Rat Liver NAD(P)H:Quinone Reductase Nucleotide Sequence Analysis of a Quinone Reductase cDNA Clone and Prediction of the Amino Acid Sequence of the Corresponding Protein*

We have determined the nucleotide sequence of a cDNA clone, pDTD55, complementary to rat liver quinone reductase mRNA (Williams, J. B., Lu, A. Y. H., Cameron, R. G., and Pickett, C. B. (1986) J. BioZ. Chern. 261,5524-5528). The cDNA clone contains an open reading frame of 769 nucleotides encoding a poly- peptide comprised of 253 amino acids with a M, = 28,564. To verify the predicted amino acid sequence of quinone reductase, we have been able to align the amino acid sequences of a cyanogen bromide digest of the purified enzyme to the sequence deduced from the cDNA clone. A comparison of the quinone reductase sequence with other known flavoenzymes did not re- veal a significant degree of amino acid sequence homology. These data suggest that the quinone reductase gene has evolved independently from genes encoding other flavoenzymes.

Quinone reductase, formerly called DT-diaphorase (NAD(P)H:quinone oxidoreductase, EC 1.6.99.2) catalyzes the two-electron reduction of quinones and quinonoid compounds to hydroquinones (1)(2)(3)(4). This flavoprotein has an apparent molecular weight of 54,000 and consists of two apparently identical subunits, each containing one FAD (3). Studies utilizing isolated hepatocytes and subcellular fractions have demonstrated that the cytotoxic effects mediated by the one-electron reduction of menadione by NADPHcytochrome P-450 reductase can be diminished by the addition of quinone reductase (5,6).
Recently, our laboratory has constructed a cDNA clone complementary to quinone reductase mRNA (7). We have utilized the cDNA clone to demonstrate that the level of quinone reductase mRNA is elevated in rats treated with 3methylcholanthrene and in persistent hepatocyte nodules induced by chemical carcinogens (7,s). Southern blot analysis of genomic DNA suggests that the quinone reductase gene is hypomethylated in nodular tissue compared to normal liver or liver tissue surrounding the nodules (7).
In the present study, we have determined the nucleotide sequence of the cDNA clone, pDTD55, and have predicted the amino acid sequence of rat liver cytosolic quinone reductase. The authenticity of the cDNA clone has been confirmed * The costa of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession numbetfs) 502640. by comparing the predicted amino acid sequence with amino acids obtained from conventional protein sequence analysis of a cyanogen bromide digest of purified enzyme. Surprisingly, we have found no significant sequence homology between quinone reductase and other flavoenzymes.

MATERIALS AND METHODS
Nucleotide Sequence Analysis of pDTD55"Appropriate restriction fragments were 5' (NcoI, Ball and HindIII) or 3' (PstI) end-labeled and subjected to the chemical cleavage method of Maxam and Gilbert (9). Purified PstI fragments of pDTD55 were cloned into M13 phage and sequenced by the Sanger dideoxy-sequencing method (10). An oligonucleotide, 17 nucleotides long, was synthesized on a Sam One Synthesizer (Biosearch) and purified by electrophoresis on a 15% polyacrylamide gel. The 17-mer was used in place of the universal primer in certain sequencing runs.
Determination of Amino Acid Composition-The amino acid composition was determined on a Beckman amino acid analyzer (model 121MB) using standard methodology.
Amino Acid Sequence Analysis-Quinone reductase was subjected to cyanogen bromide digestion (11) and the mixed digest sequenced in the Applied Biosystems gas-phase sequenator (model 870A) according to the manufacturer's specifications. High performance liquid chromatography was used to quantitate the phenylthiohydantoin derivatives produced at each step (12).
Computer Analysis-The sequence comparison program of Intelligenetics using the Align command (Needleman and Wunsch homology search algorithm (13)) and Search command (Queen-Korn algorithm (14)) were used in this study to compare the sequence of quinone reductase with other flavoenzymes.

RESULTS AND DISCUSSION
DNA Sequence Analysis of the Quinone Reductase cDNA Clone-The quinone reductase cDNA clone, pDTD55, was constructed from mRNA prepared by polysomal immunoabsorption techniques (7). The length of the cDNA insert in the clone was 1900 bp'. The strategy for DNA sequence analysis is presented in Fig. 1 (Appendix). The solid arrows represent the sites used for 5' or 3' end labeling and sequence analysis by the Maxam-Gilbert chemical sequencing procedure (91, whereas the dashed lines represent regions of the cDNA insert sequenced by the dideoxy method (10). Although the total length of the insert in pDTD55 was originally reported to be 1900 bp (7), we found that the sequence 5' to the NcoI site was a 400-bp duplication of the sequence from NcoI toward the 3' end of the clone. This duplicated sequence most likely arose during second strand synthesis when the hairpin loop generated during first strand synthesis was not made blunt end by DNA polymerase I. Therefore, the actual length of the cDNA insert in pDTD55 is 1400 bp.

Analysis of the nucleotide sequence revealed a single open
The abbreviation used is: bp, base pairs. reading frame of 759 bp encoding a protein comprised of 253 amino acids (Fig. 2, Appendix). Although there is an in-frame Met (ATG) codon 78 bp upstream from the Met codon assigned as the initiation codon, the 78-bp sequence contains two termination codons in the reading frame. Hence, the 78bp sequence is not part of the open reading frame. In addition to the coding sequence, the cDNA clone contains 113 nucleotides of the 5"untransIated region and 522 nucleotides of the 3"untranslated region. Amino Acid Sequence Analysis of Cyanogen Bromide Fragments of Quinone Reductase-Initial efforts to verify our predicted amino acid sequence by NH2-terminal sequence analysis of the purified enzyme proved unsuccessful. Numerous sequencing runs suggested that the NH2-terminal amino acid was blocked. Consequently, a cyanogen bromide digest was performed and the mixed digest sequenced utilizing the gas phase sequenator. The amino acid yields at each step from cycle 1 to 33 were quantitated and then aligned with the predicted amino acid sequence of the protein (Table I and Fig. 2, underlined amino acids). At each step in the sequencing run, the amino acids generated from conventional protein sequencing could be aligned with the amino acid sequence predicted from DNA sequence analysis. The only exception is the third amino acid, Gln, in the second underlined fragment of the deduced amino acid sequence. Conventional amino acid sequencing indicated a Lys in that position. The amino acid composition of purified DT-diaphorase also agreed with the composition predicted from DNA sequence analysis of the cDNA clone (Table 11).
Although there have been suggestions that multiple quinone reductases exist in rat liver (15), our cloning data suggest the presence of a single form (7). These findings are in contrast to the isozyme composition in mice. Prochaska and Talalay (16) have purified two forms of the enzyme from mouse liver cytosol, which are immunochemically similar. We cannot rule out, however, in the rat that multiple quinone reductases exist and are encoded by a gene or gene family that is unrelated to the gene encoding the quinone reductase characterized in this study. Finally, since all the amino acids from conventional protein sequencing could be accounted for in the deduced amino acid sequence, we also believe that each subunit of the enzyme is identical or extremely similar.
Comparison of the Amino Acid Sequence of Quinone Reductase to Other Flauoenzymes-Porter and Kasper (17,18) have made extensive comparisons of the sequence of NADPH cytochrome P-450 reductase to other flavoenzymes. They have found that the tentative FAD-binding domain of the reductase, residues 267-678, shows a high degree of similarity to ferredoxin NADP+ reductase and NADH-cytochrome bS reductase. We have compared the amino acid sequence of quinone reductase to other flavoenzymes, which include NADPH cytochrome P-450 reductase (17), NADH cytochrome bs reductase (19), ferredoxin NADP+ reductase (20), glutathione reductase (21), p-hydroxybenzoate hydroxylase (22), Desulfouibrio uulgaris flavodoxin (23), Clostridium MP flavoprotein (24), pig kidney D-amino acid oxidase (25), and Escherichia coli fumarate reductase (26). Using the alignment algorithm of Needleman and Wunsch (13), we did not observe significant amino acid sequence homology between quinone reductase and the other flavoenzymes. These findings suggest that the quinone reductase gene evolved from an ancestral gene that was distinct from the ancestral gene(s) encoding other flavoenzymes. From a functional viewpoint, the quinone reductase is unique in that it accepts electrons from both NADH and NADPH with equal efficiency, whereas many other flavoproteins are highly specific to electron donors. In addition, dicumarol is a highly specific inhibitor of quinone reductase (5,6), and it inhibits the enzyme by interfering with the electron transfer from NADPH to FAD (3). Thus, the  FAD-binding domain of the quinone reductase must be considerably different from other flavoproteins in order to exhibit the unique electron accepting properties and its high affinity for dicumarol. In summary, the quinone reductase cDNA clone characterized in this study will allow us to characterize the structural gene encoding the enzyme and to elucidate the regulatory elements that make the gene responsive to polycyclic aromatic hydrocarbons. were cloned into M13 phage and sequenced by the dideoxy-sequencing method (10). The broken line represents the direction of dideoxy sequencing. All fragments were sequenced two to three times.

3'
. " " " _ " " " The amino acid sequence was deduced by computer analysis of the DNA sequence. The underlined amino acids correspond to amino acid sequences within quinone reductase that correspond to the amino acids determined by protein sequence analysis of the cyanogen bromide digest of the purified protein (see Table I).   29  TGG GCC  AAT ACA ATC AGG GCT  CTT CTC ACC GCC  ATG GCT CCA GAA  GTT GGG GGC GGA  TCG TAG TGT CAA GCG CTG AT1 GGC TGA GCA GAG AGG ACA  TCA TTC AAC TAT   5  T h r G l y G l u P r o L y s A s p S e r G l u A s n P h e G l n T y r P r o V a l G l u S e r S e r L e u A l a T y r L y s G l u G l y A r g L e u S e r P r o A s p I l e V a l A l a G l u G l n L y s L y s L e u GAA GCT GCA GAC CTA GTG ATA TTT CAG TTC CCA TTG TAT TGG TTT GGG GTG CCC GCC  AT1 CTG AAA GGC TGG TTT GAG AGA GTG  CTT GTA GCA GGA  TTC GCC TAC ACG   353  380  407  434 G l u A l a A l a A s p L e u V a l I l e P h e G l n P h e P r o L e u T y r T r p P h e G l y V a l P r o A l a I l e Leu L y s G l y T r p P h e G l u A r g V a l L e u V a l A l a G l y P h e A l a T y r T h r  TTC TAG GTC TTT TGT ACA CTA TAA GCT TTT TTC TTC GGG CTA GCC TTG GCT AAA TGG CAT CCA ATC CTC CAC CCA  CTT GTT   1109   1244  1271  1217  GCT AT1 AGT TAC CTC TCT GTG GTT TAG GGC AGG AGG GAA  TTG CTC AAA CAA TGG CTG AGG GAC  TAA CTT  GTT TAG CAG TTA GCT AAA GCC TGT TTA TGA TCC ATC CTG   GTT TCA AT1 ACT GTG CAG TGA CTG ACA AGC  CTC GGG GGA  TTG CTC TCC AGC  TCT TCT CTG CCT TGT ACA TAG CAC ACC CAG GTC  CTG GGA AAT GAA TAC A