A Novel Alcohol Oxidase/RNA-binding Protein with Affinity for Mycovirus Double-stranded RNA from the Filamentous FungusHelminthosporium (Cochliobolus)victoriae

We have cloned and sequenced a novel alcohol oxidase (Hv-p68) from the filamentous fungusHelminthosporium (Cochliobolus)victoriae that copurifies with mycoviral double-stranded RNAs. Sequence analysis revealed that Hv-p68 belongs to the large family of FAD-dependent glucose methanol choline oxidoreductases and that it shares significant sequence identity (>67%) with the alcohol oxidases of the methylotrophic yeasts. Unlike the intronless alcohol oxidases from methylotrophic yeasts, a genomic fragment of the Hv-p68 gene was found to contain four introns. Hv-p68, purified from fungal extracts, showed only limited methanol oxidizing activity, and its expression was not induced in cultures supplemented with methanol as the sole carbon source. Northern hybridization analysis indicated that overexpression of Hv-p68 is associated with virus infection, because significantly higher Hv-p68 mRNA levels (10- to 20-fold) were detected in virus-infected isolates compared with virus-free ones. We confirmed by Northwestern blot analysis that Hv-p68 exhibits RNA binding activity and demonstrated that the RNA-binding domain is localized within the N-terminal region that contains a typical ADP-binding β-α-β fold motif. The Hv-p68 gene, or closely similar genes, was present in all species of the genusCochliobolus but absent in the filamentous fungus,Penicillium chrysogenum, as well as in two nonmethylotrophic yeasts examined. This study represents the first reported case that a member of the FAD-dependent glucose methanol choline oxidoreductase family, Hv-p68, may function as an RNA-binding protein.

It has recently become increasingly clear that cellular factors play important roles in the transcription and replication of RNA viruses (1). The double-stranded RNA (dsRNA) 1 genomes of the two mycoviruses (2,3) known to infect the plant pathogenic fungus Helminthosporium victoriae (teleomorph: Cochliobolus victoriae; synonym: Bipolaris victoriae) consists of two genes that encode a capsid protein (CP) and an RNA-dependent RNA polymerase (RDRP). The CP and RDRP genes are either present on the same dsRNA segment, as in the case of the totivirus H. victoriae 190S virus, or on two separate dsRNA segments, in the case of the putative chrysovirus H. victoriae 145S virus (Hv145SV). It is thus not surprising that these viruses subvert host proteins for their own use. We have previously reported that host enzymes (a protein kinase and a protease) are involved in post-translational modification of the Hv190SV CP (4 -6).
H. victoriae 190S virus (Hv190SV) has been extensively studied (2)(3)(4)(5)(6)(7)(8)(9), and is classified as a definitive member of the genus Totivirus in the family Totiviridae (10,11). The Hv145SV, on the other hand, has only been subjected to limited biochemical characterization (2). Because of similarity in size and number of dsRNA segments between the Hv145SV and viruses in the genus Chrysovirus in the family Partitiviridae, the Hv145SV was tentatively classified as a member of the genus Chrysovirus (12). We have recently (13) isolated a cellular protein, Hv-p68, that copurifies with viral dsRNA from H. victoriae isolates infected with both Hv190SV and Hv145SV. Additionally, Hv-p68 was demonstrated to be present as a minor component in the viral capsids (13). Our initial biochemical characterization studies of Hv-p68 indicated that it occurs in vivo as an octamer and that it is consistently present in higher levels in virus-infected H. victoriae isolates than in virus-free ones. We have also demonstrated by a gel retardation assay that Hv-p68 has RNA binding activity (13).
In the present study, we report the isolation and complete nucleotide sequence of a cDNA clone of Hv-p68. We present evidence that Hv-p68 belongs to the FAD-dependent GMC family of oxidoreductases and that it has high sequence similarity to the alcohol oxidases of methylotrophic yeasts. Furthermore, we show that overexpression of Hv-p68 is associated with virus infection and that the RNA-binding domain of Hv-p68 is localized in the N-terminal region that contains a canonical ADPbinding ␤-␣-␤ fold motif. In addition, we demonstrate by Southern hybridization analysis that the Hv-p68 gene, or closely similar genes, is present in all Cochliobolus species examined but not in the filamentous fungus Penicillium chrysogenum nor in two nonmethylotrophic yeast species.

EXPERIMENTAL PROCEDURES
Fungal Isolates-Three H. (Cochliobolus) victoriae isolates that differ in their virus content were used. Isolate A-9 (ATTC 42018), a diseased isolate known to contain both Hv190SV and Hv145SV (1), was routinely used as a source for virions and Hv-p68. A single conidial isolate of H. victoriae isolate B-2 (ATCC 42020), designated B-2ss, has recently been demonstrated to be devoid of virus and was used as a representative of a cured fungal isolate (13). The virus-free H. victoriae isolate 408, used in previous virus transmission studies (14), was used as an example of a naturally occurring virus-free isolate. Isolates of C. heterostrophus, C. zeicola, and C. sativum, provided by M. Carson (North Carolina State University), were used in Southern screening for the Hv-p68 gene. DNA from the filamentous fungus P. chrysogenum (ATCC 9480), Saccharomyces cerevisiae (strain 717; provided by Reed Wickner) and Schizosaccharomyces pombe (strain SP-Q01; Stratagene) were also included in the Southern analysis.
Isolation of mRNA and cDNA Synthesis-Total RNA was isolated from 4-day-old stationary cultures of H. victoriae strain A-9 by the guanidinium isothiocyanate/phenol method (15). Polyadenylated RNA was purified from total RNA by oligo(dT)-Sepharose affinity chromatography (Amersham Pharmacia Biotech). First and second strand cDNA were synthesized from poly(A) RNA using the SuperScript cDNA synthesis kit (Life Technologies, Inc.).
PCR Amplification of Hv-p68 cDNA Sequences-Forward and reverse degenerate primers corresponding to the N-terminal sequence of the native Hv-p68 protein (primer P1; see Fig. 1) and to amino acid sequencing data of an internal tryptic peptide (primer P2; see Fig. 1) were used along with the double-stranded cDNA synthesized to fungal mRNA (Hv-cDNA) to prime the amplification of an Hv-p68 cDNA fragment. Amplification reactions were carried out using Platinum High Fidelity Taq DNA polymerase (Life Technologies, Inc.) and cycling parameters specified for TD-PCR (94°C for 4 min, 94°C for 1 min, 60°C (⌬ Ϫ0.5°C per cycle) for 2 min, 72°C for 3 min, 30 cycles; 94°C for 1 min, 50°C for 2 min, 72°C for 3 min, 10 cycles; 72°C for 12 min). The PCR product generated (1.5 kbp) was gel-purified using GeneClean (BIO 101) and blunt-end-cloned into an EcoRV-digested pZErO vector (Invitrogen). Four clones were subjected to automated sequencing (ABI), and the sequencing information was used to design a sequencespecific primer (P3). A 750-bp fragment corresponding to the 3Ј-end of the Hv-p68 mRNA was amplified from synthesized double-stranded cDNA by PCR using an oligo(dT) NotI primer and P3, a primer based on Hv-p68 cDNA sequences at nucleotide (nt) positions 1396 -1420 (see Fig. 1). Amplification reactions were carried out using Platinum High Fidelity Taq DNA polymerase (Life Technologies, Inc.), and the cycling parameters for TD-PCR were as specified above. The amplified PCR product was gel-purified by GeneClean (BIO 101), blunt-end-cloned into an EcoRV-digested pZErO vector (Invitrogen), and subjected to automated sequencing (ABI).
Construction of a cDNA Library-Poly(A) ϩ RNA isolated from a 3-day-old stationary culture of isolate A-9 of H. victoriae was used to generate a cDNA library in the lambda ZipLox vector following the procedures for the SuperScript Lambda cloning system (Life Technologies, Inc.). The library was packaged using Gigapack II Gold (Stratagene) and amplified once to a high titer (8 ϫ 10 10 plaque-forming units/ml) in Escherichia coli cells. Hv-p68 lambda clones were isolated by screening ϳ800,000 recombinant phages with a 32 P-labeled Hv-p68 probe generated by nick-translation (Life Technologies, Inc.) of the 1.5-kbp DOP-PCR product. Replica lifts (Hybond N, Amersham Pharmacia Biotech) from plated phage were hybridized for 14 -16 h at 42°C in 50% formamide/hybridization solution (50% formamide/1ϫ Denhardt's solution/5ϫ SSC/0.5% SDS/20 mM sodium phosphate, pH 7.0) and washed under conditions of high stringency. Pure lambda clones were obtained by two successive rounds of screening with the 32 Plabeled 1.5-kbp cDNA probe. Four independent lambda clones were selected for in vivo excision into the cloning vector pZL1 as described by the manufacturer (Life Technologies, Inc.), and subjected to automated sequencing.
Nucleotide Sequencing and Analysis-All sequencing was performed by dideoxy-termination sequencing using the Rhodamine-Terminator sequencing kit (ABI) and an ABI600 automated sequencer. Automated sequencing was performed on both strands with universal primers and gene-specific "walking" primers. Sequence homology searches of Gen-Bank, Swissprot, and EMBL data bases were conducted using the BLAST and FASTA programs (16,17). Paired and multiple sequence alignments were carried out with the programs Best-Fit, GAP, PILEUP, and PRETTY (Wisconsin GCG software package (18)). Motif and signature pattern searches were conducted using the program Motif (19). Sequence alignments and phylogenetic analysis were performed using the program ClustalW (20). Trees generated by ClustalW were displayed using TreeView (21). Predictions for protein sorting signals and subcellular localization sites on the deduced amino acid sequence were performed using PSPRT II (22).
Expression of the Hv-p68 ORF in Bacteria and Affinity Purification of the C-terminal His-tagged Protein-A cDNA corresponding to the coding region of Hv-p68 was obtained by PCR amplification using genespecific primers and one of the isolated lambda Hv-p68 cDNA clones as a template. Primer sequences corresponding to the 5Ј-and 3Ј-ends of the Hv-p68 ORF and containing restriction-enzyme site sequences for cloning were: pET-p68-EcoRI FOR, 5Ј-CCGGAATTCGACGATCCCG-GACGACGAGGTTGATATTATCG-3Ј and pET-p68-NotI REV, 5Ј-AATATTCTTAGCGGCCGCCAAGCGCGATAATCCAGCAATC-3Ј. Amplification reactions were carried out using standard PCR conditions and Platinum High Fidelity Taq DNA polymerase (Life Technologies, Inc.). The PCR-amplified product (ϳ2 kbp) was gel-purified by agarose gel electrophoresis followed by extraction with GeneClean (Bio101), digested with EcoRI-NotI, and ligated into an EcoRI-NotI-digested pET22(b)ϩ vector (Novagen). The construct, pETp68, was used to transform E. coli strain BL21(DE3) cells according to the manufacturer's instructions (Novagen). For purification of the bacterially expressed Hv-p68, 100-ml cultures of transformed BL21(DE3) cells were grown in LB media at 37°C to a density of 1.0 A 600 nm . Protein was expressed by induction with 0.5 mM isopropyl-1-thio-␤-D-galactopyranoside and incubation for a period of 5 h at 30°C. Cells were collected by centrifugation at 4000 ϫ g for 5 min at 4°C, resuspended in 0.1 M sodium phosphate buffer (pH 7.2) containing 0.1% Triton X-100, and treated with lysozyme (100 mg/ml final concentration) for 15 min, at 30°C followed by treatment with DNase 1 (10 g/ml final concentration) in 0.1 M sodium phosphate buffer (pH 7.2) containing 10 mM MgCl 2 for 10 min at 30°C. Bacterial extracts were briefly sonicated and centrifuged at 14,000 ϫ g for 15 min at 4°C, and the pellet containing the inclusion bodies was resuspended in 6 M guanidinium hydrochloride in binding buffer (0.5 M NaCl, 20 mM Tris-HCl, pH 8.0). The protein was solubilized by incubation on a rocker platform for 2 h at 4°C. The His-tagged Hv-p68 was purified by Ni-NTA chromatography, under denaturing conditions, as described by the manufacturer (Novagen). The purified protein was renatured by step-dialysis with successive changes of 50 mM Tris-HCl, pH 7.5, 50 mM NaCl buffer containing 6, 4, or 2 M urea or no urea. The final protein preparation was concentrated using a Centricon-10 device (Amicon).
Bacterial Constructs for Expression of N-terminal and C-terminal Deletion Mutants of Hv-p68 -Constructs were generated in the bacterial expression vector pET22(b)ϩ containing large deletions at either the N-or C-terminal sequence of the Hv-p68 coding region. A construct for expression of a 271-amino acid N-terminal deletion mutant was generated by subcloning an SstI-NotI restriction fragment from construct pETp68 (corresponding to nt positions 813-2111 in the Hv-p68 cDNA; see Fig. 1) into a similarly digested pET22(b)ϩ vector. A 160amino acid C-terminal deletion construct was generated by subcloning the 1.5-kbp EcoRI-NotI fragment from construct pZp68 -1.5kbp (corresponding to nt positions 4 -1512; see Fig. 1) into the EcoRI-NotI-digested sites of pET22(b)ϩ. 50-ml cultures of BL21(DE3) cells transformed with the full-length Hv-p68 (pETp68) and the N-and C-terminal truncated constructs were grown in LB media at 37°C to a density of 1 A 600 nm . Protein was expressed by induction with 1 mM isopropyl-1-thio-␤-D-galactopyranoside and incubation for a period of 4 h at 37°C. Cells were collected by centrifugation at 4000 ϫ g for 10 min at 4°C and resuspended in 50 mM Tris, pH 7.5, 0.1% Triton X-100. The bacterial extracts were sonicated and centrifuged at 14,000 ϫ g for 15 min at 4°C to separate the soluble and the insoluble component (inclusion bodies). The overexpressed proteins were purified from the inclusion bodies by repeated resuspension of the pellet (in 50 mM Tris, pH 7.5, 0.1% Triton X-100) by brief sonication followed by centrifugation at 27,000 ϫ g, 10 min at 4°C. Five cycles of washing/centrifugation of the resultant inclusion bodies pellet resulted in 90 -95% removal of contaminant proteins from the Hv-p68 preparation. The inclusion bodies were finally resuspended in 500 l of 50 mM Tris buffer, pH 7.5, containing 50 mM NaCl and an equal volume of 2ϫ SDS-PAGE sample buffer, and the proteins were separated by SDS-PAGE for Northwestern analysis.
Nucleic Acid Isolation and Hybridization-Total RNA was isolated by a modification of the procedure of Chomczynski and Sacchi (15). H. victoriae isolates A-9, B-2ss, and 408 were grown in stationary cultures in Fernbach flasks containing 200 ml of potato dextrose broth medium supplemented with 0.5% yeast extract (PDBY) for 8 days at room temperature. For the induction studies, mycelium from H. victoriae isolate B-2ss was collected by centrifugation (3000 ϫ g, 10 min) and rinsed with minimal medium (MM) supplemented with 1% glucose (23). The mycelium was homogenized in a blender, and the homogenate was transferred to Fernbach flasks containing MM supplemented with 1% dextrose and grown for 3 days on a shaker incubator at 22°C. The mycelium was harvested and rinsed twice with MM, homogenized, and transferred to flasks containing fresh MM supplemented with either 1% glucose or 1% methanol and grown on a shaker incubator at 22°C for 3 days. Mycelia from PDBY and MM cultures were collected by straining through two layers of Miracloth (Calbiochem) and pulverized in liquid nitrogen and 100 ml of a guanidinium-denaturing solution/water-saturated phenol/2 M sodium acetate buffer, pH 4.0 (48:47:5), added (per 10 g of wet tissue). The suspension was incubated at room temperature for 15-20 min and centrifuged at 10,000 ϫ g for 15 min, and the supernatant was extracted by addition of 0.2 volume of chloroform (20 ml). The aqueous phase was re-extracted with chloroform/isoamyl alcohol (24:1), and the RNA was precipitated by addition of 1 volume of isopropanol. The RNA pellets were resuspended in diethyl pyrocarbonate treated water and subjected to two rounds of high salt-isopropanol precipitation at room temperature (0.5 volume of 1.2 M NaCl/0.8 M sodium citrate and 0.5 volume of isopropanol), followed by precipitation in 2 M LiCl (4 h at 4°C). For Northern hybridization, 25 g of the total RNA was electrophoresed on 1% agarose-formaldehyde gels and transferred to a nylon membrane (Hybond N, Amersham Pharmacia Biotech) by alkaline downward transfer (TurboBlotter, Schleicher & Schuell), and the blot was subjected to UV cross-linking. Blots were hybridized with [␣-32 P]dCTP-random-primed (Promega) Hv-p68 cDNA probe at 42°C for 16 -18 h in 50% formamide hybridization solution (50% formamide/1ϫ Denhardt's solution/5ϫ SSC/0.5% SDS/20 mM sodium phosphate, pH 7.0) and washed using high stringency conditions (2ϫ SSC, 0.2% SDS, 23°C for 10 min; 0.2ϫ SSC, 0.1%SDS, 58 -65°C for 30 -60 min). The Hv-p68 probe consisted of a gel-purified, GeneClean (BIO 101)-extracted EcoRI-SstI restriction fragment (850 bp) from a pZErO construct containing Hv-p68 cDNA. Genomic DNA was isolated from filamentous fungal species by a modification of the protocol described by Yoder (23). Mycelia from 200-ml shake cultures of the following filamentous fungi were grown in a complete medium lacking glucose for 4 -5 days: H. (Cochliobolus) victoriae, isolates A-9, B-2ss, and 408; C. heterostrophus; C. zeicola; C. sativum; and P. chrysogenum. Mycelia were collected by straining through two layers of Miracloth (Calbiochem), freeze-dried, and ground in liquid nitrogen. The powdered mycelia were resuspended in 20 ml of lysis buffer (50 mM Tris-HCl, pH 8.0, 150 mM EDTA, 1% Sarkosyl) containing 6 mg of proteinase K (Life Technologies, Inc.) and incubated at 50 -55°C for 60 min with occasional mixing by inversion. The volume was increased to 50 ml with lysis buffer, and the suspension was centrifuged at 10,000 ϫ g for 10 min. The supernatant was extracted with Tris buffer-saturated phenol (10 mM Tris-HCl, pH 7.5). The aqueous phase was re-extracted by chloroform/isoamyl alcohol (24:1), ethanol-precipitated (in 0.3 M sodium acetate), and the resultant pellet was dried and resuspended in 2 ml of Tris-EDTA (TE) buffer. The DNA preparation was treated with RNase A (20 g/ml) at 37°C for 20 min, the volume increased to 10 ml with TE buffer, and the DNA was re-extracted with phenol/chloroform/isoamyl alcohol (25:24:1) followed by extraction with chloroform/isoamyl alcohol (24:1). The DNA was ethanol-precipitated in the presence of 0.3 M sodium acetate, and the pellet was air-dried and resuspended in TE buffer. The DNA was quantified by agarose-gel electrophoresis and spectrophotometric analysis. For the isolation of genomic DNA from yeast, 100-ml suspension cultures of S. cerevisiae and S. pombe were grown in YPD medium at 30°C with shaking to a density of 2 A 600 nm . Cells were pelleted by centrifugation at 3000 ϫ g for 10 min, and the cell pellet was resuspended in 2 ml of lysis buffer (2% Triton X-100, 1% SDS, 100 mM NaCl, 1 mM EDTA, 10 mM Tris-HCl, pH 8.0) and 2 ml of phenol/chloroform/isoamyl alcohol (25:24:1). Three grams of acidwashed glass beads (450 -500 m, Sigma) were added, and the suspension was vortexed intermittently, at maximum speed, for 6 -10 min. Two milliliters of TE buffer was added, and the phases were separated by centrifugation at 6000 ϫ g for 10 min. The aqueous phase was re-extracted with chloroform/isoamyl alcohol (24:1) and ethanol-precipitated. The DNA pellets were dried, resuspended in 1 ml of TE buffer, and treated with 50 g of RNase A (at 37°C for 15 min), and the DNA was reprecipitated by addition of 100 l of 4 M ammonium acetate and 2.5 ml of ethanol. The DNA was resuspended in TE buffer and quantified by agarose-gel electrophoresis and spectrophotometry.
Alcohol Oxidase Assay-Alcohol oxidase activity in Hv-p68 preparations from H. victoriae and the bacterially expressed Hv-p68 was determined by a spectrophotometric assay, which measures the rate of conversion of methanol to hydrogen peroxide (24). Production of hydrogen peroxide after addition of methanol to the reaction was determined by monitoring the change in absorbance at 405 nm (at 25°C) over a 5-min period using 2,2Ј-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid (A-1888, Sigma) as substrate and horseradish peroxidase (P-8250, Sigma). Commercially purified Pichia pastoris alcohol oxidase (E.C. 1.1.3.13; A2404, Sigma) was used in parallel enzymatic assays for comparison. Protein content was measured using a Bio-Rad protein determination reagent. Activity was given as ⌬ A 405 nm /min (units) per mg of protein.
Radiolabeled in Vitro Transcripts-Radiolabeled transcripts (riboprobes) were synthesized by in vitro transcription from a pUC-construct pT7Hv190S, containing a full-length cDNA of the Hv190S dsRNA (3), and pZ-2-28 containing a near-full length cDNA of the Hv145S dsRNA-4 in the pZErO vector. The positive-strand transcripts were generated by incubation of pT7Hv190S and pZ-2-2, previously linearized with BamHI and NotI, respectively, with T7 RNA polymerase (Stratagene) at 37°C for 60 min in the presence of [␣-32 P]UTP. The transcription reactions were terminated by treatment with RQ1 RNase-Free DNase (2 units, for 15 min at 37°C; Promega), and the reaction mixtures were extracted with phenol/chloroform, ethanol-precipitated, and resuspended in diethyl pyrocarbonate-treated water. The specific activity of the riboprobes was determined by scintillation counting of the trichloroacetic acid-precipitable material.
Northwestern Blotting Analysis-Proteins were separated on 10% SDS-polyacrylamide gels and transferred to a nitrocellulose membrane (BA85, 0.45 m, Schleicher & Schuell) using a wet transfer apparatus (Bio-Rad). The membranes were then washed three times at room temperature, 20 min each time, in buffer A (10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 50 mM NaCl, 0.1% Triton X-100, 1ϫ Denhardt's reagent) to renature the proteins. Blotted proteins were probed by incubating with the 32 P-labeled transcripts (10 5 cpm) in the same buffer for 1 h at room temperature. The blots were then washed three times for 2 min each using the same buffer, air-dried, and autoradiographed. After autoradiography, blots were rinsed briefly in Tris-buffered saline-Tween 20 (TBS-T) and incubated for 1 h in blocking buffer (TBS-T, 5% nonfat dry milk). Blots were processed for immunoblotting using the polyclonal antibodies against Hv-p68 (13) and goat anti-rabbit IgG-alkaline phosphatase conjugate as secondary antibodies. Nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate (Promega) were used as substrates for colorimetric detection.

RESULTS
Isolation and Characterization of Hv-p68 cDNA Clones-cDNA clones for the Hv-p68 gene were isolated by a combination of DOP-PCR and conventional library screening. The amino acid sequencing data previously obtained by Edman degradation analysis (13) were used to design two sets of degenerate primers for DOP-PCR. Mixed pools of oligonucleotides corresponding to the N-terminal sequence of the native protein (primer P1; Fig. 1) and to an internal tryptic peptide (primer P2; Fig. 1) were used to direct the amplification of an Hv-p68 PCR product with Hv-cDNA as a template. Amplifications using conditions for TD-PCR yielded a single major PCR product ϳ1.5 kbp in size. This product was blunt-end-cloned into a pZErO vector (Invitrogen) and subjected to automated sequencing. The sequencing data for the 3Ј-end region of the 1.5-kbp fragment was used to design a sequence-specific primer for amplification of the 3Ј-end region of the Hv-p68 cDNA. A 750-bp PCR product corresponding to the 3Ј-end region of Hv-p68 was amplified using Hv-cDNA as template and the sequence-specific primer, P3 (Fig. 1), and an oligo(dT) primer. The 750-bp PCR product hybrid-ized strongly in Southern blots probed with the 32 P-labeled 1.5-kbp PCR product (data not shown). The 750-bp product was cloned into the vector pZErO and sequenced. The deduced amino acid sequence of the PCR-amplified products matched perfectly the amino acid sequence data derived from the N-terminal sequence of the gradient purified Hv-p68 protein as well as that from the internal peptide sequence (Fig. 1, amino acid residues printed in boldface letters).
To validate the clones obtained by the DOP-PCR cloning procedure and to isolate a full-length cDNA, we generated a cDNA library for C. victoriae in lambda phage. A primary screening of ϳ800,000 recombinant phages from the amplified library with the radiolabeled 1.5-kbp PCR product yielded four positive clones. These clones were independently purified by two additional rounds of plating and screening with the radiolabeled probe, and the selected lambda clones were sequenced. The sequencing data generated from all four lambda clones were consistent with that of the cDNA obtained through the DOP-PCR approach. Three of the four lambda clones had cDNA inserts with 5Ј-end truncations of the coding region. The fourth Peptides whose sequences were determined from purified Hv-p68 and which were used to design degenerate oligonucleotide pools (P1 and P2) are printed in boldface (arrows indicate the direction of primers). P3 (nt sequence in boldface) corresponds to the gene-specific primer used along with primer oligo(dT)-NotI to direct the amplification of a 750-bp 3Ј-end fragment of Hv-p68 by PCR. The tripeptide SRL (shaded box) located at the C terminus corresponds to the conserved signal, PTS1, for peroxisomal translocation. clone (2148 bp in length), however, was found to contain the entire coding region (1998 nt, including the termination codon) and an additional 37 nt derived from the 5Ј-untranslated region. The initiator ATG (starting at nt position 1, Fig. 1) is present in a favorable context according to Kozak (25) (CA-GAATGAC), and the region surrounding the ATG agrees with the consensus sequence for filamentous fungi (26) (5Ј-CAM-MATGNC, where M ϭ A or C; N ϭ A, C, G, or T). That the ATG at nt positions 1-3 corresponds to the translational start codon is supported by the Edman degradation sequencing data of the N terminus of Hv-p68 (Fig. 1). The complete nt sequence and deduced amino acid sequence for the Hv-p68 cDNA is shown in Fig. 1. The coding region of Hv-p68 codes for a 665-amino acid protein with a predicted molecular mass of 74,251 Da. This value is in agreement with our estimate of 68 kDa, based on SDS-PAGE analysis (13).
Analysis of the deduced amino acid sequence of Hv-p68 revealed the presence of three potential sites for N-linked glycosylation with consensus sequence NX(S/T) (Fig. 1, underlined). A search for cellular sorting signals with the program PSORT identified a PTS1 site at the C terminus of the deduced amino acid sequence (the tripeptide SRL shown in shaded box in Fig.  1) for translocation to the peroxisomal compartment. The Hv-p68 sequence, however, lacks all the well-established RNAbinding motifs (e.g. the RNP motif, RGG box, etc. (27,28)).
Although we have not isolated genomic clones that contain the entire Hv-p68 gene, we obtained genomic sequence information from a partial genomic clone generated by DOP-PCR amplification using the degenerate primers P1 and P2 with C. victoriae genomic DNA as a template. The single PCR product obtained (1.7 kbp) was blunt-end-cloned into the vector pZErO (Invitrogen) and sequenced. Comparison of the sequence of the genomic fragment with the corresponding sequence for the Hv-p68 cDNA indicated the coding region is interrupted by four short introns ranging in size from 50 to 59 bp (Fig. 2). Interestingly, the first intron interrupts the sequence of the coding region very close to the N terminus of Hv-p68 between amino acid positions 6 and 7 (Fig. 2). All four introns are in general agreement with the filamentous fungi consensus sequence (29) for 5Ј-splice junctions (5Ј-GTDHSY; where D ϭ A, G, or T; H ϭ A, C, or T; S ϭ C or G; Y ϭ C or T) and 3Ј-splice sites (5Ј-YAG) as well as internal putative lariat formation elements (5Ј-NNY-TNAY; where N ϭ any nt).
Sequence Comparisons-Homology search of protein data bases using the BLAST program revealed that the deduced amino acid sequence of the Hv-p68 cDNA has significant similarity with members of the large family of FAD-dependent GMC oxidoreductases (30). The highest sequence identity scores were obtained with the alcohol oxidases of the methylotrophic yeasts (with identities higher than 67%). A multiple alignment of the deduced amino acid sequences for Hv-p68 and for those reported for methanol oxidases are given in Fig. 3. Hv-p68 shares the five signature patterns (blocks A-E (19)) characteristic of the flavoproteins in the GMC oxidoreductase superfamily (Fig. 3, shaded areas). Block A comprises part of the FAD ADP-binding region with its typical ␤-␣-␤ fold motif ( Fig. 3; see also Ref. 31). The three Gly residues between the first ␤-sheet and ␣-helix as well as the acidic amino acid at the end of the second ␤-sheet are absolutely conserved in these flavoenzymes ( Fig. 3; printed in boldface), as well as in other homologous oxidoreductases (32,33). Although the functions of the other signature patterns (blocks B-E) are not yet definitively known, their putative functions have been discussed in a recent study (34). Block B comprises the flavin attachment loop (34). A conserved region near the C terminus (Fig. 3, boxed) that is shared by GMC oxidoreductases (33,35) is believed to correspond to the active site. Differences in substrate specificity of alcohol oxidases (aliphatic or aromatic alcohols) are probably due to differences in the active site of these enzymes (35).
The FAD-dependent GMC oxidoreductases have been reported to be evolutionarily related (30). The results of pairwise alignments, using the GAP program, of Hv-p68 and six members of the family of FAD-dependent GMC oxidoreductases, including four alcohol oxidases from methylotrophic yeasts are shown in Fig. 4A. Hv-p68 showed significantly high similarity and identity scores to the alcohol oxidases; the highest values were recorded for the AOX1 from P. pastoris with 75.5% similarity and 69.7% identity (Fig. 4A). Likewise, phylogenetic analysis showed that Hv-p68 forms a cluster with the alcohol oxidases from methylotrophic yeasts that is strongly supported by bootstrap analysis (Fig. 4B).
Bacterial Expression and Immunological Verification of the Hv-p68 cDNA Clones-The coding region of the Hv-p68 fused to a C-terminal His-tag was expressed in the bacterial expression vector pET22(b)ϩ. The overexpressed protein was purified on a Ni-NTA column and analyzed by SDS-PAGE and Western blotting. The bacterially expressed protein reacted strongly with antibodies against Hv-p68 (Fig. 5A), thus providing immunological verification for the Hv-p68 cDNA clone. The antiserum to Hv-p68 also reacted with P. pastoris alcohol oxidase (a commercial preparation from Sigma), but the intensity of the Western blot band was clearly lower than that obtained with the homologous antigen (Fig. 5A).
Localization of the RNA Binding Domain-The RNA binding activity of Hv-p68 has previously been demonstrated in gelretardation assays using Hv-p68 preparations purified from fungal extracts (13). To localize the RNA-binding domain of Hv-p68, we compared the RNA binding activity of bacterially expressed full-length and deletion mutants of Hv-p68 ORF using Northwestern blotting analysis (this study). Hv-p68 pu- FIG. 2. Nucleotide sequence of a partial genomic clone of Hv-p68 and positions of four introns (I-IV). The nt sequence of a partial genomic clone (1723 bp) was compared with a corresponding Hv-p68 cDNA clone (1511 bp). Transcribed nts in the genomic sequence are shown by uppercase letters (corresponding nts in the cDNA sequence are indicated by dots), whereas intron sequences are in lowercase letters. The putative 5Ј-and 3Ј-splice sites are printed in boldface; the consensus internal sites are underlined.
rified from the fungus and the Ni-NTA-purified bacterially expressed Hv-p68 were subjected to SDS-PAGE and blotted onto a nitrocellulose membrane, and the blots were incubated with radiolabeled riboprobes prepared from cloned cDNAs to Hv190SV or Hv145SV dsRNAs. As shown in Fig. 5B, strong binding signals corresponding to binding of the RNA probe to Hv-p68 were detected. Similar results were observed using either the Hv190S or Hv145S riboprobe. The bands on the autoradiograph corresponded to the prominent protein bands on Western blots of the fungal-and the bacterially expressed Hv-p68, which reacted with antibodies to Hv-p68 (Fig. 5A). The binding of the P. pastoris AOX to the Hv190S and 145S riboprobes was very weak (detectable only in an overexposed autoradiograph) compared with that observed for either the fungal-or bacterially expressed Hv-p68.
Data base searches with the deduced amino acid sequence of FIG. 3. A multiple sequence alignment of the deduced amino acid sequence of Hv-p68 and four alcohol oxidases from methylotrophic yeasts. The alignment was generated by the PILEUP program, and the consensus was produced using the PRETTY program. The amino acids comprising the five signature blocks, A-E, typical of GMC flavoproteins are shaded. The position of the conserved FAD ADP-binding region (a typical ␤-␣-␤ fold motif) near the N terminus is indicated by a horizontal line. The three glycine residues in the GXGXXG motif between the first ␤-sheet and ␣-helix and also the acidic amino acid, glutamic acid (E), at the end of the second ␤-sheet are absolutely conserved in these flavoproteins. The conserved region near the C terminus, corresponding to the active site, is boxed. The C-terminal tripeptide with consensus sequence for the peroxisomal translocation signal, PTS1, is shaded.
Hv-p68 indicated the absence of consensus sequences for the well-characterized RNA-binding motifs previously reported for RNA-binding proteins (e.g. RNP motif, RGG box, etc. (27, 28)). Earlier studies with several NAD ϩ -dependent dehydrogenases and oxidoreductases have revealed that these enzymes may also function as RNA-binding proteins (36). The structural feature that these enzymes have in common is a region containing a typical ␤-␣-␤ dinucleotide binding fold (Rossman fold). It has been proposed that the dinucleotide-binding sites and the RNA-binding domains are structurally related (36). It was of interest, therefore, to determine whether the FAD ADPbinding domain in Hv-p68 might also function as an RNAbinding domain. To localize the RNA-binding domain, we expressed N-and C-terminal truncations of the Hv-p68 protein in E. coli and compared their RNA-binding activities by Northwestern blot analysis. The bacterially expressed proteins were purified from the inclusion bodies as single major protein bands with molecular masses of ϳ54 and 62 kDa for the N-and C-terminal deletions, respectively (Coomassie-stained gel; Fig.  6). In Northwestern blots, binding of either of the RNA probes was similar for both the full-length and the C-terminally truncated Hv-p68. On the other hand, deletion of the N-terminal region of Hv-p68 completely abolished all binding to the RNA probe (Fig. 6). These results indicate that the RNA-binding domain is localized within the N-terminal region of Hv-p68.
Functional Analysis-The methanol oxidizing activity of Hv-p68 purified from fungal extracts and the Ni-NTA purified bacterially expressed Hv-p68 was tested by spectrophotometric measurement of hydrogen peroxide formation. A commercial preparation of P. pastoris alcohol oxidase was used as a control. The results of the enzyme assay are shown in Table I. Although Pp-AOX was highly active with calculated activity of 14 units/mg of protein, Hv-p68 exhibited relatively low activity. No methanol-oxidizing activity was detected with the bacterially expressed Hv-p68. We assayed several Hv-p68 preparations purified from fungal extracts, including freshly purified FIG. 4. Pairwise sequence comparison and phylogenetic analysis of Hv-p68 and representative members of the FAD-dependent GMC oxidoreductases. A, percentage of deduced amino acid sequence similarity (identity) between Hv-p68 and representative members of the family of FAD-dependent GMC oxidoreductases. Values from paired alignments were generated by the GAP program. Percentages of similarity/ identity are given above/below the diagonal line, respectively. B, phylogenetic relationships between Hv-p68 and members of the FAD-dependent GMC oxidoreductase family. Tree distances were calculated using the Neighbor-Joining method from an alignment generated by ClustalW, and the resulting tree was displayed by TreeView. The consensus tree was supported by the analysis of 1000 bootstrap replicates (bootstrap values are given at branch nodes). AOX1-Pp (GenBank accession number U96967-1) and AOX2-Pp (accession number U96968-1) alcohol oxidases of P. pastoris; AOX-Pa alcohol oxidase (accession number A11156-1) of P. angusta; MOX-Cb (accession number Q00922) alcohol oxidase of Candida boidinii; AOX1-Pm (accession number AAF02494) and AOX2-Pm (accession number AAF02495) alcohol oxidases of P. methanolica; GOX-An (accession number A35459) glucose oxidase of Aspergillus niger; and GLD-Ds (accession number AAB87896) glucose dehydrogenase of Drosophila subobscura. GOX-An was included as an outgroup to determine the proper root placement.
FIG. 5. Immunological reactivity and RNA binding activity of bacterially expressed Hv-p68. A, an Hv-p68 cDNA containing the full-length coding region was introduced into the bacterial expression vector pET22(b)ϩ to generate construct pETp68, and the overexpressed protein was extracted and purified by Ni-NTA chromatography. Hv-p68 purified from fungal extracts (Hv-p68 lane), the Ni-NTA-purified bacterially expressed protein (NTA-pET-p68 lane), and alcohol oxidase from P. pastoris (Pp-AOX lane) were analyzed by Western blotting using antibodies to Hv-p68. The Hv-p68 antiserum reacted strongly with its homologous antigen (Hv-p68) as well as with the bacterially expressed protein (NTA-purified pETp68), but its reaction with Pp-AOX was comparatively weaker. B, RNA binding activities of purified fungal Hv-p68, NTA-purified pET-p68, and Pp-AOX were tested by Northwestern blotting analysis using 32 P-labeled in vitro transcripts of cloned cDNAs to Hv190SV dsRNA and Hv145S dsRNA-4 (riboprobes). Native and bacterially expressed Hv-p68 showed strong binding to both probes. Binding of Pp-AOX to either probe was very weak, detectable only in overexposed autoradiographs.
preparations as well as preparations that had been kept frozen for several months. Although the enzymatic activity was generally low, preparations that have been subjected to repeated freezing and thawing showed the weakest activities (data not shown).
Hv-p68 Expression in Virus-infected and Virus-free H. victoriae Isolates-Northern hybridization analysis was performed with total RNA isolated from 12-day cultures of the virusinfected isolate A-9 and the virus-free isolates 408 and B-2ss (13). In Northern blots hybridized with a radiolabeled Hv-p68specific probe, an mRNA ϳ2.3 kb in size was detected in all three fungal isolates. The size of mRNA (2.3 kb) is consistent with the estimated size for the Hv-p68 cDNA (ϳ2.1 kbp). Interestingly, the level of the Hv-p68 mRNA in the virus-infected isolate A-9 was 10-to 20-fold higher than that in the virus-free isolates. This indicates that overexpression of Hv-p68 is associated with virus infection. The increased levels of Hv-p68 transcription are consistent with the higher amounts of Hv-p68 that are normally associated with cultures of the virus-infected isolate A-9 (13).
Expression of the alcohol oxidases of the methylotrophic yeasts is tightly regulated at the transcriptional level by a dual mechanism of repression by glucose and induction by methanol (33). We examined the possibility that the Hv-p68 gene may be under similar transcriptional control by growing the virus-free isolate B-2ss in a shake culture for 72 h in a minimal medium containing either 1% glucose or 1% methanol. The B-2ss cultures grown on a minimal medium supplemented with 1% methanol as the sole carbon source showed little or no apparent increase in mycelial mass compared with the cultures supple-mented with glucose, which showed significant growth during the 72-h period. Northern blot analysis with equivalent amounts of total RNA (30 g) isolated from these cultures revealed that the level of Hv-p68 transcript was not significantly different between the glucose and the methanol-supplemented cultures (Fig. 7B). Addition of methanol to the culture medium, under our experimental conditions, did not appear to influence the transcription rate of the Hv-p68 gene.
Distribution of the Hv-p68 Gene in Fungi-We examined the presence of the Hv-p68 gene or closely similar genes by Southern hybridization analysis using genomic DNA isolated from various fungal species and an Hv-p68 probe. The restriction profiles produced with restriction enzymes known to cleave the DNA within the Hv-p68 coding region or elsewhere varied with different fungal species. All Cochliobolus species tested contained the Hv-p68 gene or a closely similar gene (Fig. 8). The FIG. 7. Northern analysis of Hv-p68 mRNA transcript levels. A, total RNA (25 g) isolated from the virus-free isolates 408 and B-2ss and the virus-infected isolate A-9 were electrophoresed on a formaldehyde/agarose gel, blotted, and hybridized under high stringency conditions with a radiolabeled probe for Hv-p68. mRNA, ϳ2.25 kb in size, was detected in RNA samples from all three H. victoriae isolates. The level of Hv-p68 transcript in cultures of the virus-infected isolate A-9, however, was at least 10-fold higher than that for the virus-free isolates. B, total RNA (30 g) isolated from cultures of the virus-free isolate B-2ss grown for 3 days in minimal medium supplemented with either glucose or methanol. Similar amounts of Hv-p68 mRNA were detected in total RNA isolated from fungal cultures supplemented with either glucose or methanol, as a carbon source. The 32 P-labeled Hv-p68specific probe was generated by random-primer labeling of a 850-bp EcoRI-SstI restriction fragment from cloned Hv-p68 cDNA.
FIG. 6. Comparative binding activities of bacterially expressed full-length Hv-p68 and N-and C-terminal truncations. Binding activity to radiolabeled transcripts was determined by Northwestern analysis. Coomassie Blue-stained SDS-PAGE gel with bacterially expressed Hv-p68 purified from inclusion bodies; N-and C-terminally truncated Hv-p68 (estimated sizes of 54 and 62 kDa, respectively). Similar RNA binding was observed for full-length and C-terminally truncated HV-p68. The N-terminally truncated Hv-p68 did not show any detectable binding to either radiolabeled transcript (arrow and dotted line across autoradiographs indicate position of the N-terminally truncated Hv-p68). Northwestern assays were as described for Fig. 5.

TABLE I
Enzymatic assay for alcohol oxidase activity Oxidase activity was determined by monitoring spectrophotometrically the rate of hydrogen-peroxide formation from oxidation of methanol. Activity is given as units (1 unit  restriction profiles for C. victoriae isolate 408 were different from those generated for C. victoriae isolates B-2ss and A-9, suggesting that C. victoriae is probably not monophylatic. No genes closely similar to Hv-p68 were detected in the filamentous fungus P. chrysogenum nor in the nonmethylotrophic yeasts S. cerevisiae and S. pombe (Fig. 8). DISCUSSION We have previously isolated an RNA-binding cellular protein, Hv-p68, that copurifies with viral dsRNA from the filamentous fungus C. victoriae, and demonstrated that it accumulates to higher levels in virus-infected isolates than in virusfree isolates (13). In the present study, we isolated and completely sequenced a cDNA containing the full-length ORF of the Hv-p68 gene. Sequencing and phylogenetic analyses clearly indicated that Hv-p68 belongs to the family of FAD-dependent GMC oxidoreductases (30) and that it is most closely related to the alcohol oxidases (AOX) of methylotrophic yeasts. Unlike the genes of the closely related AOX, which are intronless, the Hv-p68 gene contains at least four introns. This finding has interesting evolutionary implications considering that Hv-p68, encoded by a gene from a filamentous fungus, has significantly high sequence identity with this group of flavoproteins from methylotrophic yeasts (sequence identities higher than 67%). Hv-p68 gene homologues may have existed prior to the divergence of yeasts and filamentous fungi. Furthermore, it is possible that methylotrophic yeasts may harbor viruses that are evolutionarily related to those infecting Cochliobolus species.
An increasing number of NAD ϩ -dependent dehydrogenases and oxidoreductases have recently been reported to possess RNA binding activities (35)(36)(37)(38)(39). The dinucleotide-binding sites of these enzymes share a typical ␤-␣-␤ fold motif, which is proposed to comprise the RNA-binding domain. The finding that RNA binding activity of Hv-p68 is localized at the Nterminal region containing the putative ADP-binding domain presents the first reported example of an FAD-dependent oxidoreductase that may function as an RNA-binding protein.
The RNA binding activity of Hv-p68 has been demonstrated by gel mobility shift experiments (13) and Northwestern blot analysis (this study). Cellular proteins that bind to viral RNA may serve as components of RDRP or may serve to bring various regions of a viral RNA template together to form transcription or replication complexes (1,40). Many viral RDRPs do not bind to viral RNA specifically or at all (1). Considering that the RDRPs of the Hv145SV (encoded by a monocistronic dsRNA) 2 and of the Hv190SV (9) are expressed independent of the CP, Hv-p68 may serve to mediate the binding of RDRP to its template RNA during virion assembly. The AOX of P. pastoris has also been shown to exhibit limited RNA binding activity (this study). Its affinity for viral transcripts, however, was significantly lower than that of Hv-p68.
In addition to RNA binding activity, Hv-p68 also exhibits phosphotransferase/kinase activities. 3 In this regard, the NAD ϩdependent glyceraldehyde-3-phosphate dehydrogenase (GAPDH) provides an excellent example for a comparable multifunctional cellular protein with diverse biological properties. GAPDH, once considered a simple classical glycolytic protein, is now known to display a number of different activities, including nuclear RNA export, DNA repair, translational control of gene expression, and phosphotransferase/kinase activities (for a review, see Ref. 41). Furthermore, several investigations suggest that GAPDH is in- volved in apoptosis and viral pathogenesis (41).
The low methanol-oxidizing activity of the gradient purified Hv-p68 suggests that either methanol is not the natural substrate or that the bulk of the protein in our purified preparations is present in an inactive form. It is well accepted that FAD binding is a crucial step in alcohol oxidase octamerization and that the FAD-containing octamer comprises the active form of AOX (42). Our previous data have shown that Hv-p68 occurs as an octamer in the Hv-p68-containing sucrose gradient fraction, and thus it is presumed to be in its active form. The release of FAD from Hv-p68 octamers during purification, however, cannot be ruled out. It is more likely that methanol is not the natural substrate for Hv-p68 as supported by the finding that Hv-p68 expression was neither induced by methanol nor suppressed by glucose when these two supplements were used as the sole carbon source in cultures of virus-free H. victoriae. It is noteworthy that, despite the high percentage of sequence identity in the coding region of AOX, the 5Ј-and 3Ј-untranslated regions show no significant similarities (33). It is thus plausible that different transcriptional activators may induce their expression. Comparison of Hv-p68 expression levels in virus-infected and virus-free isolates (this study) clearly demonstrated that overexpression of Hv-p68 is associated with virus infection.
It may not be surprising that the bacterially expressed Hv-p68 was completely inactive in the alcohol oxidase assay. Based on current knowledge, the bacterially expressed Hv-p68 is neither expected to bind FAD nor to oligomerize, and thus would lack oxidase activity. Species-specific proteins (chaperons) are believed to be involved in mediating FAD binding and oligomerization (42,43). This is supported by the finding that the AOX from H. polymorpha failed to bind FAD and to oligomerize when synthesized in the heterologous host S. cerevisiae (43). The absence of suitable chaperones in the heterologous systems could explain these observations.
The finding that Hv-p68 is overexpressed in virus-infected fungal isolates that exhibit the diseased phenotype (44) is of considerable interest. We have shown by Northern hybridization analysis (this study) that Hv-p68 mRNA levels were 10-to 20-fold higher in virus-infected isolates than in virus-free isolates. Although the identity of the natural substrate for the oxidase activity of Hv-p68 is not known, the structurally similar alcohol oxidases from methylotrophic yeasts or filamentous fungi mostly oxidize aliphatic primary alcohols (and in a few cases, aromatic alcohols) irreversibly to aldehydes, which are toxic (35,45). A buildup of such toxic intermediates when Hv-p68 is overproduced in virus-infected isolates may lead to the lytic/diseased phenotype in virus-infected C. victoriae isolates (44).
Although we have only examined a limited number of fungal species, it is apparent that all species in the genus Cochliobolus contain genes closely similar to Hv-p68. In a recent phylogenetic study of the genus Cochliobolus (family Pleosporaceae, order Pleosporales), it was determined that isolates of C. victoriae are not monophylatic (46). This finding is in agreement with our results on Southern analysis of restriction fragments from three isolates of C. victoriae (Fig. 8), which also showed that isolate 408 could be differentiated from other isolates of C. victoriae. It will be of interest to determine the distribution of the Hv-p68 gene or homologues in the fungal species belonging to the family Pleosporaceae or the order Pleosporales, which include several economically important plant pathogens. Transcriptional activation of the Hv-p68 gene or homologues leading to accumulation of toxic products, and hence a diseased phenotype (as postulated for C. victoriae) could be exploited as a novel approach for biocontrol of such important plant pathogens.