Glucose Sensing in L Cells: A Primary Cell Study

Summary Glucagon-like peptide-1 (GLP-1) is an enteric hormone that stimulates insulin secretion and improves glycaemia in type 2 diabetes. Although GLP-1-based treatments are clinically available, alternative strategies to increase endogenous GLP-1 release from L cells are hampered by our limited physiological understanding of this cell type. By generating transgenic mice with L cell-specific expression of a fluorescent protein, we studied the characteristics of primary L cells by electrophysiology, fluorescence calcium imaging, and expression analysis and show that single L cells are electrically excitable and glucose responsive. Sensitivity to tolbutamide and low-millimolar concentrations of glucose and α-methylglucopyranoside, assessed in single L cells and by hormone secretion from primary cultures, suggested that GLP-1 release is regulated by the activity of sodium glucose cotransporter 1 and ATP-sensitive K+ channels, consistent with their high expression levels in purified L cells by quantitative RT-PCR. These and other pathways identified using this approach will provide exciting opportunities for future physiological and therapeutic exploration.


Islet Isolation and FACS Sorting
Mice were killed by cervical dislocation and the pancreas injected immediately with 5 ml of collagenase V (1 mg/ml) in Ca 2+ and Mg 2+ -free HBSS. Pancreata were then dissected from the surrounding tissue and transported on ice. Following a 20 min digestion at 37 o C, the pancreas was disrupted by vigorous shaking, and the islets picked and re-picked manually into RPMI containing 10% FBS. Islets were disrupted into single cells by trituration following a 2 min incubation in Ca 2+ -free HBSS containing 0.1x trypsin/EDTA and 0.1% fatty acid-free bovine serum albumin (BSA). Cells were centrifuged at 300 g, re-suspended in supplemented RPMI and immediately sorted by flow cytometry. For islet sorting, we further subdivided non-fluorescent cells into a population that were Venus negative, larger (according to side and forward scatter) and with high background autofluorescence at 530 and 580nm, and a third population that were Venus negative, smaller and with low background autofluorescence. Mouse pancreatic islets were cultured in RPMI on glass coverslips for 48 hours. Islets and cultured colonic cells were fixed in 4% paraformaldehyde (PFA) for 30 mins and stored in 0.1% PBS-sodium azide at 4˚C until needed. Immunostaining was performed as in methods

Immunohistochemistry Methods
Cultured colonic cells were fixed in 4% paraformaldehyde (PFA) for 30 mins and stored in 0.1% PBS-sodium azide at 4˚C until needed. Freshly isolated mouse ileum and colon were fixed with PFA for 48 hours. Tissues were cryoprotected in 20% sucrose for 48 hrs and embedded in OCT prior to sectioning. Tissue sections (8 μm) and whole islets were blocked in 10 % goat serum for 1 hr and incubated with glucagon antibody (1:300, Santa Cruz Biotechnology, Inc.) for 3 hrs at room temperature, or with PYY antibody (1:100, Progen, Germany) overnight at 4˚C. Tissues were then incubated for 1 hr at room temperature with Alexa 635-(1:500), Alexa 555-(1:1000) conjugated goat anti-rabbit secondary antibodies (Invitrogen, UK), or with Alexa 633-conjugated goat anti-guinea pig secondary (1:300, Invitrogen, UK), as appropriate. Tissue samples stained with secondary antibody alone served as controls. Images were captured using a Zeiss LSM 510 META confocal microscope (Carl Zeiss, UK).
For SGLT1 staining of ileal sections an antigen retrieval step was required to recover the antigenic sites masked by paraformaldehyde fixation. This entailed heating the sections at 125 °C for 3 min in tri-sodium citrate buffer (10 mM, pH 6). The poly-L-lysine coated slides, onto which the tissue sections were mounted, were allowed to cool and subsequently rinsed with phosphate-buffered saline. Sections were blocked as described above and incubated at room temperature with an SGLT1 antibody (Mace et al., 2007) used at a dilution of 1:100. Owing to the loss of Venus fluorescence as a consequence of the antigen retrieval step, a fluorescein isothiocyanate-conjugated GFP antibody (1:100, Abcam) was also added for a 3 hour period (room temperature) to label the L-cells. To evaluate the specificity of the SGLT1 antibody, an excess of antigenic peptide was incubated with the antibody prior to tissue staining, as described previously (Mace et al., 2007).

EdU Staining
Cell proliferation was assessed using the Click-iT TM EdU assay kit (Invitrogen, UK). Cells isolated from mouse colon were cultured for a 3 day period prior to replacing growth media with fresh complete media supplemented with the nucleoside analog 5ethynyl-2'-deoxyuridine (EdU, 10µM). Cells were incubated in the presence of EdU for 24 hours and subsequently fixed with 3.7% PFA. The remaining steps of this assay were performed in accordance with the manufacturer's guidelines.

Immunofluorescence Microscopy of Cultured Cells
Cells grown on matrigel-coated glass-bottom culture dishes (Mattek Corporation, USA) for an 8-day period were fixed with 4% PFA for 30 minutes at room temperature and subsequently incubated for 1 hour in PBS (containing 0.1%, vol/vol, Triton X-100) supplemented with 10% (vol/vol) goat serum. Cells were then incubated for 3 hours at room temperature with either a PYY-, glucagon-or SGLT-1 primary antibody, each used at a dilution of 1: 100. After three 5 minute washes with PBS, cells were incubated with the appropriate Alexa 633-conjugated secondary antibody for 1 hour (room temperature). Finally, the cells were rinsed three times in PBS for 5 minutes and then covered with a glass coverslip. Control experiments were performed in parallel but with the primary antibody omitted. All of the presented images were captured using a Zeiss LSM 510 META confocal microscope (Carl Zeiss, UK).

Collection of Tongue Epithelium from Circumvallate Region
The epithelium of the tongue was separated by collagenase digestion, and the circumvallate region dissected using a stereomicroscope. mRNA was extracted as described in the methods for the treatment of FACS sorted cells.

Figure S1
A. The BAC construct for making transgenic mice was made by cloning Venus into the coding region of proglucagon (for further details see supplementary methods). B. Colocalisation of direct Venus fluorescence (green) with PYY immunofluorescence (red) in a fixed colonic slice. The scale bar represents 10 μm. C. Immunostaining for Venus (green) and SGLT1 (red) in a fixed ileal slice, showing the apical localisation of SGLT1 on the villus. The scale bar represents 10µm. D. Immunostaining for glucagon or PYY in 8-day old colonic cultures. Red, glucagon or PYY (as indicated); green, direct Venus fluorescence; blue, DAPI. DIC, Venus and DAPI overlays were generated using a 63x objective lens whereas the remaining images were generated using a 100x objective. All scale bars represent 20 µm. E. EdU incorporation (red) in 3-day old colonic cultures incubated for a further 24 hours in EdU. Blue, Hoechst; green, direct Venus fluorescence. Scale bar represents 20 µm. F. Immunostaining for SGLT1 in 8-day old colonic cultures. Red, SGLT1; green, direct Venus fluorescence. Scale bar represents 20 µm.

Figure S2. Expression of Venus in Islet Cells from Transgenic Mice
A. Colocalisation of direct Venus fluorescence (green) with glucagon immunofluorescence (red) in pancreatic islets. The scale bar represents 20 μm. B. Single cells were selected by flow cytometry according to their pulse width vs forward scatter characteristics (not shown), and by their yellow/green fluorescence (left) and forward/side scatter (right). Cell population 1 was collected by gating for Venus fluorescence in the yellow/green channels (green data points). Cell population 2 was collected by gating for a high background autofluorescence in the yellow/green channels (red data points) and higher forward and side scatter. Cell population 3 was collected by gating for a lower background autofluorescence (blue data points), and lower forward and side scatter. Cells in the 3 different gates on the yellow/green axes are labelled by the corresponding colours in the forward/side scatter plot. C. Expression of glucagon, insulin, somatostatin and pancreatic polypeptide in the three cell populations (1, green; 2, red; 3, blue) collected as in A and analysed by quantitative RT-PCR. Expression was normalised to that of β-actin in the same sample. Data represent samples from three sorts (each with islets pooled from 3-4 mice) for populations 1 and 2, and two sorts for population 3. Data are presented as geometric mean, and the error bar was calculated from the log(base 2) data.

Figure S3. Sucralose Pretreatment
Mixed colonic cultures were supplemented overnight with 20 mM sucralose (grey bars) or cultured in control unsupplemented media (white bars) and stimulated the next day in standard bath solution containing 10 mM glucose (gluc), or 10 mM glucose plus 20 mM sucralose, as indicated. GLP-1 release is expressed relative to the basal secretion in either untreated or sucralose-pretreated control wells, as appropriate. Basal secretion levels were 3.6% and 3.0% in untreated and sucralose-pretreated cells, respectively (not significantly different). Error bars represent 1 SE, n=3 samples each from a single experiment. Significance is shown relative to baseline using a one-sample t-test * p<0.05, ** p<0.01; or by comparing columns as indicated, Δ p<0.05, ΔΔ p<0.01.

L neg , Islet Cells and Enteroendocrine Cell Lines
Expression was analysed by quantitative RT-PCR and expressed as a ΔCT compared with β-actin in the same sample. Data are presented as mean ± SE, with n≥3 samples from separate mice for each value, except in the case of δ/PP cells, where n=2 (and therefore no SE). (<) marks data where fewer than 2 samples gave a detectable reading, indicating that the expression level is less than the value indicated. Expression in L pos and L neg cells from the same intestinal region were compared by Student's t-test; * p<0.05, ** p<0.01, *** p<0.001.