The physiological impact of an N‐terminal Halo‐tag on glucose‐dependent insulinotropic polypeptide receptor function in mice

The role of glucose dependent insulinotropic polypeptide (GIP) signalling in metabolic homeostasis is a hot area of research. The study of GIP receptor (GIPR) expression and trafficking, vital aspects in understanding GIPR physiology and pharmacology, has been hindered by the lack of validated GIPR antibodies, meaning that studies have relied on the presence of Gipr mRNA to determine GIPR expression in mouse and human tissue.1-3 The development of fluorophore-conjugated GIPR agonists has provided some insights into endogenous GIPR distribution and behaviour3, 4 but offers only indirect evidence of receptor localisation and may be misleading in some contexts, for example, if agonists dissociate from receptors after entering the endocytic pathway. G protein-coupled receptors (GPCRs) modified with self-labelling enzymatic tags such as SNAP, Halo and CLIP are becoming increasingly common in pharmacological research to study receptor expression and trafficking in vitro.5-7 Advances in genetic engineering have enabled the generation of mice which endogenously express proteins tagged with enzyme self-labels, such as the Glp1rSNAP/SNAP mouse model.8 This model was used to show that pancreatic islet GLP-1R constitutive internalisation was higher than reported using cell lines, emphasising the importance of studying receptor behaviours in their native environment. To this end, we developed a GiprHalo/Halo mouse model for the study of endogenous GIPR expression and trafficking. Methodology is provided in the Supplementary data. N-terminal tags facilitate cell surface-specific labelling (Figure 1A). Previous studies using heterologous expression systems show that GIPR tolerates N-terminal Halo- and SNAP-tag modifications.9, 10 To adapt this approach to natively expressed GIPR, we designed a GiprHalo/Halo mouse model, in which the sequence for the Halo-tag protein was inserted into the Gipr allele downstream of the GIPR signal peptide in exon 2, and loxP sites inserted between exon 5 (Figure 1B) via ES-cell-based gene targeting. To establish whether Halo-tagging affects physiological GIPR signalling, we investigated differences in body weight gain and handling of oral glucose administration between chow-fed Halo-tagged and wild-type genotypes. There were no significant differences in body weight between genotypes of either sex at 6 weeks (Figure 1C,D), or between 6 and 9 weeks measured in a subset of mice (Figure S1). This does not preclude central GIPR signalling disturbances as body weight differences between Gipr KO and wild-type mice are only revealed upon high-fat diet feeding.11 Plasma glucose responses to an oral glucose bolus were identical across all genotypes (Figure 1E,F). However, GiprHalo/Halo mice showed impaired sensitivity to human GIP (hGIP) administered exogenously during an intraperitoneal glucose tolerance test, requiring a 10 times higher dose to improve glucose tolerance than was needed for GiprWt/Wt and GiprHalo/Wt mice, suggestive of impaired pancreatic GIPR functionality (Figure 1G–R). The same trends were present across both sexes but did not reach statistical significance in all cases (Figures S2 and S3). In view of the reduced anti-hyperglycaemic response to hGIP in GiprHalo/Halo mice, we analysed cAMP signalling dynamics in dispersed pancreatic islets from GiprWt/Wt and GiprHalo/Halo mice transduced with the fluorescence-based cADDis cAMP biosensor and stimulated with a stepwise concentration gradient of hGIP or mouse GIP (mGIP) (Figure 2A–F). In accordance with our in vivo glucose tolerance results, islets from GiprHalo/Halo mice still responded to GIP, but there was a statistically significant reduction in potency compared to islets from GiprWt/Wt mice (mGIP EC50: GiprWt/Wt = 11.4 nM, GiprHalo/Halo = 72.2 nM [p = 0.01 by unpaired t test of logEC50s]; hGIP EC50: GiprWt/Wt = 42.4 nM, GiprHalo/Halo = 146.8 nM [p = 0.07 by unpaired t test of logEC50s]). We wondered if this impairment in cAMP signalling could be specific to endogenously expressed Halo-GIPR, which might be expected to reveal subtle differences in function unseen in heterologous systems. Therefore, we performed cAMP assays using AD293 cells transiently transfected with either untagged or Halo-tagged human or mouse GIPR. High-content cell surface labelling with the fluorescent GIP analogue GIP-TMR10 suggested no impact of the Halo-tag on delivery of the human or mouse GIPR to the plasma membrane (Figure 2G,H). However, in keeping with our ex vivo results from dispersed mouse islets, cells expressing the Halo-mGIPR displayed an impaired cAMP response to hGIP compared with untagged mGIPR (untagged mGIPR EC50 = 6.0 nM, Halo-mGIPR = 55.5 nM, p = 0.01 by paired t test of logEC50s), a finding also observed at the human GIPR albeit to a lesser extent (untagged hGIPR EC50 = 0.4 nM, Halo-hGIPR = 1.6 nM, p = 0.03 by paired t test of logEC50s) (Figure 2I,J). Thus, these results suggest that in in vitro models, N-terminal Halo-tagging of the GIPR may have either affected ligand binding or intracellular G protein signalling from the receptor. As exogenously expressed Halo-GIPR was expressed and trafficked to the cell surface in heterologous cells, we tested whether our Halo-GIPR mouse model could be used to visualise pancreatic islet GIPR expression by Halo-tag detection, a tissue chosen due to relatively high levels of Gipr expression. Unfortunately, initial attempts to label islets using Halo-probes were unsuccessful. While the use of improved Halo-probes may be a possibility for the future,9 we employed an antibody-based immunofluorescence approach which would potentially afford increased sensitivity from signal amplification; indeed, anti-Halo antibody staining revealed a significantly higher intensity of signal at the surface of islet cells in GiprHalo/Wt islets compared to GiprWt/Wt islets (Figure 2K–N). Although endogenous islet GIPR expression has previously been inferred via fluorophore-conjugated agonists,10 this is the first evidence of direct detection of endogenous GIPR in pancreatic islets using antibody labelling. Interestingly, we were unable to detect a signal in GiprHalo/Halo islets (Figure S4), suggesting that, unlike in overexpressing in vitro systems (Figure 2G,H), endogenously expressing both Halo-Gipr alleles appears to sufficiently alter receptor biosynthetic trafficking to prevent cell surface Halo-GIPR detection. This offers a potential explanation of our in vivo and ex vivo observation of impaired Halo-GIPR function in response to GIP stimulation. We generated a GiprHalo/Halo mouse model for the study of endogenous GIPR expression and trafficking. However, in and ex vivo analysis of pancreatic responses to pharmacological levels of GIP revealed that the N-terminally Halo-tagged GIPR displayed impaired function. Impaired signalling in heterologous cells transfected with the Halo-tagged GIPR despite normal cell surface expression, combined with an inability to detect cell surface Halo-expression in GiprHalo/Halo islets, suggests that impaired Halo-GIPR responses in and ex vivo could be due to both reduced cell surface receptor expression and impaired downstream signal transduction. In view of clear loss of GIP sensitivity, we suspect that normal glucose responses in GiprHalo/Halo mice during an oral glucose tolerance test reflects upregulation of GLP-1R signalling to compensate for the germline impairment of GIPR function.12 In conclusion, our results serve as a cautionary tale to the use of enzyme self-labels in endogenous GPCR labelling. While we advise against the use of GiprHalo/Halo mice to study GIPR functional responses, our success in labelling endogenous GIPR in pancreatic islets suggests that this model could still be a valuable tool in evaluating GIPR expression in tissues where this is contentious, such as adipose tissue and the CNS.1-4 The Section of Endocrinology and Investigative Medicine at Imperial College London is funded by grants from the MRC and NIHR and is supported by the NIHR Biomedical Research Centre Funding Scheme and the NIHR/Imperial Clinical Research Facility. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. The AT lab is funded by an MRC Project Grant (MR/X021467/1) and a Wellcome Trust Discovery Award (301619/Z/23/Z). ID is funded via a Medical Research Council Doctoral Training Partnership (MR/N014103/1). TM-MT is supported by the NIHR BRC and grants from MRC, Diabetes UK (DUK) and NIHR. BJ acknowledges funding support from the Medical Research Council (MR/Y00132X/1 and MR/X021467/1), the Wellcome Trust (301619/Z/23/Z), Diabetes UK, the Eli Lilly and Company LRAP Programme and Metsera Inc. AT also acknowledges funding from Diabetes UK, the Society for Endocrinology and the Eli Lilly and Company LRAP Programme. The authors thank the ES Cell & Transgenic Facility of the Laboratory of Medical Sciences, Medical Research Council for the rederivation of GiprHalo/Halo mice, Imperial College Facility for Imaging by Light Microscopy (FILM) for access to imaging facilities and technical support and Dr Aida Martinez-Sanchez and Prof Kevin G Murphy for providing animal project licence access. KWS is an employee of Eli Lilly and Company and may own company stock. SRB is an employee of Zihipp Ltd. TM-MT is a former consultant for and shareholder in Zihipp Ltd. AT has received funding from Eli Lilly and Company and Sun Pharma. BJ has received funding from Eli Lilly and Company, Metsera Inc and Sun Pharma, and acts as a consultant for Metsera Inc. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.16216. The data that support the findings of this study are available from the corresponding author upon reasonable request. Data S1. Supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.