Trefoil factor 3 (TFF3) expression is regulated by insulin and glucose

Introduction : Trefoil factors are effector molecules in gastrointestinal tract physiology. They are classi ﬁ ed into three groups: the gastric peptides (TFF1), spasmolytic peptide (TFF2) and intestinal trefoil factor (TFF3). Previous studies have shown that trefoil factors are located and expressed in human endocrine pancreas suggesting that TFF3 play a role in: a) pancreatic cells migration, b) β -cell mitosis, and c) pancreatic cells regeneration. We speculated that the presence of TFF3 in pancreas, could be associated to a possible regulation mechanism by insulin and glucose. To date, there are not reports whether the unbalance in carbohydrate metabolism observed in diabetes could affect the production or expression of TFF3. Methods : We determined the TFF3 levels and expression by immunoassay (ELISA) and semi-quantitative RT-PCR technique respectively, of intestinal epithelial cells (HT-29) treated with glucose and insulin. Also, Real Time-PCR (RTq-PCR) was done. Results : Increasing concentrations of glucose improved TFF3 expression and these levels were further elevated after insulin treatment. Insulin treatment also led to the up-regulation of human sodium/glucose transporter 1 (hSGLT1), which further increases intracellular glucose levels. Finally, we investigated the TFF3 levels in serum of diabetes mellitus type 1 (T1DM) and healthy patients. Here we shown that serum TFF3 levels were down-regulated in T1DM and this levels were up-regulated after insulin treatment. Also, the TFF3 levels of healthy donors were up-regulated 2 h after breakfast. Conclusion : Our ﬁ ndings suggest for the ﬁ rst time that insulin signaling is important for TFF3 optimal expression in serum and intestinal epithelial cells.


INTRODUCTION
Th e trefoil family, are peptides of fewer than 80 amino acids that are present along the gastrointesti-nal tract. Th e three main trefoil peptide families are the gastric peptides (TFF1), the spasmolytic peptide (TFF2) and the intestinal trefoil factor (TFF3). Th ey play a signifi cant role in the conservation of the surface integrity of oral mucosa and improve healing of the gastrointestinal tract by a process termed restitution (1). Trefoil peptides are widely distributed in intestinal epithelial cells, and are present in saliva, meconium, human breast milk, and serum (1)(2)(3). Trefoil families are protease-resistant peptides that are amply secreted onto the intestinal mucosal surface by specifi c cells of the human gastrointestinal tract. Th e TFFs share an completely conserved distinctive motif of six cysteine residues which form three disulfi de bonds and defi ne the "trefoil" domain, which is also known as a 'P' domain (4). Previous studies have shown that trefoil factors are expressed in human endocrine pancreas playing important functions in the physiology of the pancreas (5). Jackerott et al (6) reported that trefoil factors are expressed in human and rat endocrine pancreas and TFF3 could be an important role in cells migration and regeneration. In the same way, Fueger et al (7) reported that TFF3 stimulates human and rodent pancreatic islet β-cell replication. Based in these results, we speculated that the presence of TFF3 in pancreas could be associated to a possible regulation mechanism by insulin and glucose. Glucose is a crucial fuel in humans and a key metabolic substrate. It is obtained directly from the intake, and by synthesis from other substrates in the liver. Dietary glucose and glucose synthesized within the body is transported through transport proteins. Th ese transporters are classifi ed into two main groups, a) the Na + -dependent glucose co-transporters (hSGLT), and b) the facilitative Na + -independent sugar transporters (GLUT family) (8). Th e hSGLT group transport mainly glucose with diff erent kinetics, through a secondary active transport mechanism. It is known that the Na + /K + ATPase pump plays a key role in this transport glucose mediated by hSGLT. Th is form of glucose transport takes place through the luminal membrane of cells lining the small intestine and the proximal tubules of the kidneys (9). Th e fi rst type of glucose transport protein cloned was the SGLT1. It is known that glucose could produce intracellular signaling across its metabolism inside the cells. In β-cells isolated from pancreas, it has been demonstrated that glucose metabolism increases ATP production, closing the KATP channels, which results in membrane depolarization (Δψ), thus opening of voltage-dependent calcium channels (VDCC) and allowing Ca 2+ infl ux. Th e resultant rise in [Ca 2+ ]i triggers insulin secretion and others intracellular signaling (10). To date, there are not evidences whether trefoil factors could be regulated by glucose and insulin and whether hyperglycemia and/or insulin defi ciency are involved in TFF3 expression. In the present work, we evaluated the role of insulin and glucose in TFF3 expression. Diabetes disease produces a decrease in β-cell mass, mediated generally by autoimmune destruction of insulin-producing cells in type 1 diabetes and by increased rates of apoptosis secondary to metabolic stress in type 2 diabetes (11). Currently, there are not reports about the serum levels of TFF3 in diabetes mellitus type 1 (DMT1) nor diabetes mellitus type 2 (DMT2). Here we showed that growing concentrations of glucose and insulin treatment improved TFF3 expression in intestinal epithelial cells. Also, the Insulin treatment led to the up-regulation of human sodium/glucose transporter 1 (hSGLT1), which further increases intracellular glucose levels. Finally, we showed that TFF3 was down-regulated in DMT1 patients and these values were modifi ed after insulin treatment. Th us, our fi ndings suggest for the fi rst time that insulin signaling is important for TFF3 optimal expression in intestinal epithelial cells by elevating intracellular glucose levels and by mediating gene expression.

Patients
Th e patient group consisted of 26 patients with T1DM treated at the Especialidades Clínicas Laboratory from February 2012 to July 2012. We obtained blood samples from each patient before insulin treatment and 2 hours after insulin treatment. Th e control group consisted of 28 healthy male donors who received a health check by clinicians at Carabobo Central Hospital from February 2012 to June 2012. For the validation of the results, a second cohort of patients, consisting of 18 patients with T1DM treated at the same laboratory from August 2012 to November 2012, were analyzed for serum TFF3 levels. Serum samples were collected for research under an institutional review board-approved protocol. Th is study was approved by the Institutional Review Board of the Applied Biotechnology Laboratory. Written informed consent was obtained from each participants.

Cultured cells
Th e HT-29 colonic cell line (passages 10-20) was obtained from the American Type Culture Collection (ATCC, Manassas, USA) and cultured according to the supplier's instructions in Dulbecco's Modifi ed Eagle medium (DMEM) containing 1000 gr/L (5.5 mM) glucose, 2 mM glutamine, 50 IU/ ml penicillin, 50 mg/ml streptomycin and 10% heat inactivated fetal bovine serum as standard medium at 37 ºC in a water-saturated atmosphere with 5% CO 2 . Th e HT-29 cells were treated with diff erent glucose concentrations (5.5 mM, 10 mM, 15 mM, and 50 mM) for 6 h. Also, the cells were treated with 100, 150 and 200 nM of insulin for 6 h.

SDS-PAGE and immunoblotting
HT-29 cell lysates treated with glucose and/or insulin were subjected to electrophoresis on 15% SDS-PAGE according to the method of Laemmli (12). After electrophoresis, the gels were either fi xed and proteins were visualized with 0.1% Coomassie brilliant blue R250 (Sigma) in methanol:water:acetic acid (Merck) (1:8:1) or they were electro-blotted onto nitrocellulose for 4 h at 4 o C (8-10 V/cm). Th e membranes were incubated 12 h. at 4 o C with blocking solution (5% nonfat dried milk in PBS containing 0.1% Tween-20). After being blocked, the membranes were incubated for 2 h at room temperature with PBS containing 5% dried milk powder and a 1:1000 dilution of rabbit anti-human TTF3 (Santa Cruz Biotechnology, cat. Nº. sc-28927) or mouse ati-human β-Tubulin (Santa Cruz Biotechnology, cat. Nº. sc-55529). Th e membranes were washed fi ve times in PBS-Tween and incubated with the peroxidase-coupled anti-rabbit secondary antibody (1:3000; Santa Cruz Biotechnology, cat. Nº. sc-2030, for TFF3) or anti-mouse secondary antibody (1:1000; Santa Cruz Biotechnology, cat. Nº sc-2005, for β-Tubulin) in PBS-Tween containing 5% nonfat dried milk, for 2 h at room temperature. Th e membranes were washed three times in PBS-Tween and specifi c bands were visualized by luminol reagent (Santa Cruz Biotechnology, cat. Nº. sc-2048).

ELISA
HT-29 cells were grown in 6-well plates at 50% confl uence and serum-starved for 24 h. Th e cell number in wells was normalized by seeding equal quantity of HT-29 cells, previously counted and diluted at final concentration of 2x10 5 cells/ml. Cells were then treated by 6 h with glucose and/or insulin at diff erent concentrations, as described in the text. Afterwards, culture supernatants of HT-29 were collected and centrifuged at 1000×g for 15 min at 4ºC and TFF3 was quantifi ed. In some experiments, ELISA test was done from serum of healthy and diabetic donors. Briefl y, the primary antibody used was: antihuman TFF3 (Santa Cruz Biotechnology, cat. Nº. sc-28927) at room temperature for 2 h (1:1000 diluted). Th en, the plates were washed and incubated at room temperature with peroxidase-coupled antirabbit secondary antibody (Santa Cruz Biotechnology, cat. Nº. sc-2030) diluted to 1:2000 in PBS plus 1 mL/L Tween 20 for 30 min. Plates were washed and incubated with chromogenic substrate in the dark at room temperature for 10 min. Stop solution (100 μl, 0.5M H 2 SO 4 ) was added to each well. Absorbance was measured at 405nm using a microtiter plate spectrophotometer Synergy HT (BioTek Instruments, Winooski, VT, USA).We quantifi ed TFF3 by simultaneous ELISA runs using TFF3 purifi ed as calibrators.

Reverse transcription polymerase chain reaction (RT-PCR)
RNA was extracted from HT-29 intestinal cells by Trizol TM (Invitrogen) according to the manufacturer's instructions. RNA concentration and purity were measured using a spectrophotometer Synergy HT (BioTek Instruments, Winooski, VT, USA). Total RNA (1 μg) was reverse transcribed into cDNA using a commercial kit (Invitrogen Th ermo-Script TM RT-PCR System), according to the manufacturer's instructions. Control reactions to check for DNA contamination were run in parallel with samples processed without reverse transcriptase. Th e primer sequences were: β-actin sense, 5'-CACGC-CATCCTGCGTCGGAC-3'; β-actin antisense, 5'-CATGCCATCCTGCGTCTGGAC-3'; TFF3 sense, 5'-CCCGGCTGTGATTGCTGCCA-3'; and TFF3 antisense, 5'-TCCTGTGACGTGGGT-GCCAGT-3'; hSGLT1 sense, 5'-CACGCAC-CAGGAGAGGGGAACA-3' and hSGLT1 antisense, 5'-GGCATTGTCACCACCCCAGCC-3'. Aliquots (10 μl) of the polymerase chain reaction products were electrophoresed on 1.5% agarose gels and stained with SYBR Gold nucleic acid gel stain (Molecular Probes, Invitrogen TM ). Densitometric analyses were performed using the image analysis software Quantity One (Bio-Rad laboratories, Hercules, CA, USA). Briefl y, the digital image was analyzed to determine the pixel intensity of each band. Relative quantities of TFF3 and hSGLT1 mRNA among diff erent preparations were calculated as the ratio of the TFF3:β-actin and hSGLT1:β-actin pixel intensities from three independent RT-PCR experiments. Positive results were based on the presence of DNA bands of the expected size.

Quantitative Real Time PCR (RTq-PCR)
Real time PCR was performed to precisely quantify the trefoil and sodium/glucose transporter 1 mRNA expression. Total RNA was isolated as described above. Total RNA (1 μg) was reverse transcribed into cDNA using a commercial kit (Invitrogen Th ermoScript TM RT-PCR System), according to the manufacturer's instructions. PCRs amplifi cation were performed using the Sybr Green kit (Applied Biosystems), GAPDH for normalizing the threshold cycle (Ct), while H 2 O was used as negative control. All measurements were performed in triplicate. Th e most of these primers have been previously reported (13). Th e results were analyzed by using the comparative Ct method. Th is method is based on the Sample data were normalized to glyceraldehyde-3-phosphate dehydrogenase mRNA and are presented as -fold change relative to mRNA from untreated cells. Assumption that target and reference template DNA amplifi es with the same effi ciency. Only PCR experiments producing a single DNA fragment, analyzed by gel electrophoresis, were used for the statistical analysis.

Effect of glucose and insulin treatment on TFF3 levels in HT-29 cells
ELISA was used to determine whether glucose and/ or insulin treatment induce regulation of Trefoil factor 3 (TFF3) in cell culture supernatants from intestinal epithelial cells collected after glucose and insulin treatment. HT-29 cells were grown in 6-well plates at 50% confl uence and serum-starved for 24 h. Cells were treated with diff erent glucose concentrations (5.5 mM, 10 mM, 15 mM, and 40 mM) for 6 h. Afterwards, culture supernatants of HT-29 were collected and centrifuged at 1000×g for 15 min at 4 ºC, the proteins were precipitated by trichloroacetic acid (TCA) and TFF3 levels was measured by ELISA. TCA precipitation permits concentrate proteins contained in cell culture supernatants and   prevent non-specifi c reactions in the immunoassay (14). As shown in Figure 1, the levels of TFF (Fig.  1A) increased at 10 mM glucose compared with 5.0 mM, but higher glucose concentrations (15 mM and 40 mM) did not increase the expression above the level achieved with 10 mM glucose. To verify that the increased TFF3 levels in cell culture supernatants by elevated glucose concentrations is due to a specifi c intracellular signaling induced by glucose and not the osmotic property of glucose, mannitol was used instead of glucose. Mannitol did not lead to the up-regulation of TFF3 levels in cell culture supernatants ( Figure 1B) suggesting that glucose is needed for this eff ect. In the same way, insulin treatment (100 nM) led increase of TFF3 levels in all glucose concentration tested (from 5 mM to 40 mM), indeed, 100 nM insulin treatment for 6 h increased the levels of TFF3 at 5.0 mM and 10 mM glucose ( Figure 1A). At higher glucose concentrations (15 mM and 40 mM), the insulin treatment did not further increased the levels of TFF3 ( Figure 1A). As these levels were not statistically diff erent than those with 5 mM and 10 mM glucose plus insulin treatment, these fi ndings may suggest that higher glucose FIGURE 4. Inmunoblot for TFF3 and Tubulin. HT-29 cells were grown in 6-well plates at 50% confl uence and serum-starved for 24 h. Cells were treated with different glucose concentrations (5.5 mM, 10 mM, 15 mM, and 40 mM), with or without insulin (100 nM) for 6 h. (A) Whole cell lysate (20 μg) from HT-29 cell monolayers which express TFF3, were run on non-denaturing 15% SDS-PAGE and electrotransferred to a nitrocellulose fi lter. Then, proteins in the membrane were denatured, renatured and blocked overnight at 4°C. TFF3 was detected using anti-TFF3 as primary antibody (upper panel). Then, membranes were stripped and reprobed using standard Immunoblotting to determine Tubulin on the lysate (lower panel). The relative amount of product was quantifi ed by densitometric analysis of DNA bands (B). Results are mean ± SEM of three independent experiments.
concentrations could lead to some insulin resistance in this cell line. Finally, higher insulin concentrations further increased TFF3 levels in cell culture supernatants from HT-29 ( Figure 1C). Th us, insulin and glucose are critically involved in mediating the up-regulation of TFF3.

Glucose and insulin produce up-regulation of TFF3 mRNA in HT-29 cells
Th e TFF3 levels increased in cells culture supernatants could be produced mainly by: 1) up-regulation in mRNA expression, or 2) increase in half-life time of TFF3 mRNA. To test which mechanism was involved in this process, semi-quantitative RT-PCR ( Figure 2B) and Real Time-PCR (RTq-PCR, Figure  2D) were used to fi gure out. HT-29 cells were grown to 50% confl uence in 6-well plates and serumstarved for 24 h. Cells were then treated with diff erent glucose and insulin concentrations (as described above) and total RNA was extracted from HT-29 intestinal cells by Trizol TM (Invitrogen) according to the manufacturer's instructions. As shown in Figure  2, the levels of TFF3 mRNA increased at 10 mM glucose compared with 5.0 mM, but higher glucose concentrations did not increase the expression above the level achieved with 10 mM glucose. In the same way, 100 nM insulin treatment for 6 h increased the levels of TFF3 expression at 5.0 mM and 10 mM glucose. However, higher glucose concentrations (15 mM and 40 mM) plus insulin (100 nM) did not increase the expression of TFF3 above the level achieved with 10 mM glucose plus insulin. Also, we tested whether the increase showed in TFF3 levels of cell culture supernatants treated with diff erent insulin concentration (from 100 nM to 200 nM) produced an increase in TFF3 mRNA. We found that higher insulin concentrations further increased mRNA TFF3 expression in HT-29 cells (Figure 3). However higher concentration of insulin (150 nM and 200 nM) did not further increased the levels achieved at 100 nM insulin. Finally, we did western blot from HT-29 whole cells lysate to confi rm that glucose and/or insulin produce up-regulation of TFF3. We found that the levels of TFF3 ( Figure  4) increased at 10 and 15 mM glucose compared with 5.0 mM, but higher glucose concentrations (40 mM) did not increase the level achieved with 15 mM glucose. In the same way, insulin treatment (100 nM) led increase of TFF3 levels only at 5.5 mM and 10 mM glucose concentration. Taken to-gether, these results show that the TFF3 is up-regulated by glucose and insulin in HT-29 intestinal epithelial cells.

Effect of glucose and insulin treatment on sodium/glucose transporter 1( hSGLT1) expressed in HT-29 cells
To further evaluate the eff ect of glucose and insulin on TFF3 expression, we explored the epithelial cell transporter hSGLT1 expression in intestinal epithelial cells HT-29. In this work, we showed that glucose and insulin produced augment in TFF3 expression, to test whether hSGLT1 was involved in this process, semi-quantitative RT-PCR (fi gure 5B) and RTq-PCR (fi gure 5D) were done. HT-29 cells were grown to 50% confl uence in 6-well plates and serum-starved for 24 h. Cells were then treated with diff erent glucose and insulin concentrations (as described above) and total RNA was extracted from HT-29 intestinal cells by Trizol TM (Invitrogen) according to the manufacturer's instructions. As shown in Figure 5 glucose did modifi ed the hSGLT1 mRNA levels in HT-29. Aditionally, 100 nM insulin treatment for 6 h increased the levels of hSGLT1 expression in all glucose concentration tested (from 5 mM to 40 mM). Th is result suggests that Insulin could facilitate glucose entry into the intestinal epithelial cell by increasing hSGLT1 expression.

Trefoil factor 3 expression is regulated by insulin and glucose in serum of Type 1 Diabetes Mellitus (T1DM) and healthy patients
In this work, we have shown that glucose and insulin play an important role in TFF3 expression. To date, there are not reports about the correlation between chronic disease related with carbohydrate metabolism such as diabetes and serum trefoil factors. In this sense, we investigated the serum levels of TFF3 in T1DM and healthy donors. Figure 6A shows the serum TFF3 levels of all patients divided in two main groups: 1) healthy donors, and 2) T1DM patients. All samples were taken at 7:00 a.m., fasting serum was collected and frozen until ELISA was done. In the control group (healthy patients), the serum TFF3 level was 7.11 ng/ml. Th is level was signifi cantly higher than in T1DM group ( Figure  6A). Next, we investigated if TFF3 levels could be modifi ed by the breakfast, and a second sample was taken from every healthy donor 2 h after breakfast (9:00 am.). After 2-h postprandial period, the participants showed up-regulation of serum TFF3 levels (fi gure 6B). Here we hypothesized that the increase of glycemic and insulin followed by the meal consumption were the responsible of the serum TFF3 augment. Th ese results are in concordance with our previous results that glucose and insulin produced increase of TFF3 levels and expression. Finally, we tested whether TFF3 levels are regulated by insulin treatment in each patient with T1DM ( Figure 6C). Two hours after treatment with insulin (each patient received specifi c insulin dosage recommended previously by its specifi c clinician), we found that TFF3 levels were signifi cantly increased after insulin treatment. Together, these results suggest that glucose and insulin regulate TFF3 expression.

DISCUSSION
Trefoil peptides are found along the gastrointestinal tract and play a central role in the physiology of the gut. TFF are located and expressed in many others organs and might be involved in several physiology process related with intracellular signaling, gene expression, and cell cycle regulation. Previous studies have shown that trefoil factors are expressed in human and rat endocrine pancreas suggesting that TFF3 might play an important role in: a) pancreatic cells migration, b) β-cell mitosis, and c) pancreatic cells regeneration. Based in these results, we speculated that the presence of TFF3 in pancreas, could be associated to a possible regulation mechanism by insulin and glucose. In this work, we showed that glucose and insulin play an important role in TFF3 expression. According to Jackerott et al (6), TFF3 are located with most insulin and some glucagonexpressing cells in the adult human pancreas, and this location correlates with the mRNA expression in isolated human islets. In this work, we showed that glucose and insulin produce augment in TFF3 levels. In this sense, previous studies have shown that others intestinal peptides like human beta defensins 1 (hBD-1) mRNA is directly up-regulated by glucose in cultured HEK-293 cells grown in 25 mM glucose for 4 days (15). Moreover, Barnea et al reported that glucose and insulin are needed for optimal defensin expression in human cell lines (13). According to Barnea et al, Increasing concentrations of glucose enhanced hBD-1 expression and these levels were further elevated after insulin treatment. Insulin treatment also led to the up-regulation of human sodium/glucose transporter 1(hSGLT1), which further increases intracellular glucose levels. In the same way, our results suggest that TFF3 is up-regulated by glucose and insulin. To date, there are not reports about the correlation between TFF3 and glucose/insulin regulation in serum and/or intestinal epithelium. In this sense, we report for fi rst time that TFF3 expression is regulated by glucose and insulin. It is known that 5.5 mM glucose is the normal glycemic level in humans, and these levels are controlled by insulin and glucagon mainly. We cultured Intestinal epithelial cells HT-29 in Dulbecco's Modifi ed Eagle medium (DMEM) supplemented principally with 10% fetal serum bovine (FBS). Subsequently, in our cell system insulin is not present in the medium, glucose transport into each cell might be slow and the intestinal cells might need higher glucose concentrations for optimal TFF3 expression. Indeed, 100 nM insulin treatment for 6 h increased the levels of TFF3 expression at 5.0 mM and 10 mM glucose. Th is work suggest that TFF3 expression in intestinal epithelial cells HT-29 is mediated by glucose transport into the cells, for the reason that TFF3 higher expression levels were achieved with increasing glucose concentrations in the absence of insulin (Figures 1, 2 and  4). However, insulin was also found to be an important factor mediating TFF3 expression (Figures 1-4). Th e key role of insulin in TFF3 expression could be through a specifi c intracellular signaling pathway that leads to glucose transport into the cell via others glucose transporters like GLUT4 and the induction of hSGLT1 expression (Figures 5 and 6) . On the other hand, is thinkable that insulin mediates TFF3 up-regulation through the pathway leads regulation of the TFF3 promoter in intestinal cells. However, these suggestions deserve further investigation. To date, there are not reports about the correlation between chronic infl ammatory disease, such as diabetes mellitus and TFF3 regulation in serum and/ or intestinal epithelium, and there is not evidence whether the unbalance in carbohydrate metabolism observed in diabetes could aff ect the production or expression of TFF3 in intestinal epithelial cells. In this work, we reported that in DMT1 serum, the TFF3 levels are lower than control people (healthy donors), however it is not well known the origin of TFF3 in serum. On the other hand, Serum Levels of Trefoil Factor Family Proteins have been associated with others diseases as Gastric Cancer (16). In this sense, it is known that Foveolar hyperplasia, spasmolytic polypeptide (TFF2)-expressing metaplasia, and intestinal metaplasia are histologic changes observed in patients with atrophic gastritis; they express TFF1, TFF2, and TFF3, respectively (16,17,18). Finally, here we report for fi rst time that intestinal TFF3 expression and TFF3 serum levels are regulated by glucose and/or insulin.

CONCLUSION
Th e glucose and insulin treatment led increase TFF3 levels in cell culture supernatants and up-regulation of TFF3 mRNA in HT-29 cells, suggesting that insulin and glucose are critically involved in mediating the up-regulation of TFF3. Also, glucose and insulin treatment did modifi ed the hSGLT1 mRNA levels in HT-29, suggesting that Insulin could facilitate glucose entry into the intestinal epithelial cell by increasing hSGLT1 expression. Finally, the serum TFF3 levels was signifi cantly higher in T1DM group than control group (healthy patients), suggesting that glucose and insulin regulate TFF3 expression.

CONFLICT OF INTEREST
Th e authors declare no competing interests.