A Novel CCK-8/GLP-1 Hybrid Peptide Exhibiting Prominent Insulinotropic, Glucose-Lowering, and Satiety Actions With Significant Therapeutic Potential in High-Fat–Fed Mice

Glucagon-like peptide-1 (GLP-1) and cholecystokinin (CCK) are hormones released from the gut following feeding that act as important regulators of postprandial glucose homeostasis and overall energy balance (1). CCK exists in multiple molecular forms, but the carboxy-terminal octapeptide CCK-8 is well conserved among species and is the smallest form that retains the full range of biological actions (2). The most accepted biological action of GLP-1 is as an incretin hormone stimulating glucose-dependent insulin secretion to control postprandial glucose concentrations (3). Principal effects of CCK are gallbladder contraction together with the short-term regulation of energy balance, mediated through induction of satiety (4). Thus it is clear that CCK-8 and GLP-1 possess numerous complementary biological actions that suggest potential synergistic therapeutic application for obesity-related diabetes (5,6).

Over the past decade the major focus on gut hormone therapies has been directed predominantly toward single molecules that modulate individual peptide receptor targets, highlighted by the successful clinical development of long-acting GLP-1 mimetics (7). Despite this success, glycemic control and weight reduction achieved with certain types of gastric bypass surgery are noticeably superior to GLP-1 therapy (1). Thus multiple regulatory peptide hormones are inherently involved in glucose regulation and energy balance (8). It follows that combining the activity of gut hormones with complementary biological actions, such as GLP-1 and CCK-8, offers a more favorable approach for the treatment of obesity-related forms of diabetes. To enhance therapeutic utility, the design of a single hybrid peptide molecule capable of simultaneous activation of GLP-1 and CCK receptors would hold promise for obesity-related diabetes, facilitating formulation and dosing with a single molecule.

Following on from detailed interrogation of stable forms of CCK-8 for obesity-related diabetes (1,5,9–13), we constructed a novel CCK-8/GLP-1 hybrid molecule through fusion of the key amino acid sequences of the well-characterized, stable, and specific CCK-8 and GLP-1 analogs (pGlu-Gln)-CCK-8 (12) and exendin-4 (14). The bioactive domains of the parent peptides have been fused through use of a (2-[2-aminoethoxy]ethoxy)acetic acid linker molecule (15). We initially assessed enzymatic resistance, in vitro insulin secretion, and in vivo glucose-lowering, insulinotropic, and satiety actions of the parent peptides and novel (pGlu-Gln)-CCK-8/exendin-4 hybrid. We then examined the beneficial metabolic effects of a twice-daily injection regimen in high-fat–fed mice. The results reveal that sustained activation of CCK₁ and GLP-1 receptors, with a single novel (pGlu-Gln)-CCK-8/exendin-4 hybrid molecule or combined administration of the parent peptides, exerts a spectrum of beneficial metabolic effects in high–fat–fed mice that requires further consideration as a treatment option for obesity-related diabetes.

Supplementary Table 1 displays the amino acid sequence of (pGlu-Gln)-CCK-8, exendin-4, and the novel (pGlu-Gln)-CCK-8/exendin-4 hybrid molecule. All peptides (>95% purity) were purchased from American Peptide Company (Sunnyvale, CA). Peptides were characterized in-house using matrix-assisted laser desorption ionization–time of flight mass spectrometry, as described previously (16).

Effects of peptides on in vitro insulin secretion were examined using BRIN-BD11 cells, whose characteristics have been reported previously (17). BRIN-BD11 cells were seeded into 24-well plates (150,000 per well; Nunc, Roskilde, Denmark) and allowed to attach overnight at 37°C. Following 40 min of preincubation (1.1 mmol/L glucose; 37°C), cells were incubated (20 min; 37°C) in the presence of 5.6 mmol/L glucose with a range of test peptide concentrations (10⁻¹² to 10⁻⁶ mol/L). The effects of the specific GLP-1, CCK₁, and CCK₂ receptor antagonists—exendin(9-39), SR27897, and LY288513, respectively—on (pGlu-Gln)-CCK-8/exendin-4 hybrid–induced insulin secretion also were examined. After 20 min of incubation, the buffer was removed from each well and aliquots were stored at −20°C before the determination of insulin by radioimmunoassay (18).

Acute animal studies were carried out using male NIH Swiss mice (12 to 14 weeks old; Harlan Ltd., Blackthorne, U.K.) fed a standard rodent maintenance diet that contained 10% fat, 30% protein, and 60% carbohydrate, with percent of total energy of 12.99 kJ/g (Trouw Nutrition, Cheshire, U.K.). Longer-term experiments were performed using NIH Swiss mice previously fed a high-fat diet comprising 45% fat, 20% protein, and 35% carbohydrate (total energy 26.15 kJ/g; Special Diet Services, Essex, U.K.) for 130 days. Animals were housed in a 12-h light/12-h dark cycle (lights on at 0800 h and off at 2000 h) and had free access to drinking water and food. All animal experiments were conducted according to U.K. Home Office Regulations [U.K. Animals (Scientific Procedures) Act 1986].

Plasma glucose and insulin responses were evaluated after intraperitoneal (i.p.) injection of glucose alone (18 mmol/kg body weight [BW]) or in combination with test peptides (25 nmol/kg BW) in overnight (18 h) fasted normal NIH mice. Peptide doses were chosen based on our previous studies with (pGlu-Gln)-CCK-8– and exendin-4–based peptides (5). In a second series of experiments, normal mice fasted for 18 h were used to assess the effects of the respective test peptides on food intake. Mice received an i.p. injection of saline alone (0.9% [w/v] NaCl) or in combination with test peptide (25 nmol/kg) and food intake was measured at 30-min intervals.

Twice-daily (0930 and 1730 h) i.p. injections of saline vehicle (0.9% [w/v] NaCl), novel (pGlu-Gln)-CCK-8/exendin-4 hybrid (25 nmol/kg BW), (pGlu-Gln)-CCK-8 (25 nmol/kg BW), exendin-4 (25 nmol/kg BW), or a combination of both peptides (each 25 nmol/kg BW) were administered over 21 days to high-fat–fed mice. Energy intake, BW, and nonfasting plasma glucose and insulin concentrations were monitored at 3- to 6-day intervals. Intraperitoneal glucose tolerance (18 mmol/kg BW) and insulin sensitivity (15 units/kg BW) tests were performed after 21 days of treatment. At the end of the treatment period, glycated hemoglobin (Chirus Ltd.), circulating plasma glucagon and amylase, and a blood lipid profile also were assessed. All blood samples were collected before the routine morning injection of peptide. For assessment of metabolic rate and locomotor activity on day 21, mice were placed in Comprehensive Lab Animal Monitoring System metabolic chambers (Columbus Instruments, Columbus, OH) following the normal daily injection at 0930 h; consumption of O₂, production of CO_2, respiratory exchange ratio, energy expenditure, and ambulatory locomotor activity were calculated as described previously (12). HOMA-insulin resistance (IR) was determined using the equation HOMA-IR = [fasting glucose (mmol/L) × fasting insulin (ng/mL)]/22.5; HOMA-β was calculated using the equation HOMA-β = [20 × fasting insulin (ng/mL)/fasting glucose (mmol/L)] − 3.5.

Terminal blood samples were taken and pancreatic tissue was extracted and processed appropriately for isolated islet insulin release studies, histological analysis, or assessment of hormone content following acid ethanol (750 mL ethanol, 235 mL water, 15 mL concentrated HCl) extraction (12). For histological analysis, pancreata were fixed and processed as described previously (19). Tissue sections were deparaffinized, rehydrated, and probed with primary antibodies: rabbit anti-insulin antibody (1:200; Santa Cruz Biotechnology, Heidelberg, Germany) or guinea pig antiglucagon antibody (PCA2/4, 1:200; raised in-house) (19). In a separate series, pancreatic islets were isolated from respective treatment groups by collagenase digestion, as fully described previously (20). Insulin secretion was determined as described above for BRIN-BD11 cells. Following removal of the test solution, 200 µL of acid-ethanol solution (1.5% [v/v] HCl, 75% [v/v] ethanol, 23.5% [v/v] H₂O) was added, with overnight extraction of cellular insulin. Samples were stored at −20°C for measurement of insulin content by radioimmunoassay (18).

Blood samples were collected from the cut tip of the tail vein of conscious mice into chilled fluoride/heparin glucose microcentrifuge tubes (Sarstedt, Numbrecht, Germany) at the time points indicated in the figures. Blood glucose was measured directly using a handheld Ascensia Contour blood glucose meter (Bayer Healthcare, Newbury, Berkshire, U.K.). For plasma insulin analysis, blood samples were immediately centrifuged using a Beckman microcentrifuge (Beckman Instruments, Galway, Ireland) for 1 min at 13,000 × g and stored at −20°C. Plasma and pancreatic insulin were assayed by a modified dextran-coated charcoal radioimmunoassay (18). Plasma and pancreatic glucagon were measured by a commercially available ELISA kit (Merck Millipore, Germany), according to the manufacturer’s instructions. Total, HDL-, and LDL-cholesterol; triglycerides; and plasma amylase concentrations were measured using an Hitachi Automated Analyzer 912 (Boehringer Ingelheim, Mannheim, Germany).

Results are expressed as means ± SEMs, and data were compared using the unpaired Student t test. Where appropriate, data were compared using repeated measures ANOVA or one-way ANOVA, followed by the Student-Newman-Keuls post hoc test. Groups of data were considered to be significantly different if P < 0.05.

All test samples induced significant (P < 0.05 to P < 0.001) concentration-dependent elevations of insulin secretion from BRIN-BD11 cells (Fig. 1A). The combination of (pGlu-Gln)-CCK-8 plus exendin-4 and the hybrid were the most effective in terms of insulin secretion (Fig. 1A). Further studies with specific GLP-1, CCK₁, and CCK₂ receptor antagonists confirmed that the insulinotropic activity of the (pGlu-Gln)-CCK-8/exendin-4 hybrid was dependent on both GLP-1 and CCK₁ receptor activity (Fig. 1B). It is clear, however, particularly at elevated concentrations, that the insulin secretory action of the (pGlu-Gln)-CCK-8/exendin-4 hybrid peptide is very much dependent on intact GLP-1 receptor signaling pathways (Fig. 1B).

(pGlu-Gln)-CCK-8 failed to elicit significant insulin-releasing or glucose-modulating actions at the dose used (Fig. 2A and B). By contrast, exendin-4 significantly reduced (P < 0.05) overall 0- to 60-min area under the curve (AUC) plasma glucose values and increased (P < 0.05) the overall insulin secretory response (Fig. 2A and B). In addition, the combination of (pGlu-Gln)-CCK-8 with exendin-4, and the (pGlu-Gln)-CCK-8/exendin-4 hybrid, also significantly reduced (P < 0.01) the overall plasma glycemic excursion and increased (P < 0.05) the insulin secretory response (Fig. 2A and B). All peptides induced significant (P < 0.05 to P < 0.001) appetite-suppressive effects when administered to mice fasted overnight (Fig. 2C). The most effective treatments to inhibit feeding were (pGlu-Gln)-CCK-8 in combination with exendin-4 and the (pGlu-Gln)-CCK-8/exendin-4 hybrid (Fig. 2C).

All groups of high-fat–fed mice had significantly (P < 0.001) increased energy intake compared with lean controls during the 21-day study (Fig. 3A). Exendin-4 therapy had no effect on energy intake at the same dose used (Fig. 3A). However, twice-daily treatment with (pGlu-Gln)-CCK-8 alone and in combination with exendin-4, or the novel hybrid, significantly (P < 0.05 to P < 0.001) reduced energy intake from day 11 onward (Fig. 3A). All treatment regimens, barring exendin-4 alone, significantly (P < 0.05 to P < 0.001) reduced BW compared with high-fat–fed controls (Fig. 3B). Indeed, combined CCK₁ and GLP-1 receptor activation returned BWs to the levels of the lean controls from day 15 onward (Fig. 3B).

A similar scenario was evident in terms of circulating blood glucose concentrations, with all treatments, except exendin-4, significantly (P < 0.05 to P < 0.001) lowering nonfasting glucose concentrations compared with high–fat–fed controls (Fig. 3C). Circulating plasma insulin concentrations were significantly (P < 0.05 to P < 0.001) increased at numerous observation points in all treatment groups when compared with lean controls, but they were similar to those of high-fat–fed controls on day 21 (Fig. 3D). All high-fat–fed mice, barring those treated with exendin-4 alone, had significantly (P < 0.05 to P < 0.001) elevated glucagon concentrations compared with lean controls on day 21 (Fig. 3E). Indeed, mice treated with exendin-4 had significantly (P < 0.05) reduced plasma glucagon concentrations compared with mice treated with (pGlu-Gln)-CCK-8 (Fig. 3E). Plasma amylase concentrations were not different in any of the treatment groups when compared with the high-fat–fed controls on day 21 (Fig. 3F). However, exendin-4 treatment increased (P < 0.05) amylase concentrations when compared with lean controls (Fig. 3F). Glycated hemoglobin values in all treatment groups, barring exendin-4, returned to concentrations similar to those of lean controls by day 21 (Fig. 3G).

Following an intraperitoneal glucose challenge on day 21, overall plasma glucose concentrations were significantly (P < 0.05 to P < 0.001) reduced in all treatment groups compared with high-fat–fed controls (Fig. 4A). Furthermore, there was a significant (P < 0.05 to P < 0.01) decrease in overall AUC glucose concentrations in high-fat–fed mice treated with (pGlu-Gln)-CCK-8 in combination with exendin-4, or the hybrid, when compared with (pGlu-Gln)-CCK-8 or exendin-4 treatment alone (Fig. 4B). Corresponding individual plasma insulin concentrations were elevated (P < 0.05 to P < 0.01) in all high-fat–fed treatment groups compared with lean controls (Fig. 4C). This was corroborated by AUC data, which revealed all treatments induced a significantly (P < 0.05 to P < 0.001) elevated overall insulin secretory response compared with control mice (Fig. 4D).

The hypoglycemic action of insulin, in terms of percentage decrease from basal glucose values, was superior in all treatment groups compared with high-fat–fed controls, except those mice treated with (pGlu-Gln)-CCK-8 alone (Fig. 4E). However, the action of insulin was particularly enhanced (P < 0.001) in groups with dual activation of GLP-1 and CCK₁ receptors in terms of 0- to 60-min overall values (Fig. 4F). Calculation of HOMA-IR also indicated superior (P < 0.05 to P < 0.01) insulin sensitivity compared with other groups (0.50 ± 0.06 and 0.52 ± 0.04, respectively, for mice treated with (pGlu-Gln)-CCK-8 in combination with exendin-4 or the novel hybrid peptide, compared with 0.71 ± 0.04 for high-fat–fed controls). Pancreatic insulin concentrations were significantly (P < 0.05 to P < 0.01) reduced in all treatment groups, barring the combination of (pGlu-Gln)-CCK-8 plus exendin-4, when compared with lean control mice (Fig. 4G). Pancreatic glucagon content was elevated (P < 0.01) in mice treated with (pGlu-Gln)-CCK-8 compared with lean controls, whereas both other treatment groups undergoing sustained CCK₁ receptor activation had similar pancreatic glucagon concentrations compared with control mice (Fig. 4H). Interestingly, mice treated with exendin-4 alone had reduced (P < 0.01) pancreatic glucagon content compared with high-fat–fed controls (Fig. 4H).

There were no differences in O₂ consumption, CO₂ production, respiratory exchange ratio, and energy expenditure in any of the high-fat–fed groups of mice on day 21 (Supplementary Fig. 1A–D). There was, however, a significant (P < 0.05) decrease in overall energy expenditure in high-fat–fed mice when compared with lean controls, which was normalized by all treatment regimens (Supplementary Fig. 1D). Ambulatory activity, as assessed by X-beam breaks, was similar in all groups during the light phase (Fig. 5A) but was significantly (P < 0.05) decreased during the dark phase among mice receiving (pGlu-Gln)-CCK-8 or hybrid therapy when compared with lean controls (Fig. 5B). Rearing and jumping episodes during the light phase, as assessed by Z-beam breaks, were increased (P < 0.05) in mice receiving exendin-4 alone or in combination with (pGlu-Gln)-CCK-8 (Fig. 5C). During the dark phase, Z-beam breaks were significantly (P < 0.05 to P < 0.001) increased in all treatment groups compared with high-fat–fed controls (Fig. 5D).

Assessment of the blood lipid profile on day 21 revealed significant (P < 0.05 to P < 0.001) reductions of total cholesterol concentrations in all treatment groups (Fig. 6A). HDL-cholesterol was unaltered in all groups, barring the hybrid-treated mice, which had reduced (P < 0.05 to P < 0.01) levels compared with all groups expect those mice receiving (pGlu-Gln)-CCK-8 alone (Fig. 6B). Interestingly, LDL-cholesterol levels were significantly (P < 0.05 to P < 0.01) decreased in all groups compared with those of high-fat–fed controls (Fig. 6C). Plasma triglyceride concentrations were reduced (P < 0.001) by (pGlu-Gln)-CCK-8 and hybrid treatment when compared with those of high-fat–fed controls but were similar to those of lean controls (Fig. 6D). Treatment with exendin-4 alone or in combination with (pGlu-Gln)-CCK-8 did not reduce triglyceride concentrations (P < 0.05 to P < 0.01) compared with high-fat–fed controls (Fig. 6D).

Islets isolated from high-fat–fed mice had significantly (P < 0.05 to P < 0.001) reduced insulin secretory responses when compared with those of lean controls (Fig. 7A and B). By contrast, islets isolated from all high-fat–fed treatment groups had an insulin secretory capacity similar to that of lean controls in response to glucose, tolbutamide, and elevated Ca²⁺ (Fig. 7A and B). Moreover, there was no sign of pancreatic β-cell tachyphylaxis following 21 days of twice-daily administration of any of the peptide treatment regimens (Fig. 7A). The insulin secretory response of isolated islets to phorbol myristic acid was reduced (P < 0.05 to P < 0.001) in the hybrid group and among mice treated with (pGlu-Gln)-CCK-8 alone or in combination with exendin-4 when compared with lean controls (Fig. 7B). The secretory response to forskolin also was reduced (P < 0.01 to P < 0.001) in islets isolated from mice treated with (pGlu-Gln)-CCK-8 and exendin-4 when compared with lean controls, but not in mice treated with the novel hybrid or a combination of (pGlu-Gln)-CCK-8 and exendin-4 (Fig. 7B). The benefits of the latter treatments also were evidenced in vivo, as indicated by the superior HOMA-β values (Fig. 7C).

Representative images from the lean control, high-fat–fed, and all treatment groups are depicted in Fig. 8A–F. Importantly, the structure and cellular organization of islets in all treatment groups was essentially similar to those of lean controls, and there was no obvious evidence of pancreas fatty infiltration. All treatment regimens returned pancreatic islets numbers to levels similar to those of lean controls (Fig. 8G). Similarly, islet area was reduced (P < 0.05) by high-fat feeding, with decreases in small, medium, and large islets (Fig. 8H and I). All peptide treatments corrected this detrimental effect (Fig. 8H and I). Moreover, treatment with (pGlu-Gln)-CCK-8 alone or in combination with exendin-4 significantly (P < 0.01) increased islet area compared with lean control mice (Fig. 8H). Pancreatic β-cell area was largely unaffected by high-fat feeding, but all treatments increased (P < 0.001) β-cell area compared with both high-fat–fed and lean control mice (Fig. 8J).

Despite encouraging preclinical data (5,9–12), the development of CCK-based drugs for monotherapy of human obesity-related diabetes remains elusive (21). By contrast, GLP-1 mimetics have been adopted into the diabetic clinic with great vigor, although weight loss and metabolic control are not as impressive as first hoped (7). In this regard several recent studies revealed that dual agonism of CCK₁ and GLP-1 receptors has marked synergistic metabolic and weight-reduction benefits (5,6). It should be noted, however, that earlier clinical studies using infusions of native GLP-1 and CCK peptides in normal human subjects fasted overnight failed to reveal clear synergistic effects (22). This is likely related to important differences between these studies, including the use of infusion rather than bolus injection, normal fasted subjects as opposed to a freely fed diabetic animal model, and native metabolically liable peptides rather than enzymatically stable peptide forms (5,6,22). Therefore, in this research we evaluated the biological actions and therapeutic applicability of a novel stable (pGlu-Gln)-CCK-8/exendin-4 hybrid peptide in comparison with combined administration of the parent peptides.

The hybrid peptide was engineered to combine the satiety- and energy-regulating action of CCK-8 (4) with the robust insulin-releasing and antidiabetic actions of GLP-1 (3) in a single compound. Importantly, pharmacological studies reveal that CCK₁, as opposed to CCK₂, receptor activity is critical for synergy with GLP-1 receptor action (6). We confirmed through the use of specific GLP-1, CCK₁, and CCK₂ receptor antagonists that the insulinotropic effects of the hybrid were mediated via both GLP-1 and CCK₁ receptors. Indeed, our data revealed that the hybrid peptide retained full ability to activate CCK₁ and GLP-1 receptor signaling pathways involved in glucose homeostasis, insulin secretion, and appetite suppression (1). Even a combination of the parent peptides, both individually at the same concentration as the single hybrid, was not superior in efficacy to the novel hybrid. Notably, biological effects of (pGlu-Gln)-CCK-8/exendin-4 were significantly enhanced when compared with either parent peptide alone; the prominent glucose regulatory effects of the hybrid seemed to be largely dependent on GLP-1 receptor activation and appetite suppression predominantly driven by CCK₁ receptor signaling (5). In addition, the fact that CCK₂ receptor inhibition did not perturb the insulinotropic effect of (pGlu-Gln)-CCK-8/exendin-4 is encouraging given that activation of this receptor has been associated with panic and anxiety attacks (23,24), which has detracted from the therapeutic potential of previous nonspecific CCK-based drugs (25).

In harmony with previous findings (5), chronic twice-daily treatment with (pGlu-Gln)-CCK-8 alone, and particularly in combination with exendin-4, resulted in sustained and significant reductions of energy intake and BW in high-fat–fed mice. In relation to this, vagal afferents coexpress CCK₁ and GLP-1 receptors (26) and might represent a target synergistic interaction between the two hormones. Moreover, it has been suggested that the effects of CCK and GLP-1 on the central regulation of food intake operate via different but complementary autonomic circuits (27–29). Tolerability and induction of nausea have previously been suggested as important limiting factors in the development of CCK₁ receptor–activating drugs (30). In agreement with previous studies (12), however, we clearly show a lack of desensitization to the anorectic or insulin-releasing effects of (pGlu-Gln)-CCK-8, either alone or in combination with GLP-1 receptor activation. Crucially, the novel hybrid peptide (pGlu-Gln)-CCK-8/exendin-4 had equally effective beneficial metabolic actions when compared with combined administration of the parent peptides. Indeed, reductions in HbA_1c were observed only in those mice subjected to sustained activation of both CCK₁ and GLP-1 receptors (5,6). Development of specific assays to directly measure each peptide in plasma would be useful to provide more precise details of circulating half-life. Surprisingly, exendin-4 did not obviously decrease energy intake, BW, or circulating glucose concentrations at the dose used, although similar observations have been reported previously by us and others (5,31). The limited effect of exendin-4 could possibly reflect upregulation of adaptive mechanisms, although it may simply be dose, species, or animal model specific (5).

Glucose tolerance was improved to a similar extent by 21 days of twice-daily treatment with (pGlu-Gln)-CCK-8 or exendin-4 alone. Notably, there was much greater enhancement of glucose disposal in mice treated with (pGlu-Gln)-CCK-8 in combination with exendin-4 or the novel hybrid, again highlighting the distinct benefit of sustained dual activation of CCK₁ and GLP-1 receptors (5,6). This effect was associated with significantly increased insulin concentrations, indicating that benefits of combined (pGlu-Gln)-CCK-8 and exendin-4 therapy, or the novel (pGlu-Gln)-CCK-8/exendin-4 hybrid, are partly mediated by direct actions on pancreatic β-cell function (1). This was confirmed on day 21, when islets isolated from high-fat–fed mice treated with (pGlu-Gln)-CCK-8 in combination with exendin-4, or the novel hybrid peptide, exhibited a marked improvement in the insulin secretory response to glucose, each of the individual peptides, and a variety of other agents with diverse actions on β-cell signaling pathways. This corresponded with improvements in HOMA-β values. Interestingly, the action of phorbol myristic acid in vitro was not enhanced, suggesting that pathways downstream of protein kinase C that are usually activated by CCK (32) make a relatively small contribution to the effects of the hybrid peptide.

Prominent in vivo effects on β-cell function are consistent with the recognized insulin-releasing action of GLP-1 and CCK-8 at pharmacological levels (3,33). In harmony with this, pancreatic insulin concentrations where augmented by combined (pGlu-Gln)-CCK-8 and exendin-4 treatment. Pancreatic glucagon concentrations in these mice were unchanged by dual activation of CCK₁ and GLP-1 receptors, and they actually were reduced together with circulating concentrations by exendin-4 therapy. By contrast, (pGlu-Gln)-CCK-8 increased pancreatic glucagon, as noted previously (9). In addition, we observed consistent improvement of insulin sensitivity with both the hybrid peptide and (pGlu-Gln)-CCK-8 in combination with exendin-4 by the end of the treatment period. This is in agreement with the well-known insulin-sparing actions of GLP-1 (34), and evidence also suggests an integral role for CCK as a regulator of insulin action, especially under conditions of high-fat feeding (35). In this study the islet number, together with both islet area and β-cell area, were increased by all treatment modalities. However, assessment of pancreatic cell volumes would be required to fully confirm islet proliferative effects of the treatment regimens. Furthermore, notable reductions in total and LDL-cholesterol concentrations were observed with all treatment regimens, in agreement with effects of individual peptides in previous studies (5,9–11). HDL-cholesterol concentrations also were reduced by the hybrid, which was unique to this treatment modality and necessitates further investigation. While it should be noted that decreased circulating HDL-cholesterol is generally considered to be a cardiovascular disease risk factor (36), this rather one-dimensional hypothesis has been questioned lately (37). Encouragingly circulating triglyceride concentrations were markedly reduced by all treatments, except exendin-4 alone, signifying improved metabolic control and insulin resistance (38).

To further clarify possible mechanisms behind the weight reduction observed with dual CCK₁ and GLP-1 receptor activation, we assessed aspects of metabolic rate and locomotor activity. Given the prominent effects of GLP-1 and CCK on energy balance (3,26), alterations of energy expenditure may have been predicted in this study. Consistent with other studies using (pGlu-Gln)-CCK-8 (12), however, metabolic benefits in this study were not associated with changes in energy expenditure or metabolic rate. There was a mild reduction in physical activity levels during the dark phase among mice treated with (pGlu-Gln)-CCK-8 or the hybrid, which might accompany the period of reduced energy intake. In addition, differences in locomotor activity between the hybrid and the combined treatment group are intriguing. Thus the overall significance of these centrally mediated effects requires further detailed elucidation. Moreover, assessment of gene or protein expression could also aid in uncovering the mechanisms behind the beneficial effects observed with independent and combined CCK₁ and GLP-1 receptor activation. Nonetheless, taken together, these observations suggest that decreased energy intake is the driving force behind weight loss in this study. Importantly, there were no obvious signs of malaise in these mice, and weight reduction was robust and durable.

Given the recent controversy relating to detrimental GLP-1 effects on the exocrine pancreas (39), and the fact that extremely large doses of CCK have caused pancreatitis (40), further consideration of this safety aspect is required for any combinational therapeutic approach involving GLP-1 and CCK compounds. Indeed, plasma amylase concentrations were increased with exendin-4 monotherapy in this study. Nevertheless, and most important, we did not observe any evidence of pancreatic inflammation with the (pGlu-Gln)-CCK-8/exendin-4 hybrid peptide or with (pGlu-Gln)-CCK-8 alone or in combination with exendin-4, as reflected by unaltered plasma amylase concentrations and pancreatic histology. This is in agreement with similar studies using (pGlu-Gln)-CCK-8 (13) or CCK-8 and GLP-1 combination therapies (6). Moreover, the CCK₁ receptor is likely expressed at much lower levels in pancreatic acinar cells in humans as opposed to rodents (41). However, more detailed and extended studies looking at other side effects and related toxicology are needed before testing in humans.

In conclusion, this study demonstrated the novel hybrid peptide analog (pGlu-Gln)-CCK-8/exendin-4 is a dual-acting CCK₁ and GLP-1 receptor agonist with equivalent or superior therapeutic efficacy when compared with combined administration of the parent peptides. (pGlu-Gln)-CCK-8/exendin-4 has robust satiety, glucose homeostatic, and insulin secretory actions and improves glucose tolerance, insulin resistance, pancreatic β-cell function, and islet morphology in high-fat–fed mice. The synergistic interplay that exists between CCK₁ and GLP-1 receptor signaling merits further consideration as a new treatment paradigm for type 2 diabetes.

Funding. These studies were supported by the SAAD Trading and Contracting Company and University of Ulster selective research funding.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. N.I. and P.R.F. conceived the study and wrote the manuscript. V.P. analyzed and interpreted data. All authors revised the manuscript critically for intellectual content and approved the final version of the manuscript. V.P. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.