The dual incretin co-agonist tirzepatide increases both insulin secretion and glucose effectiveness in model experiments in mice

Tirzepatide is a dual GIP and GLP-1 receptor co-agonist which is approved for glucose-lowering therapy in type 2 diabetes. Here, we explored its effects on beta cell function, insulin sensitivity and insulin-independent glucose elimination (glucose effectiveness) in normal mice. Anesthetized female C57/BL/6 J mice were injected intra-venously with saline or glucose (0.125, 0.35 or 0.75 g/kg) with or without simultaneous administration of synthetic tirzepatide (3 nmol/kg). Samples were taken at 0, 1, 5, 10, 20 and 50 min. Glucose elimination rate was estimated by the percentage reduction in glucose from min 5 to min 20 (K G ). The 50 min areas under the curve (AUC) for insulin and glucose were determined. Beta cell function was assessed as AUC insulin divided by AUC-glucose . Insulin sensitivity (S I ) and glucose effectiveness (S G ) were determined by minimal model analysis of the insulin and glucose data. Tirzepatide glucose-dependently reduced glucose levels and increased insulin levels. The slope for the regression of AUC insulin versus AUC glucose was increased 7-fold by tirzepatide from 0.014 ± 0.004 with glucose only to 0.099 ± 0.016 (P < 0.001). S I was not affected by tirzepatide, whereas S G was increased by 78% (P < 0.001). The increase in S G contributed to an increase in K G by 74 ± 4% after glucose alone and by 67 ± 8% after glucose + tirzepatide, whereas contribution by S I times AUC insulin insulin (i.e., disposition index) was 26 ± 4% and 33 ± 8%, respectively. In conclusion, tirzepatide stimulates both insulin secretion and glucose effectiveness, with stimulation of glucose effectiveness being the prominent process to reduce glucose.


Introduction
A key factor for glucose disposal after intravenous glucose administration is the secretion of insulin, since insulin stimulates glucose transportation into insulin-sensitive tissue, such as skeletal muscle, with reduction of circulating glucose [1].However, there is also an important insulin-independent mechanism to reduce glucose.This was reported already in 1937 [2] and later confirmed and quantified in 1979 by Bergman and collaborators using minimal model analyses of glucose and insulin data after intravenous glucose administration [3].The existence of an insulin-independent glucose elimination which contributes to glucose metabolism has subsequently been confirmed in several studies [4][5][6].We have transposed the minimal model to an experimental model in mice and demonstrated an important insulin-independent glucose disposal also in this species [7].We recently reviewed the early history of determining insulin-independent glucose disposal and studies exploiting this process [8].
Insulin-independent glucose disposal is named glucose effectiveness (acronym S G ) and it is usually assessed by minimal model of data achieved after intravenous glucose administration [3,4,[7][8][9][10].It contributes by more than 50-70% to glucose disposal, as shown in both humans and experimental animals [5][6][7], and it is reduced in obesity and diabetes [11][12][13].Augmenting glucose effectiveness may therefore be a potential target to improve glycemia in these conditions.
It has previously been demonstrated that glucagon-like peptide-1 (GLP-1) increases glucose effectiveness [14,15].This is of clinical interest in view of the prominent positioning of GLP-1 receptor agonists as glucose-lowering therapy in type 2 diabetes [16].The other incretin hormone, glucose-dependent insulinotropic polypeptide (GIP), has also been shown to increase glucose effectiveness, as demonstrated in mice [15].
Recently, the GLP-1 and GIP receptor co-agonist tirzepatide was approved for use as a glucose-lowering agent in type 2 diabetes [17][18][19][20][21]. Tirzepatide is a 39 amino acid linear peptide which is conjugated to a C20 fatty diacid moiety via a linker which is connected to the lysine residue at position 20 [22].It activates both GIP and GLP-1 receptors with a preference for the activation of GIP receptors more than of GLP-1 receptors [23].In fact, the affinity to GIP receptors is similar to native GIP, whereas its affinity to GLP-1 receptors is ≈ 5-fold lower than for native GLP-1 [22].Tirzepatide has been shown to reduce circulating glucose during an intraperitoneal glucose tolerance test in mice [22], and to stimulate insulin secretion from mouse [22,24] and human islets [24] and in humans [22,25].It is not known, however, whether tirzepatide, like GLP-1 and GIP [15], stimulates glucose effectiveness besides its stimulation of insulin secretion.Therefore, a primary aim of this study was to explore whether tirzepatide affects glucose effectiveness in model experiments in mice following intravenous administration of glucose.A second aim was to explore the glucose-dependency of the effect of tirzepatide to stimulate insulin secretion in mice.The study was undertaken as an acute study to avoid confounding influence through changes in body weight or other metabolic actions which are initiated by tirzepatide [22,24,25].

Animals
A total of 106 female C57BL/6 J mice from Taconic, Skensved, Denmark were used (mean body weight 20.8 ± 1.6 (SD) g).In four animals, the experiments failed due to technical problems (missed injection or blood sampling); therefore the completer population consisted of 102 animals.They were 4-6 months of age, maintained in a temperature-controlled room (22 • C) on a 12:12 h light-dark cycle (light on at 7:00 AM) and fed a standard pellet diet (energy 14.1 MJ/kg with 14% from fat, 60% from carbohydrate and 26% from protein; SAFE, Augy, France) and tap water ad libitum.Female mice were used to avoid the stress of single housing, which is used in male mice, and to be in line with the previous studies on GIP and GLP-1 [15,26].We used the mice randomly during the estrous cycle since it is not known whether glucose effectiveness is affected by the estrous cycle.The study was approved by the Lund/Malmö Animal Ethics Committee (Approval No. 5.8.18-06417/2020) and performed according to Good Laboratory Practice.

Experiments
Studies were undertaken in groups of 6-8 mice on each experimental day by one experienced technician.In all individual experiments, experimental and control animals were involved to avoid bias in different results on different days.After a 2 hr fast, at 9.30 AM, anaesthesia was induced with Fluafent (i.e., a mixture of fluanisone and fentanyl citrate) and midazolam, as previously described [27].In short, 10 mg fluanisone (Key Organics, Camelford, Cornwall, UK) was dissolved in 1 ml sterile water at 70⁰C for 60 min.This solution was mixed with 1 ml of fentanyl citrate (Sigma-Aldrich, St Louis, MO, U.S.A.; 0.315 mg/ml); 100 µl of this solution were given intraperitoneally to each mouse (0.016 mg fentanyl citrate and 0.5 mg fluanisone/mouse).Midazolam (0.167 mg/mouse; Roche, Basel, Switzerland) was also given (100 µl/mouse).Fifteen minutes later, an intravenous bolus dose of D-glucose over 3 s (dissolved in saline; Sigma; 0.125, 0.35 or 0.75 g/kg; doses of glucose selected to result in small, medium or high glucose peaks) or saline was injected in a tail vein (volume load 10 µl/g) alone or together with tirzepatide (3 nmol/kg; dose selected from a previous dose response curves for tirzepatide [22] and GIP and GLP-1 [15]; tirzepatide dissolved in 40 µmol/l Tris with added 0.02% polysorbate 80; Invivo Chem, Libertyville, Il, U.S.A.).Whole blood was sampled in heparinized pipettes from the intraorbital retrobulbar sinus plexus (40 µl) immediately before glucose injection (time t = 0) and at 1, 5, 10, 20 and 50 min after injection.Glucose was measured in the blood samples, and then plasma was separated by centrifugation and stored at − 20 • C until analysis for insulin.

Assays
All individual results from the completer population were included in the final analysis and statistics.Glucose was analyzed with the glucose oxidase method using AccuChek Aviva (Hoffman-La Roche, Basel, Switzerland).Insulin was determined by ELISA (Mercodia, Uppsala, Sweden).The intra-assay coefficient of variation (CV) of the method is 4% at both low and high levels, and the interassay CV is 5% at both low and high levels.The lower limit of quantification of the assay is 6 pmol/ l.

Data analysis
In the experiments with glucose administration at 0.35 g/kg, glucose and insulin data were applied to minimal model; therefore these groups had higher number of animals than the other experimental groups.Glucose disappearance rate was estimated as the net glucose elimination rate after the glucose injection (K G , the glucose tolerance index) as the slope from 5 to 20 min after glucose injection of the logarithmic transformation of the individual plasma glucose values [8].Insulin sensitivity index (S I ) and glucose effectiveness (S G ) were evaluated with the minimal model technique as explained elsewhere [9,10,28,29].S I is defined as the ability of insulin to enhance net glucose disappearance and inhibit glucose production, whereas S G is the net glucose disappearance per se from plasma without any change in dynamic insulin.Area under the curves for glucose (AUC glucose ) and insulin levels (AUC insulin ) were calculated by the trapezoid rule and beta-cell sensitivity to glucose was estimated by AUC insulin divided by AUC glucose .Disposition index was estimated by multiplying S I times AUC insulin .

Statistical analysis
Data and results are shown as mean±SEM.Differences between experimental groups were determined at each time point (Fig. 1), at each glucose level (Fig. 2) and between two groups (Table 1) using t-test, since insulin and glucose data were normalized, as examined using the Kolmogorov-Smirnoff test.Linear regression was used for estimating the relation between AUC glucose and AUC insulin .Multiple regression analysis was used for estimating the relative contribution of S G and disposition index to K G .For all analyses, carried out using SPSS, v. 27, statistical significance was defined as P < 0.05.

Effects of tirzepatide on glucose and insulin levels (Figs. 1 and 2)
When tirzepatide was injected together with saline (i.e., without glucose) or with glucose at 0.125, 0.35 or 0.75 g/kg, there was a glucose-dependent reduction in glucose levels and enhancement in insulin levels compared to controls.Fig. 1 shows that glucose levels were significantly reduced by tirzepatide after 20 and 50 min in all tests, with significant reduction of AUC glucose (Fig. 2).Furthermore, insulin levels were significantly enhanced by tirzepatide at 1 min in the saline test, after 5 min in the 0.125 g/kg glucose test, after 1, 5 and 10 min in the 0.35 g/kg glucose test and after 1, 5, 10 and 20 min in the 0.75 g/kg glucose test (Fig. 1).AUC insulin was significantly enhanced by tirzepatide after 0.35 and 0.75 g/kg glucose (Fig. 2).Glucose disposal (K G ) after glucose injection was significantly enhanced by tirzepatide in all glucose tests (Fig. 2).

Effects of tirzepatide on the regression between AUC glucose and AUC insulin and beta cell sensitivity
Fig. 2 shows the relation between AUC glucose and AUC insulin after glucose (or saline) with or without tirzepatide.It is seen that there was a clear increase in AUC insulin in relation to AUC glucose by tirzepatide.The regression slope between the two parameters was 0.014 ± 0.004 nmol insulin/mmol glucose in glucose controls (complete function AUC insulin = 5.3 + 0.014 x AUC glucose , r = 0.474, P < 0.001) and this slope was 7fold increased to 0.099 ± 0.016 nmol/mmol by tirzepatide (P < 0.001) (complete function AUC insulin = − 21 + 0.099 x AUC glucose , r = 0.67, P < 0.001).Also the estimated beta cell sensitivity was significantly increased by tirzepatide in a glucose-dependent manner.

Effects of tirzepatide on glucose effectiveness and insulin sensitivity (Table 1)
Table 1 shows the data derived from the minimal model analyses after the intravenous administration of glucose at 0.35 g/kg with and without tirzepatide.It is seen that after glucose administration at 0.35 g/kg, glucose elimination rate was 73% higher (P < 0.001), beta cell sensitivity was 55% higher (P < 0.001), AUC insulin was 32% higher (P = 0.002) and AUC glucose was 18% lower (P < 0.001) after gluco-se+tirzepatide when compared to glucose alone.Furthermore, modeling of the glucose and insulin data revealed that insulin sensitivity (S I ) was not significantly affected by tirzepatide whereas glucose effectiveness (S G ) was significantly increased by 78% with tirzepatide (P < 0.001).Multiple regression analysis revealed that the increase in S G contributed to the increase in K G by 74 ± 4% after glucose versus by 67 ± 8% after glucose+tirzepatide, and that the corresponding contributions by S I times AUC insulin (i.e., disposition index) was 26 ± 4% and 33 ± 8%, respectively.

Discussion
This study explored the effects of the dual incretin hormone receptor co-agonist tirzepatide on glucose and insulin data after intravenous glucose administration in model experiments in mice.The main findings were that tirzepatide glucose-dependently increased glucose disposal and beta cell function and that glucose effectiveness was increased by tirzepatide.
As recently reported and reviewed, glucose disposal is mediated not only by insulin dependent but also by insulin independent mechanisms and that impairment of both these mechanisms are of relevance for the development of glucose intolerance and type 2 diabetes [5][6][7][8][9][10][11][12][13].As desrcibed previously [9,28], glucose effectiveness is calculated from the minimal model of insulin and glucose data after an intravenous glucose load.The model accounts for the effect of insulin and glucose itself on glucose disappearance after exogenous glucose injection and S G describes glucose disappearance from plasma without any change in dynamic insulin.Previous studies have shown that exogenous administration of both GLP-1 and GIP increase S G in model experiments in mice [15] and GLP-1 has also been shown to increase S G in humans [14].Comparing the results in our previous work in mice [15] with the present results shows that S G is higher after tirzepatide (0.096 min -1 ) than after both GIP (0.072 min -1 ) and GLP-1 (0.066 min -1 ).This would suggest that tirzepatide is more efficient in augmenting glucose effectiveness than either of the two incretin hormones alone.However, to establish this difference in potency requires head-to-head studies between the compounds in the same study.
Our finding that tirzepatide, which activates both GLP-1 and GIP receptors, stimulates S G would suggest that also the glucose-lowering action of tirzepatide, as of GLP-1 and GLP-1 receptor agonists, relies not only on stimulating insulin secretion but also on augmenting glucose effectiveness.This is of clinical relevance since tirzepatide is now approved for its use as a glucose-lowering medication in type 2 diabetes [17][18][19][20][21].Our estimation of the relative contribution showed that S G contributed by ≈ 67% to glucose disposal after tirzepatide.This is similar as in the group given glucose alone (contribution by ≈74%).This shows that the relative contribution by insulin secretion and glucose effectiveness to glucose disposal was not significantly altered by Fig. 2. Area under the 50 min curve (AUC) of glucose and insulin, 5-20 min glucose elimination rate (K G ), the relation between AUC glucose and AUC insulin and beta cell glucose sensitivity after intravenous injection of saline (0 glucose) or glucose (0.125, 0.35 or 0.75 g/kg) alone or with administration of tirzepatide (3 nmol/kg) in C57BL/6 J mice.The number of animals in each group is reported in Fig. 1.NA means that K G cannot be estimated without any glucose injection.Mean±SEM are shown.Asterisks indicate the probability level of random difference between the groups, as determined by t-test (*P < 0.05, **P < 0.01, ***P < 0.001).tirzepatide compared to glucose alone, which indicates that the co-agonist increases insulin secretion and glucose effectiveness in parallel for its enhancement of glucose disposal.
As discussed in a recent review [8], the molecular mechanism of the insulin-independent glucose disposal is not established in detail and still under study.It relates, however, mainly to a stimulation of glucose uptake and a suppression of hepatic glucose oputput and is probably mediated by actions in CNS, muscle and liver tissue [6].Evidence for such an assumption includes the impaired S G in subjects with liver failure [30] and the existence of a non-insulin-mediated glucose disposal in human muscles [31].However, whether the stimulation of glucose effectiveness by tirzepatide is dependent on direct effects of the drug on CNS, muscle and liver function or indirectly through neural contribution, which also is a possibility, needs to be studied.
Our present study also demonstrates a potent glucose-dependent action of tirzepatide to stimulate insulin secretion.A stimulated insulin secretion by tirzepatide was evident already under baseline conditions after saline administration, when glucose levels were ≈ 7 mmol/l.Furthermore, the glucose-dependent stimulation of insulin secretion was markedly augmented by tirzepatide, such that the slope between AUCglucose and AUC insulin after the different glucose doses was enhanced 7fold by tirzepatide, and beta cell glucose sensitivity was increased by 55% by tirzepatide after glucose administration of 0.35 g/kg.This illustrates the potent action of this novel glucose-lowering medication to stimulate insulin secretion.
There was no action of tirzepatide on S I , which suggests that during these acute tests, tirzepatide does not influence insulin sensitivity and therefore that the increase in disposition index relies solely on the stimulation of insulin secretion.A previous study has reported a stimulation of insulin sensitivity by tirzepatide through the activation of GIP receptors in adipocytes following a 14 days treatment with tirzepatide in high-fat diet fed obese mice [32].Part of this effect was dependent on reduction in body weight which occurred after tirzepatide treatment, However, there was also a stimulation of insulin sensitivity independent from reduction in body weight as evident by a stronger effect of tirzepatide versus the GLP-1 receptor agonist semaglutide, in spite of similar reduction in body weight, and was evident when comparing insulin sensitivity after tirzepatide with pair-fed controls [32].This result seems to be at variance with the findings in the present study.The potential discrepancy may be explained by varying results in two different models (high-fat diet fed mice versus normal mice) and/or varying results related to acute versus long-term treatment, i.e., weight-independent metabolic effects induced by acute versus the 14 day terzipatide treatment, such as oxidation of glucose, lipids, and branch-chained amino acids, as suggested by the authors [32].It should also be emphasized that also insulin-independent glucose disposal contributes to the estimation of glucose disposal during clamp studies, and therefore is involved in the quantification of insulin sensitivity in that model.Similarly, during a clamp study in subjects with type 2 diabetes, tirzepatide was found to stimulate insulin sensitivity after 28 weeks of treatment [25].Therefore, our results that tirzepatide does not increase insulin sensitivity seems restricted to acute and direct effects in normal animals without any confounding factors induced by long-term treatment.However, a future study on glucose tolerance and insulin sensitivity after longer period of treatment with tirzepatide in normal and high-fat fed mice would be of interest.
Tirzepatide is a co-agonist stimulating both GLP-1 and GIP receptors, and since both GLP-1 and GIP stimulate both glucose effectiveness and insulin secretion in mice [15], it would be of interest to examine the relative contribution of these actions for its metabolic effects.This is now possible to do in future studies in mice by using the well characterized GLP-1 receptor antagonists exendin-9 and recently described GIP receptor antagonists [33][34][35].
A strength of our study is that we used a well controlled design with a large number of animals and that the glucose levels under which the effects of tirzepatide was tested was well controlled.A limitation is that the study was undertaken in mice, and therefore a similar study in humans is now warranted.Another limitation is that the study does not allow conclusions on the relative contribution of activation of GLP-1 versus GIP receptors for the actions.Also, the use of a maximal dose of tirzepatide only is a limitation for establishing dose-response characteristics of tirzepatide on glucose effectiveness, but sufficient for allowing the main conclusion of this work that tirzepatide stimulates glucose effectiveness in mice.
Based on our results, we conclude that the GIP and GLP-1 receptor co-agonist tirzepatide potently increases glucose disposal in model experiments in mice through actions to augmenting both insulin secretion and insulin independent glucose effectiveness.

Table 1
Body weight, baseline glucose and insulin, area under the 50 min curve (AUC) of glucose and insulin, 5-20 min glucose elimination rate (K G ), glucose effectiveness (S G ), insulin sensitivity index (S I ), beta-cell glucose sensitivity and disposition index after intravenous administration of glucose (0.35 g/kg) alone or with tirzepatide (3 nmol/kg).Means±SEM are shown.P indicates the probability level of random difference between the groups as determined by t-test.

Fig. 1 .
Fig. 1.Glucose and insulin levels after intravenous injection of saline (0 glucosed) or glucose (0.125, 0.35 or 0.75 g/kg) alone or with administration of tirzepatide (3 nmol/kg) in C57BL/6 J mice.n indicates the number of animals in each group.Mean±SEM are shown.Asterisks indicate the probability level of random difference between the groups, as determined by t-test (*P < 0.05, **P < 0.01, ***P < 0.001).