Doxorubicinol (dxol) is the major metabolite formed in the hearts of cancer patients being treated with the widely used chemotherapeutic agent, doxorubicin (dox). The well-documented cardiomyopathy associated with dox treatment has been studied in vitro and ex vivo providing evidence that the C-13 hydroxy metabolite, dxol, might play a key role in the development of dox-induced cardiotoxicity. In this report, we have developed transgenic mice with heart-specific expression of human carbonyl reductase (HCBR), an enzyme that metabolizes dox to dxol. Dox was rapidly converted to dxol in the hearts of the transgenic expressers, which led to advanced development of both acute and chronic cardiotoxicity. Acute cardiotoxicity was evident by a 60% increase in serum creatine kinase activity and a 5-fold increase in cardiac damage measured by electron microscopy. Myofibril degeneration was the major damage observed in acute dox toxicity. Electrocardiograph telemetry, survival data, and electron microscopy were monitored during chronic dox-induced cardiotoxicity. HCBR expressers developed cardiotoxicity 6-7 weeks before the nonexpressers. The HCBR expressers survived for 5 weeks compared with 12 weeks for the controls. Electrocardiograph profiles and necropsies showed the cause of death to be the development of cardiomyopathies leading to congestive heart failure. Levels of dxol were four times higher in the HCBR expresser hearts than in the nonexpressers. Electron microscopy data showed swelling and major structural damage of the mitochondria in the HCBR expressers. These data demonstrate that the C-13 hydroxy metabolite of dox advances the development of dox-induced cardiotoxicity in an in vivo system and suggest that heart carbonyl reductase activity may contribute to dox-induced cardiotoxicity in humans.

Carbonyl reductase (secondary-alcohol:NADP+ oxidoreductase, EC 1.1.1.184) belongs to the family of short chain dehydrogenases/reductases (SDR). Carbonyl reductases (CBRs) are NADPH-dependent, mostly monomeric, cytosolic enzymes with broad substrate specificity for many endogenous and xenobiotic carbonyl compounds. They catalyze the reduction of endogenous prostaglandins, steroids, and other aliphatic aldehydes and ketones. They also reduce a wide variety of xenobiotic quinones derived from polycyclic aromatic hydrocarbons. CBR reduces the anthracycline anticancer drugs, daunorubicin(dn) and doxorubicin (dox) to their C-13 hydroxy metabolites, changing the pharmacological properties of these drugs. Emerging data on CBRs over the last several years is generating new insights on the potential involvement of CBRs in a variety of cellular and molecular reactions associated with drug metabolism, detoxication, drug resistance, mutagenesis, and carcinogenesis.

Carbonyl reductase (CBR) catalyzes the reduction of daunorubicin (DN) to its corresponding alcohol, daunorubicinol (DNOL), and changes the pharmacological properties of this cancer chemotherapeutic drug. The DN reductase associated with CBR reduces the C13 methyl ketone group and does not metabolize the quinone ring of DN. Reports comparing DN and DNOL toxicity have resulted in various conclusions depending on the cells tested. Differences in toxicity could be due to variations in several enzymes involved in DN metabolism. In this report, the effects of CBR expression on DN metabolism and cell toxicity were determined by cloning and expressing a human CBR cDNA in DN reductase-deficient myeloid erythroleukemia K562 cells. CBR activity increased 83-fold in the K562-transfected cells and was associated with a 2-3-fold reduction in DN toxicity. Maximum protection occurred at 30 nM DN where 94% of the intracellular DN was converted to DNOL within 2 h. The reduced toxicity was specific for DN. Other CBR substrates such as menadione, phenanthrenequinone, and doxorubicin were equally toxic to both the CBR expresser cells and the control cells under the conditions tested. Our results suggest that high levels of CBR in tumor cells could contribute to drug resistance. The results also suggest that reduction of DN to DNOL protects against DN toxicity by altering interaction of the drug at one or more of the many target sites.

Two cDNA sequences encoding rabbit carbonyl reductase (CBR) were cloned from a lambda gt10 rabbit liver cDNA library, The rabbit cDNAs coded for a protein with 84% identity to human CBR. Transient expression of the two rabbit cDNA sequences in COS7 cells increased both quinone reductase and aldo-keto reductase activities. These data demonstrate that CBR cDNAs from rabbit and human tissues code for similar proteins.

An association between the resistance to mitomycin C (MMC) and a decrease of NAD(P)H:quinone oxidoreductase (NQO1) activity was reported for a MMC-resistant subline, HCT 116-R30A, derived from MMC-sensitive HCT 116 cells. Eight NQO1 cDNA clones were isolated from these two sublines by reverse transcription-PCR. Two clones, pDT9 from HCT 116 and pDT20 from HCT 116-R30A, are the full length of 274 amino acids. These two clones differ by a T to C substitution at nucleotide 464, which results in a replacement of arginine 139 by tryptophan in the enzyme. NQO1 of pDT9 and pDT20 was expressed in Escherichia coli, purified, and shown to have a protein subunit of M(r) 30,000. The change of amino acid 139 resulted in a shift of isoelectric pH from 9.5 to 8.35 and a 60% decrease of activity in reducing MMC. All of the other six clones differ from pDT9 by a deletion of exon 4. On Northern blot, we detected two mRNA species of NQO1 (1.2 and 2.7 kilobases) due to alternative polyadenylation in all sublines. MMC-resistant sublines showed 75-90% mRNA expression relative to HCT 116 cells. Reverse transcription-PCR amplification of cDNA fragment of nucleotide 298-617 revealed two full-length mRNAs in HCT 116 cells but only one full-length mRNA in HCT 116-R30A cells. An exon 4 deletion mRNA was detected in both sublines. The two full-length mRNAs may be from either alleles or chimeras of the same gene and the exon 4 deletion mRNA is a result of alternative splicing. On Western blot, we detected only one M(r) 30,000 protein in all sublines. A substantial decrease of this protein in MMC-resistant sublines (5% of HCT 116) explained the 95% decrease of their NQO1 activity. Transcriptional regulation and posttranscriptional modification may be responsible for the disparity of gene expression of NQO1 and the low concentration of NQO1 protein in MMC-resistant sublines. Reversal of MMC resistance and the recovery of NQO1 in two revertants further supports the hypothesis that cellular control of NQO1 can modulate the cytotoxicity of MMC.

Human carbonyl reductase (CBR) belongs to a group of NADPH-dependent enzymes called aldo-keto reductases. The enzyme can function as an aldo-keto reductase or as a quinone reductase with potential for modulating quinone-mediated oxygen free radicals, The CBR gene was mapped by high-resolution fluorescence in situ hybridization to band 21q22.12, very close to the SOD1 locus at position 21q22.11. CBR displayed gene dosage effects in trisomy 21 human lymphoblasts at the DNA and mRNA levels. Lymphoblasts with increasing chromosome 21 ploidy also showed increased aldo-keto reductase activity and increased quinone reductase activity. Both aldo-keto reductase activity and quinone reductase activity have been shown to be associated with carbonyl reductase. The location of CBR near SOD1 and the increased enzyme activity and potential for free radical modulation in trisomy 21 cells implicate CBR as a candidate for contributing to the pathology of certain diseases such as Down syndrome and Alzheimer disease.

We have mapped the chromosomal location of four genes previously assigned to human chromosome 21--Cu/Zn superoxide dismutase (SOD1), the protooncogene ETS2, the interferon alpha/beta receptor gene (IFNAR), and the carbonyl reductase gene (CBR)--in the tammar, Macropus eugenii. The genes are localized on two separate autosomes: SOD1 and CBR map to chromosome 7 and ETS2 and IFNAR map to chromosome 3 or 4. These results provide the first example of asynteny between SOD1/CBR and ETS2/IFNAR in a mammalian species. The results suggest that either this synteny group has been disrupted in the marsupial lineage, or, alternatively, the genes located on human chromosome 21 may have been joined after the marsupials diverged from the eutherian mammals some 130-150 million years ago.

A prokaryotic expression plasmid, pKK-DT2, containing the cDNA of rat liver NAD(P)H:quinone-acceptor oxidoreductase (EC 1.6.99.2; DT-diaphorase) was constructed and used to transform Escherichia coli strain JM109. The rat liver quinone reductase was expressed in strain in JM109 and was inducible with isopropyl beta-D-thiogalactopyranoside (IPTG). The expressed rat protein was purified by affinity chromatography and had kinetic and physical properties identical with the protein purified from rat liver in that it could utilize either NADH or NADPH as the electron donor and its activity was inhibited by dicoumarol. In addition, we have generated four mutants, Arg-177----His (R177H), Arg-177----Ala (R177A), Arg-177----Cys (R177C) and Arg-177----Leu (R177L), using this expression system. Several of the mutants behaved anomalously on SDS/PAGE, but all of the mutant proteins had the expected M(r) as determined by electrospray m.s. These results and those obtained from enzyme kinetic analysis, u.v./visible absorption spectral analysis, and flavin and tryptophan fluorescence analysis of the wild-type enzyme and four mutants indicated that mutations at Arg-177 changed the conformation of the enzyme, resulting in a decrease in enzyme activity. Replacing Arg-177 with leucine altered the protein conformation and decreased FAD incorporation.

A DNA polymorphism has been found in the 3′ untranslated region of the carbonyl reductase gene (CBR). Genotypes of the members of the CEPH pedigrees have been obtained and used in linkage analysis to map the CBR gene in the linkage map of human chromosome 21. The gene maps between the interferon-α receptor (IFNAR) and the D21S55 loci.

Carbonyl reductase (NADPH: secondary-alcohol oxidoreductase; EC 1.1.1.184), a widely distributed NADPH-dependent enzyme considered as both an aldo-keto reductase and a quinone reductase, was cloned from a human liver genomic library and transiently expressed in COS7 cells. The gene contains 3142 bases comprising three exons and two introns. The absence of a CAAT and TATA box and the presence of a GC-rich island are characteristic of many "housekeeping" genes. Transient expression of the genomic gene in COS7 cells using an expression vector containing an SV40 origin of replication resulted in a greater than 50-fold increase in both menadione reductase activity and daunorubicin reductase activity, suggesting that both activities are derived from the same enzyme. Carbonyl reductase mRNA levels reflected enzyme activity levels in the transfected cells. Other parameters, such as pH profile, cofactor requirements, substrates, and inhibitors, were similar to those of carbonyl reductase purified by other investigators. Potential regulatory elements with consensus sequences for two GC boxes and the transcriptional activator protein AP-2 were present upstream of the transcriptional start site. Although the precise role of carbonyl reductase is unknown, the enzyme is involved in drug metabolism and in the reduction of activated carbonyl compounds. Its ability to act as a quinone reductase also implies a potential to modulate oxygen free radicals.

Rat liver NAD(P)H:quinone oxidoreductase cDNA was cloned and expressed in a eukaryotic cell expression plasmid containing a cytomegalovirus (CMV) promoter. Transient expression of enzyme activity and RNA transcription were measured in COS7 cells. The expressed quinone reductase has kinetic properties similar to the rat liver enzyme and is inhibited by dicourmarol, a known inhibitor of NAD(P)H:quinone oxidoreductase. Site-directed mutagenesis experiments carried out using this expression system revealed possible regions involved in NAD(P)H binding.

Tyrosyltubulin ligase (TTL) was found to be present in CHO and V79 Chinese hamster cells grown in tissue culture. The enzyme is soluble and requires potassium, magnesium, and ATP for maximum activity and requires tubulin as a substrate. TTL was analyzed through the cell cycle of V79 and CHO Chinese hamster cells. The enzyme showed two peaks of activity in V79 cells at 4 h and 7 h after mitotic selection, corresponding to the early S and mid to late S phases of the cell cycle. In CHO cells the enzyme displayed a major peak of activity at mid S and a minor peak or plateau during early S. Tubulin, as measured by (3H)colchicine binding, was shown to increase through S phase and reach a maximum late in the cycle during G2 approx. 3 h after maximum TTL activity.

Chinese hamster ovary cells were synchronized by mitotic selection and used to study the relation of poly(adenosine diphosphate ribose) synthesis to DNA synthesis and the different phases of the cell cycle. DNA synthesis was measured in cells rendered permeable to exogenously supplied nucleotides. Poly(ADPR) synthesis was also measured in permeable cells in the presence of both minimum and maximum DNA damage. The maximum DNA damage was produced by treating the cells with saturating concentrations of DNase. As anticipated, the DNA synthesis complex showed its maximum activity during S phase and showed 4–5-fold less activity during the other phases of the cell cycle. The basal level of poly(ADPR) synthesis was elevated during G1, fell to its lowest level during S phase, then increased during G2 and rose to its highest level during G1. The DNase responsive activity of poly(ADPR) synthesis was relatively constant thru the cell cycle but showed a peak at the end of S phase; then the activity decreased during the subsequent G2-M period.

Synthesis, degradation, and colchicine-binding characteristics of the microtubule protein were assayed in the cell cycle of two Chinese hamster lines. Levels of microtubule protein in the cell cycle were determined by electrophoresis of purified iodinated microtubule protein together with tritium-labeled whole cell lysates. Synthesis occurred almost exclusively in the G2 phase of the cloned aneuploid V79 line, whereas in the diploid Don line, considerable microtubule protein was synthesized in mid S phase as well as G2. In both Don and V79 cells, colchicine-binding activity displayed a pattern which was in agreement with the changing levels of microtubule protein determined with polyacrylamide gels. Degradation of the protein appeared to be intermittent. Microtubule protein synthesized in G2 phase survived into the ensuing G1 phase but was catabolized rapidly and concomitantly with initiation of DNA replication. The results show an intracyclic rhythm in microtubule synthesis and colchicine-binding activity and give support to the idea that oscillatory behavior of proteins is a common characteristic of the mammalian cell cycle.