Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

cGMP-dependent protein kinase (PKG) is a serine/threonine protein kinase and is one of the major intracellular receptors for cGMP. PKG is selectively activated by cGMP and regulates cytoplasmic Ca²⁺ concentration by several pathways (Rashatwar et al., 1987; Ruth et al., 1990; Komalavilas and Lincoln, 1994). There are two types of PKG present in eukaryotic cells: type I and type II (Lincoln and Cornwell, 1993). PKG type I is a dimer of identical subunits, each with a mass of ~78 kDa. PKG type II also exists as a dimer whose subunit mass is ~86 kDa (Gamm et al. 1995). PKG type I is widely distributed and is isolated from soluble extracts of tissues, whereas PKG type II is a particulate form of the enzyme and has a limited tissue distribution (DeJonge, 1981). PKG type I has two subtypes,  and  (Lincoln et al., 1988). Both PKG type I and type I have been cloned (Sandberg et al., 1989; Wernet et al., 1989; Tamura et al., 1996) and have been shown to be present in mammalian cells (Lincoln and Cornwell, 1993). Both PKG subtypes selectively bind cGMP and have two cGMP-binding sites per subunit. The two cGMP-binding sites are distinguished by varying affinities for cGMP analogs (Corbin et al. 1986). The N terminus of PKG type I (1-89 amino acids), also known as the I-specific region, contains the "leucine-isoleucine zipper" and major sites for autophosphorylation (Ser⁵¹, Ser⁷³ and Thr⁵⁹, Thr⁸⁵) (Aitkin et al., 1981).

cGMP has been shown to exert a variety of generally inhibitory actions on adult heart cells, including a negative inotropic effect (Nawrath, 1977) and inhibition (Hartzell and Fischmeister, 1986; Fischmeister and Hartzell, 1987; Levi et al., 1989; Wahler et al., 1990) as well as stimulation (Ono and Trautwein, 1991; Kirstein et al. 1995; Wahler and Dollinger, 1995) of L-type calcium current (I_Ca), particularly in the presence of -adrenergic stimulation of elevated cAMP levels. We have shown that basal cGMP levels, as determined by basal G-cyclase activity, are much higher in newborn compared with adult rabbit heart cells (Kumar et al., 1994). Recently, we have shown that stimulation of PKG by intracellular perfusion of 8-bromo-cGMP or 8-chlorophenylthio-cGMP significantly increased basal I_Ca in newborn but not in adult rabbit ventricular myocytes (Kumar et al., 1997). The stimulatory effect of cGMP was blocked by PKG inhibitor (KT-5823) but not by cAMP-dependent protein kinase inhibitor (5-22) or by phosphodiesterase (PDE) inhibitor [3-isobutyl-1-methylxanthine (IBMX)]. This suggested that the stimulatory effect of cGMP in newborn cells was mediated by PKG (Kumar et al., 1997). Our findings on the role of the cGMP-PKG cascade in the regulation of basal I_ca in newborn cells are unique and fundamentally different from findings by other investigators on the effects of PKG stimulation on I_ca in adult heart cells of different species. Our data indicate that individual cardiac cells alter their mechanism of response to cGMP during development. The developmental decline in the role of PKG in the regulation of calcium current could be produced by either a decline in the expression of PKG with developmental age; developmental differences in the amino acid sequence of PKG, which might alter its affinity for cGMP or its ability to phosphorylate calcium channels; or developmental differences in phosphorylation of the calcium channels phosphorylated by PKG.

The tissue-specific expression of PKG is highly variable and appears to be under physiological control. Very little information is available concerning the expression of PKG in cardiovascular tissues. Although PKG was found in adult cardiac muscle (Lincoln and Keely, 1981), it was most abundant in cardiac vasculature. Sandberg et al. (1989) have studied developmental changes in PKG mRNA and protein levels in rat heart. They showed the presence of two mRNA forms (7.5 and 6.5 kb) in rat heart and also showed a developmental decline of both forms with faster and more complete decline for the 6.5-kb form. In contrast to mRNA levels, there was no change in PKG protein levels during cardiac development. To elucidate the mechanism responsible for the developmental decline and the significance of PKG in the regulation of cardiac calcium channels, we isolated the full-length rabbit heart PKG type I cDNA, determined the sequence to examine species-specific sequence heterogeneity, and expressed the construct in mammalian cells to verify that the cloned cDNA encodes PKG. We also analyzed expressed levels of PKG type I mRNA by Northern blotting and PKG type I protein levels by Western immunoblotting to investigate the regulation of PKG expression during postnatal development of rabbit heart.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Preparation of Isolated Cells. New Zealand White adult (1.5-2 kg) and newborn (1- to 4-day-old) rabbits of either sex were used. Adult rabbits were heparinized (1000 U i.v.) and anesthetized with sodium pentobarbital (50 mg/kg i.v.). For newborn rabbits, the same drugs were given intraperitoneally. To isolate single ventricular myocytes, the heart was rapidly removed via thoracotomy with artificial ventilation and the aorta was cannulated. Single ventricular myocytes were obtained by enzymatic dissociation as previously described (Osaka et al., 1993). The cell suspension was purified before making homogenates as described earlier (Kumar et al., 1996) until the suspension had more than 80% rod-shaped (in adult) or spindle-shaped (in newborn) myocytes.

Preparation of Total Homogenates from Tissues and Isolated Cells. To prepare total homogenates from isolated ventricular myocytes (Kumar et al., 1994), isolated myocytes were homogenized in hypotonic membrane buffer [50 mM Tris-HCl (pH 7.5), 5 mM MgCl₂,1 mM EDTA, 1 mM dithiothreitol, 0.001 mM pepstatin A, 0.4 mM phenylmethylsulfonyl fluoride, 1 mM phenanthroline, and 1 mM iodoacetamide] by sonication. The homogenate was then centrifuged at 100g for 15 min to separate the unbroken cells, and the supernatant (total homogenate) was stored at 70°C in small aliquots. To make homogenates from intact ventricles, hearts were separated into atria and ventricles. The ventricles were homogenized in Tris-buffered saline (TBS) (50 mM Tris-HCl, 150 mM NaCl, pH 7.4) containing protease inhibitors with a Tekmar tissumizer. The homogenate was centrifuged at 100g for 15 min, and the supernatant (total homogenate) was stored at 70°C in small aliquots. Protein was determined by a dye method (Bio-Rad, Hercules, CA).

Preparation of RNA. For mRNA isolation, hearts were rapidly excised from anesthetized adult and newborn rabbits and plunged into cold Trizol solution (Life Technologies, Inc., Paisley, Scotland). Ventricles were carefully separated and homogenized with a polytron homogenizer. The total RNA was isolated by one-step isolation with phenol-guanidinium isothiocyanate solution developed by Chomczynski and Sacchi (1987). We used an oligo(dT)-cellulose column to isolate mRNA from total RNA for cDNA cloning and for Northern blot analysis.

Cloning and Expression of Rabbit PKG Type I cDNA. To construct polymerase chain reaction (PCR)-mediated PKG type I cDNA, the coding sequence from bovine lung cDNA was divided into four regions: bp 1-486, bp 456-1032, bp 456-1663, and bp 999-2016, each corresponding to the published sequence for the cDNA-encoding bovine lung PKG type I (Wernet et al., 1989). Poly(A+)-selected RNA (mRNA) was purified from total RNA isolated from newborn rabbit heart. The mRNA was converted to first strand of cDNA with avian malony virus reverse transcriptase. The cDNA derived from reverse transcription (RT) reactions was amplified with 1 µM PCR primers in 10 mM Tris-HCl, 50 mM KCl reaction buffer (pH 8.8), and 2.5-5 U of Taq DNA polymerase (Promega Biotec, Madison, WI). The amplification reaction was carried out under standard conditions (denaturation at 95°C for 1 min, annealing at 60°C for 2 min, and extension at 72°C for 3 min), followed by a final extension for 10 min at 72°C. The primers used for the isolation of bp 1-486 (A1-A2 fragment) were CCCGGATCCATGAGCGAGCTGGAGGAAGACTTTGCCAAG and CCCGGATCCAGGACCCATTGTGCACAGCTTCACGCC for the sense and antisense oligonucleotides, respectively. The primers for the isolation of bp 456-1032 (B1-B2 fragment) were CCCGGATCCAGAAGGCGTGAAGCTGTGCACAATGGGTCC and CCCGGATCCTTTTGCCTTAGCTTCTGCATC-TTCATATGC for the sense and antisense oligonucleotides, respectively. The primers used for the isolation of bp 456-1663 (B1-CB2 fragment) were CCCGGATCCAGAAGGCGTGAAGCTGTGCACAATGGGTCC and TGCCAGTCAGGAGTTCATACATTAGGATTC for the sense and antisense oligonucleotides, respectively. The primers used for the isolation of bp 999-2016 (C1-C2 fragment) were CCCCTCGAGTTAAGCATATGAAGATGCAGAAGCTAAGGC and CCCCTCGCGTAAGAAGTCTATGTCCCATCCTGAGTTGTC for the sense and antisense oligonucleotides, respectively. To obtain the 3'-terminal region of rabbit PKG type I cDNA (bp 1567-2119; D1-D2 fragment), we used another set of primers with sense primer from coding region GAGTATGTAGCCCCAGAGAT and antisense primer from noncoding region TGACCCCGAGCACTAAT. The amplified fragments were isolated on 1% agarose gels, purified by chromatography on Wizard minicolumn (Promega) and subcloned into EcoRI sites in pGEM-TEasy vector. The sequence of cDNA fragments was determined on both the strands by automated DNA sequencing (Applied Biosystems Prism 377 DNA sequencer, Applied Biosystems, Foster City, CA) with Applied Biosystems Prism cycle sequencing Dye Terminator ready reaction kit. The three amplified cDNA fragments (A1-A2, B1-B2, and C1-C2) were sequentially ligated (Rapid DNA ligation; Boehringer Mannheim Corp., Indianapolis, IN) into pcDNA 3.1(+) vector (Invitrogen, San Diego, CA) to yield the entire cDNA sequence.

Expression of PKG and Assay of PKG Activity. The recombinant vector was grown and purified by Midi prep kit (Promega). COS-7 cells (monkey fibroblasts) were grown to 80% confluency in 60-mm culture dishes and then transiently transfected with 5 µg of pcDNA3.1/PKG with Tfx transfection reagent (Promega). Cells were grown for 24 and 48 h at 37°C. After rinsing the monolayer with PBS, cells were harvested and homogenized with cold buffer consisting of 20 mM sodium phosphate, pH 6.8, 2 mM EDTA, 15 mM 2-mercaptoethanol, 150 mM NaCl, 2 mM benzamidine, 0.1 mM phenylmethylsulfonyl fluoride, 10 µg/ml pepstatin A, and 10 µg/ml leupeptin. The suspension was centrifuged for 10 min at 14,000 rpm to obtain cell extract. Aliquots of extract were analyzed for PKG activity and also for Western blotting with an affinity-purified polyclonal rabbit anti-PKG antibody as described later. PKG activity was assayed as described in Boerth and Lincoln (1994) with a peptide substrate (RKISASEFDRPL) selective for PKG (Colbran et al., 1992). All assays were performed in the presence of 1 µM cAMP-dependent protein kinase inhibitor peptide (5-24), 0.1 mM IBMX, and in the presence or absence of 1 µM cGMP. The difference in the phosphorylation of substrate in the presence and absence of cGMP was taken as PKG activity and expressed as pmol/min/mg protein.

Northern Blot Analysis. Northern blot analysis was carried out with a standard method (Brown and Mackey, 1997). To identify PKG type I mRNA, we used bovine PKG type I cDNA, rabbit cardiac PKG type I containing 999-2016 nucleotides (C1-C2 fragment), and rabbit cardiac PKG type I containing 456-1663 nucleotides (B1-CB2 fragment), which can identify both type I and I. Equal loading of the RNA samples was initially verified by ethidium bromide staining of replicate lanes run in the same gel. As an internal standard, we used a cDNA probe for mRNA of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) because expression of GAPDH mRNA is not developmentally regulated (Lompre et al., 1991). GAPDH cDNA probe was prepared by RT-PCR of rabbit heart mRNA with rat GAPDH primers (Stratagene, Inc., La Jolla, CA). Relative amounts of specific mRNAs in adult and newborn ventricular mRNA preparations were quantified by two-dimensional gel imaging with an Alpha Imager 2000 documentation and analysis system (Alpha Innotech Corp., San Leandro, CA) for multiple adult and newborn preparations from different hearts.

Western Blot Analysis. Immunological characterization and quantification of the amounts of PKG type I in homogenates prepared from enzymatically dissociated ventricular myocytes, or from whole ventricles was performed with a method described in Kumar et al. (1994). A polyclonal antibody that recognizes specifically PKG type I was raised against bovine PKG holoprotein in goat or rabbit. The affinity-purified antibodies for PKG type I recognize both PKG type I and I and do not recognize PKG type II and any other "standard" homologous kinases such as PKA or PKC (unpublished data). Purified PKG type I, used to identify the binding of antibody (positive control), was isolated and purified from bovine lung. We loaded 1 to 6 ng of purified PKG with 1 µg of BSA as a positive control. After electrophoretic transfer of proteins on PVDF membrane (Amersham Corp., Arlington Heights, IL), membranes were blocked (TBS containing 0.1% Tween 20 and 5% nonfat dry milk) and then incubated with an affinity-purified polyclonal PKG antibody at 1:1000 in the blocker solution overnight at 4°C. Following several washes in wash buffer (TBS containing 0.05% SDS, 0.05% Nonidet P-40, and 0.125% sodium deoxycholate), the membranes were incubated with horseradish peroxidase-conjugated secondary antibody diluted at 1:10,000 in the blocker solution for 1 h at room temperature. For the detection of bands, we used enhanced chemiluminescence detection (ECL+plus; Amersham Corp.) with Lumigen PS-3. To compare the relative amounts of PKG in newborn versus adult preparations, we use a two-dimensional gel-imaging system (Alpha Imager 2000 documentation and analysis system; Alpha Innotech Corp.) with multiple adult and newborn preparations from different hearts and purified PKG as positive control run on different tracks of the same gel. Peak area obtained for adult and newborn bands was normalized with peak area for positive control to get the relative amount of PKG present per milligram of protein of isolated ventricular myocytes.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cloning and Expression of Rabbit PKG Type I cDNA. We obtained cDNA fragments of PKG type I from rabbit cardiac mRNA (Fig. 1) with RT-PCR technique. Figure 1 shows the composed nucleotide sequence and deduced amino acid sequence of rabbit cardiac PKG type I cDNA. The cDNA obtained from rabbit heart consisted of 2095 bp (1-2095) and contained an open reading frame of 2013 bp. The coding region of rabbit cardiac PKG type I has 94% homology to published sequences of human and bovine PKG type I (Wernet et al., 1989; Tamura et al., 1996) and consisted of 671 amino acids with a molecular mass of 76,452 Da. There were seven amino acid substitutions between human (Tamura et al., 1996) or bovine (Wernet et al., 1989) and rabbit PKG type I; these substitutions are shown in Fig. 2. The leucine-isoleucine zipper responsible for dimerization of subunits and the autophosphorylation sites in the I-specific N-terminal region (1-89 amino acids) were conserved among human, bovine, and rabbit PKG type I. In the I-specific region, there was only one amino acid substitution from bovine and human to rabbit PKG type I; this was from Ser⁸⁷ (bovine and human) to Phe⁸⁷ (rabbit). The major substitutions were found in the cGMP-binding domains (amino acids 102-341). These substitutions were Val¹⁴⁰ to Ala¹⁴⁰ in the cGMP2-binding site domain and Gly²⁴⁵ to Glu²⁴⁵, Ile²⁴⁸ to Ser²⁴⁸, Val²⁷⁹ to Ile²⁷⁹ (human and bovine to rabbit) in the cGMP1-binding site domain. The amino acid at position 265 (Lys²⁶⁵) in rabbit was the same as for the bovine form but different from Thr²⁶⁵ in human PKG type I. Similarly, the amino acid at position 275 (Ser²⁷⁵) was the same as in the human form but different from Asn²⁷⁵ in bovine PKG type I. We found T/C variability at nucleotides 1617 and 1673 in the ATP binding/catalytic site domain (amino acids 341-602). The change from T to C at position 1617 would not change the encoded amino acid (Ala). However, at position 1673, a change from T to C would change the encoded amino acid Phe⁵⁵⁸ to Ser⁵⁵⁸. There was one substitution in the C-terminal region at position 651, where Ser⁶⁵¹ (bovine and human) was replaced by Gly⁶⁵¹.

View larger version (80K): [in this window] [in a new window]   Fig. 1.   Nucleotide and deduced amino acid sequences of rabbit cardiac PKG type I cDNA. Nucleotides are numbered starting with 1 at the first nucleotide in the coding region. Amino acids are shown by their three-letter code. Leucine-isoleucine residues of the dimerization domain are indicated by  and the autophosphorylation sites are indicated by . The nucleotide sequence reported in this article has been deposited in GenBank (Accession No. AF076969), EMBL in Europe, and DNA Data Bank of Japan nucleotide sequence databases.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Specific amino acid substitutions in different regions among human, bovine, and rabbit PKG type I.

View larger version (61K): [in this window] [in a new window]   Fig. 3.   Top, immunoreactivity of PKG in extracts of transfected COS cells. Lanes 1 and 2 contained 100 µg of homogenate protein from COS cells after 48 h of transfection with full-length cDNA construct of PKG type I. Lanes 3 and 4 contained 100 µg of homogenate protein from COS cells without any transfection and the cells transfected with pcDNA3.1 containing no insert, respectively, as negative controls. The lane marked as PKG contained 6 ng of purified bovine lung PKG used as positive control. The blots were probed with an affinity-purified rabbit PKG antibody at 1:1000 dilution. Immunostained bands were detected at an apparent molecular mass of 79 kDa with enhanced chemiluminescence. Bottom, PKG enzyme activity in extracts of COS cells transfected with PKG I cDNA. Activity was measured as cGMP stimulated PKG activity after 24 and 48 h of transfection and is shown as picomoles per minute per milligram protein. The extract from COS cells transfected with pcDNA 3.1 without any insert was used as control.

Northern Blot Analysis in Adult and Newborn Heart. To determine if PKG type I mRNA levels are different in adult and newborn heart, we analyzed PKG expression by Northern blot analysis of cardiac PKG with mRNA isolated from adult and newborn rabbit ventricles. Figure 4 shows representative blots in which lanes marked as newborn (NB) and adult (AD) contained 5 µg of mRNA from newborn and adult rabbit ventricles, respectively. The blot shown in Fig. 4A was probed with the bp 456-1663 region (B1-CB2 fragment) of rabbit PKG type I cDNA, which was expected to detect both types I and I mRNA, as expressed by other species. We detected only one band at 6.8 kb with this probe, the same message was identified with cDNA probes for bovine PKG type I and rabbit cardiac PKG type I (containing C1-C2 fragment) (data not shown). The results confirm the expression of only one mRNA species of PKG type I in adult and newborn rabbit ventricles. However, the relative abundance of PKG type I mRNA appears to be age-dependent with significantly more abundance in newborn compared with adult ventricles. Figure 4C shows the relative amounts of PKG type I mRNA in adult and newborn rabbit ventricles calculated by comparing the peak area of band from multiple adult and newborn animals by densitometric analysis. PKG type I mRNA levels were 3.5 times higher in newborn compared with adult bands (4851 ± 817 arbitrary units for newborn and 1389 ± 389 arbitrary units for adult, n = 4, p < .05). To eliminate the possibility of unequal loading of adult and newborn mRNA, the same membranes were reprobed (after stripping the PKG probe) with a cDNA probe for GAPDH (Fig. 4B). No significant differences were observed (4408 ± 290 arbitrary units for newborn and 3910 ± 718 arbitrary units for adult, n = 4, p > .5) in the density of adult and newborn bands probed with GAPDH as evident from densitometric analysis in Fig. 4D.

View larger version (52K): [in this window] [in a new window]   Fig. 4.   Northern blot analysis of PKG type I and GAPDH mRNAs with poly(A+) RNA from adult and newborn rabbit ventricles. Five micrograms of poly(A+) RNA isolated from AD and NB ventricles was loaded in each lane. The replicate lanes represent samples from different animals. A, blot hybridized with a B1-CB2 fragment of cDNA generated by RT-PCR from rabbit mRNA. The probe detected a message of ~6.8 kb corresponding to PKG type I mRNA in newborn and adult rabbit heart. B, Northern blot analysis of the same blot hybridized with a cDNA probe encoding GAPDH as an internal control. C, changes in PKG mRNA levels (arbitrary units of peak area) from newborn to adult rabbit ventricles. D, changes in GAPDH levels (arbitrary units of peak area) from newborn to adult.

Western Blot Analysis in Adult and Newborn Heart. We assessed the levels of PKG type I protein by Western blot analysis in homogenates prepared from whole ventricles, as well as from isolated ventricular myocytes of adult and newborn rabbit heart with a PKG type I-specific antibody. This antibody, raised against bovine lung PKG, recognizes both types I and I and does not cross-react with PKG type II and other protein kinases. To define the specificity of the antibody used in this study, we used purified bovine lung PKG type I as a positive control. We also quantified the relative amounts of PKG type I present in adult and newborn per milligram of cell/tissue protein with a gel-imaging system. Figure 5 shows a representative immunoblot in which lanes labeled as NB and lanes labeled as AD each contained 60 µg of homogenate protein prepared from ventricular myocytes isolated from different newborn and adult rabbit hearts. We loaded 3 ng of purified bovine lung PKG type I in lanes labeled as PKG. The results indicate an immunoreactive protein at an apparent molecular mass of 79 kDa in newborn and adult preparations that was aligned with the positive control of PKG. The PKG levels were much higher in newborn (n = 6) compared with adult (n = 6) preparations; PKG levels were too low to be detected in some adult preparations. Figure 5B shows the relative amount of PKG in adult and newborn rabbit ventricular myocytes calculated by normalizing the adult and newborn band density to PKG positive control. The relative amount of PKG present in newborn rabbit ventricular myocytes (39.4 ± 7.6 ng/mg protein, n = 6) was much greater (p < .05) than the PKG present in adult rabbit ventricular myocytes (1.25 ± 0.53 ng/mg protein, n = 6). Figure 6A shows a representative immunoblot in which lanes labeled as NB and lanes labeled as AD each contained 50 µg of homogenate protein from intact ventricles of different newborn and adult rabbit hearts. We loaded 2 ng of purified bovine lung PKG type I in lanes labeled as PKG. The results indicate an immunoreactive protein at an apparent molecular mass of 79 kDa in newborn and adult preparations. The PKG levels in preparations from intact ventricles were only two to three times higher in newborn (n = 6) compared with adult (n = 6) preparations. Figure 6B shows the relative amount of PKG type I in adult and newborn rabbit ventricles calculated by normalizing the adult and newborn band density to PKG positive control. The relative amount of PKG type I present in newborn rabbit ventricles (12.6 ± 0.67 ng/mg protein, n = 6) was 2.3 times greater (p < .05) than the PKG type I present in adult rabbit ventricles (5.35 ± 0.62 ng/mg protein, n = 6). By comparing Western blotting results in preparations from whole ventricles with preparations from isolated ventricular myocytes, it is clear that the difference in PKG levels between adult and newborn preparations is much greater in isolated ventricular myocytes (30-fold) compared with whole ventricles (2.35-fold).

View larger version (45K): [in this window] [in a new window]   Fig. 5.   A, immunoblot analysis of PKG in total homogenates prepared from isolated ventricular cells of adult and newborn rabbit hearts. Lanes marked as NB and AD contained 60 µg of homogenate protein from adult and newborn ventricular cells, respectively. The replicate lanes represent samples from different animals. The lanes marked as PKG contained 3 ng of purified bovine lung PKG. The blots were probed with an affinity-purified goat PKG antibody at 1:1000 dilution. Immunostained bands were visualized at an apparent molecular mass of 79 kDa with enhanced chemiluminescence. B, changes in PKG levels (nanograms per milligram protein) from newborn to adult in isolated ventricular cells. Bands visualized by enhanced chemiluminescence were quantified by measuring density and normalized to the value for purified bovine lung PKG.

View larger version (37K): [in this window] [in a new window]   Fig. 6.   A, immunoblot analysis of PKG in total homogenates prepared from intact ventricles of adult and newborn rabbit hearts. Lanes marked as NB and AD contained 50 µg of homogenate protein from adult and newborn ventricles, respectively. The replicate lanes represent samples from different animals. The lanes marked as PKG contained 2 ng of purified bovine lung PKG. The blots were probed with an affinity-purified rabbit PKG antibody at 1:1200 dilution. Immunostained bands were visualized at an apparent molecular mass of 79 kDa with enhanced chemiluminescence. B, changes in PKG levels (nanograms per milligram protein) from newborn to adult in intact ventricles. Bands visualized by enhanced chemiluminescence were quantified by measuring density and normalized to the value for purified bovine lung PKG.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The cGMP-dependent protein kinase is a major receptor protein for cGMP. We previously demonstrated that this enzyme mediates the maintenance of basal calcium current and stimulation of calcium current by cGMP in newborn rabbit ventricular myocytes but not in adult cells. We focused this study on the cDNA cloning, and expression of PKG type I, and the postnatal changes in the expressed mRNA and protein levels of PKG type I in rabbit ventricles. The salient findings of this study are summarized as follows: 1) The coding region of rabbit PKG type I showed 94% homology to published sequences of human and bovine PKG type I. 2) The deduced amino acid sequence showed major substitutions in two cGMP-binding site domains. 3) The expressed recombinant PKG was catalytically active, stimulated by cGMP, and showed the same immunoreactivity as the native PKG. 4) Expressed mRNA and protein levels for PKG type I were much higher in newborn compared with adult preparations.

We identified only one distinct species of mRNA by Northern blotting of PKG type I in mRNA preparations of adult and newborn rabbit ventricles. Thus, we performed cDNA cloning of rabbit PKG type I from ventricular tissue of newborn rabbit. The composed nucleotide sequence and deduced amino acid sequence of rabbit cardiac PKG type I cDNA showed some significant differences from published human and bovine PKG type I sequences (Wernet et al., 1989; Tamura et al., 1996). Most amino acid substitutions were found in the two cGMP-binding domains. The two cGMP-binding sites are different on the basis of their affinities for various cGMP analogs and are responsible for the cooperative nature of cGMP binding and activation. For the PKG type I form, 8-substituted analogs of cGMP (e.g., 8-bromo-cGMP and 8-chlorophenylthio-cGMP) are much more potent activators of enzyme than the native nucleotide (Wolfe et al., 1989). We previously showed that 8-bromo-cGMP and 8-chlorophenylthio-cGMP were much more potent in stimulating calcium current than the native cGMP in newborn rabbit heart cells (Kumar et al., 1997). The substitutions in cGMP-binding domains might be important in determining the affinity of PKG for cGMP binding. The cloning of human PKG type I has shown two amino acid substitutions from bovine PKG type I to human PKG type I (Lys²⁶⁵ to Thr²⁶⁵ and Asn²⁷⁵ to Ser²⁷⁵) (Tamura et al., 1996). It is interesting to note that Lys²⁶⁵ in rabbit was the same as in bovine but different from human (Thr²⁶⁵) and Ser²⁷⁵ in rabbit was same as in human but different from bovine (Asn²⁷⁵). In addition to these substitutions, we found amino acid substitutions in the amino terminal I-specific region and C-terminal region. The PKG type I-specific region of the enzyme contains the autophosphorylation sites and the leucine-isoleucine zipper, which is responsible for dimerization of two subunits. The four different autophosphorylation sites (Ser⁵¹, Thr⁵⁹, Ser⁷³, and Thr⁸⁵) observed by Aitkin et al. (1981) and the leucine-isoleucine zipper were conserved. The substitutions we found in type I-specific region (Ser⁸⁷ to Phe⁸⁷) and in C-terminal region (Ser⁶⁵¹ to Gly⁶⁵¹) may be due to species differences.

To verify that the isolated cDNA encodes PKG, we transfected COS cells with our cDNA clone. The PKG holoenzyme has been successfully expressed in COS cells. Western blotting experiments with homogenates of PKG-transfected COS cells confirmed that expressed PKG is immunoreactive and has the same molecular mass as the native PKG. The extracts of PKG-transfected cells also showed PKG enzyme activity that was stimulated by cGMP. This confirmed that expressed PKG is not only immunoactive but also catalytically active. The availability of this recombinant cardiac PKG may help to define the roles of specific cGMP-dependent phosphorylation in cardiac cells.

Northern blot analysis showed PKG type I mRNA message as a single band at 6.8 kb in both adult and newborn rabbit ventricular mRNA preparations. The message size for type I isoform in newborn and adult rabbit ventricles exceeds that of coding region by more than 2-fold, which shows that PKG type I mRNA contains a large noncoding region. Similar results were obtained by other investigators (Sandberg et al., 1989; Wernet et al., 1989; Cornwell et al., 1994). Our studies could be interpreted that PKG type I isoform is highly expressed in newborn heart but expressed at a very low level in adult heart. A developmental decrease in PKG type I mRNA also was shown during postnatal development of rat heart by Sandberg et al. (1989). However, they found two PKG type I mRNA species in rat heart at 7.5 and 6.5 kb and showed a faster and more complete decrease for 6.5- than for 7.5-kb species during development. In contrast to the studies on rat by Sandberg et al. (1989), our results show the presence of a single mRNA species (6.8 kb) for PKG type I in adult and newborn rabbit ventricles. Its expression declines during postnatal development from newborn to adult.

Western immunoblotting experiments with homogenates prepared from isolated ventricular myocytes also showed that PKG is present in appreciable amounts in newborn heart cells, and that its apparent molecular mass is similar to that of bovine lung PKG (79 kDa). The immunodetectable levels of PKG type I protein were 30-fold higher in newborn compared with adult ventricular myocytes. This 30-fold difference in PKG levels between adult and newborn ventricular myocytes was diminished to 2.35-fold when PKG levels were measured in preparations from intact ventricles of adult and newborn heart. The developmental decline in PKG protein levels (2.35-fold) from newborn to adult ventricles was comparable to the decline in PKG mRNA levels (3.5-fold) from newborn to adult ventricles. However, the developmental decline in PKG type I protein levels from newborn and adult ventricular myocytes (30-fold) was much higher than was revealed by decline of protein or mRNA levels in newborn and adult intact ventricles. This discrepancy in the levels of PKG in intact ventricles and isolated ventricular myocytes could be because PKG mRNA and protein levels detected in adult ventricles may have been derived largely from cardiac vascular tissue. The expressed mRNA levels or protein levels in whole ventricles may not represent the specific expression in cardiac myocytes because other cell types are included within the heart tissue. Ecker et al. (1989) showed that cardiac vasculature is particularly rich in PKG. In some other studies of adult heart (Walter, 1989), PKG was either not detected or was detected at extremely low levels. In contrast to our findings by Northern and Western analyses, Sandberg et al. (1989) showed no postnatal decrease in rat cardiac PKG protein during development, although they showed a sharp decrease in PKG mRNA levels. The differences in PKG levels between adult and newborn may be of physiological and pathological significance. Previous studies on aorta and heart have shown that PKG levels were affected by experimental hypertension (Coquil et al., 1987; Ecker et al., 1989) but these changes were mainly associated with the change in vascular PKG.

In summary, our findings suggest that PKG is developmentally regulated in rabbit heart and its expressed levels are much higher in newborn compared with adult ventricular myocytes. Rabbit PKG type I shows considerable homology to bovine and human PKG type I. However, several amino acid substitutions in rabbit PKG type I compared with bovine and human PKG type I may perhaps change some of the catalytic properties of PKG in rabbit heart by changing the cGMP-binding affinity. The greater expression of PKG in newborn cells compared with adult cells could be responsible for differences in the effects of cGMP on L-type calcium current in adult and newborn rabbit ventricular myocytes.