Nucleotide sequence of cDNA fragments coding for ...

Vol. 42, No. 6, September 1997 BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL Pages 1125-1 | 30

N U C L E O T I D E SEQUENCE OF cDNA FRAGMENTS CODING FOR THE FAD-, G L Y C E R O P H O S P H A T E - AND CALCIUM-BINDING DOMAINS OF HUMAN ISLET MITOCHONDRIAL GLYCEROPHOSPHATE D E H Y D R O G E N A S E

Anna NOVIALS t, Carmen FRANCO 1, Willy J. MALAISSE 2, Ramon GOMIS l

1Diabetes Unit, Hospital Clinic, Barcelona University, Villarroel 170, 08036 Barcelona, Spain and eLaboratory of Experimental Medicine, Brussels Free University, 808 Route de Lennik, B-1070 Brussels, Belgium Received May 26, 1997

SUMMARY : The cDNA fragments coding for the FAD-, glycerophosphate- and calciumbinding domains of mitochondrial glycerophosphate dehydrogenase (mGDH) were synthetized using RNA extracted from freshly isolated pancreatic islets of a normal subject and two-non-insulin-dependent diabetic patients. Single strand conformation polymorphism analysis of the PCR products yielded the same mobility as control cDNA probes. Likewise, the nucleotide sequence and corresponding amino acid sequence were identical to the normal gene bank sequence. These findings argue against the presence, in pancreatic islets, of an mGDH isoform distinct from that previously characterized in extrapancreatic organs.

INTRODUCTION The FAD-linked mitochondrial glycerophosphate dehydrogenase (mGDH) represents the key regulatory enzyme of the glycerol phosphate shuttle in pancreatic islets [1, 2]. Its activation by cytosolic Ca 2+ in glucose-stimulated islets indeed plays an essential role in the process of glucose recognition as an insutinotropic agent [3].

An inherited or acquired

deficiency of this enzyme may contribute to the preferential impairment of the pancreatic Bcell response to glucose, as distinct from other nutrient or non-nutrient secretagogues, in several animal models of non-insulin-dependent diabetes mellitus [4] and in some human patients affected by this disease [5]. The structural organization of the human mGDH gene, its chromosomal localization, and the identification by genomic cloning of pseudogenes were recently reported [6-8]. In some respects, however, the information so far available remains fragmentary, represents a matter of controversial interpretation or may require confirmation, as documented by the following considerations. First, in the sole study conducted in human islets, only normal subjects were used for the isolation of islets. Second, identification of different 5' mGDH

1039-9712/97/061125~36505.00/0 1125

Copyright 9 1997 by Academic Press Australia, All rights of reproduction in any form reserved.

Vol. 42, No. 6, 1997

BIOCHEMISTRYond MOLECULAR BIOLOGY INTERNATIONAL

mRNA regions led Ferrer et al. [6] to propose the possible presence of alternate promoters and/or the existence of variable translational efficiency in different cell types or developmental stages. This proposal was denied, however, in a subsequent publication [8]. Last, because the intrinsic catalytic properties of mGDH, as explored in homogenates of different tissues, are apparently different in islets and extrapancreatic organs [9], the possible existence of distinct isoforms of the enzyme cannot be ignored. In the present study, advantage was taken from the access to freshly isolated islets from both normal and diabetic subjects to reinvestigate the structure of the mGDH gene in the endocrine pancreas, with emphasis on the sequence of the gene coding for the FAD-, glycerophosphate- and Ca2+-binding domains of the islet enzyme.

MATERIALS AND METHODS

Pancreases were obtained from one non-diabetic and two non-insulin-dependent diabetic cadaveric organ donors, with informed consent of their families. The non-diabetic subject (A) was a 55 years old woman in traumatic coma. The first diabetic patient (13) was a 65 years old obese and hypertensive woman, who was treated with oral antidiabetic agents and underwent a cerebral stroke. The second diabetic patient (C) was a 72 years old man, previously treated by diet and suffering of a subarachnoidal bleeding. Islets were isolated by collagenase digestion of the pancreas and separated from the exocrine tissue by bovine serum albumin discentinuous density gradients [10]. Total RNA was extracted from 400 freshly isolated islets, using a Quick prep total RNA extraction reagent (Pharmacia Biotech, Uppsala, Sweden). The synthesis of eDNA from 1 gg total RNA was achieved by reverse transcription over 90 min incubation at 37~ with avian myeloblastosis virus reverse transcriptase (10 U/gl, Promega, Madison, WI, USA). Amplification of the eDNA sequences of interest was performed in an automated DNA thermal cycler (Perking Elmer Cetus, Branchburg, N J, USA) using [cr as tracer and the following primers. For the FAD-binding domain, 5'-TCCAGAGAAGCTAGCTACT-3' and 5'-GGATCAATTTAGTGCTTCTG-3' were used as primers sense and primer antisense, respectively. For the glycerophosphate-binding domain, the primer sense was 5'AGGGGATGTCCTGGCAGCAT-Y and the primer antisense 5'GCATTTATGGTATCTTCTGCC-3'. For the calcium-binding domain, the primer sense was 5'-GACAGGTATAAGAAGAGATT-3' and the primer antisense was 5'CATCAGCTGCAAAAATTCAT-3'. PCR amplification was carried out for 30 cycles over 30 s at 94~ 30 s at 57-60~ and 50-60 s at 72~ with a final extension for 5 min at 72~ Single strand conformation polymorphism analysis was conducted on 1 ~tt of each PCR product, which was mixed with 95 % formamide, 10 mM EDTA, 0.05 % bromophenol and 0.05 % xylen cyanol, heat-denaturated, loaded onto a 5 % polyacrylamide gel and rtm at either 4~ (calcium-binding domain), 25~ (FAD-binding domain) or 33~ (glycerophosphate-binding domain) at 20 W constant power. In the latter case, the gel also contained 5 % glycerol. The gels were dried and exposed to Kodak XAR films.

1126

Vol. 42, No. 6, 1997

BIOCHEMISTRYand MOLECULAR BIOLOGYINTERNATIONAL

9.

mE