Distinctive anatomical patterns of gene expression for cGMP-inhibited cyclic nucleotide phosphodiesterases

Novel alternative splice variants of cGMP-binding cGMP-specific phosphodiesterase

After our recent findings that the amino-terminal portion of rat cGMP-binding, cGMP-specific phosphodiesterase (cGB-PDE) differs from those of bovine and human cGB-PDEs, we found two forms of canine cGB-PDE cDNAs (CFPDE5A1 and CFPDE5A2) in canine lung. Each contained a distinct amino-terminal sequence, CFPDE5A1, possessing an amino-terminal portion with sequence similar to those of bovine and human, and CFPDE5A2, having one similar to that of rat. Other portions coding for the cGMP binding domains and the catalytic domain were conserved. Both CFPDE5A1 and CFPDE5A2 transcripts were detected in the cerebellum, hippocampus, retina, lung, heart, spleen, and thoracic artery. CFPDE5A1 transcripts were particularly abundant in the pylorus, whereas CFPDE5A2 transcripts were quite low in this tissue. CFPDE5A1 and CFPDE5A2 expressed in COS-7 cells had cGMP Km values of 2.68 and 1.97 microM, respectively, and both were inhibited by a low concentration of a cGB-PDE inhibitor, Zaprinast. Both CFPDE5A1 and CFPDE5A2 bound cGMP to their allosteric cGMP binding domains, and this cGMP binding was stimulated by 3-isobutyl-1-methylxanthine. Thus, two types of alternative splice variants of canine cGB-PDE have been identified and shown to have similar biological properties in vitro.

Phosphodiesterase 3 inhibitors suppress oocyte maturation and consequent pregnancy without affecting ovulation and cyclicity in rodents

During each reproductive cycle, a preovulatory surge of gonadotropins induces meiotic maturation of the oocyte in the preovulatory follicle followed by ovulation. Although gonadotropins stimulate cAMP production in somatic cells of the follicle, a decrease in intra-oocyte cAMP levels is required for resumption of meiosis in oocytes. Based on the observed compartmentalization of the cAMP-degrading enzyme, phosphodiesterase, in follicular somatic and germ cells, inhibitors of phosphodiesterase 3 were used to block meiosis in ovulating oocytes in rodents. By this strategy, we demonstrated that fertilization and pregnancy could be prevented without disturbing follicle rupture and normal estrous cyclicity. In contrast to conventional contraceptive pills that disrupt ovarian steroidogenesis and reproductive cycles, the present strategy achieves effective contraception by selective blockage of oocyte maturation and development without alterations in ovulation and reproductive cyclicity.

Expression and characterization of deletion recombinants of two cGMP-inhibited cyclic nucleotide phosphodiesterases (PDE-3)

cDNAs encoding two PDE-3 or cyclic GMP-inhibited (cGI) cyclic nucleotide phosphodiesterase (PDE) isoforms, RPDE-3B (RcGIP1) and HPDE-3A (HcGIP2), were cloned from rat (R) adipose tissue and human (H) heart cDNA libraries. Deletion and N- and C-terminal truncation mutants were expressed in Escherichia coli in order to define their catalytic core. Active mutants of both RPDE-3B and HPDE-3A included the domain conserved among all PDEs plus additional upstream and downstream sequences. An RPDE-3B mutant consisting of the conserved domain alone and one from which the RPDE-3B 44-amino acid insertion was deleted exhibited little or no activity. All active recombinants exhibited a high affinity (< 1 microM) for cyclic AMP (cAMP) and cyclic GMP (cGMP), were inhibited by cAMP, cGMP, and cilostamide, but not by rolipram, and were photolabeled with [32P]-cGMP. The IC50 values for cGMP inhibition of cAMP hydrolysis were lower for HPDE-3A than for RPDE-3B recombinants. The deduced amino acid sequences of HPDE-3A and RPDE-3B catalytic domains are very similar except for the 44-amino acid insertion not found in other PDEs. It is possible that this insertion may not only distinguish PDE-3 catalytic domains from other PDEs and identify catalytic domains of PDE-3 subfamilies or conserved members of the PDE-3 gene family, but may also be involved in the regulation of sensitivity of PDE-3s to cGMP.

Downregulation of right ventricular phosphodiesterase PDE-3A mRNA and protein before the development of canine heart failure

Phosphodiesterase III (PDE-3) inhibitors are inotropes used to treat congestive heart failure (HF). Previous studies showed PDE-3A mRNA levels were reduced in the left ventricle (LV) in dogs subjected to pacing-induced HF. The present study evaluated a time-course for RV-specific changes in PDE-3A mRNAs and proteins after pacing for 3 wk (n = 4) or in HF (4-5 wk; n = 4-6). Total RNA from LV/RV tissues was isolated for Northern analyses; cytosolic and microsomal proteins were prepared for PDE-3A immunoblots. PDE-3A mRNAs (7-8 and 10 kb) were normalized against glyceraldehyde-3-phosphodehydrogenase (GAPDH) or ribosomal 18s with similar results. PDE-3A/GAPDH ratios in 3 wk were unchanged in LV, but significantly (p < 0.05) reduced by 48% in RV vs unpaced controls (n = 8). In contrast, PDE-3A (7-8 kb)/GAPDH ratios were significantly reduced in HF by 50-59% in both ventricles. Consistent with mRNA levels, significant reductions in microsomal 135 kDa (93-96%) and cytosolic 120 kDa PDE-3A (57-69%) were seen in both ventricles in HF or in the RV at 3 wk; an LV-specific reduction (50%) in cytosolic 80 kDa PDE-3A in HF was also detected. In summary, RV-specific downregulation of PDE-3A mRNA/protein(s) at 3 wk suggests that hemodynamic rather than humoral mechanisms are responsible, and provides a molecular basis for the limited efficacy of milrinone in the progression of HF.

Cyclic nucleotide PDE-3. Quantitation of PDE-3A and -3B mRNAs in rat tissues by RNase protection assay

Type 3 cyclic nucleotide phosphodiesterase (PDE-3) isoforms exhibit a high affinity ("low K(m)") for cAMP and are specifically inhibited by cGMP and a number of pharmacological agents, which increase myocardial contractility, inhibit platelet aggregation, and increase smooth muscle relaxation. The PDE-3 family consists of at least two isozymes, PDE-3A (cardiac type) and PDE-3B (adipocyte type), with distinct tissue-specific distributions. PDE-3A mRNA is highly expressed in the cardiovascular system, whereas PDE-3B mRNA is primarily expressed in adipocytes and hepatocytes. Toward understanding potential roles of PDE-3 in diabetes mellitus, we have established a specific and sensitive RNase protection assay (RPA) for quantitating PDE-3A and PDE-3B mRNA in rat diabetic models. In fatty Zucker diabetic (ZDF) rats, PDE-3A mRNA, but not PDE-3B mRNA, was expressed in heart, whereas liver and white and brown fat tissues predominantly expressed PDE-3B mRNA. Unexpectedly, PDE-3B mRNA expression was approximately 2.5 times higher than PDE-3A mRNA in aorta from both ZDF and Sprague-Dawley (SD) rats. In contrast, expression levels of PDE-3A mRNA in heart were similar in both species. With this RPA, we were thus able to compare PDE-3A and -3B mRNA levels in different tissues as well as in different rat species.

Phosphorylation and activation of hormone-sensitive adipocyte phosphodiesterase type 3B

Phosphodiesterases (PDEs) include a large group of structurally related enzymes that belong to at least seven related gene families (PDEs 1-7) that differ in their primary structure, affinity for cAMP and cGMP, response to specific effectors, sensitivity to specific inhibitors, and regulatory mechanism. One characteristic of PDE3s involves their phosphorylation and activation in response to insulin as well as to agents that increase cAMP in adipocytes, hepatocytes, and platelets and in response to insulin-like growth factor 1 in pancreatic beta cells. In adipocytes, activation of the membrane-associated PDE3B is the major mechanism whereby insulin antagonizes catecholamine-induced lipolysis. PDE3B activation results in increased degradation of cAMP and, thereby, a lowering of the activity of cAMP-dependent protein kinase (PKA). The reduced activity of PKA leads to a net dephosphorylation and decreased activity of hormone-sensitive lipase and reduced hydrolysis of triglycerides. Activation of the rat adipocyte PDE3B by insulin is associated with phosphorylation of serine-302. The mechanism whereby insulin stimulation leads to phosphorylation/activation of PDE3B is only partly understood. In rat adipocytes, lipolytic hormones and other agents that increase cAMP, including isoproterenol, also induce rapid phosphorylation, presumably catalyzed by PKA, of serine-302 of PDE3B. The phosphorylation is associated with activation of the enzyme, most likely representing "feedback" regulation of cAMP, presumably allowing close coupling of the regulation of steady-state concentrations of both cAMP and PKA and, thereby, control of lipolysis. In the review we describe methods and strategies used in the authors' laboratories to study phosphorylation and activation of PDE3B in adipocytes and in vitro.

Development of decompensated dilated cardiomyopathy is associated with decreased gene expression and activity of the milrinone-sensitive cAMP phosphodiesterase PDE3A

BACKGROUND

Phosphodiesterase III (PDE3) inhibitors are inotropic agents used to treat congestive heart failure (CHF) and are less effective in patients with severe CHF. Little is known about relative changes in PDE3 activity or gene expression during the evolution of cardiomyopathy.

METHODS AND RESULTS

In the present study, we evaluated temporal changes in PDE3A gene expression before and after pacing-induced CHF in nine mongrel dogs. Three weeks of left ventricular (LV) pacing produced LV end-diastolic pressures of 15+/-1.7 mm Hg, whereas overt CHF at 4 to 5 weeks was associated with LV end-diastolic pressures of 24+/-1.7 mm Hg; prepacing values were 6.6+/-0.6 mm Hg. Total RNA isolated from LV tissues was analyzed on Northern blots; 10 unpaced normal hearts served as tissue controls. Signals for PDE3A mRNAs (7, 8, and 10 kb) or PDE4D (7.6 kb) were normalized against glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or ribosomal 18S RNA. Before the onset of CHF, PDE3A/GAPDH ratios were not different between the control and 3-week paced groups. In contrast, all PDE3A/GAPDH ratios were selectively reduced by 52%, and PDE3A/18S was reduced by 70% (P<.05) in CHF; PDE4D/GAPDH (or 18S) was unchanged. LV tissues from four control and four CHF dogs were also processed to isolate cytosolic and microsomal membrane protein for cAMP PDE3 activity assays. CHF was associated with a significant 54% reduction (P<.05) in microsomal but not cytosolic PDE3 activity.

CONCLUSIONS

Selective downregulation of PDE3A may account in part for the ineffectiveness of milrinone in the treatment of severe CHF.

Characterization of two recombinant PDE3 (cGMP-inhibited cyclic nucleotide phosphodiesterase) isoforms, RcGIP1 and HcGIP2, expressed in NIH 3006 murine fibroblasts and Sf9 insect cells

cDNAs encoding PDE3 [cGMP-inhibited cyclic nucleotide phosphodiesterase (cGI PDE)] isoforms, cGIP1 and cGIP2, have been cloned from rat (R) and human (H) cDNA libraries. The deduced amino acid sequences of RcGIP1 and HcGIP2 are very similar in their conserved catalytic domains but differ in their N-terminal regulatory domains [Meacci, E., et al. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 3721-3725; Taira, M., et al. (1993) J. Biol. Chem. 268, 18573-18579]. cDNAs encoding both rat adipocyte RcGIP1 and human myocardial HcGIP2 (full-length forms and truncated forms lacking much of the putative N-terminal domain) were expressed in NIH 3006 fibroblasts and in Sf9 insect cells. The recombinant proteins exhibited the expected subunit molecular mass, immunologic reactivities, and characteristics of native membrane-associated forms of the enzymes, e.g., high affinity for cAMP (Km), sensitivity to the selective cGI PDE inhibitors OPC 3689 and OPC 3911 and to cGMP. The full-length recombinants were predominantly particulate, whereas the truncated HcGIP2 forms were cytosolic suggesting that N-terminal domains contain structural determinants important for membrane association. Both fibroblast RcGIP1 and authentic adipocyte cGI PDE were phosphorylated in vitro by cAMP-dependent protein kinase; tryptic [32P]peptides released from rat adipocyte 32P-cGI PDE and 32P-RcGIP1 exhibited identical electrophoretic profiles suggesting that the same peptides are phosphorylated in both.

Expression of type II cGMP-dependent protein kinase in rat kidney is regulated by dehydration and correlated with renin gene expression

cGMP-based regulatory systems are vital for counteracting the renin-angiotensin system (RAS) which promotes volume expansion and high blood pressure. Natriuretic peptides and nitric oxide acting through their second messenger cGMP normally increase natriuresis and diuresis, and regulate renin release; however, the severe pathological state of cardiac heart failure is characterized by elevated levels of atrial natriuretic peptide that are no longer able to effectively oppose exaggerated RAS effects. There is presently limited information on the intracellular effectors of cGMP actions in the kidney. Recently we reported the cloning of the cDNA for type II cGMP-dependent protein kinase (cGK II), which is highly enriched in intestinal mucosa but was also detected for the first time in kidney. In the present study, cGK II was localized to juxtaglomerular (JG) cells, the ascending thin limb (ATL), and to a lesser extent the brush border of proximal tubules. An activator of renin gene expression, the angiotensin II type I receptor inhibitor, losartan, increased cGK II mRNA and protein three to fourfold in JG cells. In other experiments, water deprivation increased cGK II mRNA and protein three to fourfold in the inner medulla where both cGK II, and a kidney specific CI- channel shown by others to be regulated by dehydration, are localized in the ATL. Whereas additional data suggest that cGK I may primarily mediate cGMP-related changes in renal hemodynamics, cGK II may regulate renin release and ATL ion transport.

Differential expression of cGMP-inhibited cyclic nucleotide phosphodiesterases in human hepatoma cell lines

PDE3 or cGMP-inhibited cyclic nucleotide phosphodiesterase (cGI PDE) activity was detected in homogenates of HepG2, Hep3B and HuH7, but not SK-Hep-1, human hepatoma cells. In HepG2 and Hep3B cells PDE3 activity was found predominantly in particulate fractions; in HuH7, in both particulate and supernatant fractions. cDNAs encoding two human PDE3s (an 'adipocyte' type, HcGIP1, and a 'cardiovascular' type, HcGIP2) have been cloned. HcGIP1 cDNA hybridized strongly with poly(A)+ RNA species from HepG2 and Hep3B. Both HcGIP1 and HcGIP2 mRNAs were expressed in Hep3B and HuH7 cells. The nucleotide sequence of an approximately 300-bp cDNA fragment, isolated after RT-PCR cloning from HepG2 RNA, was identical to a sequence within the conserved domain of HcGIP1 cDNA, consistent with the presence of HcGIP1 mRNA in HepG2 cells.

Effects of the 3',5'-phosphodiesterase inhibitors isobutylmethylxanthine and zaprinast on NO-mediated cGMP accumulation in the hippocampus slice preparation: an immunocytochemical study

The effect of inhibition of 3',5'-phosphodiesterase (PDE) activity on the cGMP accumulation was studied in control and nitric oxide (NO) stimulated hippocampal slices incubated in vitro using immunohistochemical visualisation of cGMP. Isobutylmethylxanthine (IBMX) was used as a non-selective PDE inhibitor and zaprinast was used as a selective inhibitor of cGMP-specific PDE activity. In the absence of PDE inhibitors cGMP-immunoreactivity (cGMP-IR) was found in blood vessel walls only. After incubation with the NO-donor sodium nitroprusside (SNP) cGMP-IR was found in a few isolated varicose fibres which were distributed throughout the slice. Incubation in the presence of either 1 mM IBMX or 10 microM zaprinast resulted in cGMP-IR in small numbers of varicose fibres distributed throughout the hippocampal slice. SNP in combination with IBMX resulted in cGMP-IR in small numbers multitude of varicose fibres throughout the slice; occasionally cell somata were observed. After incubation with SNP and zaprinast cGMP-IR was found in varicose fibres, although with a more restricted distribution and less numerous than in the presence of IBMX. In the latter combination, varicose fibres were observed predominantly in the CA2/CA3 region and in the stratum lacunosum molecular of the hippocampus, and cell somata were occasionally observed throughout the hippocampus. The differential distribution of cGMP-IR in the presence of different PDE inhibitors is consistent with the notion that there are regional differences in the localization of cGMP hydrolyzing enzymes in the hippocampus.

Differential cellular pattern of gene expression for two distinct cGMP-inhibited cyclic nucleotide phosphodiesterases in developing and mature rat brain

Cyclic GMP-inhibited phosphodiesterases are characterized by sensitivity of cAMP hydrolysis to inhibition by cGMP. This phosphodiesterase family contains at least two different isoforms (PDE3A and PDE3B) encoded by distinct genes and serving tissue-specific roles in regulation of lipolysis, glycogenolysis, myocardial contractility, and smooth muscle relaxation. Our previous work indicated an abundance of these two phosphodiesterase messenger RNAs in the embryonic rat brain, and therefore, to elucidate the potential functions of these enzymes in brain development as well as in mature brain function, the present study mapped cellular patterns of gene expression for these two enzymes from embryonic day 15 to adulthood using in situ hybridization histochemistry. Phosphodiesterase 3B isoform messenger RNA is uniformly expressed in germinal neuroepithelium and mature neurons, with distribution generally reflecting cell density. Phosphodiesterase isoform 3A messenger RNA, in contrast, demonstrates striking spatiotemporal specificity of expression, with three distinct patterns being evident. Firstly, this mRNA is highly abundant in both primary and secondary neuroepithelial germinal zones. Secondly, during early postnatal development PDE3A mRNA is transiently but highly expressed in neurons localized in basal forebrain, deep cerebellar, pontine, interpeduncular and a variety of thalamic, midbrain and brainstem nuclei. Thirdly, PDE3A mRNA is focally expressed in isolated large striatal and hippocampal neurons from the perinatal period without attenuation into adulthood. In summary, two cGMP-inhibited phosphodiesterase isoforms show distinctive patterns of gene expression in brain: PDE3B gene expression is uniform without evidence of system specificity or developmental stage specificity, suggesting that this isoform has a constitutive role in neuroepithelial metabolism, while PDE3B gene expression demonstrates a high level of spatiotemporal heterogeneity, suggesting that this isoform subserves a variety of developmental stage-specific and system-specific functions.

Insulin stimulates hormone-sensitive cyclic GMP-inhibited cyclic nucleotide phosphodiesterase in rat brown adipose cells

The presence and regulation of a hormone-sensitive cyclic GMP-inhibited cyclic nucleotide phosphodiesterase (cGI PDE) in rat brown adipose cells was investigated. cDNA clones for two cGI PDE isoforms, cGIP1 and cGIP2, have been isolated. Using a rat cGIP1 (RcGIP1) cDNA probe, RcGIP1 mRNA (approximately 5.3 kb) was detected in Northern blots of both brown and white adipose RNA. cGI PDE was detected in both microsomal and plasma membrane fractions of brown and white adipose cells by Western blotting using anti-RcGIP1 peptide antibody. When cells were incubated with insulin before membrane preparation, cGI PDE activity in the microsomal fraction was increased by 2- to 2.5-fold within 10 min. Isoproterenol also stimulated the activity of cGI PDE in the microsomal fraction by 1.5-fold. In cells incubated with both insulin and isoproterenol, microsomal cGI PDE activity was similar to that in microsomal fractions isolated from cells incubated with insulin alone. These results suggest that the hormonal regulation of cGI PDE, presumably a cGIP1 isoform, in rat brown adipose cells is similar to that in white adipose cells.