Novel promoter and alternate transcription start site of the human serotonin reuptake transporter in intestinal mucosa

Abstract

Selective serotonin-reuptake inhibitors are therapies for psychological and bowel disorders, but produce adverse effects in the non-targeted system. To determine whether human serotonin-selective reuptake transporter (SERT) transcripts in the intestine are different from the brain, rapid amplification of cDNA ends, primer extension and RT-PCR assays were used to evaluate SERT transcripts from each region. Potential SLC6A4 gene promoter constructs were evaluated with a secreted alkaline phosphatase reporter assay. A novel transcript of the human SLC6A4 gene was discovered that predominates in the intestine, and differs from previous transcripts in the 5'-untranslated region. The distinct transcriptional start site and alternate promoter suggest that gastrointestinal SERT can be differentially regulated from brain SERT, may explain why the polymorphism in the previously identified promoter is associated with affective disorders, but not associated with gastrointestinal dysfunction, and suggest the intriguing possibility of the development of site-specific therapeutics for SERT regulation in the treatment of multiple disorders.

Figure 1

PCR products of 5′-RACE assays on RNA obtained from the pontine region of the human brainstem and the human colonic mucosa separated on an agarose gel. A 161 bp reaction product was amplified from cDNA of the human brainstem, and sequencing confirmed this product was the previously identified exon 1a spliced to exon 2. A 327 bp reaction product was amplified from cDNA of the human colonic mucosa. Sequencing of this product identified a novel transcript contain 144 bp immediately upstream of the previously identified exon 1b, and exon 1b spliced to exon 2 (see Fig. 3).

Figure 2

PCR products resulting from the use of three different primer sets on cDNA from the human colonic mucosa (left) and human pontine brainstem (right) separated on an agarose gel. Primer set 1 contained a forward primer corresponding to a sequence in exon 1a and a reverse primer corresponding to a sequence in the coding region of exon 2. Amplification products of 136bp and 233bp correspond to transcripts containing exon 1a spliced to exon 2, and exon 1a spliced to exon 1b spliced to exon 2, respectively. The cDNA of the colonic mucosa amplified a 136bp product but not the 233bp product, while the cDNA from brainstem amplified both products. Primer set 2 contained a forward primer corresponding to 20 bp within the 53bp sequence in the genomic DNA (Genbank accession # NC_000017) upstream of exon 1c that shares identity with the mouse exon 1c and a reverse primer corresponding to a sequence in the coding region of exon 2. No amplification products from either the colon or brain samples were detected using this primer set. Primer set 3 contained a forward primer corresponding to a sequence in exon 1c and a reverse primer corresponding to a sequence in exon 2. The sequence of the 250bp amplification product corresponds to the same novel sequence identified by RACE and primer extension. cDNA from both the colon and brain generated this amplification product.

Figure 3

A) Schematic of the location of promoter construct inserts relative to the human SERT gene used in secreted alkaline phosphatase reporter assays. Promoter activity, as measured by the activity of secreted alkaline phosphatase, of five constructs transfected into Caco-2, human intestinal epithelial cells (B) or JAR, human choriocarcinoma cells (C). Data are the mean (± SEM) of four independent experiments with triplicate values for each experiment. Construct 1, which is upstream of exon 1a, demonstrated the highest activity in both Caco-2 cells and JAR cells. In JAR cells, this activity was over 90 times greater than the promoterless basic vector. In Caco-2 cells, this activity was nearly 8 times greater than the promoterless basic vector. Constructs 2, 3, and 4 all contain various lengths of genomic DNA upstream of exon 2, and demonstrated activity in Caco-2 cells that was 5, 6, and 7 times greater than the promoterless basic vector, respectively. Construct 5, however, demonstrated no significant activity above the promoterless basic vector in the Caco cells. Conversely in the JAR cells, constructs 2, 3 and 4 demonstrated no significant activity in comparsion to the promoterless vector, while construct 5 had levels of activity 5 fold greater than the promoterless vector. *P<0.05 compared to activity of the promoterless basic vector, ANOVA for repeated measures, with Bonferronis Multiple Comparisons Test.

Figure 4

Schematic representation of the organizations of the human and mouse slc6a4 genes and the three known transcripts of each gene that occurs via alternative splicing or alternate transcriptional start sites. Data for the human gene are from the present study while data for the mouse gene are from Ozsarac et al., and compared to mouse chromosome 11 (Genbank accession # NC_000077). To emphasize the transcribed sequences and to fit within the space provided, exons (shaded boxes) are to scale with other exons but are approximately ten times the scale of introns (solid lines). Introns are in scale to each other. Scales for both exons and introns are the same between the two species. Exons 4–14 for each species are not to scale as these exons are not different between each of the three transcripts for either species.