Primary structure of human deoxycytidylate deaminase and overexpression of its functional protein in Escherichia coli.

The cDNA encoding human dCMP deaminase was isolated from a lambda ZAPII expression library using an antibody generated against highly purified HeLa cell dCMP deaminase. The cloned cDNA consists of 1856 base pairs and encodes a protein of 178 amino acids with a calculated molecular mass of 19,985 daltons. The sequence of several cyanogen bromide-cleaved peptides derived from HeLa cell dCMP deaminase are all contained within the deduced amino acid sequence. A zinc binding region is present in the enzyme, similar to that reported for cytidine deaminase (Yang, E. C., Carlow, D., Wolfenden, R., and Short, S. A. (1992) Biochemistry 31, 4168-4174). Northern blot analysis revealed a predominant messenger RNA species of 1.9 kilobases. Expression of the active protein to about 10% of Escherichia coli's total protein was achieved by subcloning the open reading frame into a high expression system using the polymerase chain reaction. Polyacrylamide gel electrophoresis revealed a prominent protein band which comigrated with affinity purified HeLa dCMP deaminase, while Western blot analysis yielded an immunoreactive band which comigrated with the single immunoreactive affinity column purified dCMP deaminase band. The enzyme which possesses a kcat of 1.02 x 10(3) s-1 was purified to homogeneity in over 60% yield. The overexpression of dCMP deaminase should permit more exacting studies on the regulation of this important allosteric enzyme which provides substrate for DNA synthesis.

of dUMP available for thymidylate synthase is finely controlled by the end products of the pyrimidine deoxynucleotide pathway.
Several observations indicate the potential importance of dCMP deaminase in DNA replication. First, the activity of the deaminase is elevated in such rapidly dividing tissues as chick and rat embryo (3), regenerating liver (4), and rat hepatomas (5). Second, HeLa cell dCMP deaminase activity is highest in late S phase and subsequently declines in the following G2 phase (6). Finally, the absence of dCMP deaminase activity in mammalian cells induces an imbalance in deoxyribonucleotide pools such that dTTP levels decrease while dCTP levels increase, resulting in enhanced mutagenesis (7). Evidence has also been presented by Jackson et al. (8) and Chiu et al. (9) that dCMP deaminase is a major contributor of dUMP for thymidylate synthase. In the past several years, dCMP deaminase has been purified to homogeneity from a variety of sources including chick embryo (lo), donkey spleen ( l l ) , T2r' bacteriophage-infected Escherichia coli (12), and HeLa cells (13). However, due to the low amounts of the enzyme present in mammalian cells, purification of the enzyme to homogeneity has proven difficult, thus limiting the availability of pure enzyme for biochemical analysis.
Recently, a dCMP deaminase gene was cloned and sequenced from T4 bacteriophage (14) and compared with the complete amino acid sequence of the enzyme isolated from T2 bacteriophage-infected E. coli (15). A homology comparison of the latter with a deduced sequence of dCMP deaminase from Saccharomyces cerevisiue has been made (16), and several regions of similarity were found. In this paper, we describe the cloning and expression of dCMP deaminase from HeLa cells, providing for the first time the amino acid sequence from a mammalian source.

Materials
The HeLa cell XZAPII cDNA library and picoblue immunoscreening kit were purchased from Stratagene (Palo Alto, CA). The nitrocellulose filters used were from Schleicher and Schuell. Hybond-N nylon membranes, [(u-~'P]~CTP and [(U-~~SI~ATP were obtained from Amersham Corp. Cyanogen bromide was purchased from Pierce Chemical Co. All solvents were prepared with Millipore nanopure water. Oligonucleotide primers were synthesized in this institution using a Millipore model 8750 DNA Synthesizer.

Methods
Peptide Sequencing-Human (HeLa cell) dCMP deaminase was purified as described previously (13) and was homogeneous as determined by SDS-PAGE.' Approximately 400 pg of the purified enzyme The abbreviations used are: PAGE, polyacrylamide gel electrophoresis; IPTG, isopropyl-l-thio-+%D-galactopyranoside; CD, DNA insert encoding dCMP deaminase; PCR, polymerase chain reaction, AP, affinity column purified; kb, kilobase(s). was cleaved with cyanogen bromide (200-fold excess over methionine) in 70% formic acid at room temperature in the dark (17). After 18 h the reaction mixture was diluted 20-fold with water and lyophilized. The cleavage products were dissolved in 400 p1 of 0.1% trifluoroacetic acid and subjected to chromatography on a C8 aquapore RP-300 reverse-phase column (0.4 X 25 cm) using a Beckman model llOA system equipped with a Chromopac CR3A data processor (Shimadzu, Kyoto, Japan). The column was eluted at a flow rate of 0.7 ml/min with a linear gradient of 0.1% trifluoroacetic acid to 60% of a solution containing 0.1% trifluoroacetic acid in 90% acetonitrile. Individual fractions were collected, and those selected for sequencing were concentrated in a Savant Speed-Vac to a small volume. Peptide sequences were determined by automated Edman degradation with a 477A Applied Biosystem sequencer/l20A phenylthiohydantoin amino acid analyzer.
Immunoscreening for Detection of dCMP Deaminase Antigem- Screening of the HeLa cell library (2 X lo6 recombinants) was performed according to the procedure of Young and Davis (18) with minor modifications as described in Stratagene's (Stratagene Cloning Systems, La Jolla, CA) picoblue immunoscreening protocol. The antisera, rabbit anti-dCMP deaminase, was prepared as described previously using homogeneous HeLa dCMP deaminase as the immunogen (13). The antisera was protein A purified and precleared of nonspecific antibodies at a dilution of 1:lOO by pseudoscreening with the XZAPII library (19). The precleared dCMP deaminase antiserum reacted only marginally with an E. coli phage lysate. For immunoscreening, the anti-dCMP deaminase serum was used at a 1:lOOO dilution in TBST buffer (10 mM Tris base, 100 mM NaCl, 0.05% Tween-20, pH 8.0). All procedures following plaque lifts were performed at room temperature. IPTG-soaked nitrocellulose filters were lifted and briefly rinsed in TBST. The filters were then blocked overnight in 1% gelatin-TBS. Afterward, the filters were washed two times for 10 min each in TBST, followed by incubation for 2 h with the deaminase antiserum. The filters were then washed three times for 10 min each and incubated with goat antirabbit-IgG-alkaline phosphatase-conjugate for 1 h at a 1:2000 dilution in TBST. Color development was initiated by incubating the filters with 0.3 mg/ml of nitro blue tetrazolium and 0.15 mg/ml of 5-hromo-4-chloro-3-indolyl phosphate in 100 mM Tris-HC1, pH 9.5, 100 mM NaCl, and 5 mM MgC12. Under these conditions, approximately 2 ng of purified HeLa dCMP deaminase could be detected by dot blotting.
DNA Sequence Analysis-The dCMP deaminase cDNA insert was excised from the XZAPII vector and subcloned into pBluescript according to the in vivo excision procedure supplied by Stratagene. The nucleotide sequence of both strands of the double-stranded dCMP deaminase cDNA was determined by the Sanger dideoxynucleotide chain termination method (20) using Sequenase (United States Biochemical Corp.). The ends of the cDNA were sequenced using T, and T? primers, and the remaining sequence was determined by primer walking using synthetic 18-to 21-mer oligonucleotides.
RNA Isolation-Total RNA from approximately 1 X 10' HeLa cells was isolated according to the procedure of Birnboim (21), with minor modifications (22). RNA was quantitated spectrophotometrically where an Also of 1 = 40 pg/ml (23). When the A2~//A2w ratio was greater than 1.8 poly(A)+ mRNA was purified using the Poly A Quik oligo dT columns supplied by Stratagene.
Northern Blot Analysis-For Northern blot analysis, 2.5 pg of total RNA or 2 pg of poly(A)+ mRNA was size fractionated on a 1.2% agarose-formaldehyde gel (24). The RNA was then transferred to Hybond-N nylon membranes and hybridized with 32P-labeled dCMP deaminase cDNA, as described previously (22).
Subcloning of dCMP Deaminase-PCR (25) was used to subclone the protein coding region of pCD12 into PET-3xc (Novagen, Madison, WI) at the NdeI and BamHI polylinker sites. The primers used for amplification were 5'"AGCACCAGTGATCGACATATGAGT-GAAGTTTCC-3', which introduces an NdeI site and contains those bases encoding the first 5 amino acids of dCMP deaminase (underlined), and 5'-CTACGAACTGTGACGGATCCTCACTGAAGCTT-9 -3 ' , which introduces a BarnHI site and encodes the last 4 amino acids and stop codon of dCMP deaminase (underlined). The PCR amplification was performed in a 50-p1 volume with 5 ng of linearized template (pCD12), 50 pmol of each primer, 200 p~ of each dNTP, and 1.5 units of Vent DNA polymerase (New England Biolabs, Beverly, MA). The sample was overlayed with 100 pl of paraffin oil and amplified for 21 cycles using a DNA thermocycler (Perkin-Elmer Cetus) and the following parameters: denature at 95 "C for 2 min, anneal at 45 'C for 2 min, extend at 72 "C for 3 min. The PCR product was separated on a 1% low melt agarose gel and identified with ethidium bromide as a 574-base pair fragment, which was then excised and purified using Magic PCR Prep resin (Promega, Madison, WI). Next, the PCR product was digested with NdeI and BamHI, purified as above, and ligated into PET-3xc resulting in pETCD.
Expression of Active dCMP Deaminase-The pETCD construct was transformed into BLZl(DE3)pLysS (26) which was grown at 37 "C in LB medium containing 50 pg/ml of ampicillin and 25 pg/ml of chloramphenicol to an A m of 0.600. The culture was then induced by the addition of 0.4 mM IPTG (26). Cells were harvested 4-h postinduction by centrifugation at 4,300 X g for 10 min at 4 "C. The bacterial pellet was resuspended in TME buffer (10 mM Tris-HCl, pH 8.0, 1 mM MgCL, 0.1 mM EDTA), centrifuged at 12,000 X g for 10 min at 4 "C, and the resultant pellet was frozen at -20 "C. The pellet was defrosted on ice and resuspended in 12 ml of buffer A (10 mM potassium phosphate, pH 7.5, 10% ethylene glycol (v/v) 0.2 mM MgC12, 0.1 mM dCTP, 0.1 mM EDTA, 20 mM 2-mercaptoethanol)/g of cells. Protease inhibitors were added (0.1 mM EDTA, 1 mM benzamidine, 1 pg/ml leupeptin, 1 pg/ml aprotinin, and 0.1 mM phenylmethylsulfonyl fluoride), and the suspension was sonicated using a Vibracel sonicator from Sonics and Materials, Inc., Danbury, CT. The sample was centrifuged at 12,000 X g for 15 min at 4 "C to obtain a soluble extract which was assayed for dCMP deaminase activity spectrophotometrically (27).
Purification of the Induced Enzyme to Homogeneity-All procedures were carried out at 0-4 "C unless specified otherwise. Frozen cells (BL21(DE3) pLys/pETCD) from 4 liters of cell culture (8-10 g) were thawed and suspended in 12 ml/g of TME containing 20 mM 2mercaptoethanol, and the protease inhibitors indicated above. This mixture was sonicated 10 times for 1 min each using a probe with 12.3-mm tip, with cooling for 2 min between each pulse, and centrifuged for 30 min at 43,000 x g. Solid ammonium sulfate was added slowly to the supernatant fraction to 30% saturation. After stirring for 10 min, the small precipitate which formed was removed by centrifugation and discarded. Additional solid ammonium sulfate was added to a final concentration of 80% saturation, and after stirring for 10 min the precipitate was collected by centrifugation. The residue obtained after decanting the supernatant was stored at -50 "C until the next step. At this point the frozen pellet was thawed, dissolved in buffer A, and dialyzed overnight against two 2-liter changes of buffer A.

DE52
Chromatography-A column of DE52 (3.4 X 10 cm) preequilibrated with 10 mM potassium phosphate, pH 7.5, was washed with 200 ml of buffer A prior to addition of the sample. The dialyzed sample was sonicated 10 times for 1 min each (to further reduce associated nucleic acid in size) with 2-min intervals of cooling in between each sonication and then applied to the column. After the sample was adsorbed, the column was washed with buffer A until the Azw decreased and the eluate was clear. The potassium phosphate concentration of buffer A was increased to 100 mM, and 50 ml was applied to the column, followed by 300 ml of buffer A containing 300 mM potassium phosphate, pH 7.5. The enzyme was eluted in the last buffer in a volume of about 200 ml and was precipitated with solid ammonium sulfate to a final concentration of 80% saturation. The precipitate was collected by centrifugation at 43,000 X g for 20 rnin and stored frozen at -50 "C.
Phosphocellulose Chromatography-The frozen pellet was thawed and dissolved in buffer A at pH 7.1 and dialyzed overnight against two 1-liter changes of buffer A, pH 7.1. A column of phosphocellulose (3.4 x 10 cm) pre-equilibrated with 10 mM potassium phosphate, pH 7.1, was washed with 150 ml of buffer A at pH 7.1, prior to addition of the dialyzed sample to the column. Afterward, the column was washed with buffer A, pH 7.1, until the A2w was negligible. The column was then developed by adding 100 ml of buffer A containing 50 mM potassium phosphate, pH 7.1, followed by 200 ml of buffer A containing 200 mM potassium phosphate, pH 7.1. The enzyme eluted in the last buffer and was concentrated in a stirred Amicon concentrator (Amicon, Inc., Beverly, MA) using a PMlO membrane. For storage the concentrated enzyme was precipitated by addition of solid ammonium sulfate to 80% saturation and centrifuged. In most cases the enzyme was over 95% pure at this stage but sometimes required another passage through a second phosphocellulose column. Small amounts of deaminase can be purified to complete homogeneity by passage of the enzyme from the first column over an affinity column developed for the Hela dCMP deaminase (13). The procedure, as described, yields about 50 mg of the pure enzyme from 4 liters of starting culture. The results of a typical purification are presented in Table I    Sodium Dodecyl Sulfate-Polyacrylnmide Gel Electrophoresis-SDS-PAGE was performed as described by Laemmli (28) using a Mini-Protean I1 Dual slab cell (Bio-Rad). The samples were diluted in a denaturing buffer of 0.06 M Tris-HC1, pH 6.8, 1% SDS (w/v), 10% glycerol (v/v), 0.14 M 2-mercaptoethanol, and 0.003% bromphenol blue) and heated to 100 "C for 4 rnin. The samples and molecular weight standards (Bio-Rad) were electrophoresed using a 12.5% (w/ v) resolving gel and a 4% stacking gel, each containing 0.1% SDS, at 100 V for 25 min followed by 200 V for an additional 50 min. After electrophoresis the proteins in the gel were stained with Coomassie Brilliant Blue R-250 or subjected to Western blot analysis.

TGTACAAGGCCCCTCTGCAACTGGAGAG~TTAATTCCTATCCCGTGAGTGGATTGT GAGAAATTCCACCCACGTGGAGACAGCTTACTGCAGCACTGTTGGTG~CGGAGCTCTT CTGTGCCCTGGCTCCATGCTTTCACCTACACAAGCATCACCTTCCTAATCACCGCGGGG CGGGGAGCGTGTGGCTGTGCCCCTTCTCTT~AATCTCATTTAATT~ATT~CATGC TCAGTACCTGTGTTGAGAGGCTTTCTTTATCCT~GATTATTACCTTTTTAAAGCCCATGT GCTCTTATATTTTCATGAGTTTTTATTTTGTCTCTGAGATTTTGTATTCCACATTCTAG G G T A T T C T G T A A T T T G G C T C C T T A C C A T A T T A T T~T C T T A T T~T C
Western Bfot Analysis-The proteins in the gel were transferred electrophoretically to nitrocellulose according to the method of Towbin et ai. (29). The transfer was performed at 250 mA for 1 h using a Mini Transblot Electrophoretic Cell (Bio-Rad). Western blot analysis was performed at room temperature as described previously (30). The nitrocellulose was blocked with 1% gelatin-TBS for 16 h followed by incubation for 2 h with the rabbit dCMP deaminase antisera and for 1 h with goat-anti-rabbit IgG-alkaline phosphatase conjugate. Between each incubation, the nitrocellulose was washed with TBST. Immunoreactive bands were identified by incubating with nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate as described above.

Zsotation and Sequence Analysis of a HeLa Cell dCMP
Deaminase cDNA-Screening of approximately 2 X lo6 recombinants from a HeLa cell XZAPII cDNA library with protein A-purified dCMP deaminase antiserum yielded three positive clones. All three were digested with EcoRI and ana-lyzed by agarose gel electrophoresis (data not shown). One of the clones, pCD12, which contained a 1.8-kb insert was selected for further analysis, since it hybridized with degenerate oligonucleotides based on the partial amino acid sequence of CNBr peptides derived from dCMP deaminase.
The complete nucleotide sequence of both strands of pCDl2 was obtained by the dideoxy chain termination method (20) and is presented with the deduced amino acid sequence of dCMP deaminase in Fig. 1. The deaminase contains a 535base pair open reading frame encoding a protein of 178 amino acids with a calculated M, of 19,985. As further evidence that pCD12 encoded dCMP deaminase, the sequences of six dCMP deaminase CNBr peptides were located in the deduced amino acid sequence of the pCD12 insert (underlined in Fig. 1). The first three bases of the insert appear to code for the aminoterminal methionine of dCMP deaminase, but since the amino terminus of the isolated HeLa dCMP deaminase was found 28s --. to be blocked, this residue could not be verified by amino acid sequencing. Interestingly, the pCD12 insert contains a very long 3"untranslated region of 1.3 kb, which includes a polyadenylation signal a t nucleotides 902-907, two putative AUrich motifs a t nucleotides 738-742 and 1639-1643, as well as the poly(A)+ tail (residues 1832-1856). Northern Blot Analysis of HeLa Cell RNA-A "P-labeled EcoRI/HindIII fragment of pCD12, which spans the entire coding region was used as a hybridization probe to determine the size of dCMP deaminase mRNA. Northern blot analysis of total RNA (Fig. 2, lane I ) and poly(A)+ mRNA (lane 2) isolated from HeLa cells revealed the labeled DNA fragment to hybridize with a major RNA band of 1.9 kb in length, which is about the size expected for a transcript of the 1.86kb insert of pCD12. I t is clearly seen from the intensity of the band in lane 2 relative to lane I that the deaminase mRNA is greatly enriched in the latter. Hybridization of the EcoRI/ Hind111 fragment to a minor band of the poly(A)+ mRNA at 0.7 kb, which is not seen with total RNA (lane I ), may be due to a degradation product of the full-length mRNA encoding dCMP deaminase.
Expression of a Functional HeLa dCMP Deaminase in E. coli-To determine the extent to which pCD12 expresses dCMP deaminase in E coli, 0.4 mM IPTG was used to promote the synthesis of this enzyme from the lac promotor preceding the insert. At best only 0.06 units/mg protein of dCMP deaminase activity was obtained, which was comparable to that found in crude HeLa cell extracts (13). Because of this low level of deaminase expression the pBluescript plasmid was obviously not useful as a vector to amplify the deaminase. This effect may be due to the apparent toxicity of dCMP deaminase when expressed a t high levels in continuously growing cells (14). A more controlled system was sought, therefore, which could be turned on as desired and was found in the PET-3xc vector developed by Studier and co-workers (26,31,32), where protein synthesis is under the control of bacteriophage T 7 transcription and translation signals. The PCR procedure was used to amplify the open reading frame of pCD12 with NdeI and BarnHI sites conveniently placed at the 5' and 3' ends, respectively. This DNA fragment was then   cloned into the NdeIIBarnHI polylinker sites of PET-3xc to generate pETCD, which was transformed into the tightly regulated expressing strain of E. coli, BL21(DE3)pLysS (26).
Nucleotide sequencing of both strands confirmed that the polymerase chain reaction faithfully reproduced the dCMP deaminase coding region of pCD12. Following IPTG induction, dCMP deaminase activity was monitored over time using the spectrophotometric assay. It was noted that the specific activity of dCMP deaminase was maximal 4 h after the addition of IPTG and declined thereafter (data not shown). Using a 4-h induction period, 83 units/ mg of dCMP deaminase were obtained. This represents a 1,400-fold amplification over that obtained with the pBluescript plasmid (pCD12). Interestingly, in the absence of induction by IPTG 19 units/mg of dCMP deaminase was obtained, indicating that PET-3xc expression from the T 7 promotor was somewhat leaky. This effect occurred despite the presence of pLysS which should have provided sufficient T 7 lysozyme to inhibit any T7 RNA polymerase that might have been produced in the absence of IPTG (26). Evidence that this activity is due to dCMP deaminase was obtained with 2'-B-~-deoxyribose pyrimidin-2-one 5'phosphate, a compound that is a potent and specific inhibitor of dCMP deaminase (13).
Soluble cell extracts prepared from IPTG-induced pETCD containing cells were applied to a 12.5% gel (Fig. 3A), and as indicated in lane 4 a prominent band comigrating with AP-HeLa dCMP deaminase ( l a n e 5 ) was obtained. Non-induced pETCD extracts ( l a n e 3 ) also contained a band with a similar M , to AP-dCMP deaminase, although the intensity of this band was far less then when the enzyme was induced by IPTG. To confirm the requirement for the CD cDNA insert, IPTG-induced cells harboring PET-3xc lacked a band in the M, range of dCMP deaminase. As further proof, that the bands in lanes 3 and 4, which comigrate with AP-dCMP deaminase (lane 5 ) were due to the deaminase, the soluble extracts were subjected to Western blot analysis (Fig. 3B). As expected, a single immunoreactive band was detected in lanes 3 and 4, which comigrated with the immunoreactive AP-dCMP deaminase band in lane 5, but none was evident in the soluble extract prepared from IPTG-induced cells harboring PET-3xc (lane 2) in contrast to those cells containing pETCD ( lanes 3-5 ) .
Because the IPTG-induced deaminase was present to the extent of 10% of the cellular protein, the purification of this enzyme should have posed little difficulty. However, as shown in the gel pattern (Fig. 5 ) , a very minor contaminant at about 37 kDa persisted even after passage through a dTMP synthase affinity column, which was effective in removing a similar band from the HeLa cell deaminase (13). However, passage of the enzyme from step 4 ( Table 1) through a second phosphocellulose column could minimize the contaminants's presence, or even better passage of the step 4 enzyme through a dCMP deaminase affinity column (13) could basically eliminate it (Fig. 4). These preparations of dCMP deaminase possessed a specific activity of about 500 units/mg of protein or a kc,, of 1.02 x lo3 s-l a t 30 "C, which is comparable to that obtained for the deaminase purified from HeLa cells (13).

DISCUSSION
This paper describes for the first time the amino acid sequence of a human dCMP deaminase as deduced from its corresponding cDNA. The only other eukaryote dCMP deaminase sequence available to date is that from s. cereuisiae (16), but in view of its unusual size (312 amino acids) relative to the human enzyme (178 amino acids) and most other deaminases described to date, its structure should be verified. Most of the homology between the sequence of the yeast deaminase and the T4 phage and human deaminases resides in 150 amino acids of the yeast enzyme's carboxyl-terminal end and of these, about 50 amino acid residues are related and fall into the linear peptide segments presented in Fig. 5. Since the yeast deaminase has not been purified it is not clear how many subunits constitute this protein or what its true size is. It would be surprising though if this enzyme turns out to be as different from the T4 phage and human dCMP deaminases, as the deduced sequence suggests (16).
Another deaminase with an anomalous molecular mass has been purified to apparent homogeneity from human spleen (33), one with an apparent molecular mass of 106 kDa. Unlike the other dCMP deaminases, which contain six subunits, this enzyme is composed of two subunits and possesses a specific activity that is 40-50 times lower than the HeLa cell deaminase. However, the spleen enzyme is still allosterically regulated by dCTP and dTTP. Whether the spleen deaminase represents an organ-specific variant of the human dCMP deaminase described in this paper remains to be determined.
There is little doubt that the cDNA isolated from the HeLa cell library encodes the same protein purified to homogeneity from HeLa cell extracts (13). Thus, several CNBr peptides isolated from the purified enzyme were identified within the cDNA sequence of the deaminase, the M , values of the isolated HeLa cell and recombinant proteins were identical, and both gave positive Western blots with antibody to the HeLa deaminase at the same migration distance following SDS-PAGE. In addition both were inhibited with the same Ki by the transition state analogue, 2'-P-~-deoxyribose-pyrimidin-2-one 5"phosphate. The only difference between the two proteins resides in the fact that the recombinant protein's amino end is free, while that from HeLa cells appears to be blocked, which is a common feature of proteins isolated from mammalian sources. The human dCMP deaminase mRNA contains an unusually long 3"untranslated region, which is about 1.3 kb in length. Transferrin receptor mRNA also contains a long 3'untranslated region, which appears to be involved in the irondependent regulation of mRNA stability (34, 35). Whether a similar role can be attributed to the 3"untranslated region of dCMP deaminase mRNA remains to be determined.
Several regions of common identity have been noted on comparison of the human, T4-phage and S. cerevisiae dCMP deaminases (Fig. 5). One of particular interest is that most likely involved in the catalytic site, which is similar to that described for cytidine deaminase (36). The latter has been shown to contain a region that could be one of high affinity for the chelation of zinc. The residues in this site believed to be involved are 1 histidine and 2 cysteines (see boxed regions in Fig. 5). This is in contrast to the case of adenosine deaminase, where it has been shown that zinc is coordinated by 3 histidines in the active site region (37). It appears to be more than a coincidence that the human, T4 phage, and yeast dCMP deaminases have 1 histidine and 2 cysteines in positions comparable to those of cytidine deaminase (Fig. 5). It has been shown recently that the phage deaminase not only contains a zinc atom in the above-mentioned catalytic site, but also one in another site that appears to be a unique zinc finger in which the zinc is coordinated between 3 histidines and 1 cysteine (38). Evidence has been obtained for the T4-dCMP deaminase to support the role of zinc as a promoter of hydrolysis similar to that found for the cytidine and adenosine deaminases, in that replacement of histidine 105 or cysteines 132 or 135 of T4 phage dCMP deaminase by other amino acids, results in the loss of zinc and coincidentally deaminase activity? It is of interest to note that the addition of Zn2+ to the HeLa deaminase affects neither activity nor allosteric activation or inhibition. In these cases M P appears to play a major role. However, in the case of the BaciElus subtilis dCMP deaminase Zn2+ is absolutely required for both effects and could not be replaced by M%+ (39). The role of zinc in such biological processes as transcription and enzyme activity is becoming increasingly obvious. It would be of interest to determine whether a mutational deficiency in Zn2+ is deleterious to these processes, in particular dCMP deaminase activity.
The importance of dCMP deaminase in DNA replication, through its provision of dUMP for thymidylate synthase, has become evident through studies with cells that are deficient in the enzyme. These cells show imbalances in their intracellular dCTP and dTTP pools (6, 40, 41), which can lead to increased mutation rates during DNA replication (42-44). Thus, the relative importance of dCMP deaminase in DNA replication heralds its importance as a potential target for chemotherapeutic agents. However, due to the relative difficulty in obtaining the active human enzyme until now, most studies on the deaminase have been performed on non-human sources. Our ability to overexpress the active enzyme will now enable the isolation of sufficient deaminase to probe its structure and to undertake the development of potential chemotherapeutic agents specifically targeted to the deaminase. It should be determined whether their use in combination with inhibitors of thymidylate synthase can potentiate the effectiveness of thymidylate synthase inhibitors.
Clinically, dCMP deaminase levels have been found to be elevated in the sera from patients with various disease states (45) and, therefore, have the potential of being used for the early detection of such diseases. As an example, several studies have clearly demonstrated that pregnant women with preeclampsia show increases in serum dCMP deaminase activity early in the onset of the disease (46-49). In another case, that of myocardial infarction (50, 51), the enzyme slowly but significantly increases in serum. The assays employed in these studies, however, are too complex or too insensitive to be clinically useful, and in addition, determine only the amount of active deaminase in serum. Now, through the large scale production of the active recombinant enzyme, we can develop a highly sensitive immunoassay, which could determine more accurately the true serum levels of dCMP deaminase protein both active and inactive.