GENOMICS 45, 185–193 (1997) ARTICLE NO. GE974936 Variable Expression of Hepatic Glucokinase in Mice Is Due to a Regulational LocusThat Cosegregates with the Glucokinase Gene J. Michael Moates,* Catherine Postic,† Jean-Francois Decaux,‡ Jean Girard,‡ and Mark A. Magnuson * , † ,1 †Department of Molecular Physiology and Biophysics and * Department of Medicine, Vanderbilt University Medical School, Nashville, Tennessee 37232; and ‡Centre de Recherche sur l’Endocrinologie Mole ´culaire et le De ´veloppement, Meudon 92190, France Received April 1, 1997; accepted July 25, 1997 neuroendocrine (NE) cells in the brain and gut, includ- The Gk activity locus affects expression of hepatic ing the pancreatic b cell, that are thought to sense glu- glucokinase (GK) in mice. Analysis of microsatellites cose fluxes (Jetton et al., 1994). The downstream GK within the mouse GK gene locus revealed two major promoter is expressed only in hepatocytes and deter- haplotypes in 19 of 22 inbred strains predictive of ei- mines the production of the hepatic GK isoform. Hepatic ther high or low hepatic GK gene expression. C3H/HeJ GK activity is determined by both transcriptional and mice, a high-activity strain, and two other wild-de- posttranscriptional regulational mechanisms. Tran- rived strains contain less common haplotypes. No cod- scription of the hepatic GK gene is affected by insulin, ing sequence differences were found in hepatic GK- glucagon, and other hormones and factors (Magnuson coding sequences from representative high and low et al., 1994; Printz et al., 1993). The amount of function- Gk activity strains, thereby excluding kinetic abnor- ally active hepatic GK is also affected at a posttranscrip- malities as the basis for hepatic GK activity differ- tional level by the binding of GK to the GK regulatory ences. Screening of Ç10 kb of potential regulatory protein, which determines both the activity (VanSchaf- DNA, including all eight known and three of four tingen, 1994) and the location of GK in hepatocytes newly identified DNase I-hypersensitive sites, by re- (Agius and Peak, 1993; Brown et al., 1997). striction enzyme fingerprinting – single-strand confor- mation analysis revealed a tetranucleotide microsatel- Mutations in the human GK gene cause maturity lite, the length of which was also predictive of the Gk onset diabetes of the young, type-2 (MODY-2) (Froguel activity phenotype. This tetranucleotide repeat is lo- et al., 1993; Vionnet et al., 1992), a form of early onset cated in the first intron of the hepatic transcription non-insulin-dependent diabetes mellitus character- unit and lies close to a newly identified liver-specific ized by an elevated plasma glucose set point (Byrne et DNase I-hypersensitive site. These results indicate al., 1994; Fajans, 1990; Velho et al., 1992). Recently, that the Gk activity alleles are a regulational locus as- both transgenic and gene knock-out mice were gener- sociated with the GK gene locus.  1997 Academic Press ated to characterize further the effects of altered GK gene expression on blood glucose homeostasis (Bali et al., 1995; Ferre et al., 1996; Grupe et al., 1995; Hari- INTRODUCTION haran et al., 1997; Niswender et al., 1997a; Terauchi et al., 1995). A variety of different inbred strains of Glucokinase (GK) plays a key role in glucose homeo- mice were used in these studies. Since small variations stasis, and the regulation of GK gene expression is com- in GK gene expression can have a direct effect on the plex. The mouse, human, and rat GK genes (Magnuson plasma glucose concentration, interpretation of the ef- et al., 1989; Postic et al., 1995; Tanizawa et al., 1992) fects of various transgenic and gene knock-out manip- contain two widely separated, cell-specific promoters ulations on glucose metabolism requires greater that give rise to at least two different GK isoforms knowledge of strain-specific differences that affect GK (Magnuson, 1990; Magnuson and Shelton, 1989). The gene expression. upstream GK promoter, which directs production of the islet GK isoform, is expressed in a variety of neural/ Variations in hepatic GK activities in different strains of mice were first identified 20 years ago. Based Sequence data from this article have been deposited with the Gen- on cross-breeding experiments in which a representa- Bank Data Library under Accession Nos. U93275 and U93276. tive high GK activity strain was mated with a low- 1 To whom correspondence should be addressed at 702 Light Hall, activity strain and in which the offspring were then Vanderbilt University School of Medicine, Nashville, TN 37232. Fax: (615) 322-7236. E-mail: mark.magnuson@mcmail.vanderbilt.edu. crossed back with the parental strains, Coleman first 185 0888-7543/97 $25.00 Copyright  1997 by Academic Press All rights of reproduction in any form reserved.