JBC Avanti Polar Lipids

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gamm, D. M.
Right arrow Articles by Uhler, M. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gamm, D. M.
Right arrow Articles by Uhler, M. D.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Volume 270, Number 45, Issue of November 10, 1995 pp. 27380-27388
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
The Type II Isoform of cGMP-dependent Protein Kinase Is Dimeric and Possesses Regulatory and Catalytic Properties Distinct from the Type I Isoforms (*)

(Received for publication, August 1, 1995)

David M. Gamm (1),  Sharron H. Francis (4),  Timothy P. Angelotti (2),  Jackie D. Corbin (4),  Michael D. Uhler (2) (3)(§)

From the  (1)Neuroscience Program, (2)Mental Health Research Institute and (3)Department of Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109 and the (4)Department of Molecular Physiology and Biophysics, Vanderbilt University School of Medicine, Nashville, Tennessee 37232

ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES

ABSTRACT

The type I cGMP-dependent protein kinases (cGK Ialpha and Ibeta) form homodimers (subunit M(r) 76,000), presumably through conserved, amino-terminal leucine zipper motifs. Type II cGMP-dependent protein kinase (cGK II) has been reported to be monomeric (M(r) 86,000), but recent cloning and sequencing of mouse brain cGK II cDNA revealed a leucine zipper motif near its amino terminus. In the present study, recombinant mouse brain cGK II was expressed, purified, and characterized. Sucrose gradient centrifugation and gel filtration chromatography were used to determine M(r) values for holoenzymes of cGK Ialpha (168,000) and cGK II (152,500), which suggest that both are dimers. Native cGK Ialpha possessed significantly lower K values for cGMP (8-fold) and beta-phenyl-1,N^2-etheno-cGMP (300-fold) than did recombinant cGK II. Conversely, the Sp- and Rp-isomers of 8-(4-chloro-phenylthio)-guanosine-3`,5`-cyclic monophosphorothioate demonstrated selectivity toward cGK II in assays of kinase activation or inhibition, respectively. A peptide substrate derived from histone f had a 20-fold greater V(max)/K ratio for cGK Ialpha than for cGK II, whereas a peptide based upon a cAMP response element binding protein phosphorylation site exhibited a greater V(max)/K ratio for cGK II. Finally, gel filtration of extracts of mouse intestine partially resolved two cGK activities, one of which had properties similar to those demonstrated by recombinant cGK II. The combined results show that both cGK I and cGK II form homodimers but possess distinct cyclic nucleotide and substrate specificities.

INTRODUCTION

Appreciation of cGMP as a distinct intracellular second messenger was closely followed by an intensive search for effector proteins in various organisms. Subsequently, a cGMP-dependent protein kinase (cGK) (^1)was discovered in arthropods(1) , leading to the eventual isolation of cGK from mammalian tissues(2, 3) . More recently, additional families of cGMP receptors have been described which include phosphodiesterases (4, 5) and ion channels(6, 7) . The existence of numerous intracellular cGMP receptors, as well as the restricted tissue distribution of cGK and the lack of well characterized physiological substrates, has hindered attempts to clearly define the physiological roles of cGK(8) . However, increases in cGMP have been shown to occur in response to physiological stimuli such as nitric oxide and natriuretic peptides(9, 10) . In certain tissues and cell types, elevation of cGMP is associated with a decrease in intracellular calcium levels, which can cause profound biological effects, including smooth muscle relaxation (11, 12) and inhibition of platelet aggregation(13) . Cyclic GMP-regulated processes have also been implicated in olfaction(7) , pancreatic enzyme and hormone secretion(14, 15) , intestinal chloride secretion(16) , and long-term potentiation in hippocampal neurons(17) .

Investigation of the cellular consequences of cGK activation is further complicated by the presence of multiple forms of the enzyme. The type I cGK (cGK I) is predominantly cytosolic and is present in relatively high concentrations in smooth muscle(18) , lung(19) , platelets(20) , and cerebellum(21) . Two isoforms of cGK I, cGK Ialpha and cGK Ibeta, have been purified and their amino acid sequences have been determined(3, 22) . These isoforms differ only in their amino termini and both form homodimers composed of 76-kDa subunits. Analyses of partial cDNA sequences of bovine cGK Ialpha and Ibeta (23) and the gene encoding human cGK Ibeta (24) provide further evidence that cGK Ialpha and Ibeta arise from an alternative mRNA splicing event(25) .

The type II cGK (cGK II) was first identified in rat and pig intestinal microvilli as an 86-kDa monomer associated with membrane fractions (26) . Molecular cloning of cGK II from a mouse brain (27) and rat intestine (28) cDNA library predicted a protein with 66 and 45% amino acid identity to the catalytic and cGMP binding domains of cGK I, respectively. In addition to intestine and brain, cGK II mRNA was found in mouse lung and kidney(27) , suggesting that cGK II may have diverse functions in multiple tissues. The limited homology between cGK I and II, along with their distinct mRNA tissue distributions, raised the possibility that these isoforms have different physical and biochemical properties, which may in turn result in functional differences.

Although initial studies suggested that cGK II is monomeric(26) , cGK II possesses an amino-terminal leucine zipper motif that is also found in the dimeric cGK I isoforms(8, 28) . Since physical interactions such as dimerization could influence the properties of cGK(29) , we sought to determine whether recombinant cGK II is a dimer using sucrose gradient centrifugation and gel filtration chromatography. In addition, comparisons of cGMP analog and peptide substrate specificities of cGK II and cGK Ialpha were performed. The results from these in vitro studies were then used to devise procedures to differentiate cGK II and cGK I activities from partially purified extracts of mouse intestine.

MATERIALS AND METHODS

Construction of cGK II Mammalian Expression Vectors, Transfection, and Purification

The pCMV.His(6)cGKII expression vector was constructed by polymerase chain reaction using the oligonucleotides GGAGATCTCCACCATGCACCATCACCATCATCATCATGGGAAATGGTTCAGTG and CTTGCTCTGCATTTCTTC (University of Michigan Biomedical Core Facilities) to generate a 220-base pair DNA fragment coding for the amino terminus of cGK II. This fragment was digested with BglII and SacI (Life Technologies, Inc.), isolated, and ligated into pCMV.cGKII (27) which had been BglII and SacI digested. The resulting construct, pCMV.His(6)cGKII, encodes a protein with the amino-terminal extension of Met-His-His-His-His-His-His. During construction of this vector, an error in the previously reported mouse sequence coding for amino acids 186-192 was discovered(27) . The previously reported codons (GGG-GAG-AAA-CTA-TCA-ACA-GGC) are actually GGG-AGA-AAC-TAT-CAA-CAG-GGG, which result in a change in amino acid sequence from GEKLSTG to GRNYQQ. This revision in amino acid sequence is in perfect agreement with the published rat cGK II sequence(28) .

For transient transfection, 15-cm plates of COS-1 or HEK-293 cells at 40% confluency were transfected using a calcium phosphate transfection method as described previously(27) . In later experiments, stably expressing HEK-293 cells were clonally isolated after selection with G-418 (750 µg/ml) (Life Technologies, Inc.).

For isolation of cGK II protein from transfected cells, homogenization buffer (500 µl) (27) containing 1 mM phenylmethylsulfonyl fluoride, 5 µg/ml leupeptin, and 1 µg/ml pepstatin A (Boehringer Mannheim) (Buffer A) was added to each 15-cm plate of pCMV.His(6)cGKII-transfected COS-1 or HEK 293 cells. The cells were scraped into separate tubes and sonicated for 5 s. After centrifuging the samples for 10 min, the supernatant was collected and imidazole (Sigma) was added to a final concentration of 10 mM. The samples were centrifuged once more and the supernatants containing His(6)-cGK II from COS-1 or HEK 293 cells were bound to nickel-affinity resin (Qiagen), washed with 8 column volumes of Buffer A containing 10 mM imidazole, and eluted with a continuous gradient of Buffer A containing 10-250 mM imidazole.

Construction of the cGK II Baculovirus Transfer Vector, Sf9 Cell Infection, and Purification

A 2.4-kilobase pair insert containing the mouse brain cGK II cDNA with an amino-terminal hexahistidine tag was released from the parent vector (pCMV.His(6)cGKII) by digestion with BglII. The cGK II insert was subcloned into the BglII site of the baculovirus transfer vector pBlueBac III (Invitrogen), creating the cGK II baculovirus transfer vector pBB.His(6)cGKII. Prior to protein production, recombinant baculovirus plaques were isolated and propagated using standard protocols as provided by the manufacturer (Invitrogen). Briefly, viruses and Spodoptera frugiperda (Sf9) cells were grown at 27 °C in TMN-FH media (PharMingen) containing 10% fetal calf serum (HyClone) or the serum-free media SFM 900 (Life Technologies, Inc.). Five days after transfection, the supernatant containing recombinant virions was collected and a single recombinant plaque was isolated by three rounds of plaque purification using agarose overlays. Recombinant plaques were identified by beta-galactosidase staining. The single recombinant plaque was propagated in a 100-ml spinner flask maintained at 27 °C. Final titer of recombinant virus was >1 times 10^8 as determined by a standard plaque assay. For protein production, 1 times 10^8 Sf9 cells were infected at a multiplicity of infection of 10. Forty-eight hours after infection, cells were collected by centrifugation, resuspended in Buffer A, and sonicated for 5 s. The expressed His(6)-cGK II protein was purified using nickel-affinity chromatography as described above. A portion of the His(6)-cGK II from both Sf9 and HEK 293 cells was dialyzed separately against a buffer containing 10 mM KPO(4), pH 7.2, which did not alter the biochemical properties of the enzyme. The His(6)-cGK II proteins expressed from Sf9 and HEK 293 cells were judged to be approximately 99% pure on silver-stained SDS-PAGE gels(30) . Protein concentrations were determined by amino acid analysis (University of Michigan Biomedical Core Facilities and Quality Controlled Biochemicals, Kovinton, MA). As is observed with the cGK I isoforms, measurements of cGK II concentration by the Bradford assay (31) using a bovine -globulin standard resulted in substantial overestimation. In contrast, protein assays using bicinchoninic acid (32) more closely approximated the true cGK II concentration.

Molecular Weight Determination of Recombinant cGK II and Native cGK Ialpha

Sucrose gradient centrifugation was performed on samples containing purified bovine lung cGK Ialpha (3.8 µg) (33) or cGK II (10 µg) in a final volume of 110 µl of KPEM buffer (10 mM potassium phosphate (pH 6.8), 1 mM EDTA, 25 mM beta-mercaptoethanol) containing 0.15 M NaCl and 140 µg of hemoglobin (Sigma). The individual samples were applied to the top of gradient tubes containing 13 ml of a linear sucrose gradient (5-20%) in KPEM buffer, 0.15 M NaCl. The gradient tubes were centrifuged in a Beckman SW41 rotor for 23 h at 4 °C. Fractions (1 ml) were collected from the bottom of each tube, assayed for cGK activity, and measured for absorbance at 280 nm(3) . In another experiment (not shown), phosphorylase b (8.2 S) (Sigma) was included with hemoglobin (4.6 S) as an additional internal standard in order to calculate the sedimentation coefficients of the cGK isoforms.

In addition, cGK Ialpha and cGK II were subjected to gel filtration chromatography using a standardized Sephacryl S-300 column (0.9 times 55.5 cm) (Pharmacia) equilibrated in KPEM buffer containing 0.1 M NaCl. Samples (225 µl) containing cGK Ialpha (3.4 µg) or cGK II (13 µg) and 3 mg of catalase (Stokes radius = 50 Å) (Sigma) were applied to the column and fractions (1 ml) were collected. Fractions were assayed for cGK activity (3) and absorbance at 280 nm. Stokes radii were determined as described by Wolfe et al.(3) . In another experiment (not shown), cGK II eluted at the same elution volume when thyroglobulin was used as an internal standard instead of catalase. The equation (34) used to calculate the molecular weights was:

where = viscosity of the medium; N = Avagadro's number; a = Stokes radius; S = sedimentation coefficient; = partial specific volume (0.738 for cGK II and 0.737 for cGK Ialpha); = density. The equation used to calculate the frictional ratios was:

where f = frictional coefficient of the sample and f(o) = frictional coefficient of a sphere (= 1.0).

[^3H]cGMP Dissociation Assay

In a sample volume of 60 µl, a final concentration of 0.03 mg/ml of the purified bovine lung cGK Ialpha or 0.045 mg/ml of recombinant cGK II was incubated with 550 µl of reaction mixture (KPE buffer (50 mM KH(2)PO(4), pH 6.8, 1 mM EDTA)) containing 0.5 mg/ml histone IIA (Sigma), 1 µM [^3H]cGMP (specific activity = 15-50 Ci/mmol) (Amersham Corp.), and 1 mg/ml bovine serum albumin. After a 1-h incubation at 30 °C (which was sufficient to saturate the binding sites), the tubes were cooled to 0-4 °C on ice. A 50-µl aliquot of the sample was added to 1 ml of cold KPE buffer and filtered for determination of the total [^3H]cGMP bound at zero time. A 100-fold excess of unlabeled cGMP was then added to the remaining mixture and 50-µl aliquots were removed at various times and added to 1 ml of KPE buffer. The sample was then filtered through a 0.45-µm nitrocellulose Millipore filter. After washing with 1 ml of KPE solution, the filter was dried and placed in a scintillation counting vial. Ten ml of Beckman Readysafe aqueous scintillant was added, and the vials were capped and shaken before counting. The time required for one-half of the bound [^3H]cGMP to dissociate (t) from each cGMP binding site of cGK Ialpha and cGK II were compared. Both cGK Ialpha and cGK II had approximately equal amounts of [^3H]cGMP bound to their respective low and high affinity cGMP binding sites under these conditions, which is consistent with saturation of the binding sites.

Determination of cGK Ialpha and cGK II K(a) Values for cGMP, cAMP, 1,N^2-PET-cGMP and Sp-8-pCPT-cGMPS

Increasing concentrations of cyclic nucleotide were added to separate tubes containing a phosphotransferase assay mixture consisting of 20 mM Tris (pH 7.5), 10 mM MgAc, 200 µM ATP, 11 nM [-P]ATP (ICN) (specific activity = 200-300 cpm/pmol), 10 mM NaF, 10 mM dithiothreitol, 0.2 mg/ml bovine serum albumin (Boehringer Mannheim), and 100 µM of the synthetic heptapeptide Arg-Lys-Arg-Ser-Arg-Ala-Glu (H2Btide) (35) as phosphate acceptor. The assay was initiated by addition of purified bovine lung cGK Ialpha (1.3 nM) or recombinant cGK II (5.8 nM) purified from HEK 293 or baculovirus-infected Sf9 cells. The phosphotransfer reaction was allowed to proceed for 30 min at 30 °C, at which time it was terminated by spotting aliquots onto P81 phosphocellulose papers (Whatman), which were washed in 10 mM phosphoric acid and counted. Enzyme activity is expressed as a percentage of the maximum micromoles of phosphate transferred per min per mg of cGK. The K(a) values for each cyclic nucleotide were determined by Eadie-Hofstee analysis (36, 37) of three to seven experiments.

Determination of the Rp-8-pCPT-cGMPS ICfor cGK Ialpha and cGK II

The cGK Ialpha (1.3 nM) or cGK II (5.8 nM) enzyme was preincubated for 10 min at 30 °C in the phosphotransferase assay mixture containing increasing concentrations of the cGK inhibitor Rp-8-pCPT-cGMPS. The reaction was initiated by the addition of 1 or 10 µM cGMP and allowed to proceed for 30 min at 30 °C, at which time the reaction was stopped as described above. IC values (constants representing 50% inhibition of activity) were obtained from three experiments. In a separate experiment, increasing concentrations of Rp-8-pCPT-cGMPS and 10 µM cGMP were both included in the phosphotransferase assay mixture and the reaction was initiated by the addition of enzyme. No difference was found between this experiment and those having a preincubation step.

Determination of K(m) and V(max)Values for Peptide Substrates of cGK Ialpha and cGK II

Increasing concentrations of peptide substrates were added to the phosphotransferase assay mixture in the presence of 20 µM cGMP. Subsequently, cGK Ialpha (1.3 nM) or cGK II (5.8 nM) was added to the mixture to initiate the reaction as described above. K(m) and V(max) values were determined by Eadie-Hofstee analysis of three to six experiments.

Analysis of cGK Activity in Mouse Intestine

Whole intestine from mouse (200 mg) was quick-frozen in liquid nitrogen, pulverized, and added to 500 µl of ice-cold Buffer A containing the following additional protease inhibitors (Sigma): chymostatin (0.1 µg/ml), aprotinin (2.0 µg/ml), phosphoramidon (1.1 µg/ml), E-64 (7.2 µg/ml), antipain (2.5 µg/ml), benzamidine (0.1 mM), and sodium metabisulfite (0.1 mM) (Buffer B). The suspension was sonicated for 10 s in an ice bath, centrifuged for 10 min at 4 °C, and the supernatant was removed. The pellet was resuspended in 500 µl of Buffer B containing 0.4% Triton X-100 and incubated on ice for 60 min. The resuspension was centrifuged 10 min at 4 °C and the resulting supernatant containing solubilized membrane proteins was removed and retained. An identical procedure was followed for isolation of solubilized membrane proteins from mouse lung. The intestine and lung samples (200 µl) were subjected to Sephacryl S-300 gel filtration chromatography as described above, with the exception that 1.2-ml fractions were collected. The gel filtration chromatography of mouse intestine was repeated using a second, independent sample with similar results. All gel filtration experiments in this study were performed using the same Sephacryl S-300 column, thereby allowing direct comparison of elution positions of cGK activity among experiments. Fractions were assayed at least twice for cGK activity using H2Btide as the phosphoacceptor as described previously(3) , with the exception that 20 µM protein kinase inhibitor peptide or 250 nM protein kinase inhibitor protein (38) was used to inhibit cAMP-dependent protein kinase activity. Intestinal fractions containing cGK activity that corresponded to the elution peaks of native bovine lung cGK Ialpha (fraction 18) and recombinant cGK II (fraction 21) were assayed twice with increasing concentrations of 1,N^2-PET-cGMP as described above to determine K(a) values. The phosphotransferase activities of the intestinal fractions were also assayed using the synthetic peptide Lys-Arg-Arg-Glu-Ile-Leu-Ser-Arg-Arg-Pro-Ser-Tyr-Arg (CREBtide) (39) as the phosphoacceptor and compared to the activities using H2Btide. This experiment was repeated three times. Enzyme activity is expressed as a percentage of the maximum micromoles of phosphate transferred to peptide substrate per min per ml of fraction.

RESULTS

In contrast to cGK I, little is known about the biochemical properties and physiological role of cGK II. In intestinal epithelial cells, cGK II may participate in the regulation of chloride absorption and secretion(40) . However, cGK II mRNA is also present in mouse brain, lung and, to a lesser extent, kidney(27) . The primary amino acid sequences of the rat intestine and mouse brain proteins are greater than 99% identical, with the slight difference likely representing species variation(27, 28) . Nonetheless, the physiological role of cGK II in these tissues is likely to differ. In the present study, critical physical and biochemical properties of cGK II and cGK I have been compared in vitro to determine whether differences exist that might suggest a distinct role for cGK II. Results from these experiments were then used to discriminate endogenous cGK activities in mouse tissue extracts.

Expression and Purification of cGK II

The cloning of cGK II cDNA from mouse brain (27) permitted the construction of the expression vector pCMV.His(6)cGKII and the baculovirus transfer vector pBB.His(6)cGKII, which produce hexahistidine-tagged cGK II in mammalian and insect cells, respectively. The plasmid pCMV.His(6)cGKII was stably transfected into HEK 293 cells and transiently transfected into COS-1 cells and the expressed cGK II protein was purified using nickel-affinity chromatography. An identical approach was used for the purification of cGK II from baculovirus-infected Sf9 cells (Fig. 1). Although earlier reports demonstrated that cGK II is membrane-associated(26) , the majority of the total cGK activities found in crude homogenates from the transfected HEK 293 and COS-1 cells and infected Sf9 cells were retained in the soluble fraction (data not shown).

Figure 1: Silver-stained SDS-PAGE of recombinant mouse brain cGK II and bovine lung cGK Ialpha. Recombinant cGK II was purified from HEK 293, COS-1, and Sf9 cells as described under ``Materials and Methods.'' The cGK II lane shown here was loaded with 300 ng of cGK II purified from Sf9 cells. The cGK Ialpha lane was loaded with 300 ng of bovine lung cGK Ialpha purified as described previously (33) . The gel was stained using the Bio-Rad Silver Stain Plus Kit. Protein standard molecular masses in kDa are indicated to the left.


Using histone f as substrate and a Mg level of 30 mM, a substrate V(max) of 1.6 µmol/minbulletmg was obtained for cGK II isolated from HEK 293 and Sf9 cells, which is nearly identical to the value obtained for cGK Ialpha purified from bovine lung (V(max) = 1.8 µmol/minbulletmg) (data not shown). Under similar conditions, de Jonge (26) obtained a V(max) of 1.5-2.0 µmol/minbulletmg for cGK II purified from rat intestinal brush-borders and a V(max) of 2.5-3.0 µmol/minbulletmg for bovine lung cGK I. These results suggest that the recombinant and native cGK II enzymes possess similar activities. Preparations of cGK II from both HEK 293 and Sf9 cells were used to determine kinetic constants in this study.

Molecular Weight Determination of cGK II

Since the amino-terminal dimerization domains of the cGK I isoforms may affect biochemical properties such as cGMP activation(3, 41) , we investigated the possibility that cGK II is also dimeric using enzyme expressed from COS-1 cells. On a 5-20% linear sucrose gradient, recombinant cGK II peak activity sedimented slightly faster than did cGK Ialpha when compared with hemoglobin as the internal marker (Fig. 2, A and B). Using phosphorylase b (8.2 S) and hemoglobin (4.6 S) as internal standards (data not shown), sedimentation coefficients of cGK Ialpha and cGK II were determined to be 7.8 S and 8.8 S, respectively (Table 1). The S value obtained for cGK Ialpha is similar to that determined previously(3, 42) .

Figure 2: Sucrose density gradient centrifugation of cGK Ialpha and cGK II. A, a mixture containing 3.8 µg of cGK Ialpha () and 140 µg of hemoglobin () or B, 10 µg of cGK II () and 140 µg of hemoglobin () was applied to a linear sucrose gradient (5-20%) and centrifuged for 23 h at 37,000 rpm. One-ml fractions were removed beginning at the bottom of each gradient tube and assayed for cGK activity and absorbance at 280 nm as described under ``Materials and Methods.'' In a separate experiment (not shown), phosphorylase b was also included with hemoglobin and cGK Ialpha or cGK II to facilitate calculation of sedimentation coefficients (Table 1). The results shown are representative of three experiments.


On a standardized Sephacryl S-300 gel filtration column, cGK Ialpha eluted virtually identically with the internal standard catalase (Fig. 3A), but recombinant cGK II eluted from the same column at a significantly higher volume (Fig. 3B). This is apparently not due to an artifact of the catalase internal standard since cGK II eluted at a similar position when thyroglobulin was used as an internal standard (data not shown). Proteins of known Stokes radii were used to construct a standard curve (not shown), which indicated that the Stokes radius of cGK II (40 Å) is smaller than that of cGK Ialpha (50 Å) (Table 1), even though the subunit M(r) of cGK II (86,000) is greater than that of cGK Ialpha (76,000). The differences in behavior of cGK Ialpha and cGK II by gel filtration is explained by a relatively lower frictional ratio calculated for cGK II (f/f(o) = 1.08) compared to cGK Ialpha (f/f(o) = 1.42). Despite the difference in Stokes radii between cGK Ialpha and cGK II, the calculated M(r) of cGK II (152,500) is similar to the calculated M(r) of cGK Ialpha (168,000), which predicts that both isoforms form homodimers. This dimerization likely occurs via interchain interactions involving an amino-terminal region which contains a consensus leucine zipper motif(43) . As both the regulatory and catalytic domains of cGK II are present on a single subunit, such protein-protein interactions involving the amino terminus are likely to influence biochemical properties such as cyclic nucleotide binding and activation(8, 29) .

Figure 3: Gel filtration chromatography of cGK Ialpha and cGK II. A, a mixture of cGK Ialpha (3.4 µg) () and catalase (3 mg) () or B, cGK II (13 µg) () and catalase (3 mg) () was chromatographed on a Sephacryl S-300 gel filtration column. Fractions of 1 ml were collected and cGK activity and absorbance at 280 nm were measured as described under ``Materials and Methods.'' The Stokes radii of cGK Ialpha and cGK II (Table 1) were determined from a standard curve (not shown) of proteins with known Stokes radii (apoferritin, 59 Å; phosphorylase b, 55 Å; aldolase, 52 Å; catalase, 50 Å; bovine serum albumin, 47.5 Å).


Dissociation of [^3H]cGMP from cGK II and cGK Ialpha

Dissociation of [^3H]cGMP from cGK Ialpha or Ibeta was shown in earlier reports to be biphasic due to the presence of two classes of cyclic nucleotide binding sites: one high affinity and one low affinity site per cGK monomer(44, 45) . Furthermore, the observation that dimeric cGK Ialpha is partially (50%) activated when only the high affinity binding sites are occupied by cGMP suggests that cGK Ialpha may respond differentially over a wide range of cGMP levels in vivo(44, 45) . To test whether the cGK II dimer also possesses two distinct dissociation t values for its cyclic nucleotide binding sites, the time course of displacement of [^3H]cGMP from cGK II by cGMP was measured and compared to that of cGK Ialpha (Fig. 4). The results demonstrate that cGK II also exhibits a biphasic pattern of [^3H]cGMP dissociation, although the absolute t values for the two sites in cGK II differ significantly from those in cGK Ialpha. The t for dissociation from the most rapidly dissociating site in cGK Ialpha was 0.5 min compared to 3 min for cGK II. During the second, slow dissociation phase, the t for [^3H]cGMP dissociation from cGK Ialpha was 25 min compared to 14 min for cGK II. These results imply that cGK II, unlike cGK Ialpha, can only exist in a partially cGMP-saturated state over a narrow range of cGMP concentrations. Furthermore, the more rapid dissociation of cGMP from the slowly dissociating (high affinity) site of cGK II compared to that of cGK Ialpha could explain in part the higher K(a) of cGK II.

Figure 4: [^3H]cGMP dissociation behavior of cGK Ialpha and cGK II. A final concentration of 0.03 mg/ml cGK Ialpha (bullet) or 0.045 mg/ml cGK II (circle) was incubated in a [^3H]cGMP binding mixture with 1.0 µM [^3H]cGMP for 1 h at 30 °C to allow saturation of cGMP binding sites with [^3H]cGMP. After 50-µl aliquots were removed for total binding (B) determination, the tubes were placed on ice, a 100-fold excess of unlabeled cGMP was added, and 50-µl aliquots were taken at different time points for binding (B) determination (see ``Materials and Methods''). Dissociation t values obtained from these experiments are discussed in the text.


Comparison of cGK II and cGK Ialpha Activation and Inhibition by Cyclic Nucleotide Analogs

Using a battery of cyclic nucleotide analogs, Sekhar et al.(46) demonstrated marked differences in the activation constants of cGK Ialpha and Ibeta despite the fact that the amino acid sequences of the cGMP binding sites in these isoforms are identical. Therefore, the activation or inhibition constants of various cyclic nucleotides and cyclic nucleotide analogs were also determined for cGK II using cGK Ialpha as a control. In these experiments, cGK Ialpha possessed an 8.7-fold lower K(a) value (0.092 µM) for cGMP than did cGK II (K(a) = 0.80 µM) (Fig. 5A, Table 2), suggesting that, like cGK Ibeta, higher concentrations of cGMP may be required in vivo to fully activate cGK II relative to cGK Ialpha.

Figure 5: Activation or inhibition of cGK Ialpha and cGK II by cyclic nucleotides and cyclic nucleotide analogs. The activities of cGK Ialpha (1.3 nM) (open symbols) and cGK II (5.8 nM) (filled symbols) were measured in the presence of increasing concentrations of cGMP (circle, bullet) or 1,N^2-PET-cGMP (down triangle, ) (A). Likewise, activation of cGK Ialpha (circle) and cGK II (bullet) by Sp-8-pCPT-cGMPS was examined (B). Activity was determined as described under ``Materials and Methods'' and expressed as the percentage of the highest cGK activity obtained. Inhibition of cGK Ialpha (circle) and cGK II (bullet) by Rp-8-pCPT-cGMPS was also determined (see ``Materials and Methods'') (C). Activity in these experiments was expressed as the percentage of cGK activity (1 µM cGMP) in the absence of inhibitor. The experiments were performed three to seven times for each cyclic nucleotide. Average K or IC values and measurements of error for each cyclic nucleotide are reported in Table 2.


In this study, the cGMP analog most selective for cGK Ialpha was beta-phenyl-1,N^2-etheno-cGMP (1,N^2-PET-cGMP), which exhibited a K(a) value (0.016 µM) 300-fold lower for cGK Ialpha than for cGK II (K(a) = 4.7 µM) (Fig. 5A, Table 2). Previous studies have shown that 1,N^2-PET-cGMP is also a potent activator of cGK Ibeta (K(a) = 20 nM)(46) , suggesting that this cGMP analog is generally selective for the cGK I isoforms versus cGK II. Other cGMP analogs that preferentially activated cGK Ialpha over cGK II include 8-iodo-1,N^2-PET-cGMP (200-fold) and 8-(2,4-dihydroxyphenylthio)-cGMP (19-fold) (data not shown). However, these analogs are substantially less effective activators of cGK Ibeta than cGK Ialpha (46) and would be less likely to selectively distinguish between cGK II and cGK Ibeta.

Only one cyclic nucleotide analog examined, the Sp-isomer of 8-(4-chlorophenylthio)-guanosine-3`,5`-cyclic monophosphorothioate (Sp-8-pCPT-cGMPS), yielded a lower (5-fold) K(a) value for cGK II than for cGK Ialpha (Fig. 5B, Table 2). The corresponding Rp-isomer of 8-pCPT-cGMPS was tested for its ability to inhibit activation of cGK II and cGK Ialpha by cGMP. The utility of this analog was demonstrated in a previous report which showed that Rp-8-pCPT-cGMPS could selectively inhibit cGK activity in intact human platelets(47) . In our study, the IC value of Rp-8-pCPT-cGMPS for cGK II was 114-fold lower than that for cGK Ialpha in the presence of 1 µM cGMP (Fig. 5C, Table 2). In the presence of 10 µM cGMP, the IC values of Rp-8-pCPT-cGMPS for cGK Ialpha and cGK II increased 3- and 11-fold, respectively (data not shown). Furthermore, millimolar concentrations of Rp-8-pCPT-cGMPS fully activated both enzymes (data not shown), a phenomenon that has been observed previously for cGK II(48) . Since substitutions at the 8-position of the guanine ring are poorly tolerated by cGK Ibeta(46) , these results suggest that the Sp- or Rp-isomer of 8-pCPT-cGMPS may be a selective activator or inhibitor, respectively, of cGK II compared to the cGK I isoforms.

Even though the cyclic nucleotide binding sites of the cGK I isoforms are relatively specific for cGMP, cAMP has been shown to cross-activate cGK in intact pig coronary arteries(49) . Therefore, the cAMP activation constants of cGK II and cGK Ialpha were compared to determine whether cGK II might also be a candidate for cross-activation in vivo. The K(a) value for cGK Ialpha was 6.4-fold lower than that for cGK II (Table 2), suggesting that cyclic nucleotide cross-activation of cGK II would require significantly higher cAMP levels in vivo.

Comparison of Peptide Substrate Apparent Kinetic Constants for cGK II and cGK Ialpha

Despite numerous attempts to find natural substrates for cGK in different species and tissues, there are only a few well characterized proteins that have been shown to be preferentially phosphorylated by cGK either in vitro or in vivo(50) . The phosphorylation sites of some of these potential substrates have been mapped, allowing the isolation or synthesis of peptides for use in studies of catalytic site specificity. Such studies offer insight into the determinants of substrate phosphorylation and, like cyclic nucleotide analog studies, potentially provide reagents for selective examination of a particular enzyme or isozyme. Five such peptides were examined in this study for their relative specificity for cGK Ialpha and cGK II. The peptide most often used in cGK studies, H2Btide (RKRSRAE), is derived from histone f and was shown by Glass and Krebs (35) to possess a 20-fold higher V(max)/K(m) ratio for cGK than for cAK. In our study, the K(m) value of synthetic H2Btide for cGK II was 5.5-fold higher than that for cGK Ialpha (Fig. 6A, Table 3). Furthermore, the V(max) value of H2Btide for cGK II was 3.7-fold lower than that for cGK Ialpha. Therefore, comparison of the V(max)/K(m) ratio of cGK II versus cGK Ialpha yields a cGK II/cGK Ialpha specificity index of 0.05, indicating that H2Btide is a poor substrate of cGK II relative to cGK Ialpha.

Figure 6: Phosphorylation of peptide substrates by cGK Ialpha and cGK II. The activities of cGK Ialpha (1.3 nM) (open symbols) and cGK II (5.8 nM) (filled symbols) were measured in the presence of increasing concentrations of the synthetic peptide substrates H2Btide (circle, bullet) and CREBtide (down triangle, ) (A). Additional assays were performed to examine the kinetics of phosphorylation of IP(3)Rtide (down triangle), Kemptide (circle) and BPDEtide (bullet) by cGK Ialpha (B) or cGK II (C). Activity was determined as described under ``Materials and Methods'' and expressed as the percentage of the highest cGK activity obtained. The experiments were performed three to six times for each peptide substrate. Average K and V(max) values and measurements of error for each peptide substrate are reported in Table 3.


A second protein that has been investigated as a potential substrate for cGK is the cAMP response element binding protein (CREB)(39) . CREBtide (KRREILSRRPSYR) corresponds to an amino acid sequence in CREB that contains a site rapidly phosphorylated by cAK. In contrast, cGK I phosphorylated CREBtide at a much slower rate than did cAK, although they possessed similar K(m) values(39) . In our study, cGK II had a 2.3-fold lower K(m) value for CREBtide than did cGK Ialpha, while the corresponding V(max) values for both enzymes were identical (Fig. 6A, Table 3). However, with a cGK II/cGK Ialpha specificity index of 2.3, CREBtide was the most selective substrate for cGK II tested and therefore a substrate of choice for detecting cGK II activity.

Three additional synthetic peptide substrates, IP(3)Rtide, BPDEtide, and Kemptide, were also analyzed for their relative selectivities for cGK Ialpha or cGK II (Fig. 6, B and C, Table 3). IP(3)Rtide (GRRESLTSFG) is derived from a sequence in the inositol 1,4,5-trisphosphate receptor (IP(3)R) which is phosphorylated on the same serine residue in vitro by both cGK I and cAK(51) . It has been suggested that phosphorylation of IP(3)R may mediate some of the effects on intracellular calcium observed with cGK stimulation(51) . Comparison of IP(3)Rtide as a substrate for cGK Ialpha or cGK II revealed a 4.8-fold lower K(m) value and a 4.4-fold lower V(max) value for cGK II, suggesting that IP(3)Rtide is an equivalent substrate for both enzymes. BPDEtide (RKISASEFDRPLR) is derived from the sequence of the bovine lung cGMP-binding cGMP-specific phosphodiesterase (cG-BPDE) that was determined to be selective for cGK as compared to cAK(52) . Kemptide (LRRASLG) (53) is derived from pyruvate kinase and is commonly employed for cAK measurement, although it can also be phosphorylated by cGK I. A 1.9-fold lower K(m) value for BPDEtide was obtained with cGK II compared to cGK Ialpha, whereas the K(m) value for Kemptide was not significantly different between the cGK isoforms. However, the lower cGK II V(max) values found using these substrates results in cGK II/cGK Ialpha specificity indices that favor cGK Ialpha. Overall, the rank order of selectivity for cGK II using these substrates was CREBtide > IP(3)Rtide > BPDEtide > Kemptide H2Btide. Therefore, assays of cGK activity using impure enzyme preparations and H2Btide as the phosphoacceptor peptide may fail to detect cGK II activity. Results from these experiments suggest that, unlike the cGK I isoforms, cGK Ialpha and cGK II interact with substrates differently, which is likely due to amino acid differences within the catalytic domains of cGK Ialpha and cGK II.

Partial Purification and Preliminary Identification of cGK Activities from Mouse Intestine

In order to study native cGK II, we utilized selected cyclic nucleotides and peptide substrates to discriminate between cGK I and cGK II from mouse tissue extracts. The cGK activities were partially resolved by Sephacryl S-300 gel filtration of solubilized membrane proteins from mouse intestine and lung. Whole intestine possesses both cGK I within smooth muscle cells as well as cGK II within the microvilli of epithelial cells(26, 28) . In contrast, only cGK I protein has been demonstrated in lung thus far, although cGK II mRNA has been detected(27) . Intestinal cGK II is membrane-bound(26, 28) , but the majority of cGK I is cytoplasmic and therefore is retained in the soluble fraction during extract preparation (see ``Materials and Methods''). However, some cGK I is likely to be found in the solubilized membrane protein fraction via membrane association and/or contamination from the soluble protein fraction.

Kinase assays were conducted in the presence of protein kinase inhibitor peptide and H2Btide substrate using the mouse lung and intestine fractions obtained from the gel filtration column (Fig. 7A). Assay of the lung fractions revealed a single peak of kinase activity at an elution volume consistent with purified cGK Ialpha (see Fig. 3A). Assay of the intestine fractions, on the other hand, revealed a single peak of kinase activity that had a prominent shoulder of activity on the leading edge. The elution position of the peak kinase activity from intestine corresponds to the elution position of purified recombinant cGK II (Fig. 3B), whereas the shoulder arose at a position comparable to that of cGK Ialpha. Autophosphorylation experiments performed on the intestine fractions confirmed the existence of a predominant 76-kDa species (which corresponds with the SDS-PAGE mobility of cGK Ialpha) in the early shoulder fractions (data not shown).

Figure 7: Apparent separation and identification of cGK Ialpha and cGK II enzyme activities from extracts of mouse intestine. A, extracts containing solubilized membrane proteins from mouse whole intestine (bullet) or lung (down triangle) were chromatographed on the same Sephacryl S-300 gel filtration column used for Fig. 3. The gel filtration chromatography was repeated using a second, independent sample of mouse intestine (not shown). Fractions (1.2 ml) collected from the column were assayed at least twice for kinase activity using the heptapeptide substrate H2Btide in the presence of the protein kinase inhibitor peptide (see ``Materials and Methods''). Activity for each fraction was expressed as the percentage of maximal kinase activity obtained. The hatched bar spans the elution volume where the majority of the purified cGK Ialpha activity was found in Fig. 3. The open bar spans the elution volume where the majority of the purified recombinant cGK II activity was found in Fig. 3. The asterisk and star demarcate intestine fractions 18 and 21, respectively, which were further assayed in B. B, mouse intestine fractions 18 (circle) and 21 (bullet) (see A), which correspond to the elution profiles of purified cGK Ialpha and purified recombinant cGK II, respectively, were assayed in the presence of increasing concentrations of 1,N^2-PET-cGMP as described under ``Materials and Methods.'' This experiment was repeated with similar results. C, the same fractions from the intestinal extract obtained in A were reassayed using either H2Btide (box) or CREBtide () as phosphoacceptors (see ``Materials and Methods''). This experiment was performed three times with similar results. Note that again, activity is expressed as a percentage of maximal kinase activity rather than as absolute kinase activity in order to better discriminate the elution profile obtained with H2Btide.


To further establish the identity of the kinases responsible for the intestine peak and shoulder activities, samples from a shoulder fraction (fraction 18) and the peak fraction (fraction 21) were assayed separately in the presence of increasing concentrations of 1,N^2-PET-cGMP (Fig. 7B). Fraction 18 exhibited a K(a) value of 0.01 µM, similar to the K(a) value of purified cGK Ialpha (0.016 µM, Table 2), whereas fraction 21 displayed a K(a) value of 1.0 µM, which approaches the value obtained with purified recombinant cGK II (4.7 µM, Table 2).

Last, the intestine column fractions used in Fig. 7A were assayed again using either 100 µM H2Btide or 100 µM CREBtide as substrate. According to results shown in Fig. 6A and listed in Table 3, CREBtide should be a significantly better substrate than H2Btide for cGK II at this concentration. The peak cGK activity was 9.6-fold higher in the presence of CREBtide than H2Btide, whereas the shoulder activity was essentially eliminated (Fig. 7C). It should be noted that Fig. 7C is plotted as a percentage of control kinase activity in order to display the results obtained using both H2Btide and CREBtide. Data from these studies indicate that cGK Ialpha and cGK II activity can be distinguished using selected cyclic nucleotide analogs and substrates. Together, these results suggest that endogenous dimeric cGK II and cGK I are responsible for the observed intestine peak and shoulder activities, respectively.

DISCUSSION

In an earlier study, cGK II isolated from intestine appeared to be particulate and monomeric(26) . However, cloning of the cGK II cDNA failed to reveal divergent regions containing significant hydrophobicity to account for the observed membrane association(28) . In addition, we found cGK II expressed in COS-1, HEK 293, and Sf9 cells to be predominantly soluble. Nonetheless, Jarchau et al.(28) reported that recombinant cGK II expressed in HEK 293 cells partitioned to the particulate fraction. The discrepancy in cGK II subcellular localization may be explained by differences in expression or extraction conditions. Also, the possibility exists that cGK II associates with membranes by interacting with specific anchoring protein(s) which may not be present in some cell lines(27) .

Previous reports have shown that cGK Ialpha and cGK Ibeta, which differ only in their amino termini, display distinct cyclic nucleotide analog specificities, K(a) values and Hill coefficients, which may in turn produce physiological consequences(3, 8) . In addition, monomerization of cGK Ibeta by proteolysis of the dimerization domain resulted in a cGMP-dependent enzyme with an increased K(a) and altered cGMP dissociation rates(29) . Therefore, conformational changes that occur upon dimerization are likely to influence cGK I activity(8, 54) . This consideration of the possible consequences of protein-protein binding, as well as the presence of a conserved leucine zipper motif in cGK II, led us to investigate whether cGK II dimerizes prior to embarking on kinetic studies of the enzyme. It should be noted that the Stokes radius determined with the recombinant brain cGK II is significantly different from that reported previously for intestinal cGK II(26) . It is possible that high expression of recombinant cGK II results in artifactual dimerization. However, both recombinant and native intestinal cGK II exhibited identical gel filtration properties in this study. An even greater discrepancy is noted for the sedimentation coefficient of cGK II, which was reported earlier to be 5.1 S (26) compared with our value of 8.8 S. Values for cGK Ialpha of 6.9 S (26) and 7.8 S (this study) are in better agreement. The finding of a lower frictional ratio for cGK II than for cGK Ialpha is indicative of a more globular form for cGK II. Our conclusion that cGK II is dimeric has direct bearing on the interpretation of subsequent results described in this study. Wolfe et al.(45) found that occupation of only one cyclic nucleotide binding site by cGMP per cGK I dimer failed to significantly activate the enzyme, whereas binding of two cGMP molecules to identical sites in each polypeptide resulted in 50% activation. Therefore, it is clear that important interchain effects can occur in dimeric cGK proteins.

Since the cGK I isoforms and cGK II possess relatively similar K(a) values, more selective cGMP analogs are needed to facilitate in vivo studies of isoform function. Our results demonstrate that 1,N^2-PET-cGMP is strongly selective for cGK I, whereas Sp-8-pCPT-cGMPS preferentially activates cGK II. Based on these data, 1,N^2-PET-cGMP is likely to discriminate between cGK Ialpha or Ibeta and cGK II in crude cell or tissue extracts and in intact tissues. Conveniently, both analogs are membrane permeant and Sp-8-pCPT-cGMPS is known to be resistant to mammalian cyclic nucleotide phosphodiesterases, although it is also a potent activator of cAMP-dependent protein kinase(55, 56) .

Cyclic GMP may not be the only cyclic nucleotide that activates cGK enzymes in vivo(49) . Even so, our data indicate that cAMP is a significantly less potent activator of cGK II compared to cGK Ialpha, implying that cGK II would require much higher cAMP levels for cross-activation. Whether such cAMP concentrations exist transiently in certain cellular compartments has not been determined. However, since autophosphorylation of cGK I can increase cAMP affinity 6- to 10-fold (57, 58) , more thorough studies of the effects of cGK II autophosphorylation are required to determine the likelihood of cross-activation.

Results from the substrate phosphorylation experiments indicate that CREBtide is the most selective substrate for detecting cGK II activity relative to cGK Ialpha. Conversely, H2Btide is by far the least selective substrate of cGK II. From these studies, it is possible to make inferences about the preferred substrate phosphorylation sequence of cGK II. Previous reports have established the importance of a series of basic residues located amino-terminal to the phosphorylation (p) site in cGK I substrates(50) . In addition, a basic residue located at the substrate p+1 position (50) and a phenylalanine at the p+4 position appear to provide selectivity for cGK I relative to cAMP-dependent protein kinase(59) . From our studies, cGK II seems to favor a typical cAMP-dependent protein kinase consensus phosphorylation sequence (Arg-Arg-X-Ser/Thr). Alternatively, our results could be explained if a basic residue positioned at the p+1 site acts as a negative determinant for cGK II phosphorylation.

The preceding studies provide evidence that cGK Ialpha and cGK II are dimeric and distinct in both their regulation by cyclic nucleotides and their recognition of substrates. Using the criteria determined herein, extracts of whole mouse intestine were shown to contain two separable cGK activities. The fractions containing these cGK activities have enzymatic properties consistent with those of cGK I and cGK II, respectively. These kinase activities were not inhibited by the cAMP-dependent protein kinase-specific protein kinase inhibitor peptide or protein, and they eluted from a standardized gel filtration column at positions corresponding to purified dimeric cGK Ialpha and cGK II.

Identification of the cGK I and cGK II activities was accomplished in part through assays of cGMP-dependent phosphorylation using either 100 µM H2Btide or 100 µM CREBtide. At this concentration, cGK I should maximally phosphorylate both substrates, although the lower V(max) value of cGK Ialpha for CREBtide resulted in diminished overall activity. Conversely, cGK II phosphorylated 100 µM CREBtide to a much greater extent than 100 µM H2Btide, as predicted from the previous experiments (Fig. 6A). This relative selectivity may be further enhanced by using the cGMP analogs that preferentially activate either cGK I (1,N^2-PET-cGMP) or cGK II (Sp-8-pCPT-cGMPS), or possibly by using Rp-8-pCPT-cGMPS, which is a more selective inhibitor of cGK II. These combinations of cyclic nucleotide analog and peptide substrate should be useful in the characterization of cGK activities in other tissues, including lung and brain. Indeed, a recent study demonstrated the presence of over 40 selective substrates of cGK in rat brain extracts(60) . Since rat brain contains both cGK I and cGK II mRNA(61) , it would be of interest to further characterize such substrates in terms of their isoform selectivity using the cyclic nucleotide analogs described herein. Ultimately, these tools may prove helpful in the investigation of the in vivo functions of the cGK isoforms as well.

FOOTNOTES

*
This work was supported by National Institutes of Health Grants GM 38788 and MH 42652 (to M. D. U.) and DK 40029 (to J. D. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

§
To whom correspondence should be addressed: Neuroscience Laboratories Building, 1103 E. Huron St., University of Michigan, Ann Arbor, MI 48104-1687. Tel.: 313-747-3172; Fax: 313-936-2690.

(^1)
The abbreviations used are: cGK, cGMP-dependent protein kinase; 1,N^2-PET-cGMP, beta-phenyl-1,N^2-etheno-cGMP; Sp-8-pCPT-cGMPS, Sp-8-(4-chlorophenylthio)-guanosine-3`,5`-cyclic monophosphorothioate; Rp-8-pCPT-cGMPS, Rp-8-(4-chlorophenylthio)-guanosine-3`,5`-cyclic monophosphorothioate; H2B, histone f; CREB, cAMP response element binding protein; IP(3)R, inositol 1,4,5-trisphosphate receptor; BPDE, bovine lung cGMP-binding cGMP-specific phosphodiesterase; E-64, trans-epoxysuccinyl-L-leucylamido-(4-guanidino)butane; PAGE, polyacrylamide gel electrophoresis; CAK, cAMP-dependent protein kinase.

ACKNOWLEDGEMENTS

We thank Adele Barres for assistance in the preparation of this manuscript and Linda Harper for her tissue culture expertise.

REFERENCES

  1. Kuo, J. F., and Greengard, P. (1970) J. Biol. Chem. 245, 2493-2498 [Abstract/Free Full Text]
  2. Lincoln, T. M., Hall, C. L., Park, C. R., and Corbin, J. D. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 2559-2563 [Abstract/Free Full Text]
  3. Wolfe, L., Corbin, J. D., and Francis, S. H. (1989) J. Biol. Chem. 264, 7734-7741 [Abstract/Free Full Text]
  4. Beavo, J. A. (1988) Adv. Second Messenger Phosphoprotein Res. 22, 1-38 [Medline] [Order article via Infotrieve]
  5. Francis, S. H., Lincoln, T. M., and Corbin, J. D. (1980) J. Biol. Chem. 255, 620-626 [Free Full Text]
  6. Fesenko, E. E., Kolesnikov, S. S., and Lyubarsky, A. L. (1985) Nature 313, 310-313 [CrossRef][Medline] [Order article via Infotrieve]
  7. Nakamura, T., and Gold, G. H. (1987) Nature 325, 442-444 [CrossRef][Medline] [Order article via Infotrieve]
  8. Francis, S. H., and Corbin, J. D. (1994) Adv. Pharm. 26, 115-170
  9. Tremblay, J., Gerzer, R., and Hamet, P. (1988) Adv. Second Messenger Phosphoprotein Res. 22, 319-383 [Medline] [Order article via Infotrieve]
  10. Waldman, S. A., Rapoport, R. M., and Murad, F. (1984) J. Biol. Chem. 259, 14332-14334 [Abstract/Free Full Text]
  11. Kobayashi, S., Kanaide, H., and Nakamura, M. (1985) Science 229, 553-556 [Abstract/Free Full Text]
  12. Lincoln, T. M. (1989) Pharmacol. Ther. 41, 479-502 [CrossRef][Medline] [Order article via Infotrieve]
  13. Haslam, R. J. (1987) in Thrombosis and Haemostasis (Verstraete, M., Vermyleu, V., Lijnen, R., and Arnout, J., eds) pp. 147-174, University Press, Leuven
  14. Rogers, J., Hughes, R. J., and Mathews, E. K. (1988) J. Biol. Chem. 263, 3713-3719 [Abstract/Free Full Text]
  15. Schmidt, H. H., Warner, T. D., Ishii, K., Sheng, H., and Murad, R. (1992) Science 255, 721-723 [Abstract/Free Full Text]
  16. Forte, L. R., Thorne, P. K., Eber, S. L., Krause, W. J., Freeman, R. H., Francis, S. H., and Corbin, J. D. (1992) Am. J. Physiol. 263, C607-C615
  17. Schuman, E. M., and Madison, D. V. (1991) Science 254, 1503-1506 [Abstract/Free Full Text]
  18. Francis, S. H., Noblett, B. D., Todd, B. W., Wells, J. N., and Corbin, J. D. (1988) Mol. Pharmacol. 34, 506-517 [Abstract]
  19. Kuo, J. F. (1975) Proc. Natl. Acad. Sci. U. S. A. 72, 2256-2259 [Abstract/Free Full Text]
  20. Walter, U. (1989) Rev. Physiol. Biochem. Pharmacol. 113, 42-88
  21. Lohmann, S. M., Walter, U., Miller, P. E., Greengard, P., and DeCamilli, P. (1981) Proc. Natl. Acad. Sci. U. S. A. 78, 653-657 [Abstract/Free Full Text]
  22. Takio, K., Wade, R. D., Smith, S. B., Krebs, E. G., Walsh, K. A., and Titani, K. (1984) Biochemistry 23, 4207-4218 [CrossRef][Medline] [Order article via Infotrieve]
  23. Wernet, W., Flockerzi, V., and Hofmann, F. (1989) FEBS Lett. 251, 191-196 [CrossRef][Medline] [Order article via Infotrieve]
  24. Orstavik, S., Sandberg, M., Berube, D., Natarajan, V., Simard, J., Walter, U., Gagne, R., Hansson, V., and Jahnsen, T. (1992) Cytogenet. Cell Genet. 59, 270-273 [Medline] [Order article via Infotrieve]
  25. Francis, S. H., Woodford, T. A., Wolfe, L., and Corbin, J. D. (1988-1989) Second Messengers Phosphoproteins 12, 301-310
  26. de Jonge, H. R. (1981) Adv. Cyclic Nuc. Res. 14, 315-333 [Medline] [Order article via Infotrieve]
  27. Uhler, M. D. (1993) J. Biol. Chem. 268, 13586-13591 [Abstract/Free Full Text]
  28. Jarchau, T., Hausler, C., Markert, T., Pohler, D., Vandekerckhove, J., de Jonge, H. R., Lohmann, S. M., and Walter, U. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 9426-9430 [Abstract/Free Full Text]
  29. Wolfe, L., Francis, S. H., and Corbin, J. D. (1989) J. Biol. Chem. 264, 4157-4162 [Abstract/Free Full Text]
  30. Gottlieb, M., and Chavko, M. (1987) Anal. Biochem. 165, 33-37 [CrossRef][Medline] [Order article via Infotrieve]
  31. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 [CrossRef][Medline] [Order article via Infotrieve]
  32. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985) Anal. Biochem. 150, 76-85 [CrossRef][Medline] [Order article via Infotrieve]
  33. Francis, S. H., Wolfe, L., and Corbin, J. D. (1991) Methods Enzymol. 200, 332-341 [Medline] [Order article via Infotrieve]
  34. Siegel, L. M., and Monty, K. J. (1966) Biochim. Biophys. Acta 112, 346-362 [Medline] [Order article via Infotrieve]
  35. Glass, D. B., and Krebs, E. G. (1982) J. Biol. Chem. 257, 1196-1200 [Abstract/Free Full Text]
  36. Eadie, G. S. (1942) J. Biol. Chem. 146, 85-93 [Free Full Text]
  37. Hofstee, B. H. J. (1942) Nature 184, 1296-1298