The people behind the development of the GIP receptor antibody have looked at the possible mechanisms (10) and focused on GIP receptor down regulation. It is known that GIP activation of its receptor is associated with recruitment of beta arrestins and that arrestins are needed for the subsequent internalization of the hormone receptor complex (46). By extended exposure of a GIP receptor expressing tissue to GIP, it would therefore be possible to create profound down regulation and therefore desensitization of the GIP receptor and impairment of the GIP sensitivity of the tissue. Indeed, this was directly demonstrated by Mohammad et al (47), who showed that an initial GIP stimulation can impair subsequent GIP stimulations, associated with disappearance of GIPR from the plasma membrane in 3T3-L1 adipocytes. This mechanism would be consistent with the remarkable lack of responses to increasing GIP concentrations, brought about by infusions of GIP, on top of the normal meal responses in healthy subjects (6). Furthermore, it was recently shown that the GIP receptor antagonist GIP (3–29)NH2 was able to restore the cell surface expression of the GIP receptor in transfected HEK293 cells after pre-incubation (and thereby agonist-induced receptor internalization) with endogenous GIP (46). Hence, it may be anticipated that antagonizing endogenous GIP actions in vivo, as can be done with both receptor antibodies and with peptide-based GIP receptor antagonists including GIP (3-29)NH2 in humans, would result in increased receptor expression on the cell surface, whereby the sensitivity of the system is regained. It is, however, still difficult to understand how GIP can activate the receptor in the presence of an antagonist, given the competitive nature of at least peptide-based GIPR antagonists (48). Nevertheless, the receptor internalization process is apparently important for GIP actions. For instance, when studied in vitro, the well-known GIP receptor mutation E354Q, which is associated with impaired glucose tolerance and increased fracture risk in postmenopausal women (49), actually shows enhanced agonist-mediated and basal 3′,5′-cyclic AMP formation and maintained arrestin recruitment, but prolonged agonist residence time, resulting in accelerated internalization and therefore impaired overall activation of the receptor signaling (50,51). This mutation is also associated with a slower recycling of internalized receptors to the cell surface, which, although it has been shown that the GIP receptor may also signal from endosomes (52), probably contributes to an overall impaired receptor function.
Thus, an effect on receptor recycling is apparently important for the actions of both GIP agonists and antagonists. But what about the effects of the GIP-GLP-1 co-agonists and their apparently beneficial metabolic actions? As previously discussed, the beneficial effect of GIP receptor activation is difficult to understand, as the effect of GIP is impaired in patients suffering from T2DM and obesity. So how can a dual-acting GIP-GLP-1 receptor agonist be better than the GLP-1 part of the combination? At first, it might be considered whether this is indeed the case. Upon closer scrutiny, the first dual GIP-GLP-1 co-agonist (NN9709, formerly MAR709 and RG7697) wasn’t terribly impressive after all, and its performance in a Phase 2 clinical trial did not differ from that of liraglutide (53). The second, tirzepatide, was clearly superior to the GLP-1 RA control, dulaglutide, in the dose-finding Phase 2 study mentioned in the beginning (7) although it was not ensured that optimal dosing had been investigated for the comparator—the fact that increasing doses of dulaglutide are currently being investigated (54) might suggest that the dose employed in the Phase 2 study was suboptimal. Nevertheless, as already mentioned, it is possible that the administration of a molecule that can activate both the GIP and the GLP-1 receptor may be beneficial in a sequential manner. Thus, the activation of the GLP-1 system might be the primary beneficial action, so that the beneficial effect of GIP may only be observed after metabolic control has been (partly) restored by GLP-1. In other words, the insulinotropic action of GIP may be regained after a GLP-1-mediated lowering of the blood glucose in agreement with the beneficial effects of intensive insulin therapy as previously mentioned (33). However, the disappointing results of adding high-dose GIP infusions to chronic liraglutide treatment (31) speak against this possibility. Another explanation could lie in a different pharmacodynamic profile of the dual agonist as compared to the individual signaling profiles of GIP and GLP-1, for instance caused by altered signaling of 1 or both of the 2 components. In fact, it has been shown that even small changes in the GIP as well as the GLP-1 molecule may change the receptor signaling towards a preferential G protein signaling with decreased arrestin recruitment and/or reduced receptor internalization (for GIP changes, see (51); for GLP-1 changes, (55, 56)). For the GIP system, such an effect would be beneficial due to a lower degree of receptor desensitization and internalization, and thereby improved therapeutic effect, given the proven downregulation of this system upon prolonged GIP administration (47,51,50). For the GLP1-1 system, receptor internalization seems independent of arrestin recruitment (57). Nevertheless, it was recently shown that N-terminal modifications of exendin-4 (a strong GLP-1 receptor agonist) will turn it into a biased agonist with a lower tendency to arrestin recruitment and/or receptor internalization and therefore with potentially greater efficacy and tolerability as a therapeutic (58). A similar alteration in GLP-1 receptor signaling profile was recently established for a dual acting GIP-GLP-1 peptide (59). It is thus possible that the observed beneficial effects in vivo of dual GIP-GLP-1 agonists—at least partly—rely on altered signaling of the molecule toward a biased signaling profile for one, or both, of the components. If both mechanisms apply to the GIP- GLP-1 co-agonists, the effect might be even greater. It is still unclear how the GIP part of the co-agonist would lead to weight loss, but if the incorporation of GIP activity in the co-agonist changes the GLP-1 signaling, then it would make sense that even the GIP part of the molecule might contribute to an enhanced weight-losing effect. It should be possible with careful molecular pharmacological experimentation to determine whether it is the influence of one part of the co-agonist (GIP) on the signaling pathways of the other part that makes a co-agonist like tirzepatide so effective, despite the overwhelming evidence that GIP, investigated in isolation, does not possess these activities. Such experiments are ongoing, and we will probably soon have at least some answers to this mind-boggling paradox.
