Mutational Analysis of Active Site Residues of Human Adenosine Deaminase*

THEJOURNAL OF BIOLOGICAL CHEMISTRY Vol. 268, No. E, Issue of March 15, pp. 546”5470,1993 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc Printed in U.S.A.

Mutational Analysis of Active Site Residues of Human Adenosine Deaminase*

(Received for publication, September 8, 1992)

Dipa Bhaumik, Jeffrey Medin$, Karen Gathy, and Mary Sue Coleman8 From the Department of Biochemistry and Biophysics, University of North Carolina, Chapel Hill, North Carolina 27599

Adenosine deaminase was overexpressed in a bacu- combined immunodeficiency (Giblett et al., 1972). These pa- lovirus system. The pure recombinant and native en- tients have no obvious gastrointestinal tract abnormalities, zymes were identical in size,Zn2+ content, and activity. but they do exhibit a dramaticlymphopenia that seems to be Five amino acids, in proximity to the active site, were a direct consequence of the absence of adenosine deaminase replaced by mutagenesis. The altered enzymes were (Coleman et al., 1978; Donofrio et al., 1978). purified to homogeneity and compared to wild-type Potent inhibitors of adenosine deaminase are lympholytic adenosine deaminase with respect to zinc content, en- in humans, and this property hasbeen exploited in the treat- zymatic activity, and kinetic parameters. All but one ment of certain leukemias, the hallmark of which is accumu- of the alterations produced significant activitypertur- lation of differentiation-arrested lymphocytes (Coleman, bations. Replacement of Cysz02produced a protein that 1983). Ground and transition stateanalog inhibitors have also retained at least 30-40% of wild-type activity.In con- proven useful in studies of the reaction mechanism of aden- trast, replacements of His17, His214,Hiszs8, and Glu217 resulted in dramatic losses of enzyme activity. None of osine deaminase. With a rate enhancementof about lo‘’, this these mutants exhibited large variations in K,. The enzyme is among the most efficient that have been described proteins produced from alterations of amino acids im- (Frick et al., 1987). A hydrate tetrahedral intermediate has plicated in metal coordination were slightly activated been postulated from a large number of chemical studies by inclusion ofZnz+ throughout purification. These (Evans and Wolfenden, 1973; Wolfenden et al., 1969; Kurz experiments confirm that in the active enzyme Zn2+ and Frieden, 1983). The most convincing evidence for this plays a critical role in catalysis, that a histidine or intermediate reaction product is from 13C NMR studies of glutamate residue plays a mechanistic role in the hy- adenosine deaminase bound to purine riboside (1,6-dihydro- drolytic deamination step, and that cysteine is not in- purine riboside), in which a change of hybridization from sp2 volved in the catalytic mechanism of adenosine deam- to sp3 is detected (Kurz and Frieden, 1987). Subsequent UV inase. These data support the roles for these amino acid and NMR studies confirmed that this inhibitor is bound as residues suggested from the x-ray structure of murine an oxygen adduct, presumably hydrated at the 1,6 position adenosine deaminase (Wilson, D. K., Rudolf, F. B., and (Jones et al., 1989). This covalent hydrate with C6 in the Quicho, F. A. (1991)Science 252, 1278-1284). adenosine deaminase-purine riboside complex has been con- firmed recently by the determination of its structure by x-ray crystallography (Wilson et al., 1991). Unexpectedly, the crys- talstructure also revealed that adenosine deaminase is a Adenosine deaminase (EC 3.5.4.4), an important enzyme of metalloenzyme that complexes 1 mol of Zn2+ per molof the purine salvage pathway, catalyzes the irreversible hydro- protein. lytic deamination of adenosine or deoxyadenosine to inosine Solution of the crystal structure of a mammalian adenosine or deoxyinosine. Adenosine deaminase is expressed at very deaminase provided knowledge of the amino acids in the high levels along the entire murine gastrointestinal tract, in active site. However, at pH4.2, where crystals were generated thymic T cells and in decidual cells of the developing mater- for x-ray analysis, adenosine deaminase is almost completely nal-fetal interface (Lee, 1973; Knudsen et al., 1988 and 1989; inactive, and at this pH thesubstrate analogue, purine ribo- Witte et al., 1991). In humans,the upper gastrointestinal tract side is only weakly bound (Wolfenden and Kati, 1991). Con- is devoid of this enzyme activity, but high levels are expressed struction of mutations in active site residues coupled with in the lower part of the tract. determination of functional consequences of each mutation The wide spectrum of adenosine deaminase activity in under conditions of optimal enzyme activity, will permit de- mammalian tissues portended an important role for purine tailed characterization of the reaction pathway and descrip- metabolism in nutrition and reproduction. However, the en- tion of enzyme intermediates. tirepurine salvage pathway, and adenosine deaminase in In this study, guided in selection of targets by the crystal particular, became the focus of intenseinterest with the structure, we have altered key amino acid residues within the observation that hereditary deficiency of the enzyme in hu- active site of human adenosine deaminase, an enzyme that is man infants is invariably associated with a form of severe highly homologous to its murine counterpart. The recombi- nant enzymes were expressed in a baculovirus system and * This work was supported in part by United States Public Health purified to homogeneity on a monoclonal antibody affinity Service Grant CA26391 (to M. S. C.). The costs of publication of this column. Kinetic characteristics,stabilities, and metal binding article were defrayed in part by the payment of page charges. This capacities of the altered enzymes were assessed and correlated article must therefore be hereby marked “aduertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact. with mechanistic models. $ Recipient of an Army Predoctoral Fellowship in Biotechnology. Present address: Laboratory of Molecular Growth Regulation, EXPERIMENTALPROCEDURES NICHD, National Institutes of Health, Bethesda, MD 20892. Materials-Oligodeoxynucleotide primers used in constructing mu- § To whom correspondence should be addressed. tations and sequencing were synthesized at the University of Ken-

5464

This is an Open Access article under the CC BY license. MutationalAnalysis of AdenosineDeaminase 5465 tucky Macromolecular Structure Facility on an Applied Biosystems stage. The infection was allowed to continue for 4 days after which 380B DNA synthesizer. Restriction endonucleases were from United the larvae were collected and frozen immediately at -70"C. The States Biochemicals and New England Biolabs. Sequenase DNA mutants H17A and H214L were overexpressed in Sf-9 and High-5 sequencing kits were obtained from United States Biochemicals. [a- cells. For cellular infection, 2.5 X 107cellsin T-175 flasks were infected 36S]dATPand [14C]adenosinewere Du Pont-New England Nuclear at a multiplicity of infection of 10 with the appropriate virus stock. products. The polyclonal antibody used in these experiments was Cells were harvested 65-h postinfection, washed twice with cold raised inrabbits in our laboratory against homogeneous human phosphate-buffered saline, and frozen at -70 "C. adenosine deaminase. The anti-adenosine deaminase monoclonal an- Purification of Recombinant Proteins-Wild-type adenosine de- tibody (NlD1) used in the study was also generated in our laboratory aminase was purified from frozen larvae by adenosine-affinity chro- and propagated in ascitesfluid (Philips etal., 1987).All other reagents matography (Medin et al., 1990). Briefly, frozen larvae (28 g) contain- were of the highest commercial grade available. ing the recombinant protein were homogenized in a buffer (10 mM Bacterial Strains and Vectors-The Escherichia coli strains used sodium acetate, pH 6.4, 2 mM EGTA, 5 mM benzamidine, 10 mM 6- for plasmid propagation were CJ236 and DH5a (Bethesda Research aminocaproic acid, 5 mM phenylmethylsulfonyl fluoride), and centri- Laboratories). The plasmid vector M13 mp18 was used for site- fuged at 30,000 X g for 30 min. Protamine sulfate was added to the directed mutagenesis of the adenosine deaminase cDNA. The bacu- crude extract and allowed to precipitate. The clarified supernatant lovirus transfer vector pAcC4, a generous gift from Cetus, was used obtained after centrifugation was loaded onto a DEAE-Sephadex in homologous recombination experiments to construct specific bac- column (25 X 5 cm). The protein was eluted with the same buffer ulovirus variants. containing 0.5 M sodium chloride and concentrated by ammonium Viruses, Cells, and Larvae-Autographa californica nuclear poly- sulfate precipitation. The precipitate, resuspended ina minimum hedrosis virus (ACMNPV strain Li, Invitrogen Corp.) and Spodoptera volume (-8 ml) of phosphate-buffered saline, pH 7.4, was applied to frugiperda (Sf-9) insect cells (Invitrogen) used were propagated in the an adenosine-Sepharose column (110 X 1.5 cm) (Schrader and Stacy, laboratory and used in the protein expression experiments. The 1977), and the fractions containing the major adenosine deaminase second insect cell line, High-5 (Invitrogen), was derived from eggs of activity were pooled and concentrated by ultrafiltration. the cabbage looper and was an alternate host for recombinant bacu- Mutant proteins were isolated by using a monoclonal antibody lovirus propagation. Both insect cell lines were cultured in Grace's affinity column. The matrix was constructed by cross-linking the media containing 10% fetal calf serum. Trichoplusia ni larvae were antibody (NlD1, the isotype with the highest affinity for human produced in this laboratory by methods previously described (Medin adenosine deaminase) to protein A-Sepharose (PAS-NlD1).The et al., 1990, 1992). immunoaffinity column was made as described by Philips et al. (1987). Site-directed Mutagenesis-To facilitate site-directed mutagenesis The isolation of the proteins on the monoclonal antibody column was of selected regions of the enzyme, the full length adenosine deaminase essentially the same as previously described (Philips etal., 1987) with cDNA was cloned into thevector M13 mp18. The cDNA wasreleased modifications in the initial extraction procedure. Frozen larvae (3-10 from the vector pBR8 (Medin et al., 1990) with the restriction g) were suspended in three volumes of cold extraction buffer contain- enzymes NcoI and HinfI. Phosphorylated linkers that convert the ing several protease inhibitors (50 mM potassium phosphate, pH 6.8, HinfI site toa NcoI site were ligated to the HinfI -endof the cDNA. 10 mM 6-aminocaproic acid, 5 mM benzamidine, 2 mM EGTA, 20 p~ This adenosine deaminase cDNA was isolated after subcloning into leupeptin, 8 p~ pepstatin, and 1 mM phenylmethylsulfonyl fluoride) the NcoI site of pAcC4 by digestion with NcoI. Phosphorylated NcoI and homogenized on ice with a Tekmar tissue homogenizer. The to EcoRI linkers were ligated to theadenosine deaminase cDNA, and homogenate was centrifuged at 30,000 X g for 30 min. The pellet was this cDNA was digested with EcoRI and ligated into the dephospho- re-extracted with half the original volume of buffer, homogenized, rylated M13 mp18 at theEcoRI site. This construct (M13 adenosine and centrifuged as before. Supernatants from both extractions were deaminase) was used in the development of all site-specific mutant mixed (crude extract) and brought to 30% ammonium sulfate satu- forms of adenosine deaminase. Site-directed mutagenesis was carried ration. The supernatant obtained after the first centrifugation was out using the method of Kunkel et al. (1987). M13 adenosine deami- further subjected to another cycle of fractionation with ammonium nase was transformed into E. coli strain CJ 236 (dut-, ung-). Single- sulfate (70%). Theprecipitate was resuspended in the original volume stranded uracil containing M13 adenosine deaminase was isolated of buffer used and incubated with pre-equilibrated protein A-Sepha- and annealed to synthetic phosphorylated mutagenic oligomers rose-monoclonal antibody matrix overnight on a rotating platform. (Table I). The second strand of DNA was synthesized using Sequen- The elution and concentration of the proteins (in 4-6 M urea) were ase version I1 and dNTPs. After ligation the reaction product con- done essentially as described before. The entire operation was carried taining the newly synthesized second strand was transformed into out at 4 "C. (dut', ung+) E. coli strain DH5a. The presence of the desired mutation For extraction of recombinant proteins from cells, frozen cells were was confirmed by DNA sequencing using dideoxynucleotides. suspended in four volumes of extraction buffer (see above, 6-amino- Expression of Recombinant Wild-type and MutantHuman Adeno- caproic acid was omitted and aprotinin was added to a final concen- sine Deaminase in Baculovirus-infectedLarvae and Insect Cells-The tration of 5 pglml) and sonicated on ice (two 30-s bursts). The mutated cDNA constructs were released from M13 mp18 by digestion sonicate was centrifuged at 30,000 X g for 30 min. The pellet was with NcoI and subcloned into pAcC4 at the NcoI site. In order to resuspended in half the volume of original buffer, homogenized, and produce the desired baculovirus strains, Sf-9 insect cells were cotrans- centrifuged again. Both the supernatantswere mixed with the protein fected with extracellular virus (ACMNPV) DNA and pAcC4 contain- A-Sepharose antibody column matrix. The rest of the procedure was ing adenosine deaminase. Recombinant viruses were identified, iso- identical to thatdescribed above. lated, and purified as in our previous study (Medin et al., 1990). Protein Determination-Protein concentrations were determined The wild-type adenosine deaminase andthe mutants H214A, by the Coomassie Blue dye binding method (Bradford, 1976) using H214N, E217A, H238A, and C262A were overexpressed in the 2'. ni bovine serum albumin as the standard and reagents from Bio-Rad. larvae (Medin et al., 1990). Briefly, 10 p1 of each viral stock (-lo9 For zinc determination experiments, protein concentrations were plaque-forming units) was injected into thelarvae at thefourth instar measured by amino acid analysis at the University of Kentucky,

TABLEI Nucleotide sequence of synthetic oligonucleotides used for isolation of mutants Base changes are underlined.

Name Synthetic primers Codon change Amino acid change H17A GAACTGCATGTCGCJCTAGACGGATCC CAC - GCC His-Ala H214A CACCGTACTGTCECGCCGGGGAGGTGGGC CAC - GCC His-Ala H214L CACCGTACTGTCCTCGCCGGGGAGGTGGGC CAC - CTC His-Leu H214N CACCGTACTGTCMCGCCGGGGAGGTGGGC CAC - AAC His-Asn E217A CACGCCGGGGCGGTGGGCTCC GAG - GCG Glu-Ala H238A GAGCGGCTGGGAECGGCTACCACACC CAC - GCC His-Ala C262A CATGCACTTCGAGATCECCCCTGGTCCAG TGC - GCC Cys-Ala 5466 MutationalAnalysis of Adenosine Deaminase Macromolecular Structure Facility. SDS-Polyacrylamide Gel Electrophoresis and Western Blotting- SDS-polyacrylamide gel electrophoresis was performed according to the method of Laemmli (1970) in a mini-gel apparatus (Bio-Rad). To visualize proteins the gels were either stained with Coomassie Bril- liant Blue or with silver staining procedures (Wray et al., 1981). Adenosine deaminase in the crude extracts was determined by West- ern blot analysis. Electrotransfer of proteins onto nitrocellulose sheets was performed according to Burnetteet al. (1981). Blots of the protein were analyzed with anti-human adenosine deaminase poly- clonal antibody (1:500, dilution) or anti-human adenosine deaminase monoclonal antibody (1:250, dilution). The second antibody used in the procedure was goat anti-rabbit or anti-mouse IgG alkaline phos- phatase conjugate (Bio-Rad). The immunoreactive proteins were detected with an alkaline phosphatase conjugate substrate kit from Bio-Rad. Enzyme Assay and KineticStudies-Adenosine deaminase activity was detected by using the radiochemical assay described before (Cole- FIG. 1. Proposed active site amino acid residues in adeno- man and Hutton, 1975) with ["Cladenosine as the substrate. One sine deaminases. Residues depicted are based on the solution of the unit of adenosine deaminase activity is defined as the amount of crystal structureof murine adenosine deaminase complexed to purine enzyme required to produce 1pmol of inosine/min at 37 "C.For each riboside (Wilson et al., 1991). These amino acids are identical in protein, kinetic measurements were done at a seriesof six concentra- human adenosine deaminase. The residues targeted for mutagenesis tions of adenosine so as to bracket the expected K,,, value. The in this study are boxed in bold lettering. concentration of enzyme ranged from 0.75 nM for the wild type to 1 pM for the least active mutant proteins. Kinetic constants were 123 4 5 6 7 8 910111213 determined by an enzyme kinetics program (Trinity software, Camp- . - . -. - " .- ~ - ton, NH). The nonlinear regression analyses from this program are reported in this paper. Thermolysin Digestion of Wild Type and Mutant Adenosine De- aminase-The structural integrity of each mutant protein was studied by performing a thermolysin digestion at a variety of temperatures. The method used by Polesky et al. (1990) was followed with slight modifications. The proteins were stable to thermolysin at an en- zyme:protein ratio of 1:50. Therefore an enzyme:protein ratio of 1:lO was used. Digestions were carried out for 45 min and were terminated by the addition of protein gel sample buffer. The digestion products were electrophoresed on a 12.5% polyacrylamide gel and visualized by staining with Coomassie Brilliant Blue. Zinc Content of Wild-type and Mutant Adenosine Deaminase- FIG. 2. Immunoblot analysis of recombinant human wild- Two methods, flame atomic absorption spectrometry and graphite type and mutant adenosine deaminases produced in insect furnace atomic absorption spectrometry, were used for zinc analysis. larvae and cells. Equivalent amounts of crude extract protein (30 Absorbance peaks by both methods were measured at 213.9 nm. All pg) or 200 ng of the purified wild-type adenosine deaminase were values shown are averages of duplicate determinations. The detection applied to 12% polyacrylamide gels and subjected to SDS-polyacryl- limits for zinc by the above two methods were 120 and 1.5 nM, amide gel electrophoresis. The proteins were transferred to nitrocel- respectively. lulose as described under "Experimental Procedures" and probed with For zinc analysis the protein samples were extensively dialyzed rabbit anti-human adenosine deaminase polyclonal antibody. Lune 1, against 5 mM HEPES buffer (passed through a Chelex-100 column) wild-type (larvae); lane 2, H238A (larvae); lane 3, E217A (larvae); in metal-free dialysis tubing prepared as suggested by Auld (1988). lane 4, C262A (larvae); lane 5, H214N (larvae); lane 6,H214A (larvae); The concentration of residual zinc in the buffer was subtracted from lane 7, H214L (larvae); lane 8, H17A (larvae); lane 9, uninfected larval those in the unknown protein samples. protein; lane 10,purified recombinant wild-type adenosine deaminase; lane 11, blank; lane 12 H17A (cells); lane 13, H214L (cells). RESULTS technique pioneered by Kunkel et al. (1987). Recombinant The goal of this study was to produce recombinant human plasmids were isolated, identified by restriction analysis, con- adenosine deaminase proteins in which amino acid residues firmed by sequence analysis, and subcloned into the transfer that have been implicated in or areclose to theenzyme active vector. site (Wilson et al., 1991) have been mutationally altered. 5 Each of the transfervector DNAs containing the adenosine amino acid residues were targeted for substitution, Hisz3*, deaminase open reading frame was then used in homologous Glu217, His214, His17, and CysZ6' (Fig. 1). Each of these was recombination with wild-type baculovirus DNA to generate converted to Ala except the residue which was changed recombinant baculovirus for each mutation. The relative lev- to Asn, Leu, and Ala. The first 4 residues in this group are els of wild-type and mutant adenosine deaminases produced predicted to be involved in ligand binding or catalysis. The in infected larvae were assessed by immunoblot analyses of CysZ6' is close to the active site pocket, but is blocked by a crude extracts. No immunoreactive protein was detected in wall of residues (Hisz3', SeP5 andAspzg5). crude extract of insect larvae infected with nonrecombinant We have previously demonstrated that wild-type human virus (Fig. 2, lane 9). adenosine deaminase is produced in very high amounts in T. Immunoreactive protein with a mobility equal to that ob- ni larvae and can be readily purified from larvae infectedwith served in a preparation of purified wild-type adenosine de- recombinant baculovirus containing the adenosine deaminase aminase (lane 10)was observed in clarified extracts contain- open reading frame under the control of the polyhedrin pro- ing mutants H238A, E217A, C262A, H214N, H214A, H214L, moter (Medin et al., 1990). Therefore, we selected the bacu- and H17A and a small amount of degraded adenosine deami- lovirus system for production of the seven mutant proteins. nase was observed in lanes 1-6 (lower band). However, the Expression and Immunodetection of Recombinant Pro- amounts of antigen present in equivalent aliquots of the teins-Each adenosine deaminase mutant was generated by extracts varied dramatically. The mutant protein that con- subcloning the cDNA into M13 and using the mutagenesis tained Ala in place of CysZ6*(lane 4) and Ala in 2 residues Mutational Analysisof Adenosine Deaminase 5467 that are predicted to interact with the substrate,H238A and in all mutant preparations except C262A, which exhibited E217A, either at position 1 of the heterocyclic ring (lane 3) high levels of adenosine deaminaseactivity. Therefore, we re- or with theincoming hydroxyl residue(lane 2), were produced examined the uninfected larvae and larvaeinfected with bac- in quantities equivalent to thewild-type recombinant protein ulovirus (that did not contain any adenosine deaminasecod- (lune 1). ing sequences)for a deaminase-likeactivity. Bothsets of Proteins containing mutations in the 2 residues that are larvae were subjected to the purificationprocedure. Indeed, a thought tobe coordinated toZn2+ (H17A andH214A, H214N deaminatingactivity (1-5 units/mg) was detectedin the and H214L),were present at lower levels (lanes 5-8). In fact, eluant from the adenosine-Sepharosecolumn fromuninfected immunodetection of antigen in the crude extracts containing larvalextract. However when infectedlarval extract was H214L and H17A (lanes 7 and 8) was difficult, indicating subjected to the purificationprotocol, the deaminating activ- that these proteins accumulated to much lower levels than ity dropped to 0.1-0.5 units per mg. Undoubtedly, viral repli- wild-type adenosine deaminase in infectedlarvae. Therefore, cation repressed host protein synthesis(O’Reilly et al., 1992). expression of these two mutants in infected Sf9 and T. ni Thisremaining deaminase-like activity was insignificant (High-5) insect cell cultures was compared. Accumulation of when compared towild-type adenosine deaminase(400 units/ both mutant proteins in thesecell lines wasequally improved mg), but might have interfered with detection of authentic (Fig. 2, lunes 12 and 13), though it was still muchlower than activity in some of the mutantenzymes. In order to eliminate for wild-type or any of the other mutant constructs (data not the insect deaminating activityfrom the mutant protein prep- shown). Thereason for the low accumulation of these mutant arations, an alternateprocedure was introduced for purifica- proteins has notbeen resolved. tion of all mutant recombinant proteins. The recombinant adenosine deaminase produced in this Monoclonal antibody (NlD1) directed against human aden- system was not totally soluble. The pellet fractions obtained osine deaminase (Philips et al., 1987) was used to construct from the initial clarificationof the crude extractsfrom wild- an immunoaffinity column. This antibody, when tested in a type and mutant proteins contained large quantities of im- Western blot against deaminating activitypurified from un- munoreactive antigen (data not shown). Attempts to solubi- infectedlarvae, showed nodetectable antigen (data not lize recombinant adenosine deaminase from the pellets (usingshown). In addition, no adenosine deaminase antigen or ac- nonionic detergents, urea, or high salt) yielded only an addi- tivity wasrecovered from uninfected larval extracts when tional 10-15% of theantigen (data not shown). The two subjected to a purification procedure using theimmunoaffin- mutants (H214L andH17A) that were present in the extracts ity column. As a further precaution against cross-contami- in low quantities exhibited correspondingly smaller amounts nation of the mutant proteins, individual monoclonal anti- of antigen in the pellet fractions,suggesting that solubility of body affinity columns were constructed for each mutant and these proteins was about the same as that of the wild-type used only in the purificationof that protein. adenosine deaminase. The results of the purification of the seven mutants are Purification of Wild-type and Mutant Adenosine Deami- shown in Fig. 3. Each of these proteins has been purified 10 nase-The wild-type adenosine deaminase was purified using to 25 times. The proteins, detected by silver staining, were a two-column procedure (including an adenosine-Sepharose produced in excellent yield and purity, except for two con- column)that we have previouslydescribed (Medin et al., structs. H214L and H17A consistently gave low yields when 1990). In the present studies the procedure was scaled to purified from either infected larvae or cells. These two mu- about 30 g of larvae from which 25-30 mg of recombinant tants have been purified 15 times from different cell prepa- protein was purified, corresponding to a yield of 25%. The rations with similar results. The two preparations shown in purified enzyme is shown in Fig. 3 (lane 1 ). This purification the figure were fromrecombinant virus-infected cells and procedurewas nextapplied to mutants C262A, E217A, exhibited significant degradationof adenosine deaminase, as H214A, H214L, and H214A. The retardationof wild-type and confirmed by Western blotting (data notshown). These data all mutant proteins by the adenosine-Sepharose matrix was suggested that the very low antigen yields for these two similar. Thus binding of substrate was not dramatically al- mutants may reflect an altered proteinstability. tered in anyof the mutant proteins. Thermolabilities of Wild-type and Mutant Proteins-We Unexpectedly, a low level of enzyme activity was detected explored thermolabilities of the proteins by comparing the patterns of productsobtained from thermolysin digestion 12345678 generated over a range of temperatures of all seven mutant and wild-type adenosine deaminases. The procedure was de- signed to detectdifferences in thermolabilityof wild-type and mutationally altered proteins. The rationalefor this approach is based on differential sensitivities of denatured and native proteins and differential thermolabilities of wild-type and mutationally altered proteins. Each of the purified proteins was subjected to digestion by thermolysin at varying temper- atures. The results of this analysis for two of the mutants is shown in Fig. 4. Across the entire temperature spectrum, the wild-type protein and the two mutants, H238A and C262A, exhibited similar digestion patterns. Other mutants, H214A, FIG. 3. SDS-polyacrylamide gel electrophoresis analysis of H214N, and E217A (not shown in this figure) showed iden- purified recombinant wild-type and mutant proteins. Each of tical digestion patterns. The two low yield mutants, H17A the denaturedproteins (150 ng) was applied to a 12% polyacrylamide and H214L, were also subjected to thermolysin digestion. gel containing SDS as described under “Experimental Procedures.” However, the limited quantities of these proteins made the After electrophoretic separation of the proteins, the gel was stained with silver (“Experimental Procedures”). Lane 1, wild-type; lane 2, analysis difficult to interpret. The overall digestion patterns H238A; [am 3, E217A; lune 4, C262A; lane 5, H214N; lane 6, H214A; and temperature dependence appearedsimilar, but with small lane 7, H214L; lane 8,H17A. amounts of these two proteins available, subtle changes inthe 5468 MutationalAnalysis of DeaminaseAdenosine wt enzyme C262A H238A TABLEI11 w * * Specific activity of proteins in different buffer system.. 45 5545 60 66 45 55 60 66 5545 60 66 " Protein Buffer VlIl*. units/mg Wild type Phosphate 400 Wild type Tris" 400 H238A Phosphate 0.12 H238A Tris" 0.09 E217A Phosphate 0.04 E217A Tris" 0.02 H214N Phosphate 0.03 H214N Tris" 0.15' H214A Phosphate 0.05 FIG. 4. SDS-polyacrylamide gel electrophoresis analysis of H214A Tris" 0.35' products of thermolysin digestion of wild-type and mutant a The addition of Zn2' during the purification did not alter activity. adenosine deaminases. Purified adenosine deaminases were di- 'Average enzyme activity obtained from five different purified gested with thermolysin as described under "Experimental Proce- adenosine deaminase preparations. dures." The reaction products were subjected to SDS-polyacrylamide gel electrophoresis, and the gels were stained withCoomassie Brilliant Blue. The temperatures of the digestion reactions are shown at the was C262A. This residue is highly conserved from humans to top of the lanes (45-66 "C).The products obtained from the analysis bacteria in all deaminases (Changet al., 1991), and it lies in of three recombinant proteins are shown: wild-type, H238A, and the crystal structure in the entrance toenzyme the active site C262A. Other mutants gave identical staining patterns. (Wilson et al., 1991). When this protein was purified using the standardtwo-column procedure, specific activities in ho- TABLEI1 mogeneous preparations rangedfrom 120 to 160 units/mg. In Kinetic parameters of wild-type adenosine deaminase contrast, when this protein was purified by using the mono- and its mutant derivatives clonal antibody column procedure, specific activities of the Relative Protein K"l Vrnex" kat kc,t/Km vmmx purified preparations ranged from 24 to 137 units/mg. The most common activities in these preparations 15in separately IrM unitslmg s-' s-' pw' % wild type purified samples of C262A was 50-90 units/mg. The lower Wild type 27.5 k 3.6 447 f 15 320 11.6 100 activity of the proteinspurified by an immunoaffinityproce- C262A 30.0 f 6.2 73.0 f 33' 51.0 1.7 16 H17A 30.0 k 1.4 13.6 f 1.6' 9.8 0.32 3.0 dure probably reflected the conditions required (4-6 M urea) H17A 45.0 f 4 0.53 f 0.01' 0.38 0.01 0.12 to elute the pure proteinfrom the column. Alteration of this H214A 21.0 f 3 0.05 -C 0.001 0.033 0.002 0.01 conserved residue may impede refoldingof the protein during H214L 30.0 k 4.5 0.08 f 0.003 0.054 0.002 0.02 the procedure used for removal of urea and result in a lower H214N 43.0 f 10 0.03 f 0.003 0.054 0.0006 0.02 enzyme activity than thatobserved in the nativeenzyme. In E217A 18.5 k 3.4 0.04 k 0.002 0.03 0.002 0.01 contrast, wild-type adenosinedeaminase purifiedby both H238A 8.0 f 0.5 0.12 f 0.001 0.1 0.01 0.03 methods exhibits identical specific activities (Philips et al. "The mutant enzyme preparations used for kinetic studies were 1987).' None of the other mutant enzymes exhibited differ- purified by using an immunoaffinity column. When these proteins were purified by conventional chromatography the following values ential specific activities as a result of conventional purifica- were estimated after subtracting the endogenous adenosine deami- tion or the immunoaffinitycolumn procedure (see Footnotea nase activity: C262A (120-160 units/mg); H241N (0.05-0.15 units/ in Table 11). mg); H214A(0.02-0.025 units/mg); H214L (0.05-0.15 units/mg); The second mutant adenosine deaminase that appeared to E217A (0.02-0.05 units/mg). exhibit activitywas H17A. Enzymatic activityvaried dramat- 'This value prepresents the average VmaXobtained from 15 individ- ically among the 15 preparations of this mutant purified on ually purified enzyme preparations. When H17A was purified, several preparations exhibited activities the immunoaffinitycolumn. The lowest activity observed was in the range of 8-28 units/mg. A second group of preparations of 0.2 units/mg and the highest activity was 28 units/mg. This H17A exhibited activities ranging from 0.1 to 0.7 units/mg. The high residue and H214 are postulated beto involved in coordinating and low activity preparations were pooled separately and analyzed. the zinc atom. Therefore we examined the possibility that variations in Zn2+content throughout thepurification proce- digestion patterns would not have been observed. dures were related to the retention of adenosine deaminase The thermolysin digestion experiments were also carried activity in these altered proteins. out by incubating the samples at 66 "C for varying times, Our standard purification scheme utilized throughout po- from 10 to 60 min (data not shown).No dramatic differences tassium phosphate buffer, pH 6.8. Since phosphate is known in digestion patterns were observed during this timescale. to chelate Zn2+, we substituted for this buffer, Tris (which Kinetic Parameters of Wild-type and Mutant Deriuatiues- does not chelate Zn2+ (Sellin and Manneevvik, 1984)) and Kinetic propertiesof the purified wild-type and mutant aden- purified wild-type, H214A, H214N, H238A, and E217A in the osine deaminases were determined using adenosine as sub- presence or absence of exogenously added Zn2+. The results strate (Table 11). The apparent K,,, values for the mutants are shown in Table 111. In the presence ofZn", either by were similar to the wild-type enzyme. Thus, alterations in exogenous addition throughout thepurification or by use of a neither the side chains of the two-Zn2+coordinating residues nonchelating buffer, enzyme activity was increased by 5- to (His17 and His2I4) nor in the N1 interacting residue (Glu2") 6-fold forthe H214 mutants. Weconfirmed that this apparent interfered with productive substratebinding. The onlysignif- stimulation of adenosine deaminase activitywas not a general icant variation in K,,, @-fold) was observed for the H238A phenomenon since thespecific activities of wild-type, H238A, variant. The VmaXand kc,, values were dramatically altered in the ' D. Bhaumik, J. Medin, K. Gathy, and M. S. Coleman, published mutant proteins (Table 11). The most active of the proteins observations. MutationalAnalysis of AdenosineDeaminase 5469 and E217A enzymes were identical to those obtainedby using murine enzyme (by 11 residues at the carboxyl end of the phosphate buffer in the purification procedures (Table111). molecule) but shares greater than 80% identical side chains Zn2+ Contentof Wild-type and Mutant Adenosine Deami- (Wiginton et al., 1984; Ingolia et al., 1985). Of the 59 amino nases-The purified recombinant proteins were analyzed for acid differences betweenthe two proteins, 26 are conservative Zn2+ contentusing two methods:graphite furnace atomic substitutions. All of the residues implicated in binding sub- absorption spectrometry and flame atomic absorption spec- strate or catalyzing the hydrolytic reaction are identical in trometry (Table IV). The graphite furnace method is more the human andmouse proteins (Changet al., 1991). Therefore, sensitive, but since small quantities of protein are required we anticipated that residues identified in the mouse enzyme (nanomolar range), the signal-to-noise ratio is smaller than would serve identical functions in both enzymes. The major with the atomic absorption method. Previous analyses, in- goal of this study was to test selected amino acid residues in cluding the x-ray crystallography studies, revealed a single the active site at a pH of optimal enzyme activity (6.8). Of Zn2+ residue at the active site. In these experiments, wild- particular interest were a Cys residue close to the active site, type enzyme, E217A, H238A, and C262A were demonstrated 3 His, and a Glu residue in the active site. to containa singleZn2+ when the atomic absorption spectrom-Chemical studies with adenosine deaminase froma number etry methodwas used. The graphite furnace method indicatedof sources have previously implicated a sulfhydryl group asa 2 mol of Zn2+ per mol of protein for these enzyme prepara- catalyticallyimportant residue, perhapsas the sourcefor tions. We suspect that this higher ratio is simply due to the protonation of N1 of adenosine (Orsi et al., 1972; Wolfenden inherent problem in accurately measuring Zn2+ from small et al., 1967; Weiss et al., 1987). Furthermore, comparison of quantities of protein. Both typesof H17A preparations (high adenosine deaminase proteinsequences from Escherichia coli and low activity) containedZn2+, though the amountsof these to humans has demonstrated the presenceof a conserved Cys proteins were so limited that accurate quantitation of molar (Chang et al., 1991) which is also positionally conserved in all ratios was not feasible. The His214 mutants also contained known AMP deaminases. Thus, it seemed highly likely that Zn2+ (with the exceptionof H214L) even though these prep- this residue would be essential for enzyme catalysis or con- arations exhibited noenzyme activity in the absenceof added formation. Zn2+,but were slightly activated (H214A and H214N) when In themouse and humanenzymes this conserved residue is Zn2+was added throughout the enzyme purification. at position 262. The interpretation of the crystal structure places this Cys out of the rangeof the active site andsuggests DISCUSSION that access bythis sulfhydryl groupto the active siteis blocked Adenosine deaminase is an enzyme of enormous interest. by 3 residues involved in substrate bindingor catalysis (Wil- It plays importantroles in nutrition and reproduction, and it son et al., 1991). When this amino acid was converted to Ala serves a protective role against accumulation of deoxyadeno- as described in this study, there was only a modest effect on sine in cells of the immune system. The catalytic power of enzyme activity or kinetic parameters. The mutant protein this enzyme is among the highestknown and extensive chem- was apparently correctly folded in uiuo, since enzyme which ical studies with inhibitors have generated testable hypothesesretained 30-40% of wild-type activity was obtained following about the active site.As expected for a protein with a central conventional purification. When the protein was exposed to role in metabolism, the primary aminoacid sequence of aden- a denaturant during purification on an immunoaffinity col- osine deaminase is highly conserved across species (Chang et umn,about 20% of native enzyme activity was normally al., 1991). The crystal structure of murine adenosine deami- recovered under urea removal conditions in which 100% of nase, complexed to the inhibitor purine riboside at pH 4.2, wild-type adenosinedeaminase activity and other mutant has recently been elucidated (Wilson et al., 1991), and active enzyme activities arerecovered. The thermolability studiesof site residues have been identified. This information, coupled the C262A mutant showed that there was no evidence of a with access to thecoding sequences of adenosine deaminases drastically altered proteolyticdigestion pattern upon heating. from other species, has made the testing of amino acid side Thus, while this CysZ6'residue may play an important, but as chains in the active sitefeasible. yet poorly understood, structuralrole in theenzyme structure, Human adenosine deaminase is slightlylonger than the it is clearly not essential for catalysis. The most surprisingrevelation about adenosine deaminase TABLEIV which occurred with the solutionof the crystal structurewas Zinc content of human adenosine deaminase, wild type, and mutants the role of Zn2+in theenzyme active site (Wilsonet al.,1991). Zn2+ Protein This featureof the enzyme was unanticipated since standard GAAS" FAAS* studies with metalchelators failed todetect any enzyme rnollrnol of protein sensitivityto such agents (Zielke andSuelter, 1971). The Wild type 1.9 k 0.1 1.0 k 0.1 central role of Zn2+ in the functioningof adenosine deaminase C262A 2.0 k 0.2 1.2 k 0.1 helps to explain in part earlier observationsof the effects on * H17A' NDd the immune system produced by Zn2+ deficiency in humans H214A 1.4 f 0.2 1.6 k 0.06 andanimals. Zinc deficiency causesatrophy of lymphoid H214L co.1 ND H214N 1.3 k 0.2 ND tissues and abnormalities in T and B cell function (Cossack E217A 1.4 k 0.3 1.2 f 0.15 and Prasad, 1991) reminiscent of the severe combined im- H238A 1.9 k 0.2 1.2 f 0.06 munodeficiency observed in children who lack adenosine de- GAAS, graphite furnace atomic absorption spectrometry. Stand- aminase (Coleman, et al., 1978). ards containing zinc ranged from 15.4 to 92.4 nM. Measured values The human recombinant adenosine deaminase also con- ranged from 18.5 to 65.2 nM. tained Zn2+as shown by analytical techniques.2 residueswere FAAS, flame atomic absorption spectrometry. Standards contain- altered that have been implicated in Zn2+binding, Hid7 and ing zinc ranged from 0.385 to 6.16 p~.Measured values ranged from His214.The H17A was an unstable recombinant protein that 1.16 to 3.08 pM. H17A, Zn2+ was detected in this preparation (designated *), but exhibited an unaltered K, for substrate binding. The His214 the small quantityof protein did not permit accurate quantitation. mutants varied in stability. Substitution of His with Ala or ND, not determined. Asn produced mutants that could be obtained in reasonable 5470 MutationalAnalysis of DeaminaseAdenosine yield. The Leu substitution, however, produced a protein that energy change in this reaction, the observed K,,, would be was asunstable as the H17A. The concentration of Zn2+ expected to change by at least 10-fold with aconcurrent available at all stages of the expression and purification of decrease in kc,, compared to the wild-type enzyme. Studies the recombinant proteins appeared to be important for recov- are currently underway to analyze the contribution of Hisz3' ery of enzyme activity. Use of a non-Zn2+-chelating buffer to binding the 60Hgroup in the transition state. and addition of Zn2+throughout the purification usually re- The results presented herein from site-directed mutagene- sulted in recovery of a small percentage of wild-type enzyme sis, when combined with previous information derived from activity (0.05-0.12%). However, the addition of Zn2+during NMR and x-ray crystallography, illustrate that in adenosine prolonged dialysis of purified enzyme preparations exhibiting deaminase anSH group (CysZ6') is important for protein less than 0.02% activity was not effective in restoring any folding, but not enzyme catalysis; that GluZl7and Hisz3' are additional activity. These mutants were not devoid of Zn2+as both essential for catalysis; that correct orientation of Zn2+is shown by the atomic absorption analyses. If Zn2+ was dis- crucial for activity. The availability of specifically targeted lodged from the active site during purification by the use of active site mutants of adenosine deaminase will now permit phosphate buffer subsequent recovery of activity was never the determination of detailed interactions of each of these observed. These studies support the notion that Hid7 and amino acid residues with the substrate and with potent inhib- His214coordinate ZnZ+. itors of this important enzyme. GluZi7is proposed, on the basis of its proximity to the N1 of purine riboside in the crystal structure, to be involved in Acknowledgments-We gratefully acknowledge scientific discus- protonation of N1 of the adenosine purine ring (Wilson et al., sions with Drs. Richard Wolfenden, Gary Pielak, Mary Barkley, and 1991). The pK. of the side chain carboxylate of glutamate is Judith Lesnaw and with graduate students Dean CarIow and Greg Dzingeleski in the Wolfenden laboratory. Dana Mossman provided 4.2. At the pH of maximal enzyme activity, 7, this residue valuable technical assistance. would be negatively charged, unless its pK, has been altered by the environment of the active site. Removal of the carbox- REFERENCES Auld, D. S. (1988)Methods Enzymol. 158, 13-14 ylate residue at this position caused a dramatic reduction in Bradford, M. M. (1976)Anal. Biochem. 72, 248-254 catalytic activity (10,000-fold) but no significant change in Burnette, W. N. (1981)Anal. Biochem. 112,195-203 Chang, Z., Nygaard, P., Chinualt, A. C., and Kellems, R. E. (1991)Biochemistry the K,,, for adenosine. Thus, the putative hydrogen bond of 30,2273-2280 the potentially dissociable proton of G1u217to the lone electron Coleman, M. S. (1983)Biosciences 33,707-712 Coleman, M. S., and Hutton, J. J. (1975)Biochem. Med. 13, 46-55 pair of N1, as proposed by Wilson et al. (1991), is probably Coleman, M.S., Donofrio, J., Hutton, J. J., Hahn, L. Daoud, A,, Lampkin, B., not important in substrate binding but is important for catal- and Dyminiski, J. (1978)J. Biol. Chem. 253, 1619Ll626 Cossak, Z., and Prasad, A. S. (1991)Int. J. Vit. Nut. Res, 61,51-56 ysis as indicated by the large decrease in kcht. The function of Donofrio, J. Coleman, M. S., and Hutton, J. J. (1978)J. Clin. Inuest. 62, 884- G~u~~~can be compensated by water in the absence of the 887 Evans, B. E., and Wolfenden, R. V. (1973)Biochemistry 12, 392-398 carboxylate side chain, since activity is not completely elimi- Frick, L., Mecneda, J. P., and Wolfenden, R. (1987) Biorgan Chem. 15,100- nated in the altered protein. 108 Giblett, E. R., Anderson, J. E., Cohen, F., Pollara, B., and Meuwissen, H. J. Hisz3' and Aspzg5 have been proposed as potential proton (1972)Lancet 2,1067-1069 donors for the ammonia leaving group at C6 of adenosine, Ingolia D. E., Yeung C.-Y., Orengo, I. F., Harrison, M. L., Frayne, E. G., Ruddlph, F. B., and'Kellems, R. E. (1985)J. Biol.,Chem.,260,13261-13267 based on distance calculations (Kati and Wolfenden, 1989; Jones, W., Kurz, L. C., and Wolfenden, R. (1989)Btochemrstry 28, 1242-1247 Wilson et al., 1991). From analysis of residue distances in the Kati, W. M., and Wolfenden, R. (1989)Biochemistry 28,7919-7927 Knudsen, T. B., Green, J. D., Airhart, M. J., Higley, H. R., Chinskey, J. M., crystal structure (crystals were grown at pH 4.2), AspZg5is and Kellems, R. E. (1988)Biol. Reprod. 39,937-951 proposed to be protonated during the course of the reaction Knudsen, T. B., Gray, M. K., Church, J. K., Blackburn, M. R., Airhart, M. J., Kellems, R. E., and Shalko, R. G. (1989)Teratology 40,615-625 via the abstraction of a proton from water. In an alternate Kunkel, T. A. Roberts, J. D., and Zakour, R. A. (1987)Methods Enzymol. 154, proposal, Wolfenden andKati (1991) have suggested that 367-382 Kurz, L. C., and Frieden, C. (1983)Biochemistry 22,382-389 Hisz3' abstracts a proton from an attacking water molecule Kurz, L. C., and Frieden, C. (1987)Biochemistry 26,8450-8457 and thus is the source of the proton for the ammonia. Only Laemmli, U. K. (1970)Nature 227,680-685 Lee, P. C. (1973)Deu. Biol. 31,227-233 one of these residues was altered in this study. When Hisz3' Medin, J. A., Hunt, L., Gathy, K., Evans, R. K., and Coleman, M. S. (1990) was changed to Alaz3', a protein was produced that exhibited Proc Natl. Acad. Sci. U. S. A. 87,2760-2764 Medin, J. A., Gathy, K., and Coleman, M. S. (1993) Methods in Molecuhr drastically reduced enzyme activity (3000-fold decline) and a Biology (Richardson, C., ed) Humana, New York, in press decreased K,,, for the adenosine substrate. These data indi- OReilly, D. R., Miller, L. K., and Lucknow, V. A. (1992)Baculouirw Expression Vectors, A Laboratory Manual, pp. 24-25, W. H. Freeman and Company, cated that Hisz3*did potentiate binding of adenosine. When New York the side chain was more hydrophobic, as in Alaz3', the sub- Orsi, B. A., McFerran, N., Hill, A,, and Bingham, A. (1972)Biochemistry 11, 3386-3392 strate was more tightly bound than in the wild-type enzyme. Philips, A. V. Robbins, D. J., Coleman, M. S., and Barkley, M. D. (1987) Likewise, the decrease in kcat from 320 to 0.1 s-' indicated Biochemistry 26, 2893-2903 Polesky, A. H., Steitz, T. A., Grindley, N. D. F., and Joyce, C. M. (1990)J. that Hisz3' may indeed be the source of the proton in the Biol. Chem. 266,14579-14591 reaction, a fact that canbe established only when we have a Schrader, W. P.,and Stacy, A. P. (1977)J. Biol. Chem. 252,6409-6415 Sellin, S., and Mannervik, B. (1984)J. Biol. Chem. 259,11426-11429 mutation of the Aspzg5since activity was not completely Weiss, P. M., Cook, P. F., Hemes, J. D., and Cleland, W. W. (1987)Biochem- abolished in this H238A mutant. istry 26,977-983 Wiginton, D. A,, Adrian, G. S., and Hutton, J. J. (1984)Nucleic Acids Res. 12, The decrease in both K,,, and kc,, observed in H238A at first 2439-2446 suggested that this residue may interact with the substrate Wilson, D. K., Rudolf, F. B., and Quicho, F. A. (1991)Science 262,1278-1284 Witte, D. P., Wiginton, D. A. Hutton, J. J., and Aronow, B. J. (1991)J. Cell. transition state more strongly than with the ground state. Biol. 115, 179-190 Wolfenden, R., and Kati, W. M. (1991)Acc. Chem. Res. 24,209-215 However, inspection of k-, and free energy change (6G kl) Wolfenden, R., Sharpless, T. K., and Allan, R. (1967)J. Biol. Chem. 242,977- values for all the mutant enzymes revealed a range from 13.7 983 kcal for H17A to 14.7 kcal for H238A. These relatively small Wolfenden, R., Kaufman, J., and Macon, J. B. (1969)Biochemistry 8,2412- 2415 differences among all of the mutantsdo not support agreater Wray, W., Boulikas, T., Wray, V. P., and Hancock, R. (1981)Anal. Biochem. 72,248-254 interaction for H238A in the transition state than any of the Zielke, C. L., and Suelter, C. H. (1971)The Enzymes, 3rd Ed., Vol. IV, pp. 47- other residues that were altered. To achieve a significant free 78, Academlc Press, New York