Figure 1.
Strategies to generate hGLP-1R knock-in and Glp-1r−/− mice.
(A) The targeting vector was injected into ES cells derived from a C57BL/6 line and implanted into BALB/c females, allowing generation of pure C57BL/6 offspring. The targeting construct was designed to insert into the downstream region of exon 1 of the mouse Glp-1r genomic locus. Chimeric male offspring were bred to C57BL/6-Tg (CAG-Flpe)2 Arte females that ubiquitously express Flp recombinase to give rise to the final hGLP-1R mouse line with option for conditional deletion. The humanized allele contains human GLP-1R cDNA with a preserved human intron 2, a region containing potential regulatory elements conserved across species. (B) Glp-1r−/− mice were generated by breeding hGLP-1R animals to Rosa26-Cre-transgenic mice. An additional breeding step was performed to remove the Cre gene, resulting in the Glp-1r−/− line.
Figure 2.
Reverse transcription for PCR validation of hGLP-1R, mGlp-1r and Glp1r−/− lines.
(A) A schematic of the gene that is expressed in the hGLP-1R mice. (B) The protein that is produced from this gene is a fusion of the mouse signal peptide (beige residues) and the human GLP-1R protein (white residues). The blue residues are those that differ between mouse and human GLP-1R. The signaling peptide is cleaved, leaving behind a human GLP-1R protein containing the C-terminal FLAG epitope (green residues). (C) cDNA was generated from total RNA isolated from islet and lung of hGLP-1R, mGlp-1r, and Glp1r−/− mice for RT-PCR. The 5′ P915 primer annealed in human exon 8, while the 3′ P913 primer annealed to the FLAG region, a unique site within the hGLP-1R gene. This PCR product is a 588 bp fragment only observed in the hGLP-1R mice. (D) A schematic of the gene that is expressed in the Glp-1r−/− mice. (E) Once the splice event occurs between human exon 2 and mouse exon 2, a frame-shift mutation nullifies downstream protein expression. (F) The final protein product in Glp-1r−/− mice is a 98-aa truncation mutant. The first 36 aa’s of the mature protein encode a fraction of the GLP-1R extracellular domain, and the remaining 40 aa’s constitute missense sequence that shows no similarity to known proteins. The RT-PCR and DNA sequence analyses of the Glp-1r−/− gene product demonstrate the Glp-1r−/− mouse does not code for a functional GLP-1R. (G) A schematic of the wild-type (mGlp-1r) gene. (H) Using the same primer pair, PCR products from Glp-1r−/− (372 bp) and mGlp-1r (275 bp) mice differ in size by 97 bp. (I) The wild-type GLP-1R protein is 463 aa’s including the signaling peptide.
Figure 3.
Glucose and mixed meal tolerance, and gastric emptying.
(A) IPGTT. EX-4 (10 nmol/kg SC) was administered 30 minutes before glucose injection (2 g/kg glucose, IP injection). Both mGlp-1r and hGLP-1R mice experienced glucose lowering in response to EX-4, while the Glp-1r−/− mice were refractory to the glucose-lowering effect of EX-4. (B) Insulin secretion during the IPGTT was measured in mGlp-1r, hGLP-1R, and Glp-1r−/− mice. (C) OGTT with 2 g/kg glucose gavage in mGlp-1r, hGLP-1R, and Glp-1r−/− mice after an overnight fast. (D) Glucose excursion and during the MMTT in mGlp-1r, hGLP-1R, and Glp-1r−/− mice after a 4 hour fast (n = 7/group). (E) Gastric emptying in response to EX-4 (10 nmol/kg SC) in mGlp-1r, hGLP-1R, and Glp-1r−/− mice. All animals were fasted overnight and then treated with vehicle (VEH, white bars) or EX-4 (black bars) prior to food intake in order to measure the rate of gastric emptying after a meal. *p<0.001 VEH vs. EX-4 in mGlp-1r mice; #p<0.001 VEH vs. EX-4 in hGLP-1R mice (n = 5–7 per group, one-way ANOVA with Bonferroni post-tests).
Figure 4.
Insulin secretion and GLP-1R protein expression in pancreatic islets.
(A) mGlp-1r and (B) hGLP-1R islets secreted insulin in response to high glucose (11.2 mM) compared to low glucose (2.8 mM) treatment, potentiated by GLP-1, OXM, GIP, and EX-4. (C) Glp-1r−/− islets also secreted insulin in response to high glucose, but no potentiation was observed with GLP-1, OXM, or EX-4 treatment. GIP-induced insulin secretion remained intact in Glp-1r−/− islets. (D) IP-WB of islets from mGlp-1r, hGLP-1R, and Glp-1r−/− mice showed FLAG expression in only the hGLP-1R islets. Hand-selected, size-matched islets were used for all insulin secretion assays, and comparisons are from high (11.2 mM) glucose: **p<0.01; ***p<0.001 (one-way ANOVA with Tukey’s multiple comparison test).
Figure 5.
Validation of GLP-1R and FLAG antibodies.
(A) mGlp-1r and (D) hGLP-1R were transiently expressed in HEK cells. cAMP accumulation in response to GLP-1 showed similar cAMP accumulation. HEK cells expressing mGLP-1R showed no (B) FLAG or (C) FLAG-tagged hGLP-1R expressing. HEK cells show stained for (E) FLAG and (F) hGLP-1R protein after transient transfection.
Figure 6.
IHC of pancreata from mGlp-1r, hGLP-1R, and Glp-1r−/− mice.
Pancreata were stained for insulin and showed positive staining in the β-cells of (A) mGlp-1r, (B) hGLP-1R, and (C) Glp-1r−/− islets. FLAG staining was also performed and (E) hGLP-1R islets stained positive for FLAG with none detected in (D) mGlp-1r and (F) Glp-1r−/− islets.
Figure 7.
IHC for human GLP-1R expression.
Islets from (A) mGlp-1r, (B) hGLP-1R, (C) Glp-1r−/− mice and (D) human pancreas were stained using an antibody specific for human GLP-1R, and only the (B) hGLP-1R islets and (D) human pancreas showed positive signal. Increased magnification showed that while (E) mGlp-1r and (G) Glp-1r−/− islets display no hGLP-1R expression, the (F) hGLP-1R islets and (H) human islets show plasma membrane staining for the human GLP-1R.
Figure 8.
GLP-1R mRNA and protein expression in vivo.
(A) qPCR analyses showed abundant expression in islets, lung, and stomach in both mGlp-1r and hGLP-1R mice. (B)–(C) Western blots using anti-FLAG affinity purification and immunoblotting showed abundant expression of the FLAG-tagged human GLP-1R in the lung and stomach of the hGLP-1R mice. Moreover, there was no FLAG-tagged GLP-1R detected in whole heart, kidney, small intestine (sm. int.), or liver of hGLP-1R mice. Using the FLAG signal as a surrogate for GLP-1R protein, no expression was observed for the receptor in Glp-1r−/− tissues, indicating that LoxP-mediated deletion of the hGLP-1R gene was successful. Purified bacterial alkaline phosphatase (BAP)-FLAG was included in gels as a control for the FLAG antibody. (D) Treatment of samples with PNGase F caused the multiple bands to collapse together from GLP-1R deglycosylation, resulting in a band of smaller molecular weight than the glycosylated form of the receptor.