Elsevier

Cellular Signalling

Volume 15, Issue 6, June 2003, Pages 565-574
Cellular Signalling

Comparison of enzymatic characterization and gene organization of cyclic nucleotide phosphodiesterase 8 family in humans

https://doi.org/10.1016/S0898-6568(02)00146-8Get rights and content

Abstract

Full-length cDNAs of human cyclic nucleotide phosphodiesterase 8B (PDE8B) were isolated. Enzymatic characteristics of a dominant variant encoding a protein of 885 residues (PDE8B1) were compared with those of PDE8A1. The recombinant PDE8A1 and PDE8B1 proteins of an entire form were produced in both cytosolic and membrane fractions of the transfected COS cells. The human PDE8B1 was a high-affinity cAMP-PDE with Km value of 101±12 nM for cAMP, which is greater than that of PDE8A1 (40±1 nM). Relative Vmax value of PDE8A1 was 57±8% compared with that of PDE8B1 (100±12%). Although PDE8A1 was moderately inhibited by dipyridamole with IC50 value of 8±2 μM, the compound antagonized the PDE8B1 activity at three-fold higher concentration (IC50=23±2 μM). The human PDE8B gene was composed of 22 exons, spanning over 217 kb. Although overall sequence identity between PDE8A1 and PDE8B1 was 68%, positions of junctions of each exon between the PDE8A1 and PDE8B1 sequences were well matched, indicating evolutionary relatedness of both genes.

Introduction

Cyclic nucleotide phosphodiesterases (PDEs) participate in termination of cyclic nucleotide signalling through the hydrolysis of cAMP and cGMP, which are second messengers regulating many cellular functions in response to biologically active agents in various tissues. A conserved PDE catalytic domain composed of about 270–300 residues is located in the carboxy-terminal region of each PDE. Currently, 11 distinct PDE families (PDEs 1–11) have been recognized in mammalian tissues according to their amino acid sequence homology, biochemical characteristics and pharmacological profiles [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11]. In some PDE families, unique regulatory domains are situated in their amino-terminal parts. PDE1 is Ca2+/calmodulin-dependent PDE hydrolysing both cAMP and cGMP. PDE2 is stimulated by cGMP and hydrolyses cAMP and cGMP. A cGMP-inhibited PDE is PDE3. PDE4 is a rolipram-sensitive cAMP-PDE. A cGMP-binding and cGMP-specific PDE is PDE5. PDE6, a cGMP-PDE, functions as a photoreceptor involved in photo-signal transduction. A cAMP-specific and rolipram-insensitive PDE is PDE7. PDE8 and PDE9 are cAMP- and cGMP-specific PDEs, respectively, and are insensitive to a nonspecific PDE inhibitor, 3-isobutyl-1-methylxanthine (IBMX). PDE10A is a high affinity cAMP-PDE but also hydrolyses cGMP. PDE11A, a recent member, is a dual substrate PDE. Half of PDE families (PDE2, PDE5, PDE6, PDE10 and PDE11) contain two GAF (for cGMP binding and stimulated phosphodiesterases, Anabaena adenyl cyclases and Escherichia coli FhlA) domains [12] in their amino-terminal regions, and other PDEs lack the domain. The phylogenic tree analysis of PDEs indicated that these PDEs containing the GAF domain constitute the GAF-PDE family [13]. In addition to these distinct PDE families, multiple splice variants, some of which are expressed in a tissue-specific manner or in response to particular stimulation, are included in most of PDE families. Thus, cyclic nucleotide hydrolysis is controlled ingeniously by the PDE superfamily.

PDE8A and PDE8B are members of the PDE8 family. cDNAs encoding PDE8A and PDE8B were first identified by an approach using bioinformatics [5], [6]. Full-length cDNAs for PDE8A1 have been cloned from human and mouse [5], [14], [15], whereas only a partial cDNA sequence coding for the human PDE8B, which encodes a protein lacking an amino-terminal sequence of ca. 170 residues compared to the human PDE8A1, is available currently [7]. Characteristic structural feature of the PDE8 sequence is the presence of a PAS domain (for Per, ARNT and Sim proteins from which this domain was originally identified [16]) in their amino-terminal part. The PAS domain is observed in many signal transduction proteins. Although involvement of the PAS domain in protein–protein interactions and small molecular ligand binding is suggested, roles of the domain of the PDE8 proteins have not yet been understood. In humans, PDE8A transcripts are expressed in various tissues including testis, spleen, colon, small intestine and ovary in decreasing order [14]. Mouse PDE8A transcripts are rich in testis, followed by eye, liver, skeletal muscle, heart, 7-day embryo, kidney, ovary and brain [6]. In contrast to PDE8A, expression of human PDE8B transcripts is confined to thyroid gland [7], suggesting some particular role in this endocrine gland. Enzymatic characteristics of the recombinant PDE8A protein is high-affinity cAMP-specific and insensitive to a nonspecific PDE inhibitor, IBMX [5], [6], [7]. Km values of the recombinant mouse PDE8A1, which is a full-length form, and the amino-terminal truncated human PDE8A lacking the PAS domain, were 55 nM [6] and 150 nM [5], respectively. However, isolation of a full-length cDNA for the human PDE8B protein and kinetic analysis of this enzyme have not yet been done.

Here, we report an entire cDNA encoding the human PDE8B1 and some splice variant cDNAs of PDE8B. Enzymatic and structural characteristics of the human PDE8A1 and PDE8B1 proteins are demonstrated. The human PDE8A and PDE8B genes are also compared for their genetic organizations.

Section snippets

Materials

E. coli DH5α [7] was used as a host for plasmid construction. Restriction endonucleases, DNA-modifying enzymes and TAKARA LA Taq were obtained from Takara Shuzo (Kyoto, Japan). Marathon-Ready cDNA (human testis and thyroid gland), Advantage 2 Polymerase Mix and ExpressHyb were purchased from CLONTECH. The TA-cloning vector pGEM-T Easy and the mammalian expression vector pcDNA4/HisMax were products of Promega (Madison, WI) and Invitrogen (Carlsbad, CA), respectively. Anti-Xpress polyclonal

Cloning of full-length cDNAs for human PDE8B

A full-length cDNA for the human PDE8B cDNA was obtained by a combination of 5′ RACE and PCR (Fig. 1A). First, 5′ RACE reactions were performed using human thyroid cDNA as a template to obtain a novel 5′-extended sequence of PDE8B, which is lacking in the AF079529 sequence. An inserted DNA fragment (420 bp) carried by pPDEH8B19 was the longest form among four 5′ RACE products obtained. In order to find a sequence further extended to the 5′ end, a search for GenBank EST database was done with

Discussion

A full-length cDNA encoding the human PDE8B1 protein and enzymatic characteristics of PDE8B1 and PDE8A1 were demonstrated. The human PDE8B1 proteins was composed of 885 residues, which is larger than the human PDE8A1 protein of 829 residues, and 68% identical to the human PDE8A1 overall. PDE8B1 was a hydrophilic protein similar to PDE8A1 as reported in Ref. [14]. Recombinant human PDE8A1 and PDE8B1 proteins produced in COS cells were observed in both cytosolic and membrane fractions. Any

References (34)

  • S.H. Soderling et al.

    Curr. Opin. Cell Biol.

    (2000)
  • D.A. Fisher et al.

    Biochem. Biophys. Res. Commun.

    (1998)
  • S.H. Soderling et al.

    J. Biol. Chem.

    (1998)
  • M. Hayashi et al.

    Biochem. Biophys. Res. Commun.

    (1998)
  • D.A. Fisher et al.

    J. Biol. Chem.

    (1998)
  • K. Fujishige et al.

    J. Biol. Chem.

    (1999)
  • K. Yuasa et al.

    J. Biol. Chem.

    (2000)
  • M.D. Houslay

    Prog. Nucleic Acid Res. Mol. Biol.

    (2001)
  • L. Aravind et al.

    Trends Biochem. Sci.

    (1997)
  • P. Wang et al.

    Gene

    (2001)
  • J. Kotera et al.

    Biochem. Pharmacol.

    (2000)
  • S.F. Altschul et al.

    J. Mol. Biol.

    (1990)
  • P. Senapathy et al.

    Methods Enzymol.

    (1990)
  • V. Anantharaman et al.

    J. Mol. Biol.

    (2001)
  • P.A. Thomason et al.

    J. Biol. Chem.

    (1999)
  • J.A. Beavo

    Physiol. Rev.

    (1995)
  • M. Conti et al.

    Prog. Nucleic Acid Res. Mol. Biol.

    (2000)
  • Cited by (58)

    View all citing articles on Scopus

    The nucleotide sequences reported in this paper have been submitted to the DDBJ data bank with the accession numbers AB085824–AB085827.

    View full text