Mari T. Minowa

Two Distinct Promoters Drive Transcription of the Human D1A Dopamine Receptor Gene

The human D1A dopamine receptor gene has a GC-rich, TATA-less promoter located upstream of a small, noncoding exon 1, which is separated from the coding exon 2 by a 116-base pair (bp)-long intron. Serial 3′-deletions of the 5′-noncoding region of this gene, including the intron and 5′-end of exon 2, resulted in 80 and 40% decrease in transcriptional activity of the upstream promoter in two D1A-expressing neuroblastoma cell lines, SK-N-MC and NS20Y, respectively. To investigate the function of this region, the intron and 245 bp at the 5′-end of exon 2 were investigated. Transient expression analyses using various chloramphenicol acetyltransferase constructs showed that the transcriptional activity of the intron is higher than that of the upstream promoter by 12-fold in SK-N-MC cells and by 5.5-fold in NS20Y cells in an orientation-dependent manner, indicating that the D1A intron is a strong promoter. Primer extension and ribonuclease protection assays revealed that transcription driven by the intron promoter is initiated at the junction of intron and exon 2 and at a cluster of nucleotides located 50 bp downstream from this junction. The same transcription start sites are utilized by the chloramphenicol acetyltransferase constructs employed in transfections as well as by the D1A gene expressed within the human caudate. The relative abundance of D1A transcripts originating from the upstream promoter compared with those transcribed from the intron promoter is 1.5-2.9 times in SK-N-MC cells and 2 times in the human caudate. Transcript stability studies in SK-N-MC cells revealed that longer D1A mRNA molecules containing exon 1 are degraded 1.8 times faster than shorter transcripts lacking exon 1. Although gel mobility shift assay could not detect DNA-protein interaction at the D1A intron, competitive co-transfection using the intron as competitor confirmed the presence of trans-acting factors at the intron. These data taken together indicate that the human D1A gene has two functional TATA-less promoters, both in D1A expressing cultured neuroblastoma cells and in the human striatum.

A novel second isoenzyme of the human UDP-N-acetylglucosamine:alpha1,3-D-mannoside beta1,4-N-acetylglucosaminyltransferase family: cDNA cloning, expression, and chromosomal assignment.

We isolated a novel cDNA encoding a second isoenzyme of UDP-N-acetylglucosamine:α1,3-D-mannoside β1,4-N-acetylglucosaminyltransferase (GnT-IV; EC 2.4.1.145). The nucleotide and deduced amino acid sequences of the cDNA were homologous to those of the previously cloned human GnT-IV cDNA (63% and 62% identity, respectively). The new cDNA is also confirmed to express GnT-IV activity, suggesting that two isoenzymes of human GnT-IV exist. Although genomic Southern analysis suggested that both genes exist in many mammalian species and the chicken, northern analysis revealed that both genes are expressed in different ways in human tissues. This is the first report concerning the gene family of an N-acetylglucosaminyltransferase in mammals.

Kinetic properties and substrate specificities of two recombinant human N-acetylglucosaminyltransferase-IV isozymes

N-acetylglucosaminyltransferase (GnT)-IV catalyzes the formation of the GlcNAcβ1-4 branch on the GlcNAcβ1-2Manα1-3 arm of the core structure of N-glycans. Two human GnT-IV isozymes (GnT-IVa and GnT-IVb) had been identified, which exhibit different expression profiles among human tissues and cancer cell lines. To clarify the enzymatic properties of the respective enzymes, their kinetic parameters were determined using recombinant full-length enzymes expressed in COS7 cells. The Km of human GnT-IVb for UDP-GlcNAc was estimated to be 0.24 mM, which is 2-fold higher than that of human GnT-IVa. The Km values of GnT-IVb for pyridylaminated (PA) acceptor sugar chains with different branch numbers were 3- to 6-fold higher than those of GnT-IVa. To compare substrate specificities more precisely, we generated recombinant soluble enzymes of human GnT-IVa and GnT-IVb with N-terminal flag tags. Both enzymes showed similar substrate specificities as determined using fourteen PA-sugar chains. They preferred complex-type N-glycans over hybrid-types. Among the complex-type N-glycans tested, the relative activities of both enzymes were increased in proportion to the number of GlcNAc branches on the Man α1-6 arm. The Man α1-6 arm of the acceptors was not essential for their activities because a linear pentasaccharide lacking this arm, GlcNAcβ1-2Manα1-3Manβ1-4GlcNAcβ1-4 GlcNAc-PA, was a substrate for both enzymes. These results indicate that human GnT-IVb exhibits the same acceptor substrate specificities as human GnT-IVa, although GnT-IVb has lower affinities for donors or acceptors than GnT-IVa. This suggests that GnT-IVa is more active than GnT-IVb under physiological conditions and that it primarily contributes to the biosynthesis of N-glycans.