl does not contribute to IndCH,COOH formation in the presence of glucose as carbon source; this makes it unlikely that, under these conditions, a major portion of IndCH,COOH is generated through the Ind-Prv or tryptamine pathways.

The diadl mutants revealed the existence of a second IndCH2CH,0 dehydrogenase activity which is subject to glucose repression but is induced when L-arabinose is used as carbon source. In diadl mutants, Iad2 activity allows the efficient conversion of IndCH,CH,O to IndCH,COOH. In wild type strains, Iad2 supplements the activity of Iadl . Furthermore, Iad2 appears to be involved in the efficient conversion of tryptamine to IndCH,COOH. The fact that Iad2 activity is absent in glucose- containing medium may explain why tryptamine has not been detected as precursor of IndCH,COOH biosynthesis in U. rnnydis in previous investigations (Wolf, 1952; Navarre and Damann, 1990).

We could show that compared to wild-type strains Aiadl mutant strains were 2-%fold reduced in IndCH,COOH formation from tryptophan when grown in arabinose-containing medium. We infer from this result that a biosynthetic route from tryptophan to IndCH,COOH proceeds via IndCH,CH20. How- ever, it is presently unclear by which route IndCH,CH,O is generated.

Aldehyde dehydrogenases represent an enzyme family that are highly conserved in amino acid sequence accross species. Sequence analysis of Iadl displayed the highest degree of similarity to aldehyde dehydrogenases from A. nidulans and A. niger.

The respective genes from A. niger and A . nidulans were shown to be required for growth on ethanol (Pickett et al., 1987: O’Connell and Kelly, 1988). Since we could demonstrate that Tad1 is required for growth on ethanol and since transcription of the iadl gene is strongly stimulated during growth on ethanol, it is likely that Iadl functions primarily in the conversion of ethanol to acetate. The fact that diadl mutants did not grow on ethanol as carbon source indicates that Iad2 is either not induced under these conditions or cannot substitute for the iadl defect.

Since U. maydis is a plant pathogen and arabinose is a con- stituent of plant cell walls, it may not be coincidence that Iad2 is activated by arabinose. Unfortunately, it is not yet known which carbon sources are metabolized by U. maydis during growth in plants. The isolation of iad2 and the subsequent construction of double-mutants deleted in iadl and i d 2 will allow further insight into the role of IndCH,CH,O dehydrogenase activity in IndCH,COOH biosynthesis and the relevance of this route to pathogenicity.

We thank Karl-Heinz Braun for performing the pathogenicity assays and Michael Bolker for critical reading of the manuscript. This work was supported by the Leibniz program. C. B. was recipient of an EMBO fellowship.

REFERENCES Achatz, G., Oberkofler, H., Lechenauer, E., Simon, B., Unger, A.,

Kandler, D., Ebner, C., Prillinger, H., Kraft, D. & Breitenbach, M. (1995) Mol. Immunol. 32, 213-227.

Altschul, S . F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J . (1990) J . MoL. B i d . 215, 403-410.

von Bahr-Lindstrom, H., Hempel, J. & Jornvall, H. (1984) EUK J . Bio- chern. 141, 37-42.

Ballance, D. J. (1 990) in Molecular atid industrial ntycology-systems und upplications ,for ,filamentous ,fungi (Leong, S. A. & Berkar, R. M., eds) vol. 8, pp. 1-29, Marcel Dekker, New York.

Banuett, E & Herskowitz, I. (1989) Proc. Nail Acatl. Sci. USA 86,

Banuett, F. (1995) Annu. Rev. Genet. 29, 179-208. Bartel, B. & Fink, G. R. (1994) Proc. Nut1 Acad. Sci. USA 91, 6649-

Bartling, D., Seedorf, M., Mithofer, A. & Weiler, E. W. (1992) EUK 1.

Bartling, D., Seedorf, M., Schmidt, R. C. & Weiler, E. W. (1994) Proc.

Bolker, M., Urban, M. & Kahmann, R. (1992) Cell 68, 441 -450. Bolker, M., Bohnert, H. U., Braun, K.-H., Gorl, J . & Kahmann, R.

Bradford, M. M. (1976) Anal. Biochem. 166, 368-379. Braun, T., Bober, E., Singh, S., Agarwal, D. P. & Goedde, H. W. (1987)

Callow, J. A. & Ling, I. T. (1973) Physiol. Plant Pathol. 3, 489-494. Christensen, J. J. (1963) Am. Phytopath. SIX. M o n o g ~ 2. Church, G. M. & Gilbert, W. (1984) Proc. Nail Acad. Sc i USA 81,

De Luca, V., Marineau, C. & Brisson, N. (1989) Proc. Nut1 Acad. Sci.

Dunn, T. J., Koleske, A. J., Lindahl, R. & Pitot, H. C. (1989) J. Hiol.

FarrCs, J., Wang, T. T., Cunningham, S. J. & Weiner, H. (1995) Biocltem-

Gillissen, B., Bergemann, J., Sandmann, C., Schroeer, B.. Bijlkcr, M. &

Glickmann, E. & Dessaux, Y. (1995) App1. Bzviron. Microbiol. 61,793-

Gordon, S. A. & Weber, R. P. (1951) Plunt Physiol. 26, 192-195. Gruen, H. E. (1959) Annu. Rev. Plant PhyJio1. JO, 405-440. Hempel, J., von Bahr-Lindstrom, H. & Jornvall, H. (19x4) Eut: J. Bio-

Hoffman, C. S. & Winston, F. (1987) Gene 57, 267-272.

5878-5882.

6653.

Biochem. 205, 417-424.

Natl Acad. Sci. USA 91, 6021 -6025.

(1995) Mol. Gen. Genet. 248, 547-552.

FEBS Lett. 215, 233-236.

1991 - 1995.

USA 86, 2582-2586.

Chem. 264, 13057-13065.

istry 34, 2592-2598.

Kahmann, R. (1992) Cell 68, 1-20.

796.

chern. 141, 21-35.

65 6 Basse et al. ( E m J . Biochem. 242)

Holliday, R. (1961) Genet. Res. Camb. 2, 204-230. Holliday, R. (1974) in Handbook of genetics (King, R. C., ed.) vol. 1,

pp. 575-595, Plenum, New York. Kahmann, R., Romeis, T., Bolker, M. & KPmper, J. (1995) Curi: @in.

Genet. Dev. 5 , 559-564. Kamper, J . , Reichmann, M., Romeis, T., Bolker, M. & Kdhrnann, R.

(1995) Cell XI , 73-83. Kaper, J. M. & Veldstra, H. (1958) Biochim. Biophys. A(,ta 30, 401-

420. Klee, H., Montana, A., Horodyski, F., Lichtenstein, C., carfinkel, D.,

Fuller, S., Flores, C., Peschon, J., Nester, E:. W. & Gordon, M. P. (1984) Proc. Nut1 Acad. Sci. USA 81, 1728- 1732.

Klee, H. & Estelle, M. (1991) Annu. Rev. Plant Fhysiol. Plant Mol. Biol.

Kobayashi, M., Suzuki, T., Fujita, T., Masuda, Ril. & Shirnizu, S. (1995)

Laernrnli, U. K. (1970) Nature 227, 680-685. Ludwig-Miiller, J. & Hilgenberg, W. (1988) Physiol. Plant. 74, 240-

Marurno, S. (1986) in Clzemistty ofplant hurmnnes (Takahashi, N., ed.)

Navarre, A. & Damann, K. E. (1990) Plant Physiol. (absfi:) 80, 1055. Nonndnly, J., Cohen, J. D. & Fink, G. R. (199.3) Pmc. Nut1 Acad. Sci.

O’Connell, M. J. & Kelly, J. M. (1988) Curl: Genet. 14, 95-103. Percival, F. W. & Purves, W. K. (1974) Plant Physiol. 54, 601-607. Pickett, M., Gwynne, D. I., Buxton, F. P., Elliott, R., Davies, R. W.,

Lockington, R. A,, Scazzocchio, C. & Sealy-Lewis, H. M. (1987) Gene 51, 217-226.

42, 529-551.

Proc. Nut1 Acad. Sci. USA 92, 714-718.

250.

pp. 9 -56, CRC Press, Boca Raton, FL.

USA 90, 10355-10359.

Puhalla, J. E. (1968) Genetics 60, 461 -474. Rowell, J. B. & DeVay, J. E. (1954) Phyfopathology 44, 356-362. Rowell, J. B. (1955) Phytopurhology 45, 370-374. Saigal, D., Cunningham, S. J., Farrhs, J. & Weiner, H. (1991) J . Bacte-

r id . 173, 3199-3208. Sambrook, J. , Fritsch, E. F. & Maniatis, T. (1989) in Molecular cloning:

a laboratory manuul, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.

Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Nut1 Acad. Sci.

Schauwecker, F., Wanner, G. & Kahmann, R. (1995) Bid . Chern. Hoppe-

Schrnitt, M. E., Brown, T. A. & Trurnpower, B. L. (1990) Nucleic Acids

Schulz, B., Banuett, F., Dahl, M., Schlesinger, R., Schafer, W., Martin,

Sheldrake, A. R. (1973) Biol. Rev. 48, 509-559. Spellig, T., Bolker, M., Lottspeich, F., Frank, R. W. & Kahmann, R.

Tsukuda, T., Carleton, S., Fotheringharn, S. & Holloman, W. K. (1988)

Turian, G. & Hamilton, R. H. (1960) Biochim. Biuphy. Actu 41, 148-

Wang, X. & Weiner, H. (1995) Biochemistry 34, 237-243. Wightman, F. & Cohen, D. (1968) in Biochemistry and physiology of

plant growth substances (Wightman, F. & Setterfield, G., eds) pp. 273-288b, Runge Press, Ottawa.

USA 74, 5463-5467.

Seyler, 376, 617-625.

Res. 18, 3091-3092.

T., Herskowitz, I. & Kahrnann, R. (1990) Cell 60, 295-306.

(1994) EMBO J . 13, 1620-1627.

Mu/. Cell. Bioi. 8, 3703-3709.

150.

Wolf, F. T. (1952) Proc. Natl Acad. Sci. USA 38, 106-111. Yamada, T., Palm, C. J., Brooks, B. & Kosuge, T. (1985) Prnc. Nut1

Atad. Sci. USA 82, 622-6526

two potential indole-3-acetaldehyde dehydrogenases in the phytopathogenic fungus ustilago maydis

Documents