PCR amplification products were generated with primers flanking the 5-HTTLPR using different genomic DNA templates and analyzed by DNA sequencing. The 484/528-bp DNA fragments for short (S) and long (L) alleles were shown as a result of PCR (Figure 1). In this study, we found an insertion of 42-bp (5′-CCTTC CAGCA TCCCC CTGCA CCCCC AGCAT CCCCC CTGCA GC-3′) fragment in L allele as compared to S allele. When analyzed using Jotan Hein Method of MegAlign program, this GC-rich sequence revealed 87.8% similarity to the previously described 44-bp fragment [11]. Additionally, it contains numerous potential binding sites for several transcription factors including TFII-1, STAT4, c-Ets-1, Elk-1, PEA3, GR-alpha, XBP-1, RXR-alpha and EBF (Figure 2) as analyzed using the Promo program [32,33].

We evaluated two human colon carcinoma cell lines, SW480 and HT-29, using RT-PCR approach in order to select the most suitable in vitro model. All mRNA transcripts examined were found to express in the two cell lines, but GRα transcripts were not detected in HT-29 (Figure 3). Therefore, SW480 was a cellular model chosen in this study as both estrogen and glucocorticoid receptors were expressed, indicating that this cell line could be employed for testing the effects of respective steroid hormones via 5-HTT.

The organization of the human 5-HTT gene promoter and map of three 5-HTTLPR luciferase reporter gene constructs, p5HTT[L860]Luc (−736 to +124 bp), p5HTT[S] Luc and p5HTT[L]Luc (−1796 to +124 bp) (GenBank accession no. X76753), were depicted (Figure 4). It was evident that the 5-HTT gene promoter contained a TATA-like motif and a number of potential transcription factor recognition sites. The 5-HTTLPR is the polymorphic tandem repeat of the 5-HTT gene promoter, defined by a length variation of a GC-rich repetitive sequence with 21-bp repeat elements found in this study. Using three different constructs for reporter gene assays, allelic-specific transcriptional activity of 5-HTTLPR was analyzed in SW480 cells, which were shown to express the human 5-HTT. As shown in Figure 5, we found the luciferase activity of the long variant promoter construct (p5HTT[L]Luc) was about two-fold greater than that of the short variant (p5HTT[S]Luc), and difference was statistically significant (P < 0.001 by Student’s t-test). The two shorter constructs possessed comparable level of luciferase activities regardless of the presence or absence of short variant of 5-HTTLPR.

In addition, we also investigated whether the steroid hormones, glucocorticoid and estrogen, regulate 5-HTT expression by employing the synthetic glucocorticoid dexamethasone, which binds specifically to the GR, and β-estradiol, which binds specifically to the ER. SW480 cell line was a proper model because it was demonstrated to express not only 5-HTT transcripts, but also transcripts for steroid hormone receptors, GRα and ERα. As shown in Figure 6, black and white columns represent reporter gene activity of the long variant allele p5HTT[L]Luc and the short variant allele p5HTT[S]Luc, respectively. No difference was found at 10⁻³, 10⁻⁶ and 10⁻⁹ M concentration when compared to 0 M dexamethasone or β-estradiol for the same allele.

5-HT is most commonly thought of as a neurotransmitter in the CNS. However, the predominant site of 5-HT synthesis, storage, and release is the EC cells of the intestinal mucosa. EC cells contain more than 90% of total 5-HT, which makes GI system the main source of 5-HT in the body. 5-HT plays various very important roles in peristaltic and secretory reflexes in response to chemical or mechanical stimuli of the gut [7]. 5-HTT plays a critical role in serotonergic neurotransmission that is responsible for the reuptake of 5-HT in the mucosa of the bowel, and is a factor that determines 5-HT activity. Additionally, the alteration of the serotonergic activity, examined by genetic variants in 5-HTT gene, especially 5-HTTLPR which has been implicated in several gut diseases, has been shown to be involved in many pathological processes such as IBS. In animal model study, 5-HTT knockout mice had diarrhea, which was associated with increased colonic motility and excretion of water in the stool [34]. Moreover, diarrhea predominant IBS (D-IBS) was associated with significant reduction in human mucosal 5-HT [35], but increased in plasma concentrations [36], implicating that 5-HTT levels were reduced consistent with genetic results. It has been reported that S/S genotype was significantly associated with D-IBS [25,37].

At present, there is no molecular information regarding the impact of 5-HTTLPR on 5-HTT transcription in cells originated from the GI tract, although certain studies were conducted using cells from blood, placenta and brain. In our study, we found the L allele containing 16 repeats has approximately 2-fold higher transcriptional activity of 5-HTT gene than S allele containing 14 repeats in SW480 cells, indicating that this region contains a positive transcriptional element or the insertion of the two extra repeats changes the binding profile of transcription factor by altering the physical distance. Our finding is in agreement with other investigations using JAR [11], lymphoblasts [12] and RN46A [13] with about 2- to 3-fold higher activity found in L than S variant. Taken together with other previous findings into account, the regulatory difference between the two alleles seems to be cell-specific; however, we could not summarize that this occurred only in 5-HTT expressed cells in this study because, after screening about 10 cell lines using RT-PCR, we could not find human cell lines not expressing 5-HTT. Rhesus monkeys have an analogous 5′-flanking region length variant generated by 21-bp insertion or deletion (rh5-HTTLPR), whereas rodents lack such length variants [9], indicating that 5-HTTLPR which is only present in humans and higher non-human primates. In this investigation, the two shorter constructs possessed comparable level of luciferase activities regardless of the presence or absence of short variant of 5-HTTLPR, suggesting that basal promoter activity could be obtained without the short variant of 5-HTTLPR.

In agreement with the reporter gene assays, lymphoblasts of L/L homozygotes have been demonstrated to yield higher 5-HTT mRNA levels than those containing at least one copy of the S allele [12]. Nevertheless, 5-HTT mRNA levels in human pons were not well correlated with the 5-HTTLPR alleles [38]. This inconsistency may be due to not only cell-type specificity, but also other complex processes in regulating the gene expression, such as epigenetics. For instance, the methylation of an upstream CpG island has been shown to be associated with 5-HTT mRNA levels as the L allele having lesser amounts of methylation than those of the S allele [39]. This effect was evaluated in lymphoblast cell line, and this may not necessary be true in other models. Moreover, these functional promoter assays could theoretically give spurious results because of not representing a closed loop of gene regulation. Effects in luciferase promoter assays generally tend to be large due to the lack of negative feedback, as the product of the reaction, the fluorescent protein luciferase, does not activate the negative feedback loops limiting its transcription in the way the protein product 5-HTT would. It is therefore important to look also at physiological cellular systems where these feedback mechanisms operate. However, investigation on 5-HTT protein expression in the cellular system, as uptake measurement or inhibitor binding, is possible, but using the radioactive method is required.

In this study, we also tested whether 5-HTT expression could be regulated by the effect of steroid hormones, glucocorticoid and estrogen, via 5-HTTLPR. However, the transcriptional effect of both steroids was not found in SW480 cells. This result was inconsistent with the previous finding [28] that demonstrated the acute glucocorticoid-dependent increased 5-HTT expression in JAR as an allele-specific promoter activation, suggesting that the glucocorticoid effect may be cell-specific. Besides, this result might account for the missing sequence of about 400 bp, which was not included in previous studies [13,40]. Moreover, a novel transcript of 5-HTT based on alternate promoter and transcriptional start site has been revealed to predominate in intestinal mucosa, which differs from previously reported transcripts [41]. This particular finding reflects a more complex picture of regulation of 5-HTT gene expression, thus highlighting the difference between GI and brain in this regard. Therefore, the steroid response element of 5-HTT gene in GI may be located in another promoter region since the recent study suggests that the putative EGF-responsive elements are presented in different regions of the two alternate promoters [42]. It is worth noting that herbal compounds, berberine and evodiamine, could influence the expression of 5-HTT via 5-HTTLPR [43], thus suggesting the potential therapeutic applications of such natural compounds.

Estrogen, the main sex hormone in women essential for the menstrual cycle, is another steroid also involved in serotonergic system. Both the neurotransmitter 5-HT and the ovarian steroid estrogen have been implicated in the modulation of mood and cognition, although the nature of their relationship has not been fully elucidated. Research using ovariectomized animals has identified estrogen-induced changes in 5-HT transmission, binding and metabolism in brain regions implicated in the regulation of affect and cognition [44]. Certain studies have shown the existence of ERβ mRNA and protein in serotonergic neurons [45,46]. Immunocytochemical localization in the mouse and rat brain also found ERβ-labeled cells, but slight ERα-labeled cells in neurons of the raphe nuclei [47,48], supporting that ERβ is more closely associated with 5-HT regulatory system than ERα. Although we were not able to detect ERβ transcripts in this study, SW480 and HT-29 were epithelial-like cells from intestinal mucosa that have been reported to express ERβ in both colon cancer and normal colon tissues [49,50]. Also, β-estradiol employed in this report is a substance for binding to both ERα and ERβ.

Acute estradiol administration in ovariectomized rats increases 5-HT and its metabolite (5-HIAA) in various brain regions as dorsal raphe and striatum [51,52], suggesting an increase in 5-HT turnover. In addition, there are studies establishing a relationship between estrogen and 5-HTT, but there often appear to be conflicting results regarding estrogen effects. Certain studies have shown that estrogen administration following ovariectomy increases 5-HT uptake [53] and the number of 5-HTT binding sites [54], whereas some have shown the reduction of 5-HT uptake [31] and 5-HTT mRNA expression [29,30] after estrogen treatment and a low estrogen state increases 5-HTT activity [55]. Estrogen has been shown to decrease mRNA [56] and the activity of monoamine oxidase [57], including increased 5-HT2A receptor mRNA [58,59].

Estrogen is mostly used in hormone replacement therapy (ERT) that is given to postmenopausal or surgically menopausal women in order to prevent osteoporosis and to treat the symptoms of menopause such as hot flashes and vaginal dryness. On the other hand, there are certain health risks for long time or overdose use including breast cancer and heart disease. Therefore, further investigations on their molecular interactions in CNS and GI system are indispensable. Also molecular understanding of the relationship between estrogen and serotonin will guide to new drug development or other novel therapeutic approaches in order to increase treatment efficiency and decrease side effects.

DNA was isolated from whole blood using FlexiGene DNA kit (Qiagen, Hilden, Germany), and subsequently screened by polymerase chain reaction (PCR) using Taq PCRx DNA polymerase (Invitrogen, Carlsbad, CA, USA) and primers specific to the 5-HTTLPR [11] as shown in Table 1. The product was generated as a 484/528 bp fragment for short (S) and long (L) alleles, respectively. PCR amplification was carried out in a final volume of 25 μL consisting of genomic DNA as a template, 0.2 mM each deoxyribonucleotide, 0.2 μM of sense and antisense primers, 2× PCRx enhancer solution, 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgSO₄ and 1.25 U of Taq DNA polymerase. After denaturing for 2 min at 95 °C, the desired DNA fragment was amplified for 35 cycles of 95 °C for 30 s, 60 °C for 30 s, and 68 °C for 1 min with the final extension at 68 °C for 10 min.

The extension products from PCR analysis step were subsequently purified from excess unincorporated dye terminators by ethanol precipitation according to the manufacturer’s instructions (ABI Sequencing kit, ABI, Foster City, CA, USA) and subjected to sequence analysis by the ABI Prism 310 Genetic Analyzer (ABI).

The sequencing data of S allele was compared to L allele by MegAlign program (DNASTAR, Madison, WI, USA). Transcription factor binding sites of the different sequences were analyzed by Promo program [32,33].

In this study, plasmid constructions for reporter gene assays were carried out as previously described [60,61]. Briefly, we produced three reporter plasmid constructs containing S, L and no 5-HTTLPR DNA fragments using two forward primers, −1796F/KpnI and −736F/KpnI, and reverse primer, +124R/HindIII, which are listed in Table 1. All fragments were amplified using Taq PCRx DNA polymerase (Invitrogen), and ligated into the promoterless luciferase expression vector, pGL3-Basic (a generous gift from Dr. Robert K. Yu, Institute of Molecular Medicine and Genetics, Medical College of Georgia, Georgia Health Sciences University, GA, USA) with KpnI and HindIII enzyme sites (New England Biolabs, Ipswich, MA, USA). All plasmid were verified by restriction mapping and DNA sequencing.

The SW480 and HT-29 human colon carcinoma cell lines were kindly provided by Dr. Apiwat Mutirangura (Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand) and Dr. Weerah Wongkham (Faculty of Science, Chiang Mai University, Chiang Mai, Thailand), respectively. They were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), and cells were maintained at 37 °C in a humidified atmosphere at 5% CO₂.

Total cellular RNA was isolated from SW480 and HT-29 cells using Trizol reagent (Invitrogen) following a procedure given by the manufacturer. The amount of RNA was determined by absorbance at 260 nm. Prior to RT-PCR reaction, about 1 μg of the total RNA was treated with 10⁻⁵ U deoxyribonuclease (DNase) I (Invitrogen) for 15 min at 25 °C. RT-PCR was performed using SuperScript™ III One-Step RT-PCR system with Platinum Taq DNA polymerase (Invitrogen) and specific primers listed in Table 2 for 5-HTT transcripts [62], GRα transcripts [63], ERα transcripts [64] and either glyceraldehyde 3-phosphate dehydrogenase (GAPDH) [65] or β-actin [66] transcripts as positive controls. The reaction was performed in a final volume of 25 μL, containing about 100 ng of DNase I-treated RNA, 0.4 μM of sense and antisense primers, 1× reaction mix (containing 0.2 mM dNTP, 1.6 mM MgSO₄) and SuperScript™ III RT/ Platinum Taq mix. cDNA synthesis was carried out at 55 °C for 30 min. After initial denaturation for 2 min at 95 °C, the desired DNA fragment was amplified for 40 cycles of 95 °C for 30 s, annealing for 30 s at 60 °C for 5-HTT, 52 °C for GRα, 58 °C for ERα, GAPDH and β-actin, and 68 °C for 2 min with the final extension at 68 °C for 10 min.

Transient transfections were performed using Lipofectamine™ 2000 (Invitrogen) following a procedure described by the manufacturer. Briefly, cultured SW480 cells at about 90% confluence in a ninety six-well plate were co-transfected with 0.2 μg of each reporter plasmid construct and 0.02 μg of the pRL-CMV vector (a generous gift from Dr. Robert K. Yu), which expresses Renilla luciferase under the control of the CMV promoter, to monitor the transfection efficiency. The cultured medium was changed in the following day, and cells were analyzed for the firefly and Renilla luciferase activities after an additional 24 h in culture. Alternatively, after transfection 24 h, either dexamethasone or β-estradiol (Sigma, St. Louis, MO, USA) was added in concentrations 10⁻³, 10⁻⁶ and 10⁻⁹ M. After an additional 24 h in culture, cells were analyzed for the firefly and Renilla luciferase acivities using Dual-Glo™ luciferase assay system (Promega, Madison, WI, USA) following the manufacturer’s instructions, and chemiluminescence was read using a VICTOR³ luminometer (PerkinElmer, Waltham, MA, USA). In each transfection experiment, untransfected cells were included as controls for a background level of firefly and Renilla luciferase. The pGL3-Basic vector was also transfected in parallel as a negative control.

The firefly luciferase activities were normalized to Renilla luciferase activities. Triplicate transfections were performed in each experiment. Luciferase levels were reported as fold elevation in activity over that seen in transfections with the promoterless pGL3-Basic vector. The data represented the mean ± S.D. of three independent experiments and were analyzed by the Student’s t-test. Differences at P < 0.05 were regarded as significant.