Postprandial Stimulation of GIP-Secretion 2 2 3 3 4 4 Frank Reimann, Eleftheria Diakogiannaki, Catherine E Moss, Fiona M
Total Page：16
File Type：pdf, Size：1020Kb
*Manuscript Click here to download Manuscript: Reimann et al GIP secretion 24sep19.docx Click here to view linked References 1 1 Postprandial stimulation of GIP-secretion 2 2 3 3 4 4 Frank Reimann, Eleftheria Diakogiannaki, Catherine E Moss, Fiona M. Gribble 5 6 5 7 6 From the Wellcome Trust/MRC Institute of Metabolic Science (IMS), 8 9 7 University of Cambridge, United Kingdom 10 8 11 9 12 10 13 11 14 15 12 16 13 17 14 Corresponding authors: 18 15 Frank Reimann and Fiona M. Gribble 19 20 16 Wellcome Trust/MRC Institute of Metabolic Science (IMS) 21 17 MRC Metabolic Diseases Unit 22 18 University of Cambridge, 23 19 Addenbrooke's Hospital, Hills Road 24 20 Cambridge CB2 0QQ, 25 26 21 Tel: 0044-1223-746796 or 336746 27 22 email: [email protected] or [email protected] 28 23 29 24 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 1 64 65 25 Abstract: 1 2 3 26 Glucose-dependent insulinotropic polypeptide (GIP) is a gut hormone secreted from the 4 5 27 upper small intestine, which plays an important physiological role in the control of glucose 6 7 8 28 metabolism through its incretin action to enhance glucose-dependent insulin secretion and 9 10 29 has also been implicated in postprandial lipid homeostasis. GIP is secreted from 11 12 13 30 enteroendocrine K-cells residing in the intestinal epithelium, which sense a variety of 14 15 31 components found in the gut lumen following food consumption, resulting in a circulating 16 17 18 32 plasma GIP signal dependent on the nature and quantity of ingested nutrients. In this review, 19 20 33 we explore the mechanisms underlying the control of GIP secretion, which have been 21 22 34 23 identified through combinations of in vivo, in vitro and molecular approaches. 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 2 64 65 35 Introduction: 1 2 36 3 4 5 37 GIP was originally isolated from the proximal small intestine as a peptide inhibiting gastric 6 7 38 acid secretion, and named accordingly gastric inhibitory peptide [1]. However, based on the 8 9 10 39 sequence homology of this peptide to other secretin family members known to stimulate 11 12 40 insulin secretion it was hypothesised that GIP might actually be an incretin, a hormone 13 14 41 15 released from the intestine in response to oral, but not intravenous, glucose, that stimulates 16 17 42 insulin secretion and indeed it was subsequently shown that exogenous GIP co-administration 18 19 43 with glucose had an insulinotropic effect [2]. The effect on gastric acid production, whilst 20 21 22 44 clear in dogs, was only seen at very high (supraphysiological) concentrations in man, which 23 24 45 resulted in its renaming to Glucose-dependent insulinotropic polypeptide, thus keeping the 25 26 27 46 GIP acronym (see [3] for a historical review). 28 29 47 Soon after the amino acid sequence of the GIP peptide was published [1] (a minor correction 30 31 32 48 to the sequence reducing it from 43 to 42 amino acids was published later [4]), antisera were 33 34 49 raised and used to identify the cellular origin of this hormone. Enteroendocrine cells (EECs) 35 36 50 found in the duodenum and slightly less frequently in the jejunum [5] staining for GIP were 37 38 39 51 originally thought to be a population known as D1-cells, but were subsequently identified as 40 41 52 K-cells, both named to distinguish them from other EECs which differ in their vesicular 42 43 44 53 appearance in “ultrastructural” electron-microscopic preparations, with some other EECs 45 46 54 already named after the hormones associated with them, such as the gastrin producing G-cell 47 48 49 55 [6, 7]. In recent years the classification of EECs into many different cell types based on 50 51 56 vesicular appearance after diverse staining techniques and correlation with one predominant 52 53 57 hormone has, however, been challenged, based on at the time unexpected co-localization of 54 55 56 58 hormones, such as GIP and the sister incretin glucagon-like peptide-1 (GLP-1) occasionally 57 58 59 in the same cells [8, 9], and even more promiscuous hormone expression profiles observed in 59 60 61 62 63 3 64 65 60 transgenic animals expressing genetic tags in specific hormone expressing cells [10, 11]. 1 2 61 Indeed, EECs are now thought to change their expression profiles whilst maturing along the 3 4 5 62 crypt villus or crypt-surface epithelial axis [12-14]. Nonetheless, hormones have preferential 6 7 63 expression profiles along the proximal to distal gut axis and GIP is found predominantly in 8 9 10 64 the very proximal small intestine, the duodenum and jejunum [15, 16]. In mice the overlap 11 12 65 with GLP-1 in the same cells is relatively limited, such that most duodenal GLP-1 expressing 13 14 66 15 cells do not co-express GIP and vice versa [17, 18] and we will continue to use the K-cell 16 17 67 terminology for GIP-expressing cells throughout this review, in which we will focus on the 18 19 68 sensing of secretory stimuli by these cells. 20 21 22 69 23 24 70 Sensing of macro-nutrients by K-cells 25 26 27 71 28 29 72 GIP levels in the circulation rise rapidly following food ingestion, or when nutrients are 30 31 32 73 infused into the duodenum. Infusion studies in humans have demonstrated that GIP secretion 33 34 74 is strongly determined by the rate of nutrient infusion (e.g. of glucose), suggesting that 35 36 75 circulating GIP levels predominantly reflect the rate of nutrient delivery from the stomach 37 38 39 76 into the duodenum, and hence the rate of gastric emptying [19] although the glucose 40 41 77 absorption rate in the duodenum/jejunum will also affect the signal [20]. As described below, 42 43 44 78 a range of different nutrient and non-nutrient components have been found to modulate GIP 45 46 79 secretion in different model systems. 47 48 49 80 50 51 81 Carbohydrates: 52 53 82 54 55 56 83 Oral, but not intravenous glucose was shown to be a strong stimulant of endogenous GIP 57 58 84 secretion as early as 1974 [21], which led to its classification as an incretin, as exogenous 59 60 61 62 63 4 64 65 85 GIP had also been shown to stimulate insulin secretion [2]. In the following decade it became 1 2 86 clear that carbohydrate that were substrates for sodium-dependent glucose transporters 3 4 5 87 (SGLT; Slc5a1), such as glucose, galactose and even non-metabolisable glucose analogues 6 7 88 such as -methyl-glucose pyranoside (MDG) were good GIP stimulants and that inhibition 8 9 10 89 of active sodium-dependent glucose uptake with phloridzin prevented glucose stimulated GIP 11 12 90 release [22]. We revisited this once we had established mixed intestinal epithelial cultures 13 14 15 91 and fluorescently labelled K-cells [23]. When cultures were co-treated with forskolin and 16 17 92 IBMX to elevate cytosolic cAMP levels, glucose and MDG were reliable stimulants of GIP 18 19 20 93 secretion (2-3-fold stimulation compared to forskolin/IBMX alone) and these responses were 21 22 94 abolished by phloridizin. SGLT1 was restricted to the apical surface of K-cells [23] and thus 23 24 25 95 ideally placed to respond to luminal changes in glucose concentration – comparable to our 26 27 96 previous observations in GLP-1 secreting L-cells [24, 25] - and we concluded that direct 28 29 97 depolarization through coupled glucose and sodium uptake into K-cells resulted in activation 30 31 2+ 32 98 of voltage gated Ca -channels triggering GIP secretion. This mechanism is supported by the 33 34 99 fact that SGLT1 knock-out animals lacked the rapid rise in plasma GIP levels normally 35 36 37 100 observed after glucose ingestion [26]. Interestingly, even at later time points (up to 2 h after 38 39 101 oral glucose challenge), no GIP response was observable in SGLT-1 knock-out animals, 40 41 42 102 contrasting with GLP-1 responses, which whilst compared to wild-type animals are reduced 43 44 103 at early time points (<15 minutes after glucose ingestion) are exaggerated at later time points 45 46 104 (1 and 2 h), probably reflecting delivery of glucose to the more distal intestine where it could 47 48 49 105 stimulate a greater number of L-cells presumably after being converted to SGLT-1 50 51 106 independent stimuli such as short-chain fatty acids by the microbiota [27]. This different 52 53 54 107 effect of GIP inhibition, but GLP-1 elevation following an oral glucose tolerance test is also 55 56 108 observed in human volunteers pre-treated with a mixed SGLT-1/2 inhibitor [28]. 57 58 59 109 60 61 62 63 5 64 65 110 Despite good evidence for a predominantly SGLT-1 dependent mechanism, other possible 1 2 111 glucose sensing pathways have been suggested to play a role in GIP secretion.