Aryl-alcohol dehydrogenase from the white-rot fungus Phanerochaete chrysosporium. Gene cloning, sequence analysis, expression, and purification of the recombinant enzyme.

A cDNA clone encoding a ligninolytic aryl-alcohol dehydrogenase (AAD; EC 1.1.1.91) from the white-rot basidiomycete fungus Phanerochaete chrysosporium was isolated and characterized. The nucleotide sequence obtained reveals an open reading frame encoding a protein of 385 amino acids. Substantial homology (49.3% identity and 67.3% similarity, respectively) was observed between AAD and an open reading frame sequence present on chromosome III of Saccharomyces cerevisiae. A Southern blot analysis showed the presence of multiple AAD gene-related sequences in P. chrysosporium and in other white-rot fungi including Bjerkandera adusta and Fomes lignosus. Northern blot analyses are in line with the view that the levels and appearance of AAD mRNA correlate with the level and appearance of AAD activity and that, under conditions of nitrogen limitation, the AAD mRNA levels are higher than in carbon limited cultures. This is consistent with the regulation of the enzyme by carbon or nitrogen limitation being at the level of transcription. Moreover, the appearance of AAD-specific transcripts correlates with the appearance of lignin peroxidase-specific transcripts in the same cultures. This co-appearance is in line with the proposed synergistic interaction of the two enzymes in lignin biodegradation, which suggests a similar regulation. The AAD encoding cDNA was expressed in Escherichia coli to yield high levels of active enzyme, and the recombinant enzyme was purified by using metal chelate affinity chromatography.

degradation, subsequent studies also indicated the need for additional enzymes (reviewed by Kirk and Farrell (1987), Schoemaker et al. (19891, Pease and Tien (1991), Cullen and Kersten (1992), and Fiechter (1993)). The constant oxidation of lignin by peroxidases will end up in a very oxidized state of the polymer and subunit or degradation products thereof, with many aldehyde-, quinone-, and possibly also acidic groups present. As a consequence, peroxidase catalysis will stop. Thus, aside from oxidative reactions, reductive ones are also likely to be needed for lignin biodegradation to occur (Schoemaker et al., 1989). Such activities were found in white-rot fungi some decades ago, but their activities were not linked to lignin biodegradation at that time (Farmer et al., 1959;Zenk and Gross, 1965). In 1988 a model was put forward  in which the symphonic action of oxidative ,peroxidases and reductive, intracellular enzymes is needed for complete degradation of the lignin biopolymer. These reductive activities have recently been identified in R chrysosporium Constam et al., 1991). A reducing activity was detected in agitated as well as non-agitated, nitrogen-limited cultures of R chrysosporium 2 days after inoculation, reaching its maximum after 6 days using veratraldehyde as a substrate, but in carbon-limited, agitated cultures, the veratraldehyde reduction rates reached were considerably lower. The activity of this aryl-alcohol dehydrogenase (AAD)' appeared synchronously with the ligninolytic activity, and the production ofduring secondary metabolism points to its possible involvement in lignin biodegradation. The enzyme showed one major band with an apparent molecular mass of 47 kDa, whereas gel filtration experiments suggested a molecular mass of 280 kDa. Polyclonal antibodies raised against the highly purified 47-kDa protein were able to immunoprecipitate the activity, indicating that it is part of the enzyme. Further biochemical characterization of the purified enzyme showed a broad specificity toward aromatic compounds and the ability to also reduce dimeric compounds, underlining the direct involvement of the enzyme in lignin degradation. Leisola et al. (1988) proposed the possible need of AAD in the biodegradation of veratryl alcohol, which is one of the most simple lignin model compounds, and later on Shoemaker et al. (1989) and  expanded this view to other model compounds and to lignin in general.
With a view toward investigating the regulatory mechanism(s) controlling AAD production and assessing the importance of AAD in lignin biodegradation, we have cloned and The abbreviations used are: AAD, aryl-alcohol dehydrogenase; ORF, open reading frame; PAGE, polyacrylamide gel electrophoresis; TBS, Tris-buffered saline; IPTG, determined the nucleotide sequence of an AAD cDNA. In addition, we have expressed the cloned cDNA in Escherichia coli to yield high levels of active enzyme.
adusta, and C. versicolor were grown at 30 "C. I? ostreatus and F: lignosus were cultivated at 25 "C. Two media, modifications of that used by Kirk et al. (1978), were used. The nitrogen-limited medium contained 0.22 g liter" (2.4 mM) ammonium tartrate as a nitrogen source and 10 g liter" glucose as a carbon source. The carbon-limited medium contained 2 g liter" glucose and 0.66 g liter" (7.2 mM) ammonium tartrate. Basidiospores were produced as described by Gold and Cheng (1979).
Antibody Screening of P chrysosporium hgtll-based cDNA Expression Libraries-The P. chrysosporium cDNA libraries used have been described before (Walther et al., 1988;Pribnow et al., 1989). They were originally prepared using RNA from 6-day-old carbon-limited cultures of P. chrysosporium BKM-F-1767 (Walther et al., 1988) or from 5-and 6-day-old nitrogen-limited cultures of P chrysosporium OGClOl (Pribnow et al., 1989). Positive clones were detected using an immunoscreening procedure (Young et al., 1983b;Sambrook et al., 1989) with anti" antibodies  diluted 1:500, essentially as described above for the Western blot analysis. The background could be reduced by preadsorption of the antiserum with total E. coli proteins immobilized on CNBr-activated Sepharose (Sambrook et al., 1989). In a first cycle, several positive phage clones were isolated from a total of 100,000 clones tested. These were then purified in a second cycle. Atotal of 14 phage clones isolated from library 2 proved to be positive in the second cycle, whereas only 2 of the clones from library 1 reacted positively. Antibody-positive phage clones were used to prepare lysogens of E. coli Y1090 (Young et al., 1983a). Crude cell extracts were prepared from induced lysogens and analyzed in Western blots. The fusion proteins produced showed a molecular mass in the range of 140-157 kDa . In order to verify the clones, antibodies bound to fusion proteins in situ were eluted at pH 2.8 (Olmsted, 1981;Snyder et al., 1987). After neutralization, the affinity-purified antibodies were tested in Western blots using crude extracts of P. chrysosporium and purified AAD. Antibodies eluted from three individual clones from library 2 reacted all positively with the dehydrogenase, but antibodies purified on fusion proteins of clones obtained from library 1 did not detect the dehydrogenase in Western blots. High titer phage lysates were prepared according to Loenen and Brammar (1980), and phage DNA was isolated according to the procedure of Davis et al. (1986).
Preparation and Sequence Determination of Cyanogen Bromide Fragments-Purified AAD was electrophoresed on a 12% SDS-polyacrylamide gel (Neville, 1971) and subsequently transferred electrophoretically onto a Bio-Rad polyvinylidene difluoride membrane using 10 m M CAPS buffer, pH 11,20% methanol (LeGendre and Matsudaira, 1989) for 45 min at 500 mA. The blot was stained with 0.025% Coo-massie Brilliant Blue R-250 in 40% methanol and destained in 50% methanol, and the 43-kDa band was cut from the blot and subsequently treated with cyanogen bromide in an acidic medium (Frank et al., 1993). The amino acid sequence determinations were performed on a Knauer model 810 protein sequencer (Dr. Ing. Herbert Knauer GmbH, Berlin, Germany). For the on-line isocratic HPLC determination of phenylthiohydantoin amino acids, the method of Frank (1989) was used.
Expression of AAD cDNA in E. coli-The E. coli expression vector pTrc99A (Pharmacia Biotech Inc.) (Amann et al., 1988) containing the strong tac (trc) promoter was used for overexpression of the AAD cDNA in E. coli XL1-Blue cells (Bullock et al., 1987). For AAD activity tests, plasmid-containing E. coli XL1-Blue cells were grown at 30 "C in 2 X YT medium supplemented with 0.4 M sucrose as described by Bowden and Georgiou (1990). Cultures of 500 ml were grown at 30 "C to an A,, of 0.3 and induced by the addition of isopropyl-1-thio-P-D-galactopyranoside (IPTG) to a final concentration of 2 mM. At various times, cells were harvested by centrifugation, resuspended in 0.04 volume of ice-cold 50 m M Tris-HC1, pH 7.5 and disrupted by sonication (six times, 15 s each). After centrifugation for 10 min at 11,000 x g and 4 "C, the enzyme activity was analyzed by following the oxidation of NADPH to NADP+ in the presence of veratraldehyde as described earlier .  , 1992). 1 ml of sonicated extract was diluted with 3 ml of BC 100/40 and applied to the column. The column was washed with 15 ml of BC 100/40, followed by elutions with 7.5 ml each of BC 1000/100 and BC 1000/2000. For enzymatic assays the column eluates were used directly, and for SDS-PAGE analysis the eluates were dialyzed against 10 m~ sodium phosphate buffer, pH 7.2,0.9% NaCl and subsequently concentrated using Strata-CleanTM resin (Stratagene Cloning Systems) (Nielson et al., 1993). The enzyme activity was assessed by analyzing the reaction products formed by HPLC .
Polymerase Chain Reaction-For PCR amplification, 10 pg of each primer, 10 ng of DNA, 10 p1 of a 2 m M dNTP solution, 10 pl of 10 x Taq polymerase buffer (100 m~ Tris-HC1, pH 9.0,500 m~ KC], 15 m M MgCI,, 1% Triton X-100, 0.1% gelatin; ANAWA, Wangen, Switzerland), and 2.5 units of Taq polymerase (ANAWA) were combined and the volume adjusted to 100 $. Denaturation was for 1.5 min at 94 "C, annealing for 2 min at 50 "C, and polymerization for 2 min at 72 "C using a Perkin-Elmer GeneAmp 9600 system. The cycle was repeated 30 times. After a 10-min incubation at 72 "C, 10 pg of carrier glycogen (Boehringer Mannheim) were added and the reaction extracted with 100 pl of a 1:l mixture of phenol and chloroform. The DNA was precipitated with ethanol and subsequently digested with BglII and EcoRI and fractionwere used: primer 1,5'-AGCGGATAACAAT"CACACAGGA, primer 2, ated on a low melting temperature agarose gel. The following primers

5"CTGCTGAGATCTCTGGGGGCGGATAGC.
Subcloning and DNA Sequencing-Deletion subclones were created using restriction enzymes by following the procedures compiled by Sambrook et al. (1989). Plasmids were transformed into E. coli as described by Chung and Miller (1988). Nucleotide sequencing was carried out according to Tabor and Richardson (1987) using Sequenase (U. S. Biochemical Corp.) or T7 DNA polymerase (Pharmacia) following the protocols provided by the supplier. The cloned PCR fragment was sequenced using six different oligonucleotides (18-mers) as primers. For Maxam-Gilbert sequencing (Maxam and Gilbert, 19801, restriction fragments were extracted from low jelling temperature agarose gels (Pharmacia) (Frischauf et al., 1980) and end-labeled using DNA polymerase I (Klenow fragment) and [a-32PldCTP (Sambrook et al., 1989). The software package of the University of Wisconsin Genetics Computer Group (UWGCG) (Deverew et al., 1984) was used for storing and analyzing the DNA sequence data.
DNA Isolation and Southern Blot Analysis-Total DNA from P. chrysosporium, B. adusta, C. versicolor, P. ostreatus, and F: lignosus was isolated essentially as described by Raeder and Broda (1988). For Southern blot analysis (Southern, 1975), the fragments were separated in a 0.8% agarose gel and transferred onto a Genescreen PlusTM membrane (DuPont NEN) according to the procedure recommended by the manufacturer. Hybridization was carried out overnight in the presence of 50% formamide, 1% SDS, 1 M NaCl, and 10% dextran sulfate at 42 "C. The blot was washed twice for 15 min each at room temperature in 2 x SSC, 0.1% SDS followed by a wash step at 55 "C in 0.1 x SSC, 0.1% SDS for 30 min. For the preparation of radioactively labeled probes, the method of Feinberg and Vogelstein (1984) was used.
Isolation of RNA and Northern Blotting-The pellets from a 600-ml I? chrysosporium culture were washed with distilled water, filtered, and frozen immediately in liquid nitrogen and subsequently opened using a mortar. The powder obtained was then thawed in a solution of 4 M guanidinium isothiocyanate, 0.5% sodium lauroylsarcosine, 25 mM sodium citrate, pH 7.0, 0.1 M /3-mercaptoethanol (Teen et al., 1987). After removing the cell debris by centrifugation, the RNA was purified by pelleting through a CsCl cushion (5.7 M) for 24 h at 33,000 rpm (15 "C) using a Beckman SW41 rotor, The pellet was dissolved in diethyl pyrocarbonate-treated water and the RNA precipitated with 2.5 volumes of ethanol. For Northern blot analysis, 30 pg of total RNA were et al. (1982). "ansfer of the RNA to a Genescreen Plus" membrane separated on a 6% formaldehyde, 1% agarose gel according to Maniatis was carried out as described in the protocol supplied by the manufacturer. Hybridization using the AAD cDNA as a probe was performed as described above for the Southern blots. The blot was washed as follows: 2 x 5 min in 2 x SSC, 1% SDS at room temperature; 2 x 15 min in 2 x SSC, 0.1% SDS at 65 "C; 2 x 15 min in 0.1 x SSC, 0.1% SDS at room temperature. Lignin peroxidase-related transcripts were monitored using a synthetic oligonucleotide (5'-TTAGTTGGGGGACGGCGGCG) corresponding to the end of the protein-encoding region of the CLG4 cDNA (De Boer et al., 1987). The blot was prehybridized with 1 M NaCl, 1% SDS, 10% polyethylene glycol 6000, and 10 pg m1-I salmon sperm DNA at 55 "C for 5 h . Hybridization was carried out at 50 "C for 24 h using the kinased oligonucleotide. The blot was washed as follows: 5 min in 2 x SSC, 0.1% SDS at room temperature; 30 min in 2 x SSC, 0.1% SDS at room temperature; 15 min in 2 x SSC, 0.1% SDS at 42 "C.

Isolation of AAD Encoding cDNAs-Two r! chrysosporium
hgtll-based cDNA libraries were used. One of them was initially prepared using RNA isolated from 6-day-old carbon-limited cultures (library 1; Walther et al., 19881, while the other one had been prepared using RNA from 5and 6-day-old nitrogen-limited cultures (library 2; Pribnow et al., 1989). Both libraries were screened through two cycles using polyclonal antibodies directed against AAD . To compare the immunopositive phage clones at the molecular level, the phage DNAs were isolated, digested with EcoRI, and analyzed in a Southern blot. For this comparison the insert fragment of one of the clones was labeled. While 12 of the clones isolated from library 2 all turned out to be related at the nucleotide sequence level (data not shown), the clones isolated from library 1 did not react positively.

Nucleotide Sequence Analysis of an AAD Encoding cDNA-
The phage clone containing the largest insert was used for further studies. The 1.3-kilobase pair insert fragment was subcloned into the pUC18 vector (Yanisch-Perron et al., 1985) previously digested with EcoRI and treated with alkaline phosphatase resulting in plasmid pAM1. The complete DNA sequence of the insert was determined on both strands using a set of overlapping restriction fragments of the AAD cDNA subcloned into pUC18 according to the sequencing strategy outlined in Fig. 1B. While most of the sequence data were obtained by the dideoxy method (Sanger et al., 1977;Tabor and Richardson, 1987), some regions yielded ambiguous results when the two strands were compared. To clarify the sequences at these sites, they were analyzed by the Maxam-Gilbert method (Maxam and Gilbert, 1980). The AAD cDNA sequence and the predicted translation product are shown in Fig. 1C. The sequence comprises 1367 nucleotides including a poly(A) tail and contains an open reading frame encoding a protein of 385 amino acids. Since the phage clone codes for a fusion protein, the exact reading frame could be inferred (Young and Davis, 1983a) and proved to be identical with the experimentally determined open reading frame. The ATG codon closest to the 5'-end of the cDNA is at position 28 (Fig. 1C). I t is within the sequence A C A G C m A A , which is similar to the ATG context of a range of filamentous fungal genes (Gurr et al., 1987) and   includes the important -3 A residue of the Kozak consensus sequence (Kozak, 1987). However, since the NH,-terminal amino acid sequence of the AAD protein is unknown, the methionine residue at position 28 is not necessarily the starting amino acid. The size of the putative AAD monomer has previously been estimated to be roughly 47 kDa, corresponding to a protein chain of about 427 amino acids and thus to a mRNA of at least 1280 nucleotides in length. This is a n overestimation, however, since the E. coli-produced AAD originating from the cloned cDNA comigrated at around 43 kDa with highly purified Z? chrysosporium AAD in front of the 45-kDa ovalbumin marker in SDS-polyacrylamide gels (see Fig. 5B) and thus the cDNA analyzed appears to be complete as judged by this fact. This is   also supported by the observation that the sequence can be expressed in an active form in E. coli (see below).

A L E K V A E E I G A K S I T S V A I A Y L M Q K F P Y V F P I V G G R K V E H
Correlation between the cDNA-encoded AAD and the Fungusderived Enzyme-In an attempt to analyze the amino terminus of the AAD protein, the highly purified dehydrogenase from l? chrysosporium was blotted onto a polyvinylidene difluoride membrane and the 43-kDa AAD protein on the blot was subjected to a phenylisothiocyanate degradation. However, this treatment failed to produce the expected phenylthiohydantoins (data not shown), indicating that the amino terminus is blocked. In order to demonstrate a correlation between the predicted AAD protein encoded by the cDNA and the fungusderived enzyme, the blot with the 43-kDa AAD from l? chry- pg of total RNA were analyzed. A , AAD gene probe; B, lignin peroxidase gene probe. Lanes 1,2, and 3, RNA from 2-, 4-, and 6-day-old nitrogenlimited cultures, respectively; lanes 4, 5, and 6, RNA from 2-, 4-, and 6-day-old carbon-limited cultures, respectively. Probing was done either using the radiolabeled AAD cDNA or a 32P-labeled synthetic oligonucleotide corresponding to the CLG4 lignin peroxidase cDNA (De Boer et al., 1987). sosporium was treated with cyanogen bromide in an acidic medium. This procedure led to the scission of the protein at the various methionine residues. The newly created amino termini were subsequently reacted with phenylisothiocyanate according to the procedure of Frank et al. (19931, the phenylisothio-hydantoin amino acids liberated were analyzed, and two more cycles of degradation and analysis were subsequently carried out. The pattern of the amino acids liberated in the second cycle was particularly striking. The predicted NH, termini were Ile, Thr, Lys, Ala, Val, Gly, Asp, Arg, and Phe, and our qualitative analysis revealed Ile, Lys, Ala, Val, Gly, Arg, and Phe, with Ile being the predominant amino acid (data not shown). Furthermore, the very abundant amino acid leucine was not seen in the second or in the first cycles, indicating that the analysis was specific. Taken together these data provide evidence that the protein encoded by the cDNA and the fungus-derived enzyme correlate as far as this qualitative amino acid fingerprinting method is concerned. Extracts were analyzed in a 12.5% SDS-polyacrylamide gel (Neville, 1971). Lane I , prestained molecular size markers (Bio-Rad); lane 2, AAD from I? chrysosporium (phenyl-Superose peak fraction); lanes 3 and 4, extracts from XL1-Blue cells harboring pTrc99A with a control insert; lanes 5 and 6, extracts from XL1-Blue cells harboring pTrc99A-AAD. Lanes 4 and 6 represent extracts from IPTG induced cells.

Similarity between AAD and an ORF Sequence Encoded by
Chromosome 111 of Saccharomyces cerevisiae-The inferred AAD protein sequence was compared with those of other known proteins using the UWGCG BESTFIT and ALIGN programs (Devereux et al., 1984). The best agreement was obtained with an ORF sequence (YCR107w) present on chromosome I11 of S. cerevisiae (Oliver et al., 1992). This comparison revealed 49.3% identity and 67.3% similarity between the two sequences (Fig.  2). The S. cerevisiae sequence shows similarity with a Nicotiana tabacum auxin-induced mRNA (Oliver et al., 1992). AAD Gene-related Sequences in Different White-rot Fungi-Since the AAD activity is assumed to be important for lignin biodegradation, the presence ofAAD within the group of whiterot fungi is expected. Therefore, several lignin degrading basidiomycetes including B. adusta, F! ostreatus, I? lignosus, and C. versicolor were tested for the presence of AAD-related DNA sequences. The fungal DNAs were digested with ClaI and analyzed in a Southern blot using the AAD cDNA as a probe (Fig. 3) C. versicolor and €? ostreatus showed rather faint bands, and it is difficult to predict at this point whether these bands reflect related DNA sequences or not. Interestingly, several bands hybridized in cases where strong hybridization signals were obtained. Some of the bands may have been caused by restriction fragment length polymorphisms. Such a situation has been observed in the case of the lignin peroxidase gene family (Raeder et al., 1989;Gaskell et al., 19911, and this complication was circumvented by the use of basidiospores, which are known to be monokaryotic (Gold and Cheng, 1979;AIic and Gold, 1985). The hybridization pattern seen in Fig. 3 (lune I ) was also obtained using DNA from a basidiospore-derived culture of €? chrysosporium. Nevertheless multiple bands hybridizing to the probe are apparent, possibly indicating the presence of multiple AAD genes. In F? chrysosporium multiple genes have been observed for lignin peroxidases (Raeder et al., 1989;Huoponen et al., 1990), for Mn(I1)-dependent peroxidases (Pribnow et al., 1989; Pease a t al., 1989), and for cellulases (Covert et al., 19921, and such a multicopy family may exist for AAD as well. Expression ofAAD RNA by I? chrysosporium-To investigate the regulation of AAD synthesis in I? chrysosporium, the corresponding RNA levels were determined in both carbon-and nitrogen-limited cultures. Total RNA was isolated 2, 4, and 6 days after inoculation and analyzed in Northern blots using the labeled AAD cDNA as a probe. It was evident from this analysis that carbon-limited cultures showed lower levels of AAD transcripts than nitrogen-limited cultures (Fig. 4A). In nitrogenlimited conditions at days 4 and 6, almost equal amounts of AAD RNA were detected, whereas 2 days after inoculation no dehydrogenase RNA could be detected indicating that the gene is expressed during secondary metabolism. The same situation applied to carbon-limited cultures. Such cultures showed no AAD RNA at day 4. However, on day 6 low levels of AAD transcripts could be detected. The levels ofAAD RNAin carbonlimited cultures were more than 10 times lower than the corresponding levels found in nitrogen-limited cultures. The patterns of appearance of the lignin peroxidase-specific transcripts (Fig. 4 B ) and the ones of the AAD-specific transcripts correlate well (see also Reiser et al., 1993). This suggests that the accumulation of the dehydrogenase is controlled similarly to lignin peroxidase. The appearance of AAD activity in I? chrysosporium cultures has been investigated previously . The pattern seen at the level of the activity parallels that seen at the RNA level. Thus, AAD enzyme production appears to be regulated at the level of transcription by nitrogen or carbon limitation.

Heterologous Expression of the AAD cDNA in E. coli-To
unequivocally demonstrate that an AAD-specific sequence had been cloned, attempts were made to express the cDNA in E. coli. For this purpose, the cDNA was subcloned into the pTrc99A expression vector (Amann et al., 1988) to yield plasmid pTrc99A-AAD. In this plasmid the expression of the AAD cDNA is controlled by the regulatable E. coli tac promoter (De Boer et al., 1983). The DNA sequence including the Shine-Dalgarno sequence and the beginning of the protein encoding region is shown in Fig. 5 A . Plasmid pTrc99A-AAD was transformed into E. coli XL1-Blue (Bullock et al., 19871, and the cells were cultivated a t 30 "C in 2 x YT medium. IPTG was added to the cells during the early exponential phase to induce the production of A A D , and the cells were collected 3 h after induction and analyzed by SDS-PAGE (Fig. 5B) and by Western blotting (Fig.  5C). It is evident from this analysis that an additional band of around 43 kDa appeared in IPTG-induced cultures of E. coli XL1-Blue cells harboring plasmid pTrc99A-AAD (Fig. 5B, lane  2) but not in cells harboring pTrc99A with a control insert (Fig.  5B, lane 4). In addition, this band comigrated with purified AAD (Fig. 5B, lane 6) and it reacted specifically with anti" antibody in Western blots (Fig. 5C, lane 6).
Having demonstrated abundant production of AAD in E. coli XL1-Blue cells, AAD activity was determined to see if the protein was functional. The cells were grown at 30 "C in 2 x YT medium containing 0.4 M sucrose to an A,,, of 0.3, and the enzyme production was induced by adding 1 m~ IPTG. Aliquots were removed before induction and at various times after induction. After opening the cells using sonification, the NADPH oxidizing activity in the supernatant was measured. Significant levels of such a n activity were detected before induction, but 4-5 times more activity was present in IPTG-induced cultures, indicating that the AAD was active and that NADPH oxidizing activities unrelated to AAD are also present in crude extracts of E. coli (data not shown). 100 ml of a culture produced up 19 units of active AAD (data not shown).
Purification of Recombinant AAD by Metal Chelate Aftinity Chromatography-To investigate more directly the activity of the E. coli-derived A A D , the recombinant AAD was purified by metal affinity chromatography (Hochuli et al., 1988). This single-step strategy was facilitated by the addition of 6 histidine residues at the carboxyl terminus of the protein according to the strategy depicted in Fig. 6A. In this construct, the last amino acid of the AAD protein was changed from Lys to Arg and 8 additional amino acids (His),-Arg-Ser were spliced on (Fig.  6B). Crude cell extracts ofE. coli XL1-Blue harboring pTrc99A-AAD6His were prepared and applied onto a Ni2+-NTA-agarose column. In the presence of 40 m~ imidazole, most of the soluble bacterial proteins did not bind to the Ni2+-NTA column, whereas a significant portion of the AAD6His fusion protein was retained as judged from the SDS-gel analysis shown in Fig.  6C (lane 5 ) and from a Western blot analysis (data not shown). A step of 1 M KC1 and 100 mM imidazole (BC 1000/100) released substantial amounts of the fusion protein (Fig. 6C, lane 6 ) and a final wash of 1 M KCl, 2 M imidazole (BC 1000/2000) eluted the remaining 10% of the fusion protein (data not shown). Compared to the highly purified enzyme from the fungus (lane 7), the fusion protein had a slightly reduced mobility. This difference is most likely due to the 10 additional amino acids at the amino terminus (see Fig. 5 A ) and to the 8 additional residues at the carboxyl terminus. Small amounts of other proteins were also eluted (lanes 3 and 6), and trace amounts of a 43-kDa protein were also present in the BC 1000/100 eluate of uninduced cell extracts (lane 3 ) . Up to 75% of the applied AAD activity was recovered in the BC 1000/100 eluate (data not shown), and up to 600 pg of highly purified AAD fusion protein were obtained per 100 ml of culture. The specific activity of the fusion protein was in the order of 15-25 unitdmg, which is about half to one-third of that measured for the fungus-derived enzyme , indicating that the affinity tail at the carboxyl terminus and/or the extra amino acids at the amino terminus may have affected the enzyme. The protein present in the BC 1000/100 eluate from induced cells was capable of reducing veratraldehyde to veratryl alcohol as shown by HPLC analysis (data not shown) indicating that bona fide AAD activity was present in this fraction.
us with the cDNA library and to J. Nagle for help with the DNA