Identification and characterization of transcriptional control region of the human beta 1,4-mannosyltransferase gene

Abstract

All asparagine-linked glycans (N-glycans) on the eukaryotic glycoproteins are primarily derived from dolichol-linked oligosaccharides (DLO), synthesized on the rough endoplasmic reticulum membrane. We have previously reported cloning and identification of the human gene, HMT-1, which encodes chitobiosyldiphosphodolichol beta-mannosyltransferase (β1,4-MT) involved in the early assembly of DLO. Considering that N-glycosylation is one of the most ubiquitous post-translational modifications for many eukaryotic proteins, the HMT-1 could be postulated as one of the housekeeping genes, but its transcriptional regulation remains to be investigated. Here we screened a 1 kb region upstream from HMT-1 open reading frame (ORF) for transcriptionally regulatory sequences by using chloramphenicol acetyl transferase (CAT) assay, and found that the region from -33 to -1 positions might act in HMT-1 transcription at basal level and that the region from -200 to -42 should regulate its transcription either positively or negatively. In addition, results with CAT assays suggested the possibility that two GATA-1 motifs and an Sp1 motif within a 200 bp region upstream from HMT-1 ORF might significantly upregulate HMT-1 transcription. On the contrary, the observations obtained from site-directed mutational analyses revealed that an NF-1/AP-2 overlapping motif located at -148 to -134 positions should serve as a strong silencer. The control of the HMT-1 transcription by these motifs resided within the 200 bp region could partially explain the variation of expression level among various human tissues, suggesting availability and importance of this region for regulatory role in HMT-1 expression.

Keywords: Dolichol-linked oligosaccharide; Gene expression; Mannosyltransferase; N-glycosylation; Transcription.

Fig. 1

Expression profiles of the HMT-1 in various human tissues by semi-quantitative RT-PCR. a RT-PCRs were conducted using first-stranded cDNAs derived from human skeletal muscle, brain, lung, liver, kidney, heart, spleen, thymus (included in Human MTC Panel I cDNAs, Clontech) and pancreas, testis, prostate, placenta, ovary, small intestine, colon, leukocyte (included in Human MTC panel II cDNAs, Clontech) as templates, as described in “Materials and methods”. After 30 and 34 cycles of PCRs, 10 μL of each was analyzed by 1 % agarose gel electrophoresis. b Control PCRs for glycerol-3-phosphate dehydrogenase (G3PDH) with gene-specific primers were also performed under the same condition, except by 22 and 26 cycles

Fig. 2

Determination of initiation site for the HMT-1 transcription by primer extension analysis. a The flanking region around initiation codon (boxed) of the HMT-1 was illustrated. The 5′-biotinylated primer A1-12RB used in the experiment was designed to anneal at +95 to +118 position downstream from initiation codon of the HMT-1 transcript (a horizontal arrow). b Product after primer extension reaction was subject to electrophoresis. Two numbers on the right side (−20 and −15) indicate the estimated positions where products, respectively, extending from the primer to −20 and −15 positions from initiation codon will migrate. By reverse transcriptase, single-stranded cDNA fragments with length of 137 base (a horizontal arrow in b) were yielded. This means that the primer was extended up to −19 position upstream the start codon (a vertical arrow in a), as indicated by a horizontal broken arrow in (a)

Fig. 3

Identification of potential cis-acting motifs within the 1 kb region upstream from the HMT-1 ORF. Nucleotide sequence located in the 1 kb region upstream from the initiation codon of the HMT-1 was determined from NUP-1000 subclone, as described in “Materials and methods”. Identification of potential cis-acting motifs was performed using TF SEARCH software and TRANSFAC database on the web, or GENETYX-MAC ver.12 software (GENETYX). Each position where a motif was located is indicated by arrow. One TATA box and three CAAT boxes are highlighted in grey color

Fig. 4

CAT assays for the 1 kb region upstream from the HMT-1 ORF. a Schematic representation of human genomic clone 165E7 (GenBank accession No. AC007011) derived from chromosome 16p13.3. It contains 5′-upstream region and exons 1–12 out of 13 HMT-1 exons near one end. As illustrated, long PCR with A1-G1F and A1-13R primers produced 6.5 kb fragment from HeLa genomic DNA, which was subsequently cloned into pUC118 vector as ET-8 construct (as described in “Materials and methods”). b For the first round of CAT assays, the 1 kb region upstream from the HMT-1 ORF was re-amplified from ET-8 by standard PCR using indicated primers, subcloned into pMAMneoCAT vector as NUP-1000 construct. Then, using NUP-1000, truncated subclones NUP-800, -600, -400 and -200 were constructed by the same procedure as preparation for NUP-1000, described in “Materials and methods”. c After preparing lysates from transfectants with NUP construct series and performing CAT assays, thin layer chromatography (TLC) of CAT reactions was carried out for approximately 30 min, using ethyl acetate as solvent. The negative control on TLC analysis (“Negative” in figure) was prepared by CAT reaction with lysate from HeLa cells, while positive indicator for acetyl-chloramphenicol (“Standard” in figure) was provided from the kit. d Results of CAT activities obtained after normalization are shown. As the CAT gene on pMAMneoCAT vector exhibits leaky expression due to distal RSV promoter and inducible expression due to proximal MMTV LTR promoter with dexamethasone (Dex), background levels of CAT activities with mock transfectants in the absence and presence of Dex have been also individually validated in parallel, as shown in (c) and (d). d Relative CAT activity with each construct was calculated against that with NUP-200 as 100 %. Each CAT activity was measured, normalized according to a formula shown in “Materials and methods” and averaged in at least five independent experiments. Each bar in the graph indicates standard error (SE)

Fig. 5

CAT assays for the 200 bp region upstream from the HMT-1 ORF. a Redrawing of potentially cis-acting elements localized within the 200 bp region upstream from the HMT-1 ORF. Based on these positions, the region is divided into eight segments, as indicated by vertical broken lines. Seven constructs where a segment was deleted one by one from upstream side were additionally prepared for the second round of CAT assays (NUP-156, -132, -112, -70, -57, 41 and -33). b Results of the second round of CAT assays with these constructs are shown. Relative CAT activity with each construct was calculated against that with NUP-200 as 100 %. Each CAT activity was measured, normalized according to a formula shown in “Materials and methods” and averaged in at least five independent experiments. Each bar in the graph indicates standard error (SE)

Fig. 6

CAT assays for mutagenized NF-1/AP-2 and NF-kB/AP-2 overlapping motifs. a Detailed information of nucleotide sequences for NF-1/AP-2 and NF-kB/AP-2 overlapping motifs. In upper figure, putative binding sites of NF-1, AP-2 and NF-kB are indicated by horizontal arrows and precise positions of nucleotide residues matching with their consensus sequences (boxed) are represented by vertical lines. In lower figure, positions of substitutional mutations introduced into NUP-156 and NUP-132 constructs are shown in small letters. b Results of CAT assays with four mutants, as well as those with NUP-156 and NUP-132 as controls. Relative CAT activity with each construct was calculated against that with NUP-132 as 100 %. Each CAT activity was measured, normalized according to a formula shown in “Materials and methods” and averaged in at least five independent experiments. Each bar in the graph indicates standard error (SE)