Identification of the ATP Binding Domain of Recombinant Human 40-kDa 2′,5′-Oligoadenylate Synthetase by Photoaffinity Labeling with 8-Azido-[α-32P]ATP

Next Section Abstract Three isoforms of the interferon-inducible 2′,5′-oligoadenylate (2-5A) synthetase that require double-stranded RNA have been isolated and cloned. However, identification of the amino acid(s) of 2-5A synthetase directly interacting with ATP is crucial to the elucidation of the mechanism of the enzymatic conversion of ATP to 2′,5′-oligoadenylates by 2-5A synthetase. Recombinant human 40-kDa 2-5A synthetase has been expressed as a glutathione S-transferase fusion protein in E. coli and purified to near homogeneity in milligram quantities. The azido photoprobe, 8-azido-[α-32P]ATP, has been used to identify the ATP binding domain of the recombinant human 40-kDa 2-5A synthetase. Specific covalent photoincorporation of 8-azido-[α-32P]ATP into the 2-5A synthetase, tryptic digestion of the covalently 32P-labeled enzyme, isolation of the photolabeled phosphopeptide by metal (Al3+) chelate chromatography, and high pressure liquid chromatography identified a 32P-pentapeptide, which has been assigned to the ATP binding domain of 2-5A synthetase. The radioactive pentapeptide has the sequence D196FLKQ200 in which the photoprobe, 8-azido-[α-32P]ATP, chemically modified the amino acid lysine 199. The catalytic importance of Lys199 was further established by mutation of lysine 199 to arginine 199 and histidine 199 using site-directed mutagenesis. The K199R and K199H recombinant human 40-kDa 2-5A synthetase mutants bind 8-azido-ATP and the allosteric activator, poly(I)·poly(C) but are enzymatically inactive. These photoaffinity labeling and mutation data strongly suggest that lysine 199 is essential for the formation of a productive 2-5A synthetase-ATP-double-stranded RNA complex for the enzymatic conversion of ATP to 2-5A.

2-5A synthetase is a nonprocessive enzyme, as evidenced by the formation of a series of 2Ј,5Ј-oligoadenylates of the formula, 2Ј,5Ј-A(pA) n in which n Ն 1 (10,11). Two nucleotide binding sites on 2-5A synthetase have been suggested as requirements for formation of 2-5A, i.e. an acceptor site and a donor site. The donor site binds ATP, 2-5A, NAD ϩ , or other nucleotides with AMP moieties (10 -12). The acceptor site binds ATP, thereby providing the substrate for 2Ј-adenylation (10,11). Kinetic studies suggest a high affinity (K d ϭ 9 M) and a lower affinity (K d ϭ 1000 M) binding site (12). Therefore, identification of one or both of these nucleotide binding sites would enhance our understanding of the nonprocessive mechanism by which 2-5A synthetase, as a nucleotidyltransferase, converts ATP to 2-5A. To elucidate the catalytic mechanism of 2-5A synthetase, it is essential to understand the interactions between 2-5A synthetase and ATP and to identify the amino acid(s) at the ATP binding domain. The study described here has focused on the identification of ATP binding domain(s) of 2-5A synthetase by photoaffinity labeling technology using 8-azido-ATP as a substrate analog of ATP to bind to the ATP binding domain of the 40-kDa 2-5A synthetase. The covalent photoinsertion of 8-azido-[␣-32 P]ATP into 2-5A synthetase is highly specific and saturable, which provides the basis for the use of this photoprobe to identify the amino acids in the ATP binding domain of 2-5A synthetase, which are covalently modified by the azido * This work was supported by United States Public Health Service Grants R01-AI-34765 (to R. J. S.) and P30-CA12227 and a Dissertation Fellowship from the Graduate School, Temple University (to N. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
In a previous study, we reported that highly purified 100-kDa 2-5A synthetase from rabbit reticulocytes was covalently photolabeled with 8-azido-[␣-32 P]ATP in the presence and absence of dsRNA (21). However, an insufficient amount of highly purified enzyme prevented identification of the ATP binding domain of 2-5A synthetase. In this report, we describe the expression and purification of milligram quantities of recombinant human 40-kDa 2-5A synthetase as a GST⅐2-5A synthetase fusion protein. Following cleavage of the GST⅐2-5A synthetase with human thrombin, highly purified recombinant human 40-kDa 2-5A synthetase was covalently photolabeled by 8-azido-[␣-32 P]ATP. The photolabeled peptide was digested with trypsin and purified by Al 3ϩ affinity chromatography and reverse phase HPLC. Microsequencing of a 32 P-peptide identified the single radioactive chemically modified amino acid, lysine 199. The catalytic importance of lysine 199 was established by construction of two mutants in which lysine 199 of the 40-kDa 2-5A synthetase was replaced with arginine or histidine using site-directed mutagenesis. The K199R and K199H mutant 2-5A synthetases bind ATP and the allosteric activator, poly(I)⅐poly(C) but were catalytically inactive. These findings are presented as a contribution to the understanding of the cellular functions of 2-5A synthetase. Taq  Expression of Recombinant Human 40-kDa 2-5A Synthetase--Plasmid pTLE-1 containing human-40 kDa 2-5A synthetase cDNA was provided by Dr. Judith Chebath (Weizmann Institute of Science). The DNA fragment containing the coding region of the human 40-kDa 2-5A synthetase cDNA was isolated from pTLE-1 by restriction digestion with XbaI and EcoRI and subcloned into M13 mp18 digested with the same enzymes. The recombinant phage was used to prepare uracilcontaining single-stranded DNA, which was used as the template for in vitro mutagenesis using T7 DNA polymerase (U.S. Biochemical Corp.) (22). The primer, 5Ј-GATGAGGGTAATAACATATGATGGATCTCAG-3Ј, was synthesized to generate a NdeI site at the first ATG codon of the 40-kDa 2-5A synthetase coding region. After identification of the recombinant M13 phage with the NdeI site mutation, the human 40-kDa 2-5A synthetase cDNA insert was released by restriction digestion with NdeI and EcoRI, filled in with Klenow fragment, and subcloned into the expression vector pGEX-2T linearized with SmaI to generate pNK14. The correct reading frame was confirmed by restriction digestion, and the coding sequence of 40-kDa 2-5A synthetase was confirmed by dideoxynucleotide sequencing using the Sequenase 2.0 kit (U.S. Biochemical Corp.).

Materials-Restriction enzymes and
Expression and Purification of Recombinant Human 40-kDa 2-5A Synthetase-pNK14, which fuses the 40-kDa 2-5A synthetase cDNA to the 3Ј terminus of glutathione S-transferase cDNA, was transformed into Escherichia coli HMS174 cells (Novagen). The E. coli was streaked on an LB 15% agar plate containing 50 g/ml ampicillin. A single colony was inoculated into 50 ml of LB medium containing 50 g/ml ampicillin followed by incubation at 37°C with shaking at 200 rpm/min, overnight (New Brunswick shaker). The culture was inoculated into 500 ml of LB medium with ampicillin (50 g/ml). When the absorbance of the culture at 600 nm was 0.8, isopropyl-1-thio-␤-D-galactopyranoside was added to 0.3 mM. The culture was incubated with shaking for an additional 3 h before harvesting (23, 24). Unless otherwise specified, all protein purification procedures were performed at 4°C. The E. coli cells were collected by centrifugation (6000 ϫ g, 5 min) and suspended in 10 volumes of PBS containing 0.1 mM phenylmethylsulfonyl fluoride and 10 mM ␤-mercaptoethanol. The cells were broken in a French pressure cell (500 psi, twice). The cell lysate was centrifuged (25,000 ϫ g, 1 h). After centrifugation, the supernatant (40 mg protein/ml, 100 ml) was loaded onto a 2-ml GSH-Sepharose column, flow rate at 0.2 ml/min. The column was washed with 20 ml of PBS. The GSH-Sepharose resin was resuspended in 2 ml of PBS containing human thrombin (1 unit/ml). The GSH-Sepharose suspension was incubated with constant mixing at 25°C, 4 h. The recombinant human 40-kDa 2-5A synthetase was released into PBS buffer by centrifugation (12,000 ϫ g, 5 min). The supernatant was removed and stored at Ϫ70°C. To remove glutathione S-transferase (GST) and glutathione, the recombinant human 40-kDa 2-5A synthetase was bound to S-Sepharose (1-ml column with a capacity of 10 mg of 2-5A synthetase), which was equilibrated in buffer S (20 mM Tris-HCl, pH 7.4, 20 mM Mg(OAc) 2 containing 50 mM KCl). After washing the column with 5 ml of buffer S containing 50 mM KCl, the recombinant human 40-kDa 2-5A synthetase was displaced with 5 ml of buffer S containing 200 mM KCl. The eluant was dialyzed against 200 ml of buffer S twice for 1 h each at 4°C. The protein solution was stored at Ϫ70°C (0.5 mg of protein/ml, 6 ml). The recombinant human 40-kDa 2-5A synthetase, as analyzed by 10% SDS-polyacrylamide gel electrophoresis (25) and Coomassie Blue staining was nearly homogeneously pure.
Competition Experiments-Competition photoaffinity labeling assays were performed with 0.15 mM 8-azido-[␣-32 P]ATP (specific activity, 0.5 Ci/nmol) as described (12). The concentration of ATP in the reaction mixtures was varied from 0 to 15 mM. Photolabeling and quantification were as described above.
Trypsin Digestion of the Photolabeled Recombinant Human 40-kDa 2-5A Synthetase-The precipitated 32 P-photolabeled protein was dissolved in 80 l of 8 M urea. To the protein solution was added 1100 l of 1% ammonium acetate, pH 7.9, and 100 l of TPCK-treated trypsin solution (1 mg/ml in 1% ammonium acetate, pH 7.9) (20). The pH of the digestion mixture was adjusted to 7.9 with 1 M ammonium hydroxide, and the mixture was incubated at 37°C for 12 h. An additional 100 l of trypsin solution was added, and the mixture was incubated for 150 min at 37°C.
Purification of Photolabeled Peptide by Immobilized Al 3ϩ Affinity Chromatography-Iminodiacetic acid-Sepharose 6B resin (1 ml) was washed and equilibrated with 20 ml of glass-distilled water (20). A column (0.75 ϫ 4 cm) was prepared and equilibrated with 10 ml of 50 mM AlCl 3 . The tryptic digestion mixture (1.5 ml) was mixed with 1.5 ml of buffer A (1% ammonium acetate, 1 M NaCl, pH 5.8) and 150 l of 10% acetic acid. The mixture was transferred onto the immobilized Al 3ϩ affinity column. Two-milliliter fractions were collected, and the column was washed with buffer A (1 ml/min) until radioactivity in the eluant decreased to background level (about 20 ml). One-milliliter fractions of the 32 P-photolabeled peptide were collected from the column with 10 mM potassium phosphate, pH 7.4, and detected by their absorbance at 214 nm. The 32 P was determined by Cerenkov radiation. Fractions 20 -25 containing radioactivity were pooled and lyophilized to 200 -300 l in a vacuum concentrator.
HPLC Analysis of the Photolabeled Tryptic Peptide from Al 3ϩ Affinity Chromatography by Reverse Phase HPLC-The photolabeled 32 P-peptides that were fractionated by Al 3ϩ affinity chromatography were further purified by HPLC using C18 reverse phase column chromatography (elution buffer A: 0.1% trifluoroacetic acid, pH 2; elution buffer B: 80% acetonitrile, 0.1% trifluoroacetic acid, pH 1.5. The peptides were displaced with the following gradient: 0 -10 min, 0% B; 10 -50 min, linear gradient to 40% B; 50 -80 min, linear gradient to 75% B; 80 -85 min, linear gradient to 90% B). One ml of buffer A was added to the lyophilized photolabeled 32 P-peptide solution and it was injected onto the column. One-ml fractions were collected (flow rate, 1 ml/min). Absorption at 214 nm was monitored, and the radioactivity of an aliquot of each fraction was measured by Cerenkov radiation. The radioactive peptide peak was sequenced by the PTH-derivative method (14).
Site-directed Mutagenesis of Human 40-kDa 2-5A Synthetase-Four oligonucleotides were designed and synthesized to mutate amino acid Lys 199 to Arg or His (primer 1, 5Ј-CATCG-AGGAGTGCACCGA-3Ј; primer 2, 5Ј-AGCTTGGTGGGGCGCTG(C/A)(T/C)GCAGGAAGTCTCT-CTG-3Ј; primer 3, 5Ј-CAGCGCCCCACCAAGCT-3Ј; primer 4, 5Ј-TCG-CTCCCAAGCATAGACC-3Ј). The first round of PCR was performed using plasmid pNK14 as the template with primers 1 and 2 and primers 3 and 4, respectively. The cycles were (i) 95°C for 5 min; (ii) 95°C for 1 min, 52°C for 45 s, 72°C for 45 s (30 cycles); and (iii) 72°C for 10 min. PCR products were purified by low melting agarose gel electrophoresis. The second round of PCR was performed with the two PCR products, primers 1 and 4 using the cycles described above. The second round PCR product was digested with restriction enzymes PstI and KpnI to generate the 120-bp fragment, which was ligated with a 5000-bp fragment of pNK14 digested with PstI and a 900-bp fragment of pNK14 digested with PstI and KpnI. The resulting clones were screened by dideoxynucleotide sequencing for the mutation desired. Individual constructs were transformed into E. coli HMS 174 cells for expression of mutant human 40-kDa 2-5A synthetases.
Expression and Purification of the Lys 199 Mutant Recombinant Human 40-kDa 2-5A Synthetase-Milligram quantities of the Lys 199 mutant 40-kDa 2-5A synthetases were expressed and purified according to the procedures described above for the wild type 2-5A synthetase.
Binding and 2-5A Synthetase Activity Assays with the K199R Mutant Recombinant Human 40-kDa 2-5A Synthetase-The binding of 8-azido-ATP was assayed by photoaffinity labeling using 8-azido-[␣-32 P]ATP as described above. Assays for the catalytic activity of the K199R mutant 40-kDa 2-5A synthetase were performed under the same conditions used for the wild type enzyme. The binding of poly(I)⅐poly(C) was assayed using poly(I)⅐poly(C)-agarose. Purified 2-5A synthetase or mutant 2-5A synthetase was incubated with poly(I)⅐poly(C)-agarose for 30 min at 4°C. After washing of poly(I)•poly(C)-agarose with 0.5 ml of PBS three times, 40 l of protein loading buffer was added to the poly(I)•poly(C)-agarose to remove the bound 2-5A synthetase. The mixture was heated at 95°C for 5 min, and the dsRNA binding proteins bound to the poly(I)⅐poly(C)-agarose were fractionated by 10% SDS-PAGE. The K199H mutant 2-5A synthetase was purified by the same procedures.
Competition Experiments with the K199R Mutant-Competition photoaffinity labeling of the K199R mutant were performed with 0.15 mM 8-azido-[␣-32 P]ATP (specific activity, 0.5 Ci/nmol) and increasing concentrations of ATP as described above for the wild type 2-5A synthetase.
Peptide Sequencing-Amino acid sequencing was completed by the Amino Acid Sequencing Core Facility, Temple University School of Medicine and by the Protein Chemistry Laboratory, University of Pennsylvania School of Medicine.

Expression and Purification of Recombinant Human 40-kDa
2-5A Synthetase-The GST fusion protein strategy increased the yield of recombinant human 40-kDa 2-5A synthetase significantly compared with that obtained by direct expression of 2-5A synthetase, to approximately 1% of the total E. coli cellular protein. About 1 mg of nearly homogeneously purified recombinant human 40-kDa 2-5A synthetase was purified from 5 liters of E. coli culture (Fig. 1A, lane 2). The specific activity of the purified recombinant human 2-5A synthetase (cleaved from GST by human thrombin) in crude extracts of E. coli was 0.04 nmol/g protein/h; the specific activity of the nearly homogeneously purified 2-5A synthetase had a specific activity of 28.0 nmol/g protein/h. The specific activity of the highly purified 2-5A synthetase from E. coli was about 2000-fold greater than that in crude extracts of IFN-treated HeLa cells. Recombinant human 40-kDa 2-5A synthetase is allosterically activated by poly(I)⅐poly(C) (Fig. 1B, lanes 2 and 4). In the absence of poly(I)⅐poly(C), there was little or no synthesis of 2-5A (Fig. 1B,  lanes 1 and 3). The 2-5A synthesized by the recombinant human 40-kDa 2-5A synthetase activates murine RNase L to generate specific cleavage products after incubation with L929 cell extracts for 1 h (data not shown).
Competition Experiments-Photoaffinity labeling of recombinant human 2-5A synthetase by 8-azido-[␣-32 P]ATP can be specifically inhibited by ATP (Fig. 5, A and B), indicating that photoaffinity labeling by 8-azido-ATP is saturable and highly specific for the ATP binding domain. Fifty percent saturation of photolabeling was achieved with 0.1 mM ATP in the absence of poly(I)⅐poly(C).
Photoaffinity Labeling of Recombinant Human 2-5A Synthetase, Identification of the Photolabeled Peptide, and Identification of the Modified Amino Acid-Based on the covalent photoaffinity labeling of 2-5A synthetase, recombinant human 40-kDa 2-5A synthetase (300 g of protein) was photolabeled with 0.1 mM 8-azido-[␣-32 P]ATP for 30 s and further saturated with 0.1 mM 8-azido-ATP. After perchloric acid precipitation and a methanol wash to remove unincorporated 8-azido-[␣-32 P]ATP, the protein was hydrolyzed with TPCK-treated trypsin. The photolabeled peptide was isolated and partially purified by immobilized Al 3ϩ chromatography, a metal ion chromatography procedure that has been successfully applied to the purification of phosphopeptides (20). Ninety percent of the 32 P-peptide(s) retained by the Al 3ϩ column was displaced by 10 mM phosphate buffer in fractions 20 -25 (Fig. 6A). The purified photolabeled peptide was rechromatographed by reverse phase HPLC. A single radioactive peptide peak was eluted in frac- tions 53 and 54 (Fig. 6B). The peak of 32 P and peptide eluted with fractions 5-10 has been reported to represent unincorporated 32 P (13). Sequence analysis identified the radioactive pentapeptide as 196 DFLKQ 200 . The amino acid residue corresponding to lysine 199 was not detected due to covalent crosslinking and is unambiguously identified as the site of covalent modification. The same pentapeptide sequence and identification of lysine 199 was obtained in two independent experiments with or without poly(I)⅐poly(C) ( Table I).
Schematic Diagram of the Site-directed Mutagenesis of Lys 199 of Recombinant Human 40-kDa 2-5A Synthetase to Arg or His-The arginine mutant recombinant human 40-kDa 2-5A synthetase was constructed as shown in Fig. 7. Two rounds of PCR of the pNK14 plasmid produced a 120-bp PstI-KpnI DNA fragment, which was ligated with a 900-bp KpnI-PstI fragment and a 5000-bp PstI fragment. The second PstI site is within the ampicillin resistance gene. Therefore, the triple ligation could only occur in the 5Ј to 3Ј direction. The lysine 199 to arginine 199 mutation was inserted into human 40-kDa 2-5A synthetase cDNA. The sequence of the lysine 199 to arginine 199 mutant was confirmed by DNA dideoxy sequencing (Fig. 8). The primer containing either C/A or T/C bases at the site of mutation allowed for the simultaneous generation of K199H and K199R mutations. Plasmid pNK16 containing the K199R mutant human 40-kDa 2-5A synthetase cDNA allows the expression of GST fusion protein. The purification of the K199R mutant 2-5A synthetase was performed as described under "Experimental Procedures." The yield of the K199R 2-5A synthetase was the same as that obtained for the wild type 2-5A synthetase. Similar results were obtained with the K199H mutant (data not shown).
Recombinant Mutant K199R Human 40-kDa 2-5A Synthetase: Binding of 8-Azido-ATP, Binding of Poly(I)⅐Poly(C) and 2-5A Synthetase Activity-To determine the role of lysine 199 in the enzymatic conversion of ATP to 2-5A synthetase, Lys 199 was mutated to Arg 199 . The K199R mutant did not show catalytic activity compared with the wild type 2-5A synthetase in the presence of 50 g/ml poly(I)•poly(C) (Fig. 9A, compare lanes  2 and 4). In the absence of poly(I)⅐poly(C), there was no enzyme activity (Fig. 9A, lanes 1 and 3). Wild type 40-kDa 2-5A synthetase and the K199R mutant 2-5A synthetase were photolabeled equally by 8-azido-[␣-32 P]ATP (Fig. 9B). This strongly suggests that the change from Lys 199 to Arg 199 did not affect the binding of 8-azido-[␣-32 P]ATP. The wild type 2-5A synthetase and the K199R mutant bound equally well to poly(I)•poly(C)-agarose (Fig. 9C, compare lanes 6 and 7). Lanes 1, 2, and 3 show GST, wild type 2-5A synthetase, and the K199R mutant incubated in the absence of poly(I)⅐poly(C), respectively. Lanes 6 and 7 show the same amount of proteins in the presence of poly(I)⅐poly(C). GST does not bind poly(I)⅐poly(C)-agarose (Fig. 9C, lane 5). These data suggest that the dsRNA allosteric domain of human 40-kDa 2-5A synthetase was not affected by a point mutation of Lys 199 to Arg. Therefore, Lys 199 appears to be essential for the enzymatic formation of 2-5A from ATP. Similar results were obtained for the K199H mutant (data not shown).
Competition Experiments with Recombinant Human K199R Mutant 40-kDa 2-5A Synthetase by ATP-Although the K199R mutant 2-5A synthetase did not show enzyme activity (Fig. 9A), The photolabeled peptides were retained by Al 3ϩ affinity chromatography purification and specifically displaced from the column by phosphate buffer. B, reverse phase HPLC profile of immobilized Al 3ϩ -purified tryptic peptides from 40-kDa 2-5A synthetase photolabeled with 8-azido-[␣-32 P]ATP. The peptides were separated using H 2 O and acetonitrile as the gradient system. UV absorbance at 214 nm (---) and radioactivity (q) are plotted against fraction number.  (2) a Picomole yields of PTH-derivatives are shown in parentheses. b Pentapeptide from experiment 1 in the absence of poly(I) ⅐ poly(C). c Pentapeptide from experiment 2 in the absence of poly(I) ⅐ poly(C). d Pentapeptide from experiment 3 in the presence of poly(I) ⅐ poly(C). e X is Lys-199 as deduced from the primary amino acid sequence of the 40-kDa 2-5A synthetase (7).
f Not detectable.
the covalent photoinsertion of 8-azido-[␣-32 P]ATP into the K199R mutant was the same as observed for wild type 2-5A synthetase. Therefore, it was essential to determine if ATP could mimic the photoinsertion of 8-azido-[␣-32 P]ATP into the K199R mutant. The addition of ATP (0 -15 mM) showed a similar competition of 8-azido-[␣-32 P]ATP (Fig. 10, A and B) to that observed with wild type 2-5A synthetase in the absence of poly(I)⅐poly(C) (Fig. 5, A and B). This strongly suggests that the photoinsertion of 8-azido-[␣-32 P]ATP occurred at the ATP binding domain of the K199R mutant recombinant human 40-kDa 2-5A synthetase. DISCUSSION 2-5A synthetase is an important enzyme in the interferoninducible antiviral and antiproliferative pathways. However, the identification of the ATP binding domain of 2-5A synthe-tase has not been reported. Previous studies from this laboratory have shown that the azido photoprobe, 8-azido-[␣-32 P]ATP was covalently linked by UV light to the highly purified rabbit reticulocyte 100-kDa 2-5A synthetase (12). However, insufficient quantities of 2-5A synthetase prevented identification of the amino acid(s) at the ATP binding domain. By fusion of the 40-kDa 2-5A synthetase cDNA to GST followed by transformation of the construct into E. coli, we have successfully purified milligram quantities of near homogeneously pure recombinant human 40-kDa 2-5A synthetase, which, following UV irradiation, 0°C, 20 s in solution, resulted in the isolation of the 32 P-pentapeptide, 196 DFLKQ 200 . Microsequencing of the amino acids in this pentapeptide revealed that lysine 199 was chemically modified by the azido photoprobe. Therefore, with the availability of sufficiently highly purified 2-5A synthetase, we  (Table II).
Further evidence that lysine 199 in the 40-kDa 2-5A synthetase is essential for the formation of a productive 2-5A synthetase-ATP-dsRNA complex needed for the enzymatic synthesis of 2-5A has been further provided by point mutation methodology in which lysine 199 was replaced by either arginine 199 or histidine 199. The K199R and K199H 2-5A synthetase mutants are enzymatically inactive, even though they bind poly(I)⅐poly(C) and 8-azido-[␣-32 P]ATP. A second point mutation in the recombinant human 2-5A synthetase, i.e. Lys 199 to His, was obtained using the same mutagenesis techniques. The ability of the K199H mutant to bind 8-azido-[␣-32 P]ATP was the same as that observed for the K199R mutant and the wild type 2-5A synthetase; however, as with the K199R mutant, the K199H mutant did not show enzyme activity (data not shown). Within the pentapeptide, 196 DFLKQ 200 , there are hydrophobic and hydrophilic amino acids. For example, phenylalanine 197 might interact with the planar adenine ring of ATP via van der Waals forces, where each aromatic ring provideshydrophobic interactions. Similarly, the ⑀-amino group of lysine 199 could ionically interact with the 5Ј-triphosphoryl group of ATP. Molecular modeling of DFLKQ with ATP supports this suggestion (data not shown). Phenylalanine 197 is the conserved amino acid that is common in five 2-5A synthetases near the lysine-ATP binding domain (Table II). Therefore, mutants of the recombinant human 40-kDa 2-5A synthetase in which Phe 197 has been replaced by tyrosine and alanine are currently under study in this laboratory.  1 and 2), or K199R mutant 2-5A synthetase (lanes 3 and 4) was incubated with [␣-32 P]ATP in the absence (lanes 1 and 3) or presence or 50 g/ml poly(I)⅐poly(C) (lanes 2 and 4) at 30°C for 2 h. The 2-5A products were dephosphorylated and separated on polyethyleneimine-cellulose TLC. A scanned image of the autoradiogram of the TLC is shown. Authentic 2-5A dimer and trimer cores and P i are indicated. B, wild type (lane 1) or K199R mutant 2-5A synthetase (lane 2) was photolabeled with 8-azido-[␣-32 P]ATP at 0°C for 30 min, followed by UV irradiation for 30 s. The photolabeled proteins were analyzed by SDS-PAGE. An image of the autoradiogram of the dried gel is shown. C, GST, wild type 2-5A synthetase, or K199R mutant 2-5A synthetase was incubated in the absence (lanes 1, 2, and 3, respectively) or presence of 50 g/ml poly(I)⅐poly(C)-agarose (lanes 5, 6, and 7, respectively) at 0°C for 1 h. After washing the poly(I)⅐poly(C)-agarose was analyzed by 10% SDS-PAGE. Protein markers are shown in lane 4. Wild type and K199R mutant 2-5A synthetase are indicated by arrows. The data presented in this study can be compared with the report of Ghosh et al. (26). Using a nested set of deletion mutants from the carboxyl terminus of the murine 2-5A synthetase, clones encoding 320 and 304 nucleotides were produced. However, the encoded proteins bound to dsRNA but did not have enzymatic activity. These data indicate that the peptide region between 342 and 304 was essential for 2-5A synthetase activity. Furthermore, a lysine residue is located at position 333. Because many protein kinases have lysine residues that are essential for enzyme activity at their ATP binding domains, lysine 333 was mutated to arginine 333. However, the K333R mutant 2-5A synthetase had the same enzyme activity as the wild type enzyme.
In summary, the data presented demonstrate that 8-azido-[␣-32 P]ATP is covalently linked to the ATP binding domain in solution. X-ray crystallography studies with the recombinant human 40-kDa 2-5A synthetase have been initiated in this laboratory to provide three-dimensional data on the ATP binding domain in the solid state. 3 Because of the 100-fold difference in the binding constants for the two putative binding sites for ATP (12), it will be necessary to identify the lower affinity ATP binding site on 2-5A synthetase by the x-ray crystallographic studies currently under way. RNase L and PKR have been reported to have tumor suppressor activity in dominant negative mutant studies (27,28). It is possible that the K199R mutant will have dominant negative effects on the wild type 40-kDa 2-5A synthetase upon overexpression in human cells. The K199R mutant may also have the same effect on the 69/71-kDa 2-5A synthetase because the same catalytic domain is present in these 2-5A synthetase isoforms. In vivo studies examining dominant negative activity of the mutant 2-5A synthetase will shed light on the cellular functions of antiviral and antiproliferative activities of the 2-5A synthetase/RNase L system.