Kinetic and Spectroscopic Analyses of Mutants of a Conserved Histidine in the Metallophosphatases Calcineurin and lambda  Protein Phosphatase*

(Received for publication, December 16, 1996, and in revised form, June 13, 1997)

Pamela Mertz , Lian Yu , Robert Sikkink and Frank Rusnak Dagger

From the Section of Hematology Research and the Department of Biochemistry and Molecular Biology, Mayo Clinic and Foundation, Rochester, Minnesota 55905

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

Calcineurin belongs to a family of serine/threonine protein phosphatases that contain active site dinuclear metal cofactors. Bacteriophage lambda  protein phosphatase is also considered to be a member of this family based on sequence comparisons (Lohse, D. L., Denu, J. M., and Dixon, J. E. (1995) Structure 3, 987-990). Using EPR spectroscopy, we demonstrate that lambda  protein phosphatase accommodates a dinuclear metal center. Calcineurin and lambda  protein phosphatase likewise contain a conserved histidine that is not a metal ligand but is within 5 Å of either metal in calcineurin. In this study the conserved histidine in calcineurin was mutated to glutamine and the mutant protein analyzed by EPR spectroscopy and kinetic methods. Parallel studies with an analogous lambda  protein phosphatase mutant were also carried out. Kinetic studies using paranitrophenyl phosphate as substrate showed a decrease in kcat of 460- and 590-fold for the calcineurin and lambda  protein phosphatase mutants, respectively, compared with the wild type enzymes. With a phosphopeptide substrate, mutagenesis of the conserved histidine resulted in a decrease in kcat of 1,300-fold for calcineurin. With the analogous lambda  protein phosphatase mutant, kcat decreased 530-fold compared with wild type lambda  protein phosphatase using phenyl phosphate as a substrate. EPR studies of the iron-reconstituted enzymes indicated that although both mutant enzymes can accommodate a dinuclear metal center, spectroscopic differences compared with wild type proteins suggest a perturbation of the ligand environment, possibly by disruption of a hydrogen bond between the histidine and a metal-coordinated solvent molecule.


INTRODUCTION

Calcineurin, also known as protein phosphatase 2B, consists of a 58-kDa catalytic subunit, calcineurin A, and a 19-kDa regulatory subunit, calcineurin B. It is a serine/threonine protein phosphatase whose activity is regulated by Ca2+/calmodulin. Calcineurin is the target of the immunosuppressant drugs cyclosporin A and FK506 (1, 2). These drugs bind to intracellular proteins, termed immunophilins; cyclophilin is the binding protein for cyclosporin A, and FK506 binds to the FK506-binding proteins. The complex of immunosuppressant drug and immunophilin in turn binds to and inhibits the phosphatase activity of calcineurin. Calcineurin inhibition prevents the transcriptional activation of the interleukin 2 gene in helper T cells, leading to suppression of the immune response.

Calcineurin is a member of the class of serine/threonine protein phosphatases, whose members include protein phosphatases 1 (PP1)1 and 2A (PP2A), phosphatases essential for a number of signal transduction pathways in eukaryotic cells (3, 4). Another protein phosphatase from bacteriophage lambda , lambda PP, also belongs to this family (5). In addition, a number of less characterized enzymes containing the "phosphoesterase" consensus motif of this family, DXH(X)nGDXXD(X)nGNHD/E, have been identified via protein sequence comparisons (6, 7). It has been hypothesized that this motif provides a scaffold for an active site dinuclear metal center (7, 8), similar to the dinuclear metal centers in PP1 (9, 10) and calcineurin (11-13). A variety of experimental evidence indicates that this cluster in calcineurin is an Fe3+-Zn2+ center.

Although little is known about the catalytic mechanism of the serine/threonine protein phosphatases, several pieces of experimental data indicate that the dinuclear metal center is a key component of the active site. First, x-ray crystallographic data of calcineurin and PP1 indicate that the dinuclear metal center has a ligand environment nearly identical to that of mammalian and plant purple acid phosphatases, enzymes that contain either dinuclear Fe-Fe or Fe-Zn centers that have been demonstrated to be essential for catalytic activity (14, 15). Second, these crystallographic studies indicate that the product of the reaction, phosphate, and the product analog, tungstate, directly coordinate both metal ions (10, 11, 16). Third, redox titrations of either the Fe3+-Zn2+ (12) or Fe3+-Fe2+ forms2 of calcineurin indicate a correlation between enzyme activity and the oxidation state of the bound metal ions.

In addition to the dinuclear metal center, there are several conserved amino acids within the active site which are likely to contribute to catalysis. One of these residues in calcineurin is histidine 151 (numbering based on the rat calcineurin Aalpha sequence (18)). His-151 is not a ligand to either metal but is within 5 Å of both metal ions and is conserved in other metallophosphoesterases such as PP1 (histidine 125) and lambda PP (histidine 76) (7). In one crystal structure, His-151 was modeled to participate in a hydrogen bond to a metal-coordinated solvent molecule (13). The importance of this residue has been demonstrated by site-directed mutagenesis of both lambda PP and PP1 which found substantial effects on catalytic activity and/or protein stability.

In this study the conserved histidine residue in calcineurin A, His-151, was changed to glutamine by site-directed mutagenesis. After reconstitution with calcineurin B, the calcineurin H151Q heterodimer was purified to homogeneity. EPR spectroscopy was used to assess whether mutagenesis affected the ligand environment of the dinuclear metal center. Kinetic studies were also carried out using either pNPP or [P]-RII peptide as substrates. The analogous lambda PP mutant, H76N, was characterized by EPR in a similar fashion and assayed using either pNPP or phenyl phosphate. For both enzymes, mutagenesis resulted in decreases in kcat of 102-103. EPR spectroscopy of iron-reconstituted lambda PP confirms that this enzyme accommodates a dinuclear metal center as predicted for members of the metallophosphatase family. EPR also indicates that mutagenesis does not prevent assembly of the dinuclear metal center in either calcineurin H151Q or lambda PP(H76N). Nevertheless, differences observed in the EPR spectra of wild type versus mutant enzymes indicate a perturbation of the ligand environment, possibly by disruption of a hydrogen bond between the histidine and a metal-coordinated water molecule.


EXPERIMENTAL PROCEDURES

Materials

Competent BL21(DE3) cells were obtained from Novagen (Madison, WI). Bovine serum albumin, DEAE-Sepharose CL-6B, bovine pancreatic DNase I, type II, pNPP, phenyl phosphate, and Sephacryl S300 were purchased from Sigma. Coomassie Plus Protein Assay Reagent and M guanidine hydrochloride were from Pierce Chemical Co. [gamma -32P]ATP (~3,000 Ci/mmol) was purchased from Amersham Corp. Calmodulin was prepared from bovine brain (19, 20) and coupled to Affi-Gel active ester agarose (Bio-Rad) for use in calmodulin affinity chromatography. PM30 and YM30 membranes and Centricon-30 concentrators were purchased from Amicon (Beverly, MA). Tryptone, yeast extract, and Luria Bertani medium were purchased from Difco. Sephadex NAP-25 columns were purchased from Pharmacia Biotech Inc. The protein expression vectors used were pRCNAT77 (21) and pRCNBT775-3 (22), encoding the genes for calcineurin A and B, respectively, and the plasmid, pBB131 (23, 24), encoding the gene for N-myristoyl transferase. The construction of pT7-7 plasmids containing the genes for wild type and H76N mutant of lambda PP is described elsewhere (5, 25). The Wizard Maxipreps and Wizard Minipreps DNA purification kits and T4 polynucleotide kinase were purchased from Promega (Madison, WI). The Geneclean II kit was purchased from BIO 101, Inc. (Vista, CA). The oligonucleotide required for site-directed mutagenesis was synthesized by the Mayo Clinic Molecular Biology Core Facility. RII peptide (DLDVPIPGRFDRRVSVAAE) and [31P]-RII peptide (DLDVPIPGRFDRRVS(p)VAAE) were synthesized by the Mayo Clinic Protein Core Facility.

Methods

Protein concentrations were determined by the Bradford assay using the Pierce Coomassie Plus Protein Assay Reagent with bovine serum albumin as a standard (26). Alternatively, calcineurin concentrations were determined by UV-visible spectrophotometry (27) using epsilon 281 = 50,000 M-1 cm-1. Protein concentration values determined using this extinction coefficient agreed within 10% with concentrations determined by amino acid analysis.

Site-directed Mutagenesis

The H151Q mutation of calcineurin A, CN(H151Q), was created using a 5 Prime right-arrow 3 Prime MORPH site-specific plasmid DNA mutagenesis kit (Boulder, CO). The primer 5'-GCCTACATTCCTGGTTTCCAC-3', complementary to the coding strand of calcineurin A, was used for mutagenesis with the underlined bases representing the codon of the mutated residue. This primer was phosphorylated by T4 kinase and used according to the manufacturer's instructions. Half-mutant plasmid DNA was subsequently transformed into Escherichia coli strain BMH 71-18. Plasmid DNA isolated from colonies was screened for the desired mutation by DNA sequencing of the entire calcineurin A gene and yielded the expression plasmid pCNAT77(H151Q).

Expression and Purification of CN(H151Q)

The plasmid pCNAT77(H151Q) was transformed into competent BL21(DE3) cells. The growth of these cells, crude extract isolation, reconstitution with myristoylated calcineurin B, and purification of the calcineurin heterodimer were performed as described previously (21, 22).

Expression and Purification of Wild Type lambda PP

Expression of wild type lambda PP was performed as described (5). All purification steps were performed at 4 °C. After growth and induction with isopropyl beta -D-thiogalactopyranoside, the cells were harvested by centrifugation at 3,400 × g for 30 min, washed with 250 ml of 0.1 M Tris-Cl, pH 7.5, and recentrifuged at 4,200 × g for 20 min. The cells were resuspended in 25 mM Tris-Cl, pH 8.0, 20% glycerol, 1 mM EGTA (TGE buffer) and lysed by three passages through a French pressure cell operating at 16,000 p.s.i. The cell lysate was subsequently centrifuged at 39,000 × g for 3 h. The supernatant (40 ml) was batch adsorbed onto 150 ml of DEAE-Sepharose CL-6B preequilibrated with TGE buffer. The resin was washed in a fritted funnel with 300 ml of TGE and lambda PP eluted with TGE buffer containing 0.1 M NaCl. Fractions containing lambda PP were pooled and precipitated by the addition of ammonium sulfate to 50% saturation. After centrifugation at 34,800 × g, the protein pellet was resuspended in TGE buffer + 0.5 M NaCl and applied to a phenyl-Sepharose column (20 × 1-cm diameter) previously equilibrated with TGE buffer + 0.5 M NaCl. The column was washed with 200-300 ml of the same buffer and then with 250 ml of 20 mM Tris-Cl, pH 7.5. The enzyme was eluted with 250 ml of 50 mM Tris-Cl in 50% glycerol, pH 7.5. Fractions were assayed using pNPP as a substrate, pooled, and stored at -70 °C in 50 mM Tris-Cl and 50% glycerol, pH 7.5.

Expression of lambda PP(H76N)

The lambda PPT77(H76N) plasmid (25) was transformed into BL21(DE3) cells and single colonies used to inoculate 10 ml of Luria Bertani medium/ampicillin (0.1 mg/ml) for overnight culture at 37 °C. Overnight cultures were then used to inoculate 5 liters of 2 × YT/ampicillin medium (10 g/liter yeast extract, 20 g/liter tryptone, 10 g/liter NaCl, 0.05 g/liter ampicillin) in a New Brunswick Bioflo 3000 fermentor. Cells were grown overnight at 22 °C maintaining aeration at 30% of air saturation to a cell density that gave an absorbance at 595 nm of ~9. Glucose was added to a final concentration of 0.4%, and the cells were induced with 1 mM isopropyl beta -D-thiogalactopyranoside. Another aliquot of glucose was added to a final concentration of 0.4% when the cell density corresponded to an absorbance at 595 nm of 16. The cells were harvested 20 h postinduction by centrifugation at 3,400 × g for 20 min. The cell pellet was resuspended in ~2 ml of 50 mM Tris-Cl, pH 7.5/g of cells, wet weight. To this resuspension, 0.4 mg/ml lysozyme, 23 mM EDTA, and 0.05% Triton X-100 were added sequentially with stirring on ice for 30 min followed by a freeze/thaw process to lyse the cells. To reduce the viscosity, MgCl2 (20 mM), DNase (0.1 unit/ml final concentration), and 2% protamine sulfate (1/6 total volume) were added sequentially with stirring on ice. After centrifugation at 10,000 × g for 1 h, the protein was purified as described above for wild type lambda PP.

Circular Dichroism Measurements

Circular dichroism spectra were recorded at 25 °C on a Jasco J-710 circular dichroism spectrometer. A quartz cell of 0.0202-cm path length was used for all measurements. Mean residue ellipticities were calculated from the relationship theta m = theta obs/(10Crl) where theta m is the mean residue ellipticity, theta obs is the observed ellipticity in millidegrees, Cr is the mean residue molar concentration, and l is the path length of cell in cm. theta m is measured in degree cm2 dmol-1. Samples of calcineurin (18 µM) and CN(H151Q) (24 µM) were examined in 10 mM Hepes, pH 7.5, 1 mM MgCl2, 0.1 M EGTA, 0.2 mM DTT.

pNPP Assays

Calmodulin-dependent phosphatase activity of calcineurin and CN(H151Q) was measured using pNPP as a substrate at 30 °C in 25 mM MOPS, pH 7.0, 1.0 mM MnCl2, 0.1 mM CaCl2, 1 µM calmodulin, and 15-23 nM wild type calcineurin or 710 nM CN(H151Q). Wild type lambda PP and lambda PP(H76N) activities were measured at 30 °C in 100 mM Tris-Cl, pH 7.8, 10 mM DTT, 1 mM MnCl2, and 0.64 nM wild type lambda PP or 860 nM lambda PP(H76N). After incubation for 5 min at 30 °C, reactions were started by the addition of pNPP. Specific activity was measured by following the increase in absorbance at 410 nm with time using epsilon 410 = 7,180 M-1 cm-1 at pH 7.0 and 14,400 M-1 cm-1 at pH 7.8 based on a pKa of 7.17 and epsilon 410 = 17,800 M-1 cm-1 for the p-nitrophenolate anion. The concentration of pNPP was varied from 2 to 100 mM, and the kinetic parameters kcat and Km were determined by fitting the data to the Michaelis-Menten equation using a nonlinear least squares analysis method.

[P]-RII Peptide Assays

RII peptide was phosphorylated with [gamma -32P]ATP to a specific activity of 833 µCi/µmol using the catalytic subunit of bovine cardiac cyclic-AMP dependent protein kinase and purified as described (28). Assays were done as described (28) in 100 mM MOPS, pH 7.0, 1 mM MnCl2, 0.1 mM CaCl2, 0.5 mM DTT, 0.1-28 µM calmodulin, and 1.0 mg/ml bovine serum albumin. [32P]-RII peptide was diluted with [31P]-RII peptide and substrate added in final concentrations from 9 µM to 1 mM. Calcineurin concentrations were 10 nM for wild type and 7.7-15 µM for CN(H151Q). Data were obtained in duplicate and fitted to the Michaelis-Menten equation by a nonlinear least squares analysis method.

Phenyl Phosphate Assays

Assays were performed by determining the amount of inorganic phosphate released during hydrolysis of phenyl phosphate as described (29). Assays were done in 100 mM Tris-Cl, pH 7.8, containing 1 mM MnCl2 and 10 mM DTT. Phenyl phosphate concentrations were varied from 1 to 70 mM for lambda PP and 1 to 90 mM for lambda PP(H76N). Wild type protein concentrations ranged from 7.0 to 630 nM, whereas lambda PP(H76N) concentrations ranged from 2.2 to 12 µM. Enzyme was incubated 30 °C, 5 min, and reactions started by the addition of phenyl phosphate. At various times from 0.5 to 7.0 min, 50 µl of the reaction was taken and added to 800 µl of a solution containing a 3:1 ratio of 0.045% malachite green hydrochloride to 4.2% ammonium molybdate in 4 N HCl. After 1 min, 100 µl of 34% sodium citrate was added and the absorbance at 660 nm measured. Free phosphate was determined from a standard curve prepared using solutions of KH2PO4. Kinetic parameters for lambda PP and lambda PP(H76N) were determined as described above.

Reconstitution of CN(H151Q) and lambda PP(H76N) with Iron

About 0.4 mg/ml CN(H151Q) in 20 mM Tris-Cl, pH 7.5, 100 mM KCl, 1.0 mM magnesium acetate, 1.0 mM DTT, 0.1 mM EGTA, or 0.6 mg/ml lambda PP(H76N) in 50 mM Tris-Cl, pH 7.5, 100 mM NaCl, 10% glycerol was added to septum-sealed vials and made anaerobic by flushing repeatedly with oxygen-free argon. BME was added to a final concentration of 0.715 M and Fe(NH4)2(SO4)2·6H2O added dropwise to a final concentration of 0.375 mM. The solutions were flushed with argon and incubated an additional 17 h at room temperature. Samples were concentrated using YM30 membranes in an Amicon filtration cell to ~ 2 ml; buffer exchanged over Sephadex NAP-25 gel filtration columns equilibrated with 50 mM MOPS, pH 7.0, 1 mM BME; and concentrated using Centricon-30 membranes to ~250 µl. Samples were then transferred to quartz EPR cuvettes and frozen in liquid nitrogen. Iron-reconstituted forms of wild type calcineurin and lambda PP were prepared in a similar manner except that incubation proceeded at 4 °C for 17 h, and both enzymes were desalted into 100 mM Tris-Cl, pH 7.5, 1 mM BME prior to the final concentration step.

Addition of Orthophosphate to CN(H151Q)

To the iron-reconstituted EPR sample of CN(H151Q), a solution of 0.5 M potassium phosphate, pH 7.5, was added anaerobically to a final concentration of 20 mM, incubated for 5 min at room temperature, and frozen in liquid nitrogen.

EPR Analysis

EPR spectra were recorded using a Bruker ESP300E spectrometer operating at 9 GHz (X-band) microwave frequency equipped with an Oxford Instruments ESR 900 continuous flow cryostat for temperature regulation. Background cavity resonances were subtracted from all spectra.

Metal Analysis

Metal analysis was performed by the Mayo Clinic Metals Laboratory using inductively coupled plasma emission spectrometry.


RESULTS

Preparation of Recombinant CN(H151Q) and lambda PP(H76N) Proteins

The relationship of His-151 in calcineurin relative to the dinuclear metal center can be seen in Fig. 1. His-151 was mutated to a glutamine by site-directed mutagenesis to investigate the effect on enzyme activity and assembly of the dinuclear metal cofactor. The mutant calcineurin A subunit was expressed in E. coli in a fashion identical to that of the wild type calcineurin A subunit and reconstituted with myristoylated calcineurin B to generate the mutant protein CN(H151Q). The presence of the proper codon as well as the lack of inadvertently introduced mutations in the entire calcineurin A gene were confirmed by DNA sequence analysis. Although the yield of CN(H151Q) was less than that obtained for the wild type reconstituted protein, enough material could be obtained and purified to homogeneity for biochemical and spectroscopic (EPR) analyses. A typical purification yielded approximately 1 mg of CN(H151Q) protein/liter of cell culture.


Fig. 1. Active site of calcineurin A. The iron and zinc atoms of the binuclear metal center are shown with their respective ligands. Histidine 151 and aspartate 121 are also modeled in the active site in a hydrogen bond network with a metal-coordinated water molecule. Metal coordination was determined from the 2.1 Å resolution crystal structure (13).
[View Larger Version of this Image (16K GIF file)]

The analogous residue in lambda PP, identified by primary sequence comparisons as His-76 (7), was mutated to an asparagine residue (25). In this study, the lambda PP(H76N) protein was purified to homogeneity as described under "Methods" to yield approximately 17 mg/liter of culture.

Circular dichroism analysis of wild type calcineurin and CN(H151Q) provided evidence for a native-like conformation of CN(H151Q); CD spectra from 200 to 250 nm of both recombinant wild type calcineurin and CN(H151Q) are comparable to spectra of bovine calcineurin (30) (data not shown).

Phosphatase Activities of Wild Type Calcineurin and CN(H151Q)

Kinetic parameters using pNPP and [P]-RII peptide as substrates for wild type calcineurin and CN(H151Q) in the presence of 1 mM MnCl2 are compared in Table I. Using either substrate, the values of kcat for the mutant enzyme were significantly lower than the kcat values for wild type enzyme. Thus, the kcat for CN(H151Q) using pNPP as a substrate, 5.6 × 10-2 s-1, is 460-fold lower than recombinant wild type calcineurin prepared in an identical fashion. Using [P]-RII peptide, the kcat values for wild type calcineurin and CN(H151Q) were 1.2 × 101 s-1 and 9.0 × 10-3 s-1, respectively, a difference of 1,300-fold. Using [P]-RII peptide, Km values for both forms of calcineurin were the same within the error of the measurement. However, a 10-fold decrease in Km was observed for CN(H151Q) compared with wild type calcineurin using pNPP as a substrate.

Table I. Kinetic parameters for recombinant wild type and H151Q calcineurin using pNPP and [P]-RII peptide as substrates

Kinetic parameters were determined as described under "Experimental Procedures." Values for kinetic constants were determined by a least squares analysis with the uncertainties in the coefficients noted. Data were from at least two separate determinations.

Enzyme kcat Km Relative kcata

s-1
pNPP hydrolysis
  Wild type calcineurin 2.6  × 101 ± 3 23  ± 7 mM 1
  CN(H151Q) 5.6  × 10-2 ± 1 × 10-3 2.2  ± 0.20 mM 460
[P]-RII hydrolysis
  Wild type calcineurin 1.2  × 101 ± 1 × 10-1 110  ± 40 µM 1
  CN(H151Q) 9.0  × 10-3 ± 1 × 10-4 130  ± 60 µM 1,300

a Relative kcat compared with wild type calcineurin for each substrate listed.

Phosphatase Activities of Wild Type lambda PP and lambda PP(H76N)

Kinetic parameters using pNPP and phenyl phosphate as substrates for wild type lambda PP and lambda PP(H76N) are compared in Table II. These parameters were also determined by inclusion of 1.0 mM MnCl2 in assay buffers. Using pNPP as substrate, the kcat values for wild type lambda PP and lambda PP(H76N) were 3.9 × 102 s-1 and 6.6 × 10-1 s-1, respectively. This represents a 590-fold difference, which is comparable to the decrease in kcat observed for calcineurin for the analogous substitution. The difference in kcat for lambda PP is less than the approx 105 fold difference found previously (25) mostly because of a 30-fold higher activity measured for the lambda PP(H76N) protein isolated in the present study. Using phenyl phosphate as substrate, the kcat value for lambda PP was 1.7 × 101 s-1, compared with 3.2 × 10-2 s-1 for lambda PP(H76N), a difference of 530-fold. The Km values were very similar for both mutant and wild type proteins using either substrate.

Table II. Kinetic parameters for recombinant lambda PP and histidine mutant using pNPP and phenyl phosphate as substrates

Kinetic parameters of lambda PP and lambda PP(H76N) were determined as described under "Experimental Procedures" with pNPP and phenyl phosphate as substrates. Previously reported parameters of wild type and mutant lambda PP using pNPP as a substrate are shown for comparison. Where indicated, the values represent the fit coefficient from a least squares analysis with the uncertainties in the coefficients noted. Data were from at least two separate determinations.

Enzyme kcat Km Relative kcata Ref.

s-1 mM
pNPP hydrolysis
  Wild type lambda PP 3.9  × 102 ± 1 × 101 11  ± 1 1 This work
2.0  × 103 10 25
  lambda PP(H76N) 6.6  × 10-1 ± 4 × 10-2 16  ± 3 590 This work
2.2  × 10-2 4.1 25
Phenyl phosphate hydrolysis
  Wild type lambda PP 1.7  × 101 ± 2 14  ± 5 1 This work
  lambda PP(H76N) 3.2  × 10-2 ± 3 × 10-3 22  ± 6 530 This work

a Relative kcat compared with wild type lambda PP for each substrate listed.

EPR Analysis of Iron-reconstituted Wild Type Calcineurin and CN(H151Q)

Although not a metal ligand, H151Q in calcineurin is close enough to either metal ion such that mutagenesis might perturb the environment surrounding the metal cluster. To investigate the effect of mutagenesis on the dinuclear center, wild type calcineurin and CN(H151Q) were reconstituted with iron to generate a mixed valence Fe3+-Fe2+ cluster as a spectroscopic probe of the active site and analyzed by EPR spectrometry. The EPR spectra of wild type calcineurin, CN(H151Q), and CN(H151Q) in the presence of 20 mM potassium phosphate are compared in Fig. 2. Several features are evident in the EPR spectra including a minor high spin Fe3+ species with g values of 9.2 and 4.3 (about 700-2,000 Gauss region, not shown for clarity), a minor radical species centered at gav = 2.0, and a component with gav < 2.0 representing the major paramagnetic species.


Fig. 2. EPR spectra of iron-reconstituted wild type recombinant calcineurin and CN(H151Q) in the presence and absence of phosphate. A, EPR spectrum of Fe3+-Fe2+ form of wild type recombinant calcineurin. The sample contained 0.18 mM calcineurin in 100 mM Tris-Cl, pH 7.5, 1 mM BME. The spectrum is identical to iron-reconstituted bovine calcineurin in 50 mM MOPS, pH 7.0 (12). B, EPR spectrum of Fe3+-Fe2+ form of CN(H151Q) at 3.6 K. The protein concentration was 0.077 mM, and sample buffer was 50 mM MOPS, pH 7.0, 1 mM BME. C, EPR sample of CN(H151Q) from B after the addition of 20 mM phosphate. For all samples, the spectrometer conditions were: microwave power, 20 milliwatts; microwave frequency, 9.453 Ghz; modulation amplitude, 20 G at 100 kHz; temperature, 3.6 K.
[View Larger Version of this Image (14K GIF file)]

Simulations of the EPR signal in Fig. 2A to an S = 1/2 species yielded g values of 1.93, 1.77, and 1.64. This signal is identical to the signal observed previously in bovine brain calcineurin reconstituted with iron and arises from a dinuclear iron center in the mixed valence (Fe3+-Fe2+) oxidation state (12). The spin Hamiltonian He that describes the magnetic properties of the dinuclear iron center is
<B><IT>H</IT></B><SUB>e</SUB>=D<SUB>1</SUB>((<UP>S</UP><SUB>1z</SUB><SUP>2</SUP>−<UP>S</UP><SUB>1</SUB>(<UP>S</UP><SUB>1</SUB>+1)/3)+E<SUB>1</SUB>/D<SUB>1</SUB>(<UP>S</UP><SUB>1x</SUB><SUP>2</SUP>−<UP>S</UP><SUB>1y</SUB><SUP>2</SUP>))+&bgr;<B><IT>S</IT></B><SUB>1</SUB> · <B><IT>g</IT></B><SUB>1</SUB> · <B><IT>H</IT></B> (Eq. 1)
+D<SUB>2</SUB>((<UP>S</UP><SUB>2z</SUB><SUP>2</SUP>−<UP>S</UP><SUB>2</SUB>(<UP>S</UP><SUB>2</SUB>+1)/3+E<SUB>2</SUB>/D<SUB>2</SUB>(<UP>S</UP><SUB>2x</SUB><SUP>2</SUP>−<UP>S</UP><SUB>2y</SUB><SUP>2</SUP>))+&bgr;<B><UP>S</UP></B><SUB>2</SUB> · <B><UP>g</UP></B><SUB>2</SUB> · <B><IT>H</IT></B>
+J<B><UP>S</UP></B><SUB>1</SUB> · <B><UP>S</UP></B><SUB>2</SUB>
where H is the external magnetic field, beta  is the Bohr magneton, J is the exchange-coupling constant, and Di, Ei, and gi (i = 1, 2) represent the zero field splitting terms and g tensors for each metal ion, respectively. Antiferromagnetic coupling (J > 0) between the high spin ferric (S1 = 5/2) and high spin ferrous (S2 = 2) ions of the cluster yields a ground state with S = 1/2 which gives rise to this EPR signal.

Fig. 2B shows the EPR spectrum of iron-reconstituted CN(H151Q). An EPR signal representative of a dinuclear iron center in the mixed valence state is evident and indicates that a dinuclear metal center can be assembled in the mutant enzyme. This signal, however, is broader than the signal observed from the mixed valence center in wild type enzyme. Metal analyses of the EPR sample found 1.5 mol each of iron and zinc/mol of protein, consistent with the formation of a dinuclear iron center but also indicating the possible presence of a mixed metal Fe-Zn center or adventitious zinc, either of which would contribute to the EPR spectrum in the gav < 2.0 region.

Further proof that the signal in Fig. 2B results from an active site metal center was demonstrated by adding 20 mM potassium phosphate to the sample. The addition of phosphate led to a noticeable sharpening of the EPR signal (Fig. 2C), whereas none of the other species was affected, suggesting that phosphate coordinates to one or both of the metal ions of the dinuclear iron center.

EPR Analysis of Iron-reconstituted lambda PP and lambda PP(H76N)

In a fashion similar to that for calcineurin, lambda PP and lambda PP(H76N) were reconstituted with iron to generate a spectroscopic probe of the active site metal cluster. lambda PP as purified contains very little iron, zinc, or manganese as determined by metal analysis using inductively coupled plasma emission spectrometry (<=  0.05 mol of iron, 0.09 mol of zinc, and 0.01 mol of manganese/mol of protein). Likewise, lambda PP(H76N) also contained low amounts of these metals (<=  0.3 mol of iron, 0.09 mol of zinc, and 0.01 mol of manganese/mol of protein). Reconstitution of wild type lambda PP and lambda PP(H76N) with iron yielded samples that exhibited low temperature EPR spectra with gav < 2.0 (Fig. 3, A and B). Metal analysis of both EPR samples found 1.74 iron/mol of protein and 0.20 zinc/mol of protein for wild type lambda PP, and 1.7 iron/mol of protein and 0.05 zinc/mol of protein for lambda PP(H76N). Thus both lambda PP and lambda PP(H76N) can accommodate dinuclear metal clusters. As in the case of calcineurin, the EPR spectrum of the lambda PP(H76N) mutant is different from the spectrum of wild type lambda PP.


Fig. 3. EPR spectra of iron-reconstituted wild type lambda PP and lambda PP(H76N). A, EPR spectrum of Fe3+-Fe2+ form of lambda PP taken at 6.4 K. The sample contained 0.36 mM lambda PP in 100 mM Tris-Cl, pH 7.5, 1 mM BME. Spectrometer conditions: microwave power, 1.0 milliwatt; microwave frequency, 9.454 GHz; modulation amplitude 10 G at 100 kHz. B, EPR spectrum of Fe3+-Fe2+ form of lambda PP(H76N) at 6.6 K. The sample contained 1.92 mM protein in 50 mM MOPS, pH 7.0, 1 mM BME. Spectrometer conditions were the same as in A except the microwave power was 0.5 milliwatt.
[View Larger Version of this Image (16K GIF file)]


DISCUSSION

Recent crystallographic models of calcineurin (11, 13), PP1 (9, 10), and purple acid phosphatase (16, 31) have identified a conserved histidine in each active site which, although not coordinated to either metal of the dinuclear metal center, is within approx  5 Å of both metal ions. In this study we have mutated the corresponding residue of calcineurin (His-151) to glutamine to explore its significance in catalysis and effect, if any, on the active site dinuclear metal center. An analogous histidine in the bacteriophage lambda  protein phosphatase has been previously modified to asparagine (lambda PP(H76)) (25). The H76N mutation resulted in a 105-fold reduction in kcat toward pNPP, a 40-fold increase in the Km for Mn2+, the divalent metal ion activator used in assay buffers, yet little change in Km for substrate. In a study by Lee and colleagues (32), the comparable residue in PP1 (His-125) was mutated to a number of residues. In that study it was found that most of the substitutions resulted in the production of insoluble protein except for two mutations, H125A and H125S, where a fraction of the protein was soluble and could be purified by affinity chromatography. Although neither PP1 mutant exhibited any detectable phosphatase activity, the upper limit for activity and/or fold reduction relative to wild type PP1 was not reported.

Similar to the results noted in the PP1 study, the level of expression of H151Q soluble protein was also lower than that found for wild type calcineurin A. Hence, the yield of the CN(H151Q) heterodimer was lower (approx 4-fold) than wild type reconstituted enzyme prepared in an identical fashion.3 As with wild type calcineurin (21), growth of E. coli expressing the calcineurin A subunit at lower temperatures (23 °C) improved the yield of soluble protein in crude extract. Reconstitution of the mutant calcineurin A subunit with calcineurin B allowed purification of sufficient protein for biochemical studies and spectroscopic analysis using EPR.

To assess the affect of mutagenesis on catalytic activity, two different substrates were used for calcineurin, pNPP and [P]-RII peptide.4 Furthermore, MnCl2 was included in buffers to obtain the maximum activity for both enzymes (5, 33, 34). Manganese is known to incorporate into one or both metal sites in PP1 (9, 10). Preliminary data indicate that a spin-coupled dinuclear Mn2+ cluster is also assembled in lambda PP.5

In terms of catalytic activity, mutagenesis of His-151 to glutamine resulted in a 460-fold decrease in kcat using pNPP as substrate. In comparison, the kcat for CN(H151Q) decreased 1,300-fold compared with wild type calcineurin using [P]-RII peptide as substrate. For either substrate, Km was either not affected or slightly improved by mutagenesis. The kcat values for wild type calcineurin measured in this study are comparable or slightly higher than previously reported (24, 34, 35); these differences may reflect slight variability between rat versus bovine isoforms.

Using pNPP as substrate, lambda PP(H76N) also exhibited little difference in substrate Km and a decrease in kcat of 590-fold relative to wild type lambda PP. This difference is more than 100-fold lower than the 105 noted previously (25) primarily because of a 30-fold higher activity measured for the lambda PP(H76N) mutant in this study but also because of a slightly lower activity of the wild type enzyme used in this study (approx  5-fold; see Table II). In fact, variability in kcat has been documented for wild type lambda PP (5, 25, 36); we have also observed about a 2-3-fold difference in activity for several different preparations of wild type lambda PP prepared by similar procedures. Kinetic parameters of wild type lambda PP and lambda PP(H76N) for a second substrate, phenyl phosphate, paralleled the results using pNPP. Thus lambda PP(H76N) exhibited a decrease in kcat of 530-fold and a Km similar to wild type enzyme.

Based on primary sequence homologies, lambda PP is thought to be a member of the family of metallophosphatases including calcineurin, PP1, and PP2A (6). We now show that lambda PP can accommodate a binuclear metal center as predicted for enzymes containing the phosphoesterase motif. Therefore, it is likely that His-76 in lambda PP and His-151 in calcineurin have similar functions in phosphate ester hydrolysis. Proposed roles for this conserved histidine include an active site nucleophile, a role in orienting substrate, a role in general acid catalysis involving protonation of the leaving group, or a role in general base catalysis by deprotonation of an iron-coordinated solvent molecule. All of these would be consistent with the 102-103-fold decrease in activity observed for wild type versus mutant enzymes.

The fold decreases for His-151/His-76 are comparable to the decrease of approx 103 found for serine-to-alanine and serine-to-leucine mutations for the nucleophilic serine residue of alkaline phosphatase (37). However, experiments with purple acid phosphatase, a member of the metallophosphatase superfamily with a strikingly similar active site as calcineurin (8) demonstrated that hydrolysis proceeded by direct transfer of the substrate phosphoryl group to solvent (38). By analogy, therefore, it seems unlikely that this histidine participates in nucleophilic catalysis. Similar substrate Km values for mutant and wild type enzymes also argue against a necessary role in substrate binding.

In the crystal structure of calcineurin with phosphate bound at the active site, His-151 is within H-bonding distance of the most solvent-exposed oxygen atom of phosphate, close enough to assist in leaving group protonation (11). Similar orientations are observed in the crystal structure of human PP1gamma with bound tungstate ion (10), rabbit PP1alpha with phosphate modeled in the active site (9), and in the purple acid phosphatase structures with phosphate and tungstate bound at the active site (16). If His-151/His-76 were involved in proton donation to the leaving group, the kcat for each mutant enzyme relative to wild type should show a marked dependence using substrates that have leaving groups of disparate pKa values. In fact, only a ~3-fold difference in relative kcat (wild type versus CN(H151Q)) for pNPP versus [P]-RII peptide is observed, even though the products of the reaction have acidities that differ by >106 (p-nitrophenol has a pKa of 7.2 compared with approx 14 for serine). With lambda PP, there is no difference between relative kcat values of wild type versus H76N using pNPP and phenyl phosphate (pKa = 9.95). It seems likely therefore that this histidine is not required for protonation of the leaving group.

Although the kcat values for CN(H151Q) and lambda PP(H76N) represent significant decreases compared with wild type enzyme, they are 105-107-fold greater than the noncatalyzed rate of hydrolysis.6 A significant amount of this remaining catalytic activity is most likely derived from the presence of the active site dinuclear metal center, which has been proposed to lower the pKa of a coordinated water molecule, the putative nucleophile in the reaction. His-151/His-76 could be functioning in concert with this solvent molecule to either position a lone pair on the oxygen atom for optimum in-line attack on the phosphorus atom of the substrate or to serve as a general base to take up a proton concomitant with solvent nucleophilic attack. At least in one crystal structure model of calcineurin, the Nepsilon atom of His-151 was H-bonded to one of two solvent molecules coordinated to the iron atom (13). In the crystal structure of PP1 with microcystin bound, the Nepsilon atom of the analogous histidine, His-125, was also within H-bonding distance of a water molecule, but that water was modeled >=  3.2 Å away from the metal ions (9). Further evidence for this model is provided by mutagenesis studies of PP1 examining the influence of a conserved aspartic acid residue, Asp-95, on catalytic efficiency. This conserved aspartate residue is part of the phosphoesterase consensus motif (6, 7). PP1 residue Asp-95 is within H-bonding distance of the conserved histidine, and mutagenesis to asparagine resulted in a 71-fold decrease in activity compared with wild type using phosphorylase a as substrate (32). The analogous mutant in lambda PP, D52N, resulted in a 36-fold reduction in activity using pNPP as substrate (25). The corresponding residue in calcineurin, Asp-121, is also within H-bonding distance of His-151 (Fig. 1). Thus, the interaction of this conserved histidine/aspartate pair with a solvent molecule is analogous to the catalytic aspartate/histidine/serine motif of serine proteases and could be thought of as a "catalytic tetrad" with the metal ion serving as a Lewis acid to lower the pKa of the nucleophile.

If His-151 participates in a hydrogen bond with a metal-coordinated solvent, mutagenesis will disrupt this interaction and is likely to affect the spectroscopic properties of the dinuclear metal center. We have reconstituted calcineurin and lambda PP with iron to generate an active site dinuclear iron center for use as a spectroscopic probe of the active site. The Fe3+-Fe2+ oxidation state of this cluster gives rise to a signature EPR resonance with gav < 2.0 (14, 15) which is sensitive to changes in the metal environment via perturbation of zero-field splitting (Di, Eii) and spin coupling (J) constants in Equation 1 (39). The EPR spectrum of iron-reconstituted CN(H151Q) exhibited g values consistent with the formation of a Fe3+-Fe2+ center, indicating that the H151Q mutant enzyme is still able to support a dinuclear metal center. However, the overall shape of this spectrum was quite different from that of wild type calcineurin. In comparison, the EPR spectrum of the iron-reconstituted lambda PP(H76N) also exhibited the characteristic gav < 2.0 signal with a shape distinct from the corresponding spectrum of wild type lambda PP.

The fact that phosphate addition to iron-reconstituted CN(H151Q) caused a change in the shape of the EPR resonance indicates that it arises from an active site metal center. Interestingly, in both wild type calcineurin and purple acid phosphatase, phosphate binding to the mixed valence cluster led to a broadening of the corresponding EPR signal, a result of a decrease in the spin coupling constant, J, caused by the phosphate ion bridging the two metal ions of the cluster (39).2 With CN(H151Q), on the other hand, phosphate caused a sharpening of the EPR resonance. Further spectroscopic analysis is required to understand the structural basis for these differences.

These results demonstrate that the conserved histidine in the metallophosphatases calcineurin and lambda PP is an essential component of the active site since disruption led to significant decreases in activity. Loss of activity may have resulted from removal of an active site base and the disruption of an essential H bond to a metal-coordinated solvent molecule. Future experiments to confirm this are in progress.


FOOTNOTES

*   This work was supported by National Institutes of Health Grant GM46865.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Dagger    To whom correspondence should be addressed: Mayo Clinic and Foundation, 200 First Street, S.W., Rochester, MN 55905. Tel.: 507-284-4743; Fax: 507-284-8286.
1   The abbreviations used are: PP1, protein phosphatase 1; PP2A, protein phosphatase 2A; lambda PP, lambda  protein phosphatase; pNPP, p-nitrophenyl phosphate; [P]-RII peptide, the phosphorylated peptide derived from the RII subunit of cyclic AMP-dependent protein kinase; lambda PP(H76N), lambda PP with asparagine in place of histidine at position 76; CN(H151Q), calcineurin with glutamine in place of histidine at position 151; DTT, dithiothreitol; MOPS, 3-(N-morpholino)propanesulfonic acid; BME, beta -mercaptoethanol.
2   Yu, L., Golbeck, J., Yao, J., and Rusnak, F. (1997) Biochemistry, in press.
3   P. Mertz, L. Yu, R. Sikkink, and F. Rusnak, unpublished results.
4   Although pNPP is a substrate for calcineurin (33), the Km of 10-20 mM is significantly greater than the Km of approx 100 µM for [P]-RII peptide. Furthermore, calcineurin activity using [P]-RII peptide is inhibited by the immunosuppressant drug complexes cyclosporin A·cyclophilin and FK506·FKBP, whereas it is slightly stimulated when assayed using pNPP (1, 40). Nevertheless, calcineurin phosphatase activity toward [P]-RII peptide is progressively inhibited in the presence of increasing concentrations of pNPP with an IC50 equivalent to the Km for pNPP (data not shown). These results indicate that both pNPP and [P]-RII peptide utilize the same active site, and hence a comparison of their kinetic parameters is valid.
5   L. Yu and F. Rusnak, unpublished results.
6   The rate constant for nonenzymatic hydrolysis of pNPP, kuncat = 7 × 10-8 s-1 at pH 7.4, 39 °C (41). Hydrolysis of serine phosphate occurs with a rate constant kuncat of 6.7 × 10-10 s-1 at 25 °C (17).

ACKNOWLEDGEMENTS

We acknowledge gratefully Jack Dixon's laboratory for providing the plasmids containing the clones for lambda PP wild type and lambda PP(H76N). We also thank Jim Griffith and colleagues at Vertex for providing the coordinates of calcineurin, John Kuriyan and co-workers for PP1 coordinates, and Melissa Snyder and Sergei Venyaminov for help with circular dichroism experiments.


REFERENCES

  1. Liu, J., Farmer, J. D., Jr., Lane, W. S., Friedman, J., Weissman, I., and Schreiber, S. L. (1991) Cell 66, 807-815 [Medline] [Order article via Infotrieve]
  2. Schreiber, S. L., and Crabtree, G. R. (1992) Immunol. Today 13, 136-142 [CrossRef][Medline] [Order article via Infotrieve]
  3. Cohen, P., and Cohen, P. T. W. (1989) J. Biol. Chem. 264, 21435-21438 [Free Full Text]
  4. Shenolikar, S., and Nairn, A. C. (1991) Adv. Second Messenger Phosphoprotein Res. 23, 1-121 [Medline] [Order article via Infotrieve]
  5. Zhuo, S., Clemens, J. C., Hakes, D. J., Barford, D., and Dixon, J. E. (1993) J. Biol. Chem. 268, 17754-17761 [Abstract/Free Full Text]
  6. Koonin, E. V. (1994) Prot. Sci. 3, 356-358 [Abstract/Free Full Text]
  7. Lohse, D. L., Denu, J. M., and Dixon, J. E. (1995) Structure 3, 987-990 [Medline] [Order article via Infotrieve]
  8. Rusnak, F., Yu, L., and Mertz, P. (1996) J. Biol. Inorg. Chem. 1, 388-396 [CrossRef]
  9. Goldberg, J., Huang, H.-b., Kwon, Y.-g., Greengard, P., Nairn, A. C., and Kuriyan, J. (1995) Nature 376, 745-753 [CrossRef][Medline] [Order article via Infotrieve]
  10. Egloff, M.-P., Cohen, P. T. W., Reinemer, P., and Barford, D. (1995) J. Mol. Biol. 254, 942-959 [CrossRef][Medline] [Order article via Infotrieve]
  11. Griffith, J. P., Kim, J. L., Kim, E. E., Sintchak, M. D., Thomson, J. A., Fitzgibbon, M. J., Fleming, M. A., Caron, P. R., Hsiao, K., and Navia, M. A. (1995) Cell 82, 507-522 [Medline] [Order article via Infotrieve]
  12. Yu, L., Haddy, A., and Rusnak, F. (1995) J. Am. Chem. Soc. 117, 10147-10148
  13. Kissinger, C. R., Parge, H. E., Knighton, D. R., Lewis, C. T., Pelletier, L. A., Tempczyk, A., Kalish, V. J., Tucker, K. D., Showalter, R. E., Moomaw, E. W., Gastinel, L. N., Habuka, N., Chen, X., Maldonado, F., Barker, J. E., Bacquet, R., and Villafranca, J. E. (1995) Nature 378, 641-644 [CrossRef][Medline] [Order article via Infotrieve]
  14. Kurtz, D. M., Jr. (1990) Chem. Rev. 90, 585-606
  15. Vincent, J. B., Olivier-Lilley, G. L., and Averill, B. A. (1990) Chem. Rev. 90, 1447-1467
  16. Klabunde, T., Sträter, N., Fröhlich, R., Witzel, H., and Krebs, B. (1996) J. Mol. Biol. 259, 737-748 [CrossRef][Medline] [Order article via Infotrieve]
  17. Schray, K. J., and Benkovic, S. J. (1973) in The Enzymes (Boyer, P. D., ed), pp. 201-238, Academic Press, New York
  18. Saitoh, Y., Maeda, S., Fukunaga, K., Yasugawa, S., Sakamoto, Y., Uyemura, K., Shimada, K., Ushio, Y., and Miyamoto, E. (1991) Biomed. Res. 12, 215-218
  19. Gopalakrishna, R., and Anderson, W. B. (1982) Biochem. Biophys. Res. Commun. 104, 830-836 [Medline] [Order article via Infotrieve]
  20. Dedman, J. R., and Kaetzel, M. A. (1983) Methods Enzymol. 102, 1-8
  21. Haddy, A., and Rusnak, F. (1994) Biochem. Biophys. Res. Commun. 200, 1221-1229 [CrossRef][Medline] [Order article via Infotrieve]
  22. Sikkink, R., Haddy, A., MacKelvie, S., Mertz, P., Litwiller, R., and Rusnak, F. (1995) Biochemistry 34, 8348-8356 [Medline] [Order article via Infotrieve]
  23. Duronio, R. J., Rudnick, D. A., Johnson, R. L., Linder, M. E., and Gordon, J. I. (1990) Methods: Companion to Methods Enzymol. 1, 253-263
  24. Kennedy, M. T., Brockman, H., and Rusnak, F. (1996) J. Biol. Chem. 271, 26517-26521 [Abstract/Free Full Text]
  25. Zhuo, S., Clemens, J. C., Stone, R. L., and Dixon, J. E. (1994) J. Biol. Chem. 269, 26234-26238 [Abstract/Free Full Text]
  26. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 [CrossRef][Medline] [Order article via Infotrieve]
  27. Gill, S. C., and von Hippel, P. H. (1989) Anal. Biochem. 182, 319-326 [Medline] [Order article via Infotrieve]
  28. Hubbard, M. J., and Klee, C. B. (1991) in Molecular Neurobiology: A Practical Approach (Wheal, H., and Chad, J., eds), pp. 135-157, Oxford University Press, Oxford
  29. Lanzetta, P. A., Alvarez, L. J., Reinach, P. S., and Candia, O. A. (1979) Anal. Biochem. 100, 95-97 [Medline] [Order article via Infotrieve]
  30. Wolff, D. J., and Sved, D. W. (1985) J. Biol. Chem. 260, 4195-4202 [Abstract]
  31. Sträter, N., Klabunde, T., Tucker, P., Witzel, H., and Krebs, B. (1995) Science 268, 1489-1492 [Medline] [Order article via Infotrieve]
  32. Zhang, J., Zhang, Z., Brew, K., and Lee, E. Y. C. (1996) Biochemistry 35, 6276-6282 [CrossRef][Medline] [Order article via Infotrieve]
  33. Pallen, C. J., and Wang, J. H. (1983) J. Biol. Chem. 258, 8550-8553 [Abstract/Free Full Text]
  34. Chan, C. P., Gallis, B., Blumenthal, D. K., Pallen, C. J., Wang, J. H., and Krebs, E. G. (1986) J. Biol. Chem. 261, 9890-9895 [Abstract/Free Full Text]
  35. Martin, B., Pallen, C. J., Wang, J. H., and Graves, D. J. (1985) J. Biol. Chem. 260, 14932-14937 [Abstract/Free Full Text]
  36. Barik, S. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 10633-10637 [Abstract]
  37. Butler-Ransohoff, J. E., Rokita, S. E., Kendall, D. A., Banzon, J. A., Carano, K. S., Kaiser, E. T., and Matlin, A. R. (1992) J. Org. Chem. 57, 142-145
  38. Mueller, E. G., Crowder, M. W., Averill, B. A., and Knowles, J. R. (1993) J. Am. Chem. Soc. 115, 2974-2975
  39. Day, E. P., David, S. S., Peterson, J., Dunham, W. R., Bonvoisin, J. J., Sands, R. H., and Que, L., Jr. (1988) J. Biol. Chem. 263, 15561-15567 [Abstract/Free Full Text]
  40. Swanson, S. K.-H., Born, T., Zydowsky, L. D., Cho, H., Chang, H. Y., Walsh, C. T., and Rusnak, F. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 3741-3745 [Abstract]
  41. Kirby, A. J., and Jencks, W. P. (1965) J. Am. Chem. Soc. 87, 3209-3216

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.