Identification of the metal-binding sites of restriction endonucleases by Fe2+-mediated oxidative cleavage.

Fenton chemistry [Fenton (1894) J. Chem. Soc. 65, 899-910] techniques were employed to identify the residues involved in metal binding located at the active sites of restriction endonucleases. This process uses transition metals to catalytically oxidize the peptide linkage that is in close proximity to the amino acid residues involved in metal ligation. Fe2+ was used as the redox-active transition metal. It was expected that Fe2+ would bind to the endonucleases at the Mg2+-binding site [Liaw et al. (1993) Biochemistry 32, 7999-4003; Ermácora et al. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 6383-6387; Soundar and Colman (1993) J. Biol. Chem. 268, 5264-5271; Wei et al. (1994) Biochemistry 33, 7931-7936; Ettner et al. (1995) Biochemistry 34, 22-31; Hlavaty and Nowak (1997) Biochemistry 36, 15515-15525). Fe2+-mediated oxidation was successfully performed on TaqI endonulease, suggesting that this approach could be applied to a wide array of endonucleases [Cao and Barany (1998) J. Biol. Chem. 273, 33002-33010]. The restriction endonucleases BamHI, FokI, BglI, BglII, PvuII, SfiI, BssSI, BsoBI, EcoRI, EcoRV, MspI, and HinP1I were subjected to oxidizing conditions in the presence of Fe2+ and ascorbate. All proteins were inactivated upon treatment with Fe2+ and ascorbate. BamHI, FokI, BglI, BglII, PvuII, SfiI, BssSI, and BsoBI were specifically cleaved upon treatment with Fe2+/ascorbate. The site of Fe2+/ascorbate-induced protein cleavage for each enzyme was determined. The Fe2+-mediated oxidative cleavage of BamHI occurs between residues Glu77 and Lys78. Glu77 has been shown by structural and mutational studies to be involved in both metal ligation and catalysis [Newman et al. (1995) Science 269, 656-663; Viadiu and Aggarwal (1998) Nat. Struct. Biol. 5, 910-916; Xu and Schildkraut (1991) J. Biol. Chem. 266, 4425-4429]. The sites of Fe2+/ascorbate-induced cleavage for PvuII, FokI, BglI, and BsoBI agree with the metal-binding sites identified in their corresponding three-dimensional structures or from mutational studies [Cheng et al. (1994) EMBO J. 13, 3297-3935; Wah et al. (1997) Nature 388, 97-100; Newman et al. (1998) EMBO J. 17, 5466-5476; Ruan et al. (1997) Gene 188, 35-39]. The metal-binding residues of BglII, SfiI, and BssSI are proposed based on amino acid sequencing of their Fe2+/ascorbate-generated cleavage fragments. These results suggest that Fenton chemistry may be a useful methodology in identifying amino acids involved in metal binding in endonucleases.

Characterization of the second metal site on avian phosphoenolpyruvate carboxykinase.

Chicken liver phosphoenolpyruvate carboxykinase (PEPCK) requires two divalent cations for activity. One cation activates the enzyme through a direct interaction with the protein at site n(1). The second cation, at site n(2), acts in the cation-nucleotide complex that serves as a substrate. The Co(3+)(n(1))-PEPCK and Cr(3+)(n(1))-PEPCK complexes were used to examine the kinetic, mechanistic, and binding properties of the n(2) metal. EPR studies performed on the Co(3+)(n(1))-PEPCK-GTP complex yielded a stoichiometry of 1 mol of Mn(2+) bound per mole of Co(3+)(n(1))-PEPCK-GTP with a K(D) of 5 microM. PRR studies show a significant enhancement for the Co(3+)(n(1))-PEPCK-Mn(2+)(n(2))-GDP complex. A change in enhancement in the presence of PEP suggests that PEP interacts with the second metal ion. The distance between Mn(2+) at site n(2) on PEPCK and the cis and trans protons and the (31)P of PEP are 7.0, 7.5, and 4.8 A, respectively, as measured by high-resolution NMR. PRR studies of the Co(3+)(n(1))-PEPCK-Mn(2+)(n(2))-GTP and Co(3+)(n(1))-PEPCK-Mn(2+)(n(2))-GDP complexes as a function of frequency (omega(I)) were used to estimate the hydration number of the n(2) metal to be between 0.5 and 0.7. The metal-metal distance for the M(n(1))-PEPCK-M(n(2))-GTP complex is approximately 8.3 A, and the distance for the M(n(1))-PEPCK-M(n(2))-GDP complex is 9.2 A. The change in the metal-metal distance suggests a conformational change at the active site of PEPCK occurs during catalysis. The Co(3+)(n(1))-PEPCK complex was incubated with Co(2+), GTP, and H(2)O(2) to create a doubly labeled and inactive Co(3+)(n(1))-PEPCK-Co(3+)(n(2))-GTP complex. The Co(3+)(n(1))-PEPCK-Co(3+)(n(2))-GTP complex was digested by LysC, and two cobalt-containing peptides were purified using RP-HPLC. Amino acid sequencing of the second cobalt-containing peptide points to the region of Tyr57-Lys76 of PEPCK. Asp66, Asp69, and Glu74 are all feasible ligands to the site n(2) metal.

Chromium(III) modification of the first metal binding site of phosphoenolpyruvate carboxykinase.

Chicken liver phosphoenolpyruvate carboxykinase (PEPCK) is activated by Cr2+ as the sole activator under anaerobic conditions. PEPCK was modified with Cr3+, starting with either Cr2+ or Cr3+. Cr3+ has the distinct advantage of being a paramagnetic cation that could serve as a paramagnetic probe. Activators Mn2+, Mg2+, and Co2+ protect against Cr3+ incorporation. EPR, CD, and fluorescence studies indicate that Cr3+ was incorporated into the cation binding site of PEPCK. The water proton relaxation rate (PRR) and fluorescence binding studies showed that Cr3+(n1)-PEPCK forms enzyme-substrate complexes similar to those observed for the Mn2+(n1)-PEPCK complex (n1 represents the metal "enzyme binding site" as opposed to the metal "nucleotide binding site"). Cr3+(n1)-PEPCK requires an additional divalent cation for activity, an indication of two metal sites on PEPCK. Cr3+(n1)-PEPCK retains 15% residual activity as compared to unmodified PEPCK and demonstrates normal Michaelis-Menten kinetics. This is the first report of an active Cr3+-modified enzyme complex.

Formation and characterization of an active phosphoenolpyruvate carboxykinase-cobalt(III) complex.

Avian mitochondrial phosphoenolpyruvate carboxykinase (PEPCK) was incubated with Co2+ and H2O2 to form a stable Co3+-PEPCK complex. PEPCK, similarly incubated with H2O2 and either Mg2+ or Mn2+, resulted in no significant loss in activity over 30 min. PEPCK, incubated with Co2+ and H2O2 at pH 7.4, showed rapid inhibition as observed by a 40% decrease in activity after 5 min. The loss of activity is linear with the incorporation of cobalt into PEPCK, resulting in 15-25% activity for the stoichiometric Co3+-PEPCK complex. The incorporation of and inhibition by Co3+ is protected by PEP and GTP (ITP). Treatment of the Co3+-PEPCK complex with beta-mercaptoethanol results in a loss of cobalt and full recovery of activity. The reduction and reactivation are protected by PEP and GTP (ITP). EPR, PRR, circular dichroism, and fluorescence studies all indicate that Co3+ has been selectively incorporated into the cation site of PEPCK, resulting in a catalytically active enzyme-cation species. The substrates form Michaelis complexes with Co3+-PEPCK, and the catalytic reaction occurs as a second sphere complex as previously suggested [Lee & Nowak (1984) Biochemistry 23, 6506); Duffy & Nowak (1985) Biochemistry 24, 1152]. Proteolytic digestion of the Co3+-PEPCK complex and isolation of the cobalt-containing peptide by reverse phase HPLC were performed to identify the location of the cation binding site. From mass, amino acid composition, and sequence analyses of the isolated cobalt-peptide, the region Thr276-Lys301 is responsible for metal chelation. This very homologous region, located in the central portion of PEPCK, contains two highly conserved aspartic acids, Asp295 and Asp296, that are the only feasible metal binding ligands.

Affinity cleavage at the metal-binding site of phosphoenolpyruvate carboxykinase.

Chicken liver phosphoenolpyruvate carboxykinase (PEPCK) was rapidly inactivated by micromolar concentrations of ferrous sulfate in the presence of ascorbate at pH 7.4. Omitting ascorbate or replacing the Fe2+ with Mn2+ or Mg2+ gives no inactivation. Mn2+, Mg2+, or Co2+ at 100-fold molar excess over Fe2+ offered complete protection from Fe2+/ascorbate-induced inactivation. The substrates PEP and GTP, but not OAA, GDP, or CO2, offered full protection from inactivation. The addition of 5 mM EDTA stopped further inactivation of the enzyme. Thermodynamic studies indicate that the inactive enzyme no longer binds Mn2+ but still had high affinity for GTP indicating that the inactivation process was specific for the metal site. A decrease in cysteine content was observed over time following PEPCK treatment with Fe2+ and ascorbate. The apparent first-order rate constant for free sulfhydryl loss (0.085 +/- 0.005 min-1) is similar to the apparent first-order rate constant for inactivation (0.067 +/- 0.005 min-1). Amino acid composition analysis revealed that cysteic acid was generated upon Fe2+/ascorbate addition to PEPCK. Native chicken liver PEPCK has an Mr of 67 kDa. SDS-PAGE of the inactivated enzyme showed the presence of two new bands at 31.7 and 35.3 kDa indicating that PEPCK was specifically cleaved at a single site. The rate of cleavage was slower than the rate of inactivation and fully inactivated enzyme was only 50% cleaved. The Fe2+/ascorbate-catalyzed inactivation was not solely due to protein cleavage. The protein fragments generated by cleavage were separated by C4 reverse phase HPLC. The cleavage exposed a new N-terminus which was identified to be the 35.3 kDa C-terminal half of PEPCK. Sequencing of the fragments indicated that the site of cleavage was between Asp296 and Ile297. These results indicate that Asp296 is involved in metal chelation. This agrees with previous studies [Hlavaty, J. J., & Nowak, T. (1997) Biochemistry 36, 3389-3403] that suggested that Asp295 and Asp296 are involved in metal binding.

Nonisothermal stability assessment of stable pharmaceuticals: testing of a clindamycin phosphate formulation.

The stability of an antibiotic formulation (clindamycin phosphate in dextrose), which is stable at room temperature, was assessed by nonisothermal kinetic analysis at elevated temperatures. A preliminary study, conducted to establish apparent rate order, verified the appropriateness of a first-order kinetic model. The test formulation was then heated linearly from 70 to 90 degrees C over 12 hr. Data (drug concentration, temperature, and time) were fitted to the first-order model using nonlinear least-squares regression. Arrhenius parameter estimates obtained from three nonisothermal trials, and rate constants at 25 degrees C derived by extrapolation, demonstrated acceptable reproducibility and were in agreement with values derived from isothermal experiments at 30, 45, 55, 65, and 75 degrees C. First-order rate constants obtained from studies conducted for 20 months at 25 degrees C were in accord with isothermal and nonisothermal results.