Glucagon-like peptide-1 loaded phospholipid micelles for the treatment of type 2 diabetes: improved pharmacokinetic behaviours and prolonged glucose-lowering effects

Abstract

Glucagon-like peptide-1 (GLP-1) and GLP-1 receptor agonists are actively pursued as therapeutic agents for type 2 diabetes mellitus (T2DM). However, the therapeutic utility of GLP-1 is limited due to its rapid inactivation by dipeptidyl peptidase IV, and many GLP-1 receptor agonists suffer from innegligible adverse effects. In the present study, in order to develop long-acting GLP-1 derivatives with improved hypoglycemic activity, native GLP-1 (7-36)-NH₂ was loaded into sterically stabilized phospholipid micelles (SSM), affording GLP-1-SSM. In vitro stability test and in vivo pharmacokinetic study demonstrated that the association of GLP-1 with SSM led to enhanced stability and drug utilization without affecting its insulinotropic and glucose-lowering activities. Single and multiple glucose tolerance tests confirmed that GLP-1-SSM was a long-acting antidiabetic agent comparable or even better than exendin-4. More importantly, preclinical studies found out that a chronic twice daily treatment of GLP-1-SSM in type 2 diabetic db/db mice suppressed body weight gain and food uptake, decreased HbA1c value, and restored the glucose tolerance ability. Collectively, our results suggest that GLP-1-SSM is a promising therapy for the treatment of T2DM and deserves further investigation.

---

Introduction

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder with an increasing number of patients worldwide.¹ Characterized by insulin resistance, T2DM leads to inadequate insulin secretion in response to glucose intake and finally manifests hyperglycemia.² Glucagon-like peptide-1 (GLP-1) is a glucoincretin hormone released from intestinal L-cells upon ingestion of glucose or food intake. In contrast to most other glucose-lowering drugs, GLP-1 stimulates insulin secretion via a glucose-dependent mechanism and plays a physiological regulatory role, resulting in enhanced satiety and reduced energy intake. Therefore, GLP-1 can simultaneously lower blood glucose and body weight.^3,4 Other beneficial effects from GLP-1 includes increased beta-cell mass, stimulated beta-cell proliferation and delayed gastric emptying, all of which make GLP-1 a promising therapeutic target for treating T2DM.^5–7

However, the therapeutic application of GLP-1 is hampered by its short circulating half-life (t_1/2 < 2 min) due to its rapid inactivation by dipeptidyl peptidase IV (DPP-IV) and renal elimination.⁸ To address this issue, extensive research efforts have been undertaken to develop DPP-IV inhibitors^9–13 and long-acting GLP-1 receptor agonists. Two GLP-1 analogues, exendin-4 (Exenatide, Byetta) and Liraglutide (Victoza), have been clinically approved for the treatment of T2DM.^14,15 However, both of them suffer from innegligible immunogenicity/toxicity and adverse effects such as nausea and acute pancreatitis.^16,17 Our group has also focused on the development of long-acting GLP-1 derivatives through rational chemical modification. The conjugation of fatty chain, dicoumarol and coumarin with GLP-1 has successfully led to the discovery of several hybrids with enhanced stability and prolonged glucose-lowing activity.^18–21 Recently, novel drug delivery systems based on sterically stabilized phospholipid micelle (SSM) have been developed for controlled release of bioactive polypeptides which originally possess short t_1/2.^22,23 The SSM is composed of the water-soluble, biodegradable and biocompatible phospholipid, 1,2-distearoyl-sn-glycero-3-phosphatidylethanolamine-N-[methoxy(polyethylene glycol)-2000 (DSPE-PEG₂₀₀₀). In aqueous media, these PEGylated phospholipid molecules spontaneously self-assemble above their critical micelle concentration (CMC). Importantly, the extremely low CMC (∼0.5–1 μM) of SSMs compared with other surfactants ensure their stability after dilution caused by in vivo administration.^22,23 In practice, SSMs have been successfully applied to modify various peptides such as neuropeptide Y, vasoactive intestinal peptide and pancreatic polypeptide.^24–26 After being associated with these amphipathic and enzyme-labile peptides, the PEG palisade within SSM protects the peptides from rapid enzymatic cleavage and prevents the rapid in vivo clearance of the micelles by reticulo-endothelial system and opsonization. As a result, the in vivo circulating half-lives of these peptides were remarkably increased. Furthermore, the association of peptides with SSM in aqueous media avoids the formation of large, heterogeneous and immunogenic aggregates.^22,23 As far as GLP-1 to be concerned, the elegant work from Lim et al. for the first time demonstrated that GLP-1 self-associated with PEGylated SSM to form a nanomedicine GLP-1-SSM with increased α-helicity and promising anti-inflammatory activity against acute lung injury.²⁷ Inspired by the above results, we hypothesize that GLP-1-SSM could also be used as a long-acting antidiabetic agent, which remains unexplored.

In the present work, we report the preparation and characterization of GLP-1 (7-36)-NH₂ (GLP-1) loaded SSM in order to achieve improved circulating half-life and anti-diabetic activity. The in vitro stability and in vivo pharmacokinetic behaviors of GLP-1-SSM were explored. In addition, the insulin secretion profile and long-term anti-diabetic activities were investigated. Most importantly, we tested, for the first time, the sub-chronic treatment (twice daily) effects of GLP-1-SSM in type 2 diabetic db/db mice. Our results clearly demonstrate the therapeutic effects of GLP-1-SSM as a new strategy for the treatment of T2DM.

Results and discussion

Determination of GLP-1-SSM second structure

GLP-1-SSM was prepared as previously reported.²⁷ The secondary structure of GLP-1 and GLP-1-SSM were investigated using circular dichroism spectroscopy in saline. GLP-1 adapted a random coil conformation in saline as indicated by the relatively flat curve (Fig. 1). Upon being associated with SSM, two peak minima centered at approximately 208 and 222 nm were identified, indicating a predominantly α-helical secondary structure for GLP-1-SSM. This was in agreement with the previously reported characteristics of GLP-1-SSM.²⁷

Fig. 1 Representative circular dichroism spectra of GLP-1 (blue) and GLP-1-SSM (green) in saline.

In vitro plasma stability

To evaluate the in vitro stability of GLP-1-SSM toward DPP-IV and other metabolic enzymes, GLP-1-SSM was incubated with rat plasma at 37 °C over 24 h and GLP-1 was used as the control. As expected, GLP-1 was rapidly degraded in rat plasma due to the liable N-terminal sequence (Ala⁸-Glu⁹), and the half-life of GLP-1 in vitro was about 0.2 h (Fig. 2). In contrast, the half-life of GLP-1-SSM in vitro was remarkably prolonged to about 10.4 h. These results clearly demonstrate that the SSM micelles effectively inhibit the degradation of GLP-1 mediated by DPP-IV and other metabolic enzymes.

Fig. 2 In vitro stability of GLP-1 and GLP-1-SSM in rat plasma. GLP-1 or GLP-1-SSM (100 ng mL⁻¹) was incubated in rat plasma at 37 °C for 24 h, and their degradation behaviors were monitored using LC-MS/MS. Data were shown as means ± SD, n = 3.

In vivo pharmacokinetics

To further evaluate the in vivo behavior of GLP-1-SSM, pharmacokinetic test was performed in SD rats. Exendin-4, a marketed long-acting GLP-1 analogue, was used as the control. As shown in Fig. 3 and Table 1, after being subcutaneously administrated, the plasma concentrations of exendin-4 increased rapidly and peaked within 1 h (t_max = 0.50 ± 0.04 h), followed by decreasing to baseline at 6 h post injection. The plasma GLP-1 concentrations in rats subcutaneously injected with GLP-1-SSM (25 nmol kg⁻¹ of GLP-1) increased and reached the first peak at 1 h, followed by a “platform release” pattern within 4 h, and then by a gradual decrease to baseline within 24 h. Furthermore, as listed in Table 1, the mean residence time of GLP-1-SSM (7.57 ± 0.14 h) doubled compared with that of exendin-4 (3.72 ± 1.01 h), suggesting the potential of GLP-1-SSM formulation to be developed as a twice-a-day agent. In addition, the AUC_inf value of GLP-1-SSM was found significantly higher (3.4 fold) than that of exendin-4 (Table 1). Taken together, these results indicate that SSM not only significantly improves the stability of GLP-1 in vivo but also increases the drug utilizations.

Fig. 3 In vivo pharmacokinetic profiles of exendin-4 and GLP-1-SSM in vivo. SD rats were subcutaneously injected with exendin-4 (15 nmol/rat) or GLP-1-SSM (15 nmol GLP-1/rat). The plasma concentration of exendin-4 or GLP-1-SSM at indicated time points were shown as means ± SD, n = 3.

Table 1 Pharmacokinetic parameters of exendin-4 and GLP-1-SSM in SD rats^a

Samples	Tmax (h)	Cmax (ng mL^−1)	AUCinf (ng h mL^−1)	MRT (h)
a Data are shown as means ± SD (n = 3).
Exendin-4	0.50 ± 0.04	770.15 ± 56.36	2292.86 ± 104.73	3.72 ± 1.01
GLP-1-SSM	1.02 ± 0.11	702.50 ± 24.75	7794.58 ± 399.36	7.57 ± 0.14

---

Insulinotropic activity in SD rats

The insulinotropic activity of GLP-1-SSM was evaluated by using intraperitoneal glucose tolerance test (IPGTT) in SD rats. As shown in Fig. 4a, both GLP-1 (25 nmol kg⁻¹) and GLP-1-SSM (25 nmol kg⁻¹ GLP-1) treatments led to significantly elevated plasma insulin concentrations compared with the non-treatment saline group after oral glucose administration (8 g kg⁻¹). The insulin levels in the GLP-1 treatment group raised to a maximum of 38.27 ± 2.82 mIU L⁻¹ after 15 min and returned to baseline after 90 min. In comparison, the insulin levels in the GLP-1-SSM treatment group increased to 40.21 ± 2.76 mIU L⁻¹ after 15 min and slowly returned to baseline after 120 min, indicating a better insulin secretion-inducing ability than GLP-1, which was further confirmed by the higher calculated AUC_insulin value of GLP-1-SSM compared with that of GLP-1 (Fig. 4b).

Fig. 4 Insulinotropic activity tests of GLP-1 and GLP-1-SSM by IPGTT experiments in SD rats. Male SD rats were (a) plasma insulin concentration (mIU L⁻¹) in each group at indicated time points. (b) The insulin AUC_{0–120 min} value for each group. Means ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001.

Glucose-lowering ability in Kunming mice

The glucose-lowering ability of GLP-1 SSM was measured by IPGTT to evaluate whether the hypoglycemic activity of GLP-1 was kept after being loaded into SSM. Three groups of Kunming mice were administrated with glucose (2 g kg⁻¹) in combination with saline (vehicle control), GLP-1 (25 nmol kg⁻¹) or GLP-1-SSM (25 nmol kg⁻¹ GLP-1), and the blood glucose levels of each group were measured at indicated time points. As illustrated in Fig. 5a, upon glucose administration, the mean blood glucose level in the saline group rapidly increased to a maximum of 14.9 ± 1.5 mmol L⁻¹ after 15 min and slowly decreased to 9.1 ± 0.6 mmol L⁻¹ after 60 min. GLP-1 or GLP-1-SSM treatment resulted in a significantly enhanced glucose tolerance. The highest glucose levels at 15 min were found to be 9.3 ± 0.6 and 8.6 ± 0.9 mmol L⁻¹, respectively, for GLP-1 and GLP-1-SSM treatment groups. In addition, the in vivo glucose-lowering effects of GLP-1-SSM was self-explanatory from its lowest calculated AUC_glucose value (619.3 ± 56.7) compared with that of the GLP-1 treatment and saline groups (715.1 ± 84.8 and 1129.5 ± 44.0, respectively) (Fig. 5b).

Fig. 5 Glucose-lowering abilities of GLP-1 and GLP-1-SSM as determined by IPGTT experiments. Kunming mice were administrated with glucose (2 g kg⁻¹) in combination with saline (vehicle control), GLP-1 (25 nmol kg⁻¹) or GLP-1-SSM (25 nmol kg⁻¹ GLP-1). (a) Blood glucose levels (mmol L⁻¹) in each group at indicated time points. (b) The glucose AUC_{0–120 min} value for each group. Means ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001.

Long-term glucose stabilizing activity in db/db mice

The above results suggest that the SSM modification doesn't affect the insulinotropic and hypoglycemic activities of GLP-1. Next, the glucose stabilizing activity of GLP-1-SSM was further studied in db/db mice to test whether the SSM formulation resulted in a prolonged hypoglycemic duration. For this purpose, db/db mice were intraperitoneally (i.p.) pretreated with GLP-1-SSM, exendin-4 or saline (vehicle control), followed by IPGTT experiments. As illustrated in Fig. 6a and b, pretreatment of GLP-1-SSM or exendin-4 30 or 120 min before IPGTT led to similar glucose-lowering effects. As shown in Fig. 6c, the glucose-lowering effect of exendin-4 when administered 240 min before IPGTT was compromised, but the glucose-lowering effect of GLP-1-SSM under the same condition was relatively unaffected, indicating a more prolonged glucose stabilization ability of GLP-1-SSM. Finally, GLP-1-SSM was shown to remain prominent glucose-stabilizing effect even when administered 360 min prior to IPGTT (Fig. 6d). Taken together, these results suggest that GLP-1-SSM has comparable glucose-lowering activity and longer glucose-stabilizing ability as compared with exendin-4.

Fig. 6 Long-term glucose stabilizing profiles of exendin-4 and GLP-1-SSM in db/db mice. (a–d) Mice were pretreated (i.p.) with GLP-1-SSM (25 nmol kg⁻¹ GLP-1), exendin-4 (25 nmol kg⁻¹) or saline 30 (a), 120 (b), 240 (c), and 360 (d) min prior to the IPGTT experiments (1 g kg⁻¹ glucose). The blood glucose levels in each group at indicated time points were shown as means ± SD, n = 6. (e) GLP-1-SSM (25 nmol kg⁻¹ GLP-1), exendin-4 (25 nmol kg⁻¹) or saline was i.p. injected 0.5 h prior to glucose load, and the glucose challenges (2 g kg⁻¹) were performed at time points 0, 6 and 12 h. The blood glucose levels (mmol L⁻¹) were shown as means ± SD, n = 6.

In addition to the above single glucose tolerance tests, a modified multiple OGTT was further applied to evaluate the long-term hypoglycemic effect of GLP-1-SSM. GLP-1-SSM, exendin-4 or saline (vehicle control) was i.p. injected 0.5 h prior to glucose load, and the glucose challenges were performed at time points 0, 6 and 12 h to mimic the human diet of three meals a day. Blood glucose levels in mice treated with saline rapidly increased to over 20 mmol L⁻¹ and peaked at 0.5 h after every glucose load (Fig. 6e). Exendin-4 showed excellent glucose-lowering activity during the first glucose load period (0–6 h). However, during the following two glucose load cycles, the glucose stabilization ability of exendin-4 decreased dramatically, as indicated by the high peak values at around 6.5 and 12.5 h. On the other hand, GLP-1-SSM maintained blood glucose levels under 14 mmol L⁻¹ during the whole glucose load period. The glucose lowering effect of GLP-1-SSM was compromised after the last glucose load, but was still more effective than exendin-4. The above results further prove that, compared with exendin-4, GLP-1-SSM has a longer glucose lowering activity, which could last over 8 h.

Chronic treatment effects in vivo

The effects of chronically i.p. injected GLP-1-SSM (25 nmol kg⁻¹ of GLP-1, twice daily, 35 days) on body weight gain and food intake were studied in db/db mice to further examine the potential therapeutic utility of GLP-1-SSM. As shown in Fig. 7a, compared with the wild-type nondiabetic mice, body weight of db/db mice increased dramatically during 35 days. Both exendin-4 and GLP-1-SSM suppressed the body weight gain in db/db mice to a similar extent, as compared with the saline group. In addition, administration of either exendin-4 or GLP-1-SSM reduced food intake in db/db mice during the whole treatment period (Fig. 7b). Furthermore, the HbA1c value, a sensitive index of glycemic control, in wild-type and db/db mice was measured before and after the treatment period. As illustrated in Fig. 7c and d, while HbA1c stayed low in wild-type mice, it increased significantly in db/db mice with saline treatment. In contrast, db/db mice injected with either exendin-4 or GLP-1-SSM exhibited prominent reduction in HbA1c after the five-week treatment.

Fig. 7 Chronic treatment effects of exendin-4 and GLP-1-SSM in db/db mice. Mice were i.p. administrated with 25 nmol kg⁻¹ of GLP-1-SSM twice daily for 35 days. (a, b) Body weight gain and food intake amount at indicated time points. (c, d) HbA1c values before and after the 35 day treatment. Means ± SD, n = 6, *p < 0.05, **p < 0.01, ***p < 0.001.

Three IPGTT tests were performed before, during and after the chronic study (day – 1, 18 and 36) to evaluate whether a long-term GLP-1-SSM treatment improved the glucose tolerance of db/db mice. Before the chronic treatment, no difference in glucose tolerance ability was found among db/db mice in each group, and the blood glucose levels in db/db mice were significant higher than that in wild-type mice (Fig. 8a). After 17 days' treatment, the blood glucose levels in mice treated with exendin-4 or GLP-1-SSM were lower than that of saline treated mice during IPGTT, but still higher than that of the wild-type mice (Fig. 8b). To our delight, after the whole chronic treatment period, the glucose tolerance ability of mice treated with exendin-4 or GLP-1-SSM was found to be comparable with that of the wild-type mice (Fig. 8c). The above results demonstrate that a chronic treatment of GLP-1-SSM successfully restores the glucose tolerance ability of db/db mice.

Fig. 8 Glucose tolerance tests in db/db and wild-type mice during the chronic GLP-1-SSM treatment. IPGTT tests were performed at day −1 (a), day 18 (b) and day 36 (c) with 1 g kg⁻¹ glucose challenge, the blood glucose levels (mmol L⁻¹) were shown as means ± SD, n = 6.

Histologic analyses

All the above results demonstrate that the SSM modification not only retains the insulinotropic and hypoglycemic activities of GLP-1, but also results in a significantly prolonged glucose-lowing effect which was even superior to exendin-4. It is well known that GLP-1 increases beta-cell mass and stimulates beta-cell proliferation. Therefore, histological assays were employed to examine the effects of GLP-1-SSM on pancreatic islets in db/db mice. After the above 35 days' chronic treatment, representative histological sections from the three groups of mice are illustrated in Fig. 9a. Shrunken pancreatic islet cells were observed in untreated db/db mice. In contrast, higher pancreatic islet cell numbers and larger pancreatic islet cell areas were found in mice treated with exendin-4 or GLP-1-SSM (Fig. 9b and c). Finally, the influence of GLP-1-SSM on insulin-positive beta cells in mice chronically treated with exendin-4 or GLP-1-SSM was investigated by morphometric analysis. Representative pancreatic islet sections from each group are illustrated in Fig. 9d. It was evident that the number of insulin-positive beta cells in mice treated with exendin-4 or GLP-1-SSM significantly increased as compared with that in saline group (Fig. 9e).

Fig. 9 Effects of GLP-1-SSM and exendin-4 on pancreatic islets in db/db mice. (a) Representative images of histologic samples. (b, c) The number and area of pancreatic islets in each group. Results are shown as the mean ± SD, n = 6. *p < 0.05, **p < 0.01, ***p < 0.001. (d) Representative images of insulin-positive beta cells from groups treated with saline (left), exendin-4 (middle) and GLP-1-SSM (right). (e) The number of insulin-positive beta cells in each group. Results are shown as the mean ± SD, n = 6. *p < 0.05, **p < 0.01, ***p < 0.001.

Conclusions

Therapeutics based on GLP-1 receptor agonists contribute to effective glucose control, promoted beta-cell regeneration, and body weight loss. However, the short in vivo half-life of GLP-1 limits its therapeutic potential.²⁸ Therefore, numerous research efforts are now focused on the development of stable GLP-1 agonists.^29,30 Up to now, five GLP-1 receptor agonists have been approved in Europe and United States. However, the immunogenicity/toxicity concerns of these agonists cannot be overlooked, and their therapeutic utility is hampered by adverse effects.³¹

GLP-1-SSM was previously reported as a novel formulation of GLP-1 with enhanced receptor agonist activity and anti-inflammatory potency against acute lung injury.²⁷ In the current work, the antidiabetic effect of GLP-1-SSM was studied. After being associated with micelles, the α-helical secondary structure of GLP-1 was enhanced (Fig. 1), and the stability of GLP-1-SSM in vitro and in vivo was significantly improved (Fig. 2 and 3). Moreover, GLP-1-SSM showed better insulinotropic activity and antidiabetic effect than GLP-1 (Fig. 4 and 5). All these results demonstrate that, the increased α-helicity of GLP-1 in GLP-1-SSM successfully increases the hypoglycemic activity of GLP-1. Therefore, SSM formulation of GLP-1 proves to be also an effective strategy for developing novel agents for the intervention of T2DM.

Importantly, GLP-1-SSM was confirmed to be a long-acting glucose-lowering agent as indicated by both glucose stabilizing test and multiple intraperitoneal glucose tolerance test (Fig. 6). The hypoglycemic durations of GLP-1-SSM were greater than that of exendin-4.

Preclinical studies were further conducted. It was found that, both the body weight gain and food intake in db/db mice were reduced after chronic administration of GLP-1-SSM for 35 days, and HbA1c values were significantly decreased (Fig. 7). Considering that HbA1c is a sensitive index of glycemic control, our results indicate the promising role of GLP-1-SSM in clinical studies. Moreover, chronic treatment of GLP-1-SSM successfully increased the glucose tolerance ability of db/db mice (Fig. 8). Finally, histologic and immunohistochemical assays revealed that long-term treatment with GLP-1-SSM could preserve pancreatic beta cells mass and function in db/db mice (Fig. 9).

In summary, the present study, for the first time, evaluated the antidiabetic activity of GLP-1-SSMs. In vitro and in vivo studies demonstrate that GLP-1-SSM has improved antidiabetic activity, enhanced pharmacokinetic behavior, long-acting antidiabetic ability, and long-term treatment beneficial effects, all of which indicating its potential role as a clinical long-acting hypoglycemic agent for the treatment of type 2 diabetes.

Experimental

Materials

Exendin-4 and GLP-1 (7-36)-NH₂ were purchased from GL biochem (Shanghai, China). DSPE-PEG₂₀₀₀ was purchased from Corden Pharma Switzerland LLC (Liestal, Switzerland). Rat insulin ELISA kit was purchased from R&D Systems (Minneapolis, MN). HbA1c kit was purchased from Glycosal (Deeside, UK). All other reagents, unless otherwise indicated, were purchased from Sigma-Aldrich Co. (Saint Louis, MO) and were used as received.

Experimental animals

Male Kunming mice (9 weeks old) and Sprague-Dawley rats (body weight of 200–250 g) were purchased from the Comparative Medical Center of Yangzhou University (Jiangsu, China). Male C57BL/6J-m⁺/⁺ Lepr^db (db/db) mice (6–8 weeks old) and wild-type nondiabetic C57BL/6J mice (6–8 weeks old) were obtained from Model Animal Research Center of Nanjing University (Jiangsu, China). All animal care and experimental protocols were approved by an ethical committee at Jiangsu Normal University and were in accordance with the Laboratory Animal Management Regulations in China, the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (revised 2011). Animals were kept in a thermostatically controlled room (25 ± 2 °C) and relative air humidity (50 ± 10%) with a reverse 12 h day/12 h dark cycle. Tap water and standard laboratory chow were provided ad libitum throughout the study. Animals were allowed to acclimate for one week before the experiment began.

Preparation of GLP-1 in SSM

GLP-1-SSM was prepared as previously reported.²⁷ Briefly, weighed amount of DSPE-PEG₂₀₀₀ (M_w ≈ 2800) was added in 2 mL saline to achieve a concentration of 10 mM and vortexed for 5 min until completely dissolved. The dispersion was allowed to equilibrate at 25 °C for 1 h in the dark to form blank SSM (the CMC of DSPE-PEG₂₀₀₀ was 1 μM). Stock solution of GLP-1 (67 μM) was prepared by dissolving 0.44 mg GLP-1 in 2 mL saline just before use. Then, quantitative GLP-1 stock solution was added to SSM, and further incubated for 2 h at 25 °C in the dark.

Determination of peptide secondary structure

The secondary structure of GLP-1 and GLP-1-SSM was determined by circular dichroism spectroscopy using Jasco J-810 Spectropolarimeter (Jasco, Oklahoma City, USA).^27,32 Each sample solution was recorded between 190–250 nm at room temperature using contained 1 mm path length fused quartz cuvettes with a bandwidth of 1 nm and 1 s response time. All spectra were recorded after an accumulation of three runs and were corrected for blank SSM scans and buffer.

In vitro plasma stability test

The stability of GLP-1-SSM in vitro was assessed in rat plasma.^18,19 Rat plasma was collected from adult male Sprague Dawley rats and was stored at −20 °C until use. GLP-1-SSM was incubated in rat plasma at 37 °C with an initial concentration of 1000 ng mL⁻¹, and GLP-1 (1000 ng mL⁻¹) was used as the control. A 100 μL sample was removed from the incubation solution at 0, 0.25, 0.5, 1, 2, 4, 6, 8, 12, 16, 20, and 24 h time points and subjected to solid-phase extraction on a Waters Oasis HLB 96-well plate. Then 20 μL of the extract was injected into the LC-MS/MS system. The signal of GLP-1 was detected through multiple reaction monitoring (MRM) and using Applied Biosystems Sciex API-4000 instrument (Foster City, CA) to carry out electrospray ionization mass spectrometry (ESI-MS). The analytical conditions of RPLC were identical to our previously described method.^18,19

In vivo pharmacokinetics assessment

The pharmacokinetic behavior of GLP-1-SSM was assessed in male SD rats (n = 3, 200–250 g).^20,21 The rats were subcutaneously injected with exendin-4 (15 nmol/rat) or GLP-1-SSM in saline (15 nmol GLP-1/rat) followed by an overnight fast. Sterilized saline solution was used as the vehicle. Blood samples (approximately 0.1 mL) were collected from tail vein into K₂EDTA-containing microcentrifuge tubes at the following time points: 0, 0.5, 1, 2, 4, 6, 8, 12, 24 h. After centrifugation at 4 °C, plasma was collected and stored at −20 °C until ready for GLP-1 analysis. Plasma concentrations of GLP-1 were measured using a previously described LC-MS/MS assay. Two volumes of acetonitrile (100 μL) containing an internal standard were added into 50 μL of plasma to precipitate plasma proteins. The sample was mixed by vortex, and the precipitated proteins were removed by centrifuge. The supernatant (10 μL) was used for analysis.

Insulin secretion assay

The insulin secretion assays were carried out by i.p. injecting GLP-1-SSM (25 nmol kg⁻¹ GLP-1) into male SD rats (200–250 g).^18,19 Saline and GLP-1 (25 nmol kg⁻¹) were injected i.p. as controls. Briefly, the SD rats were fasted overnight (12 h) and randomly allocated to three groups (n = 6), which were administered with saline, GLP-1 or GLP-1-SSM, respectively, in combination with intraperitoneal glucose load (8 g kg⁻¹) at 0 min. Blood samples were collected by tail vein incision at −10, 0, 5, 10, 15, 30, 45, 60, 90, and 120 min. Blood samples were then assayed for insulin levels using a Rat Insulin ELISA kit.

Glucose-lowering ability evaluation in normal Kunming mice

The glucose-lowering effect of GLP-1-SSM was evaluated using a modified IPGTT on male Kunming mice (9 weeks old).²¹ Briefly, 36 mice were randomly divided into three groups (n = 6). The mice were fasted overnight but allowed free access to drinking water before i.p. injection. Each group of Kunming mice were administrated with glucose (2 g kg⁻¹) in combination with saline (control), GLP-1 (25 nmol kg⁻¹) or GLP-1-SSM (25 nmol kg⁻¹ GLP-1) at 0 min. A drop of blood was drawn from tail veins at −30, 0, 15, 30, 45, 60 and 120 min. The blood glucose levels were determined using a one-touch blood glucose monitor (Ascensia, Breeze 2, Bayer, Germany).

Long-term glucose stabilizing activity evaluation in db/db mice

The hypoglycemic efficacies of the GLP-1-SSM were measured using a modified IPGTT in type 2 diabetes db/db mice.³³ Briefly, 18 h fasted db/db mice were randomly divided into twelve groups (n = 6). To evaluate the long term glucose lowering effect of GLP-1-SSM in db/db mice, the mice were administrated with saline (control), exendin-4 (25 nmol kg⁻¹), or GLP-1-SSM (25 nmol kg⁻¹ GLP-1) at predetermined times (−30, −120, −240, −360 min) before the glucose challenge (1 g kg⁻¹). A drop of blood was drawn from tail veins at 0, 15, 30, 60 and 120 min. The blood glucose levels were determined using a one-touch blood glucose monitor (Ascensia, Breeze 2, Bayer, Germany).

In order to assess the long-term ability of GLP-1-SSM to reduce glucose levels, a modified multiple OGTT in db/db mice was used.^20,21 To better mimic the diabetic patient diet and the postprandial blood glucose by the meal ingestion, glucose and GLP-1-SSM were respectively administered orally and i.p. Briefly, the db/db mice (n = 6) were fasted overnight (18 h). Male db/db mice were administered with saline (vehicle control), exendin-4 (25 nmol kg⁻¹) or GLP-1-SSM (25 nmol kg⁻¹ GLP-1) 0.5 h before the first oral glucose load (2.0 g kg⁻¹). A drop of blood (1 μL) was collected from tail veins at 0, 0.25, 0.5, 0.75, 1, 2 and 3 h, and blood glucose levels were determined using a blood glucose monitor (Ascensia, Breeze 2, Bayer, Germany). After the first OGTT, the next glucose loads were administered at 6 and 12 h, respectively. The blood collected time intervals were the same for each glucose load.

Chronic treatment studies

Effects of chronic GLP-1-SSM administration were tested in C57BL/6J-m⁺/⁺ Lepr^db (db/db) mice, as a further probe of potential therapeutic utility.³⁴ Six-week-old male db/db mice were assigned into three groups with matched HbA1c (n = 6). Based on the pharmacokinetic profile of GLP-1-SSM, one group was administered with GLP-1-SSM (i.p., 25 nmol kg⁻¹ of GLP-1, twice daily for 35 days), and another treatment group was administered with exendin-4 (i.p., 25 nmol kg⁻¹, twice daily for 35 days). Saline was used as the vehicle control and was injected i.p. twice daily into the third group. Wild-type nondiabetic C57BL/6 mice (n = 6) were further used for comparison. Body weight and food consumption were monitored daily. HbA1c was measured at day 0 and at day 36 during the treatment cycle.

To evaluate the effects of GLP-1-SSM on diabetic state more precisely, mice in all groups were subjected to IPGTT at day −1, 18 and 36. Briefly, mice were fasted overnight (18 h) and 1 g kg⁻¹ glucose was administered i.p. at 0 min. Blood samples were taken from the tip of the tail of each animal. Blood glucose levels were determined using a blood glucose monitor (Ascensia, Breeze 2, Bayer, Germany) and were measured at the following time points: 0, 15, 30, 45, 60, 120 min after glucose administration.

Histologic analyses

To examine the effects of GLP-1-SSM on the number and area of pancreatic islets, the pancreatic tissues of db/db mice after 4.10 treatment were isolated, washed by 0.9% NaCl and fixed in 10% formalin (phosphate-buffered) overnight.^35,36 After dehydration, the tissues were embedded in paraffin and were sliced into 4 mm sections, which were stained with hematoxylin–eosin (H&E) for histopathological assessment. The number and total area of islets in HE-stained samples were measured using Olympus DP2-BSW digital camera software (Olympus, Center Valley, PA).

To investigate the influence of GLP-1-SSM on insulin-positive beta cells, the pancreatic tissues of db/db mice after 4.10 treatment were isolated, fixed in 4% paraformaldehyde (phosphate-buffered) saline overnight at 4 °C.^37,38 Then, the tissues were embedded in paraffin blocks, and embedded tissue was sliced into 3 mm sections. For immunohistochemistry, sections were treated with sodium citrate buffer followed by microwave heating, and a 30 min H₂O₂/methanol (3/97, v/v) treatment to block endogenous peroxidase. Sections were further washed with 0.01 mM phosphate buffer for 10 min and immunostained with the primary antibody (Insulin (C27C9) Rabbit mAb, Danvers, USA). After washing, they were incubated with secondary antibody (anti-rabbit secondary antibodies K5007, Dako, Japan). Stained sections were observed under a fluorescence microscope (Nikon Eclipse Ti-SR, Nikon, Japan) and digital images were collected.

Statistical analysis

All values were presented as mean ± SD, n referring to the number of mice in each group. The data obtained from different genotypes were analysed by Student's t-test or ANOVA, as appropriate. All of the statistical tests were performed with the GraphPad Prism software, version 5.0 (GraphPad Software, San Diego, CA).

Acknowledgements

This work is supported by the National Natural Science Foundation of China (81602964 and 81602960), Natural Science Foundation of Jiangsu Province (BK20150243 and BK20161028), Natural Science Foundation of the Higher Education Institutions of Jiangsu Province (14KJB350003 and 16KJB350002), Jiangsu Key Laboratory of New Drug Research and Clinical Pharmacy (KF-XY201404), Natural Science Foundation of JSNU (14XLR001), PAPD of Jiangsu Higher Education Institutions, and Startup Funding for Introduced Talents of Nanjing Medical University (KY109RC1602).

Notes and references

1. B. Manandhar and J. M. Ahn, J. Med. Chem., 2015, 58, 1020–1037.
2. K. C. Copeland, J. Silverstein, K. R. Moore, G. E. Prazar, T. Raymer, R. N. Shiffman, S. C. Springer, V. V. Thaker, M. Anderson, S. J. Spann and S. K. Flinn, Pediatrics, 2013, 131, 364–382.
3. L. L. Baggio and D. J. Drucker, Gastroenterology, 2007, 132, 2131–2157.
4. J. Lau, P. Bloch, L. Schaffer, I. Pettersson, J. Spetzler, J. Kofoed, K. Madsen, L. B. Knudsen, J. McGuire, D. B. Steensgaard, H. M. Strauss, D. X. Gram, S. M. Knudsen, F. S. Nielsen, P. Thygesen, S. Reedtz-Runge and T. Kruse, J. Med. Chem., 2015, 58, 7370–7380.
5. A. Astrup, R. Carraro, N. Finer, A. Harper, M. Kunesova, M. E. Lean, L. Niskanen, M. F. Rasmussen, A. Rissanen, S. Rossner, M. J. Savolainen, L. Van Gaal and N. N. Investigators, Int. J. Obes., 2012, 36, 843–854.
6. J. J. Meier, Nat. Rev. Endocrinol., 2012, 8, 728–742.
7. M. Seghieri, E. Rebelos, A. Gastaldelli, B. D. Astiarraga, A. Casolaro, E. Barsotti, A. Pocai, M. Nauck, E. Muscelli and E. Ferrannini, Diabetologia, 2013, 56, 156–161.
8. V. A. Fonseca, R. Alvarado-Ruiz, D. Raccah, G. Boka, P. Miossec, J. E. Gerich and E. F. C. G.-M. S. Investigators, Diabetes Care, 2012, 35, 1225–1231.
9. C. F. Deacon and H. E. Lebovitz, Diabetes, Obes. Metab., 2016, 18, 333–347.
10. S. H. Havale and M. Pal, Bioorg. Med. Chem., 2009, 17, 1783–1802.
11. P. Craddy, H. J. Palin and K. I. Johnson, Diabetes Ther., 2014, 5, 1–41.
12. R. Gupta, S. S. Walunj, R. K. Tokala, K. V. Parsa, S. K. Singh and M. Pal, Curr. Drug Targets, 2009, 10, 71–87.
13. N. Mulakayala, U. Reddy Ch, J. Iqbal and M. Pal, Tetrahedron, 2010, 66, 4919–4938.
14. J. B. Buse, M. Nauck, T. Forst, W. H. Sheu, S. K. Shenouda, C. R. Heilmann, B. J. Hoogwerf, A. Gao, M. K. Boardman, M. Fineman, L. Porter and G. Schernthaner, Lancet, 2013, 381, 117–124.
15. M. J. Davies, S. C. Bain, S. L. Atkin, P. Rossing, D. Scott, M. S. Shamkhalova, H. Bosch-Traberg, A. Syren and G. E. Umpierrez, Diabetes Care, 2016, 39, 222–230.
16. R. R. Henry, J. Rosenstock, D. K. Logan, T. R. Alessi, K. Luskey and M. A. Baron, Diabetes Care, 2013, 36, 2559–2565.
17. N. C. Nyborg, A. M. Molck, L. W. Madsen and L. B. Knudsen, Diabetes, 2012, 61, 1243–1249.
18. J. Han, X. Huang, L. Sun, Z. Li, H. Qian and W. Huang, Biochem. Pharmacol., 2013, 86, 297–308.
19. J. Han, L. Sun, Y. Chu, Z. Li, D. Huang, X. Zhu, H. Qian and W. Huang, J. Med. Chem., 2013, 56, 9955–9968.
20. J. Han, L. Sun, X. Huang, Z. Li, C. Zhang, H. Qian and W. Huang, Br. J. Pharmacol., 2014, 171, 5252–5264.
21. J. Han, L. D. Sun, H. Qian and W. L. Huang, Chin. J. Nat. Med., 2014, 12, 613–618.
22. A. Banerjee and H. Onyuksel, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol., 2012, 4, 562–574.
23. A. Banerjee and H. Onyuksel, Pharm. Res., 2012, 29, 1698–1711.
24. A. Banerjee and H. Onyuksel, Nanomedicine, 2013, 9, 722–728.
25. A. Kuzmis, S. B. Lim, E. Desai, E. Jeon, B. S. Lee, I. Rubinstein and H. Onyuksel, Nanomedicine, 2011, 7, 464–471.
26. V. Sethi, I. Rubinstein, A. Kuzmis, H. Kastrissios, J. Artwohl and H. Onyuksel, Mol. Pharm., 2013, 10, 728–738.
27. S. B. Lim, I. Rubinstein, R. T. Sadikot, J. E. Artwohl and H. Onyuksel, Pharm. Res., 2011, 28, 662–672.
28. L. A. Leiter, M. C. Carr, M. Stewart, A. Jones-Leone, R. Scott, F. Yang and Y. Handelsman, Diabetes Care, 2014, 37, 2723–2730.
29. L. P. Miranda, K. A. Winters, C. V. Gegg, A. Patel, J. Aral, J. Long, J. Zhang, S. Diamond, M. Guido, S. Stanislaus, M. Ma, H. Li, M. J. Rose, L. Poppe and M. M. Veniant, J. Med. Chem., 2008, 51, 2758–2765.
30. S. Son, S. Y. Chae, C. W. Kim, Y. G. Choi, S. Y. Jung, S. Lee and K. C. Lee, J. Med. Chem., 2009, 52, 6889–6896.
31. A. Lund, F. K. Knop and T. Vilsboll, Eur. J. Intern. Med., 2014, 25, 407–414.
32. S. B. Lim, I. Rubinstein and H. Onyuksel, Int. J. Pharm., 2008, 356, 345–350.
33. S. Y. Chae, Y. G. Chun, S. Lee, C. H. Jin, E. S. Lee, K. C. Lee and Y. S. Youn, J. Pharm. Sci., 2009, 98, 1556–1567.
34. D. Chen, J. Liao, N. Li, C. Zhou, Q. Liu, G. Wang, R. Zhang, S. Zhang, L. Lin, K. Chen, X. Xie, F. Nan, A. A. Young and M. W. Wang, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 943–948.
35. C. F. Gotfredsen, A. M. Molck, I. Thorup, N. C. Nyborg, Z. Salanti, L. B. Knudsen and M. O. Larsen, Diabetes, 2014, 63, 2486–2497.
36. B. Yang, X. Li, C. Zhang, S. Yan, W. Wei, X. Wang, X. Deng, H. Qian, H. Lin and W. Huang, Org. Biomol. Chem., 2015, 13, 4551–4561.
37. M. J. Jurczak, H. Y. Lee, A. L. Birkenfeld, F. R. Jornayvaz, D. W. Frederick, R. L. Pongratz, X. Zhao, G. W. Moeckel, V. T. Samuel, J. M. Whaley, G. I. Shulman and R. G. Kibbey, Diabetes, 2011, 60, 890–898.
38. C. W. Park, Y. Zhang, X. Zhang, J. Wu, L. Chen, D. R. Cha, D. Su, M. T. Hwang, X. Fan, L. Davis, G. Striker, F. Zheng, M. Breyer and Y. Guan, Kidney Int., 2006, 69, 1511–1517.

---

Footnote

† These authors contributed equally to this work.

---