ZY15557, a novel, long acting inhibitor of dipeptidyl peptidase‐4, for the treatment of Type 2 diabetes mellitus

Abstract

Background and Purpose

Dipeptidyl peptidase (DPP)‐4 inhibitors increase levels of glucagon‐like peptide‐1 (GLP‐1) and provide clinical benefit in the treatment of type 2 diabetes mellitus. As longer acting inhibitors have therapeutic advantages, we developed a novel DPP‐4 inhibitor, ZY15557, that has a sustained action and long half‐life.

Experimental Approach

We studied the potency, selectivity, efficacy and duration of action of ZY15557, in vitro, with assays of DPP‐4 activity. In vivo, the pharmacodymamics and pharmacokinetics of ZY15557 were studied, using db/db mice and Zucker fatty rats, along with normal mice, rats, dogs and non‐human primates.

Key Results

ZY15557 is a potent, competitive and long acting inhibitor of DPP‐4 (K_i 5.53 nM; K_off 3.2 × 10⁻⁴·s⁻¹, half‐life 35.8 min). ZY15557 treatment inhibited DPP‐4 activity, and enhanced active GLP‐1 and insulin in mice and rats, providing dose‐dependent anti‐hyperglycaemic effects. Anti‐hyperglycaemic effects were also observed in db/db mice and Zucker fatty rats. Following oral dosing, ZY15557 significantly inhibited plasma DPP‐4 activity, determined ex vivo, in mice and rats for more than 48 h, and for up to 168 h in dogs and non‐human primates. Allometric scaling predicts a half‐life for ZY15557 in humans of up to 60 h.

Conclusions and Implications

ZY15557 is a potent, competitive and long acting DPP‐4 inhibitor. ZY15557 showed similar DPP‐4 inhibition across different species. ZY15557 showed excellent oral bioavailability in preclinical species. It showed a low plasma clearance (CL) and large volume of distribution (V_ss) across species, resulting in an extended half‐life.

Abbreviations

DPP

dipeptidyl peptidase

GLP‐1

glucagon‐like peptide‐1

H‐Gly‐Pro‐AMC

Gly‐Pro‐7‐amido‐4‐methylcoumarin hydrobromide

MDA

malondialdehyde

MLP

maximum life‐span potential

Na‐CMC

sodium carboxymethyl cellulose

OGTT

oral glucose tolerance test

TG

triglyceride

Introduction

Diabetes is a global health threat. In 2013, 382 million people had diabetes and this number is expected to rise to 592 million by 2035, with the majority of them having Type 2 diabetes mellitus (Guariguata et al., 2014). Type 2 diabetes has become a challenge for healthcare despite the availability of many treatments (Chatterjee and Davies, 2015). Current therapies have limitations due to side effects (particularly weight gain and hypoglycaemia) or contraindications that limit their use, and the need for new drugs for sustained glycaemic control is still largely unmet (Bailey, 2015).

Secretion of the gut‐derived incretin hormone, glucagon‐like peptide‐1 (GLP‐1), is decreased in Type 2 diabetes, while its insulinotropic effect is preserved. This conclusion is based on studies showing that administration of excess GLP‐1 to diabetic patients restores glucose‐induced insulin secretion, and beta‐cell sensitivity to glucose (Holst, 2004). Pharmacological enhancement of the incretin effect is now an established way to treat type 2 diabetes. GLP‐1 receptor agonists improve glucose‐stimulated insulin release and provide good glycaemic control, but need to be given by subcutaneous injection as GLP‐1 agonists are orally less bioavailable (Kim and Egan, 2008). Though their efficacy is similar to that of oral antidiabetic agents, nausea and vomiting are the adverse events that reduce patient compliance to therapy (Zaccardi et al., 2016). On the other hand, intravenous infusions of GLP‐1 agonists reduce fasting glucose to a normal range without any risk of gastrointestinal side effects (Sun et al., 2012). These differential responses to GLP‐1/GLP‐1 receptor agonists when administered intravenously versus subcutaneously suggest that the subcutaneous environment may be detrimental for these peptide agents, or that these agents elicit side‐effects through GLP‐1 receptors on autonomic nerves in subcutaneous adipose tissue (Nauck et al., 2013).

Orally administered inhibitors of dipeptidyl peptidase (DPP)‐4 represent an alternative way to enhance incretin effect for the treatment of Type 2 diabetes (White, 2008). The enzyme DDP‐4 is responsible for degradation of the incretins, GLP‐1 and gastric inhibitory polypeptide (Barnett, 2006). Oral DPP‐4 inhibitors demonstrate glycaemic control without gastrointestinal side effects. It has also been suggested that DPP‐4 may be involved in linking adipose tissue and the metabolic syndrome by acting as an adipokine that may impair insulin sensitivity in an autocrine and paracrine fashion (Sell et al., 2013). A new generation of DPP‐4 inhibitors are being developed for a longer duration of action (Jain et al., 2015; Sheu et al., 2015). ZY15557 is a novel, orally available, and long acting DPP‐4 inhibitor. Here, we report its in vitro profile of action and in vivo pharmacological effects in animal models of diabetes.

Methods

Animals

All animal care and experimental procedures were approved by Institutional Animal Ethics Committee (as per CPCSEA, Committee for the Purpose of Control and Supervision of Experiments on Animals, Government of India) of Zydus Research Centre, which is an AAALAC accredited facility. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny et al., 2010; McGrath and Lilley, 2015). For the in vivo studies, male C57BL/6J mice (9–10 weeks), male C57BL/KsJ–db/db mice (6–8 weeks), male Zucker fatty rats (9–11 weeks), male Wistar rats (9–11 weeks), male Sprague Dawley rats (9–11 weeks), adult male beagle dogs (3–4 years, 15–16 kg), and adult male rhesus macaques (Macaca mulatta, 5–6 year, 9–10 kg) were procured from Zydus Research Centre, and maintained under standard laboratory conditions. Mice and rats were housed in polypropylene cages, with corncobs as a bedding material. A maximum of six mice or three rats were housed in a single cage. Beagle dogs and monkeys were individually housed in stainless steel cages. Housing rooms were maintained at 25 ± 3°C, humidity 30 ‐ 70% and 12/12 h light and dark cycle with food and water provided ad libitum, in a pathogen free environment.

C57BL/6J mice are a mildly hyperglycaemic strain, mostly used in diabetes and obesity research (Toye et al., 2005). Zucker fatty rats are leptin receptor‐deficient rats that are insulin‐resistant, while db/db mice are leptin receptor‐deficient and obese, and display many features of Type 2 diabetes (Augstein and Salzsieder, 2009). Wistar rats and Sprague Dawley rats are commonly used in most of the pharmacological and toxicological research. Beagle dogs are generally used to assess efficacy and safety before testing the new chemical entities in non‐human primates. Beagle dogs are closer to humans than rodents, hence are used in these experiments (Lui et al., 1986). Generally, rhesus macaques (M. mulatta) is the non‐human primate species used to assess the effects of new chemical entities, because of their similarity to humans, in terms of metabolism and other physiological characteristics (Phillips et al., 2014).

General procedures

For in vivo pharmacodynamic studies, the DPP‐4 inhibitors were administered by oral gavage. The oral formulation was an aqueous suspension in 0.5% (w/v) sodium carboxymethyl cellulose (Na‐CMC) in water, with 5% (v/v) Tween 80 and 5% (v/v) PEG400 added.

For the in vitro and in vivo experiments, blood samples were collected from the retro‐orbital plexus of mildly anaesthetized mice and rats, and from the saphenous vein of conscious dogs and monkeys. Isofluorane (4% in oxygen from a precision vaporizer) anaesthesia was used for mice and rats. Mice and rats were killed individually by adjusting the isofluorane flow rate or concentration to 5%, and isofluorane exposure was continued for one min after cessation of breathing.

In vitro enzyme inhibition assay

Human recombinant DPP‐4, DPP‐8 and DPP‐9 enzymes were purchased from BPS Bioscience (San Diego, CA). In vitro enzyme inhibition assays were performed by mixing 10 μL of ZY15557 working stock solutions, 50 μL of the DPP‐4 substrate Gly‐Pro‐7‐amido‐4‐methylcoumarin hydrobromide (H‐Gly‐Pro‐AMC), and 10 μL human recombinant DPP‐4 enzyme in 30 μL of assay buffer (20 mM Tris HCl, 100 mM NaCl, 1 mM EDTA at pH 8.0). The final incubation (100 μL) in each well contained 1 ng DPP‐4 enzyme and 5 μM substrate. The plates were incubated at 37°C for 30 min, and fluorescence was measured at 360/460 nm using a Synergy HT multi detection microplate reader (BioTek Instruments, Winooski, VT, USA).

Inhibition of endogenous DPP‐4 in serum from mouse, rat, dog, monkey and human was measured similarly, with 10 μL serum was the source of enzyme. Inhibition of DPP‐8 and DPP‐9 was studied with the same assay procedure and substrate, using recombinant enzyme; in these assays, each final incubation, per well, contained 20 and 10 ng of enzyme for DPP‐8 and DPP‐9 respectively.

For association kinetics studies, the enzyme reaction were carried out using the procedure described above with recombinant human DPP‐4, except that fluorescence intensity was measured every 20 s for a total of 1200 s (Tahara et al., 2009). The association rate constant was determined by fitting the data to a one‐phase exponential association equation.

For dissociation kinetics studies (Thomas et al., 2008), DPP‐4 was extracted from confluent Caco‐2 cells. Briefly, after 5 min of incubation with lysis buffer (120 mM Tris HCl, 100 mM NaCl, 1 mM EDTA, 0.04 U·mL⁻¹ aprotinin, 0.5% Nonidet P40, pH 8.0), cells were centrifuged at 20 817 × g at 4°C for 30 min, and the supernatant was stored at −80°C until use. Dissociation kinetics of the inhibitors (ZY15557, sitagliptin or omarigliptin) from the DPP‐4 enzyme was determined after a 1 h preincubation of Caco‐2 cell extracts with high inhibitor concentrations (250 nM ZY15557, 250 nM sitagliptin and 100 nM omarigliptin). The enzymic reaction was started by adding the substrate H‐Gly‐Pro‐AMC after a 3000‐fold dilution of the preincubation mixture with assay buffer. Under these conditions, the difference in DPP‐4 activity at a certain time point in the presence or absence of an inhibitor reflects the amount of this inhibitor still bound to the DPP‐4 enzyme. Maximal reaction rates (fluorescence units s⁻¹) in 10 min intervals were measured and calculated using formula [(V_control − V_inhibitor)/V_Control], where, V_Control = Velocity of control, V_Drug = Velocity of drug.

The dissociation rate constant was obtained by fitting the data to a one‐phase exponential decay equation in the GraphPad Prism program.

The duration and reversibility of DPP‐4 inhibition by ZY15557 was further evaluated by dialyzing enzyme‐inhibitor complex under sink conditions for 24 h to deplete the inhibitor, followed by assessment of DPP‐4 activity. Briefly, the reaction mixtures contained 20 μL inhibitor ZY15557 or sitagliptin (final concentration 300 nM) or vehicle, 20 μL recombinant human DPP‐4 or human serum DPP‐4 enzyme, in 160 μL assay buffer (20 mM Tris HCl, 100 mM NaCl, 1 mM EDTA, pH 8.0). The reaction mixture was dialysed against assay buffer for 24 h at room temperature (water bath, 100 r.p.m.) under sink conditions. Following 24 h dialysis, the reaction mixture from the donor side was recovered, and residual DPP‐4 activity in it was determined by mixing a 50 μL aliquot with 50 μL substrate solution (final substrate concentration 5 μM). Each well contained either 1 ng recombinant DPP‐4 or 10 μL serum, and 5 μM substrate.

Acute effects in C57 mice, Wistar rats, and Zucker fatty rats

Male Wistar rats and C57 mice were orally treated with ZY15557 (0.03, 0.1, 0.3, 1 and 3 mg·kg⁻¹) or vehicle, after overnight fasting. After 1 h following drug administration, an oral glucose (5 g·kg⁻¹) challenge was given to the animals. Approimately 10 min following the glucose challenge, animals were bled by retroorbital puncture and plasma was separated for determination of DPP‐4 activity, and levels of insulin using elisa (Crystal Chem Inc., IL, USA). A sample was collected in tubes containing 5% EDTA solution with 10 μM sitagliptin to prevent degradation of GLP‐1; plasma was separated and analysed for active GLP‐1 by elisa (Epitope Diagnostics Inc, San Diego, California, USA). Separately, DPP‐4 assay was performed by mixing 5 μL plasma to 35 μL assay buffer (20 mM Tris HCl, 100 mM NaCl, 1 mM EDTA in water, pH 8.0). After a 5 min preincubation, reaction was initiated by addition of 40 μL assay buffer containing 100 μM substrate H‐Gly‐Pro‐AMC. After incubation for 20 min, fluorescence was measured.

In a separate experiment, overnight fasted animals C57 mice and Zucker fatty rats were administered a glucose load (5 g·kg⁻¹, p.o.) at 15 min after vehicle or ZY15557 administration. Blood glucose was estimated using a glucometer (One‐Touch Ultra‐Blood Glucose Meter, Life Scan Inc., Milpitas, CA) at 0, 15, 30, 60 and 120 min after glucose administration. The calculation was based on values from water‐control animals, who were given water instead of glucose solution. The difference between the AUC glucose of vehicle control group and water control group was considered as 100% suppression. Based on it, percentage reduction in AUC glucose at each dose level in treated animals was determined relative to the vehicle‐treated group.

Chronic effects in db/db mice

Eight weeks old male db/db mice were randomized based on the random (non‐fasted) glucose levels and treated with ZY15557 at 3 and 10 mg·kg⁻¹, once daily, for 10 weeks. At the end of the study, all animals were bled by retroorbital puncture and % HbA1c was measured by Tina‐quant® Haemoglobin A1C Gen. 3 kit (Roche Diagnostics Ltd., Switzerland), and fasting glucose were measured by the method mentioned above. Other biochemical parameters included plasma insulin using elisa (Crystal Chem Inc., IL, USA), plasma adiponectin (K1002–1, B‐Bridge International, Inc., USA) and leptin (K1006–1, B‐Bridge International, Inc., USA).

Triglyceride (TG) and total cholesterol were estimated in liver tissues of the chronically treated animals (Patel et al., 2013). Liver tissue was homogenized in a heptane–isopropanol–Tween mixture (3:2:0.01 by volume) and centrifuged at 1500 x g for 15 min at 4°C. The supernatant (the upper phase containing extracted TGs) was collected and evaporated using a nitrogen evaporator. TG and cholesterol content was determined using kits (Pointe Scientific, USA) and expressed as mg·g⁻¹ of liver.

Lipid peroxidation was assessed by measuring malondialdehyde (MDA) formation, using the thiobarbituric acid (TBA) assay (Wlostowski et al., 2008). To 0.2 mL of the tissue homogenate, 0.2 mL of 8.1% sodium dodecyl sulfate, 1.5 mL of 20% acetic acid, 1.5 mL of 0.8% TBA, and 0.6 mL distilled water were added and vortexed. The reaction mixture was placed in a water bath at 95°C for 1 h. The resultant samples were centrifuged at 10 621 × g for 6 min, and absorbance of the supernatant was measured at 532 nm. The amount of MDA was calculated using its molar extinction coefficient 1.56 × 10⁵ M⁻¹·cm⁻¹ and reported as nmoles of MDA mg⁻¹ protein.

Pharmacokinetic studies

The intravenous and oral pharmacokinetics of ZY15557 was studied in mice (i.v.: 1 mg.kg⁻¹; p.o.: 3 mg.kg⁻¹), rats (i.v.: 1 mg.kg⁻¹; p.o.: 2 mg.kg⁻¹), dogs (i.v.: 0.5 mg.kg⁻¹; p.o.: 3 mg.kg⁻¹) and monkeys (i.v.: 0.5 mg.kg⁻¹; p.o.: 3 mg.kg⁻¹) (route and dose in parentheses). The i.v. dose formulation was prepared in normal saline (0.9%), and it was administered as a slow bolus (dose volume: 5 mL·kg⁻¹ in mice and rats via the tail vein; 0.5 mL·kg⁻¹ in dogs and monkeys via saphenous veins). The oral dose suspension formulation for mice, dogs and monkeys comprised 0.5% (w/v) Na‐CMC in water with 1% (v/v) Tween‐80 added; and for rats, it comprised 0.5% (w/v) Na‐CMC in water with 5% (v/v) Tween 80 and 5% (v/v) PEG‐400. The oral dose was administered by gavage (dose volume: 10 mL·kg⁻¹ in mice; 5 mL·kg⁻¹ in rats; 1 mL·kg⁻¹ in dogs and monkeys). Blood samples were collected predose by retroorbital puncture and at 0.083 (i.v. only), 0.25, 0.5, 1, 2, 4, 6, 8, 24, 48, 72, 96, 120, 144 and 168 h post‐dose, and processed to plasma by centrifugation. ZY15557 concentration was quantified in the plasma samples using a sensitive and specific liquid chromatography tandem mass spectrometry method with a lower limit of quantitation of 0.1 ng·mL⁻¹.

Interspecies allometric scaling to predict human pharmacokinetics

Allometric scaling of the preclinical pharmacokinetic data was performed using standard approaches (Mahmood, 2007). Volume of distribution (V_ss) at steady‐state was scaled by simple allometry, and clearance (CL) was scaled by different three approaches: simple allometry, maximum life‐span potential (MLP) based scaling, and brain weight based scaling.

Simple allometry for clearance and V_ss was performed by fitting the species‐specific preclinical clearance parameters and the respective animal body weights to the equation:

$$\mathrm{Y}=a{\mathrm{W}}^b,\text{which}\mathrm{can}\mathrm{be}log-\text{transformed as}\text{Log}\mathrm{Y}=\mathrm{Log}a+b\mathrm{Log}\mathrm{W}$$

where, Y represents clearance or volume of distribution, W is body weight, and a and b are the coefficient and exponent of the allometric equation respectively.

For MLP‐based scaling of clearance, the product of the MLP and the measured preclinical clearance was generated for each species, and the product was used for allometric scaling, where MLP is given by:

$$\mathrm{MLP}\left(\text{in years}\right)=185.4{\left(\text{Brain Weight}\right)}^{0.636}{\left(\text{Body Weight}\right)}^{-0.225}$$

Similarly, for brain weight‐based scaling of clearance, the product of brain weight and measured preclinical clearance was generated for each species, and the product was used for allometric scaling.

In all cases, the derived constants, a and b, from the allometric scaling equation were used to back‐calculate predicted estimates of clearance (CL_human,pred) and volume of distribution (V_{ss human,pred}) for a 70 kg human. These predicted values were then used to predict the plasma half‐life (t _{1/2, human, pred}) of ZY15557 in humans based on the equation shown below:

$$t_{1/2,\text{human},\text{pred}}=0.693\left({\mathrm{V}}_{\mathrm{ss}\text{human},\text{pred}}\right)/{\mathrm{CL}}_{\text{human},\text{pred}}$$

Data and statistical analysis

The data and statistical analysis complied with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015). In all in vivo studies, biochemical analysis of individual samples was carried out by operators, who were blinded to the treatments. During all in vitro studies, samples were randomized in 96 well plates. Sample identity was known to the investigator (not blinded), because all results had fixed numerical values that were not open to interpretation.

Quantitative results are expressed as the mean ± SEM. Statistical significance was determined by one‐way ANOVA for multiple comparisons with post hoc Dunnett's test. Post hoc tests were run only when F achieved P < 0.05. P < 0.05 was considered as significant. There was no exclusion of any data in all studies. All in vitro and pharmacology analyses data were analysed by using GraphPad Prism version 6.01 (GraphPad Software, San Diego, CA, USA). Pharmacokinetic parameters of ZY15557 were determined using the non‐compartmental analysis module of WinNonlin® software (Version 5.3).

Materials

Human recombinant DPP‐4, DPP‐8 and DPP‐9 enzymes were purchased from BPS Bioscience (San Diego, CA). ZY15557 (Figure 1) and sitagliptin phosphate (Kim et al., 2005) were synthesized at Zydus Research Centre, Ahmedabad, India. All other chemicals, including the DPP substrate H‐Gly‐Pro‐AMC, were purchased from Sigma‐Aldrich (St. Louis, MO, USA).

Figure 1.

Chemical structure of ZY15557

Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Southan et al., 2016), and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (Alexander et al., 2015a,b).

Results

In vitro activity of ZY15557

ZY15557 inhibited DPP‐4 activity in vitro (Figure 2A) with a K_i of 5.53 ± 0.06 nM. The corresponding K_i for sitagliptin was 9.83 ± 3.02 nM. The K_i values for inhibition of human recombinant DPP‐8 and DPP‐9 by ZY15557, shown in Table 1), demonstrated that ZY15557 was more than 6000‐fold selective for recombinant DPP‐4 relative to DPP‐8, and 3000‐fold selective relative to DPP‐9. ZY15557 potently inhibited serum DPP‐4 activity across all the species tested (mouse, rat, dog, monkey and human), with K_i values varying only two‐fold between the species (Table 2). In the same assay, the K_i values of sitagliptin for inhibition of mouse, rat, dog, monkey and human serum DPP‐4 varied more than ten‐fold (Table 2). A Lineweaver‐Burk plot of the data (Figure 2B) for ZY15557 showed that the reaction velocity for DPP‐4 inhibition decreased with increase in substrate concentration, indicating that ZY15557 inhibited the DPP‐4 enzyme by competition at the substrate binding site.

Figure 2.

In vitro DPP‐4 inhibition (A), Lineweaver‐Burk analysis plot showing inhibition of DPP‐4 by ZY15557 (B), and association (C), and dissociation (D) kinetics, for inhibition of DPP‐4 by ZY15557. Data are expressed as means ± SEM (n = 3). For association kinetics experiments, substrate was preincubated with ZY15557 (0–300 nM) and enzyme reaction was initiated by adding rat plasma. Fluorescence intensity was measured every 20 s for total of 1200 s. For dissociation kinetics experiments, ZY15557 (250 nM), sitagliptin (250 nM) and omarigliptin (100 nM) was preincubated with Caco‐2 cell extract and enzyme reaction was initiated by adding substrate. Fluorescence intensity was measured every 10 min.

Table 1.

Selectivity of ZY15557 as inhibitor of DPP isoforms

Enzyme	Ki value
DPP‐4	5.53 ± 0.06 nM
DPP‐8	37.48 ± 0.63 μM
DPP‐9	19.48 ± 0.98 μM

Values are the mean ± SEM (n = 3)

Table 2.

Effect of ZY15557 on DPP‐4 inhibition across species

DPP‐4 source	Ki value (nM)
	ZY15557	Sitagliptin
Mice	11.8 ± 0.4	80.7 ± 3.4
Rat	13.7 ± 1.0	6.9 ± 3.9
Dog	6.1 ± 1.1	9.3 ± 0.2
Non‐human primate	8.1 ± 1.3	11.6 ± 0.2
Human	5.1 ± 0.1	29.4 ± 1.7

Values are mean ± SEM (n = 3)

In enzyme kinetic studies, ZY15557 showed a clear, time‐dependent approach toward steady state, which is characteristic of slow‐binding inhibition. The association rate constants (K_on) of ZY15557 and sitagliptin were found to be 38.6 × 10⁴ M⁻¹·s⁻¹ and 13.5 × 10⁴ M⁻¹·s⁻¹ (Figure 2C). Both ZY15557 and sitagliptin, tested at 30 nM, showed sustained DPP‐4 inhibition over time. The dissociation rate constants (K_off) for ZY15557, sitagliptin and omarigliptin were found to be 3.23 × 10⁻⁴, 17.04 × 10⁻⁴ and 4.49 × 10⁻⁴ s⁻¹, respectively (Figure 2D), indicating that dissociation of bound ZY15557 from the DPP‐4 enzyme was 5.3 times slower than sitagliptin and similar to omarigliptin. The half‐lives of the enzyme‐inhibitor complex (EI half‐life) for ZY15557, sitagliptin and omarigliptin were 35.8, 6.8 and 25.7 min respectively. In dialysis experiment, ZY15557 showed more than 6 fold inhibition of DPP‐4 than Sitagliptin in rat and human sera (Table 3).

Table 3.

DPP‐4 inhibition remaining (%) following dialysis of enzyme‐inhibitor complex

DPP‐4 source	Sitagliptin (300 nM)	ZY15557 (300 nM)
Human Recombinant	0 ± 0	45.2 ± 1.7
C57 mice serum	45.0 ± 1.0	55.20 ± 0.9
Sprague Dawley rat serum	7.8 ± 0.2	50.20 ± 1.2
Human serum	7.2 ± 0.3	65.20 ± 2.3

Values are the mean ± SEM (n = 3)

Effect of ZY15557 on plasma GLP‐1, insulin and DPP‐4 inhibition in C57 mice and Wistar rats

At 10 min after glucose administration and 70 min after treatment with a range of doses of ZY15557 (0.03 ‐ 3 mg·kg⁻¹), C57 mice showed a dose‐dependent decrease in plasma DPP‐4 activity, compared with the control values (Figure 3A). In the same animals, there was a dose‐related increase in active GLP‐1 levels (Figure 3B), and insulin levels were 2.3 ± 0.2, 2.3 ± 0.2, 3.1 ± 0.1, 3.4 ± 0.2 and 4.4 ± 0.3 ng·mL⁻¹, respectively, Vs vehicle control showing 2.4 ± 0.2 ng.mL⁻¹ (Figure 3C). In Wistar rats, treatment with the same dose range of ZY15557 showed a similar profile of responses with a decrease in plasma DPP‐4 activity (Figure 3D), increase in active GLP‐1 levels (Figure 3E), and increased insulin levels (Figure 3F).

Figure 3.

Effect of a single oral administration of ZY15557 (0.03‐3 mg kg^‐1) to C57BL/6J mice and Wistar rats. (A) Plasma DPP‐4 activity, (B) Plasma active GLP‐1 levels, and (C) Plasma insulin levels, following OGTT in male C57BL/6J mice. (D) Plasma DPP‐4 activity, (E) Plasma active GLP‐1 levels, (F) and Plasma insulin levels following OGTT in male Wistar rats. ZY15557 was administered 60 min before the oral glucose load. Plasma DPP‐4 activity, plasma active GLP‐1 and plasma insulin were measured 10 min after glucose load. Data are expressed as means ± SEM; n = 6 per group for all the studies. * P < 0.05, significantly different from vehicle control; one way ANOVA followed by Dunnett's post hoc test.

Reduction in glucose excursion after oral glucose tolerance test by ZY15557 in C57 mice and Zucker fa/fa rats

Single dose oral treatment with ZY15557 (0.01 ‐ 3 mg.kg^‐1) during the oral glucose tolerance test (OGTT) in C57 mice, resulted in reduction in the AUC for glucose (mM*120 min) (Figure 4A), which was significant at 0.3 mg.kg^‐1 and above. Similar treatment with ZY15557 of Zucker fa/fa rats showed similar reductions (Figure 4B) in AUC glucose during OGTT but, in these animals, starting at a much lower dose (0.01 mg kg^‐1).

Figure 4.

Effect of ZY15557 on OGTT in C57BL/6J mice and Zucker fa/fa rats. Suppression of glucose AUC (%) in C57 mice (A) and in Zucker fa/fa rats (B). A single oral dose of ZY15557 was administered 15 min before the glucose load. Data are expressed as means ± SEM; n = 5 for C57 mice and Zucker fa/fa rats. * P < 0.05, significantly different from vehicle control; one way ANOVA followed by Dunnett's post hoc test.

Effect of ZY15557 on glycated haemoglobin and glucose in db/db mice

As shown in Table 4, after 10 weeks of treatment, ZY15557 (3 or 10 mg·kg⁻¹) decreased % HbA1c, with the higher dose (10 mg·kg⁻¹) bringing the glycated haemoglobin down to the level of non‐diabetic C57 control mice. ZY15557 also reduced glucose levels and plasma insulin levels in both treatment groups. The random glucose levels in the treatment groups were significantly higher than in non‐diabetic control mice and the high levels of insulin were decreased to nearly normal normal values after treatment with 10 mg·kg⁻¹ ZY155557. The hyperlipidaemic disease phenotype of the db/db mice was also improved by treatment with ZY15557 (reducing hepatic TGs by 18 and 24% and hepatic cholesterol by 9 and 16% with 3 and 10 mg·kg⁻¹ respectively), which was accompanied by a dose‐related reduction in hepatic oxidative stress (in terms of liver MDA). ZY15557 treatment caused a dose‐related reduction in plasma leptin (12 and 22% reduction with 3 and 10 mg·kg⁻¹ of ZY15557), and improvement in plasma adiponectin (10 and 36% increase with 3 and 10 mg·kg⁻¹ of ZY15557 respectively) in the db/db mice.

Table 4.

Effect of 10 weeks treatment with ZY15557 on biochemical parameters in db/db mice

Biochemical parameter	Vehicle Control (db/db mice)	ZY15577 (3 mg·kg−1)	ZY15577 (10 mg·kg−1)	Non diabetic control (C57 mice)
Plasma glucose (mM)	17.4 ± 2.7†	10.9 ± 1.6* †	10.1 ± 1.3* †	4.4 ± 0.4*
Blood HbA1c (%)	7.7 ± 0.1†	5.6 ± 0.2* †	4.9 ± 0.2*	4.9 ± 0.1*
Plasma insulin (ng·mL−1)	2.4 ± 0.1†	2.1 ± 0.2†	1.7 ± 0.1*	1.0 ± 0.1*
Hepatic triglycerides (mg·g−1 of liver)	23.3 ± 1.6†	19.1 ± 1.0†	17.7 ± 1.3*	13.2 ± 1.5*
Hepatic cholesterol (mg·g−1 of liver)	12.9 ± 0.7†	11.8 ± 0.4†	10.9 ± 0.9†	5.5 ± 0.7*
Plasma adiponectin (ng·mL−1)	58.3 ± 3.9†	63.9 ± 6.3†	79.0 ± 4.5*	90.3 ± 4.7*
Plasma leptin (ng·mL−1)	22.9 ± 1.2†	20.2 ± 0.9†	17.9 ± 1.1*	15.7 ± 1.0*
MDA (nM·mg−1 of protein)	1.9 ± 0.2†	1.5 ± 0.2†	1.2 ± 0.2*	0.8 ± 0.1*
Body weight (g)	42.6 ± 1.2†	43.5 ± 0.9†	39.6 ± 0.5†	32.0 ± 0.8*

Data shown are means ± SEM (n = 7 for db/db mice and n = 6 for non‐diabetic control C57BL/6J mice.

^* P < 0.05, significantly different from vehicle control;

^† P < 0.05, significantly different from non‐diabetic control.

Pharmacokinetics and PK‐PD correlation of ZY15557 in mice, rats, dogs and primates

The pharmacokinetics of ZY15557 was determined after single dose p.o. and i.v. administration to male C57 mice (i.v.: 1 mg·kg⁻¹; p.o.: 3 mg·kg⁻¹), male Sprague–Dawley rats (i.v.:1 mg·kg⁻¹; p.o.: 2 mg·kg⁻¹), beagle dogs (i.v.: 0.5 mg·kg⁻¹; p.o.:3 mg·kg⁻¹) and cynomolgus monkeys (i.v.: 0.5 mg·kg⁻¹; p.o.: 3 mg·kg⁻¹). Plasma concentrations and pharmacokinetic parameters for ZY15557 in these species are summarized in Table 5. ZY15557 showed slower clearance in monkeys than in rats and mouse, resulting in slower elimination in the higher species. The absolute oral bioavailability of ZY15557 was high, and comparable across the species (about 80‐90%). Following oral dosing in animals, ZY15557 had a long duration of action as an inhibitor of plasma DPP‐4 in vivo (Figure 5) with measurable levels in plasma for up to 168 h, in all species.

Table 5.

Preclinical pharmacokinetic parameters of ZY15557

	Mouse	Rat	Dog	Monkey
Vss, i.v. (L·kg−1)	5.12	5.32 ± 0.41	2.94 ± 0.10	4.94 ± 0.44
CLi.v. (mL·min−1·kg−1)	12.07	19.12 ± 1.91	3.00 ± 0.11	2.87 ± 0.24
t 1/2, i.v. (h)	44.41	28.13 ± 2.93	52.89 ± 7.68	61.16 ± 8.74
Oral Bioavailability (%F)	86	91.21 ± 2.93	78.43 ± 4.61	80.93 ± 12.13

Values shown are mean ± SD (n = 5, except for i.v. data in dog where n = 2). Only mean values are shown for mouse because of sparse sampling design.

Figure 5.

Inhibition of DPP‐4 enzyme activity and plasma concentration of ZY15557 in rats (A), dogs (B) and non‐human primates (C) after a single oral administration of ZY15557 (2 mg·kg⁻¹ for rats, 3 mg·kg⁻¹ for dogs and non‐human primates). Data are expressed as mean ± SEM (n = 5 for rats, n = 4 for primates and n = 3 for dogs).

Allometric scaling for prediction of human pharmacokinetics

Using three methods of allometric scaling, we were able to calculate predicted CL (mL·min⁻¹·kg ⁻¹) and predicted t _1/2 (h) values for humans (Table 6). The log–log plots of volume of distribution (V_ss) and clearance (CL) versus body weight are shown in Figure 6. They include the extrapolated or predicted values for humans. Based on simple allometry, the exponent b for V_ss was 0.940 (Figure 6A), and for CL it was 0.719 (Figure 6B); these values are within the acceptable range (0.8–1.1 for V_ss; 0.55–1.3 for CL).

Table 6.

Predicted human pharmacokinetic parameters based on allometric scaling

	Vss (L·kg−1)	CL (mL·min−1·kg−1)	t 1/2 (h)
Simple scaling	3.37	1.94	19.97
MLP‐based scaling	NA	1.07	36.32
Brain weight‐based scaling	NA	0.63	61.53

Volume and clearance estimates are back‐calculated, predicted values for a 70 kg human, using the slope and intercept derived from allometric scaling of the preclinical pharmacokinetic parameters.

Figure 6.

Log–log plots of (A) mean steady‐state V_ss versus body weight (BW), (B) mean CL versus BW, (C) product of mean clearance and MLP versus BW, and (D) and product of mean clearance and brain weight versus body weight for ZY15557 in mouse, rat, dog and monkey, along with the predicted values for humans. Regression lines along with the best‐fit equations are shown.

Discussion

Poor patient compliance is a major problem in the therapy of Type 2 diabetes, in response to adverse effects like hypoglycemia, weight gain, and gastrointestinal distress. DPP‐4 inhibitors enhance the biological activity of incretin hormones, and represent a successful approach for the treatment of Type 2 diabetes. They are a preferred therapeutic option for the treatment of Type 2 diabetes because they are weight‐neutral and do not cause hypoglycemia or gastrointestinal intolerance. Currently available DPP‐4 inhibitors are once‐daily treatments. Diabetic patients do not readily adhere to recommended dosing regimen, which is a concern for once‐daily treatments (Garcia‐Perez et al., 2013). Poor compliance is a major reason for uncontrolled diabetes and higher morbidity and mortality in these patients. We, therefore, focused our attention on the discovery and development of a DPP‐4 inhibitor with a long half‐life, so that skipping a dose would not be detrimental to the overall effect of the medicine. ZY15557 is a potent inhibitor of DPP‐4, a member of the prolyl peptidase family, which also includes DPP‐8 and DPP‐9. Inhibition of these enzymes is related to toxicities on many organs and to profound immunotoxicity in rats and dogs, indicating that poor selectivity between DPP isoforms, such as DPP‐8 and DPP‐9, might cause the toxicity noted for this class of compounds (Lankas et al., 2005). Hence, earlier efforts in development of DPP‐4 inhibitors were directed towards designing reversible inhibitors that would have a relatively low biological half‐life or dissociate rapidly from the enzyme (Burkey et al., 2008). So, the first generation of DPP‐4 inhibitors were designed as reversible inhibitors that rapidly dissociated from the enzyme and/or had short pharmacokinetic half‐lives. Alternatively, it was possible that toxicity observed in rodents might not be reliably predictive of toxicities related to inhibition of DPP‐8 or 9 in humans (Burkey et al., 2008). Proper scaling of data from pharmacokinetic and pharmacodynamic models in preclinical studies can make relatively reliable predictions from the in vitro to the in vivo data and from animals to humans (Zhang et al., 2012). Nevertheless, ZY15557 is highly selective as an inhibitor of DPP‐4, as our experiments showed 6000‐ to3000‐fold less effectiveness against DPP‐8 and DPP‐9. Trelagliptin (SYR‐472) is the long‐acting DPP‐4 inhibitor in clinical use, and we have observed that DPP‐4 inhibition by ZY15557 (K_i values of 5.1, 6.1 and 13.7 nM for human, dog and rat DPP‐4 respectively) is comparable with that by trelagliptin, with IC₅₀ values of 4.2, 6.2 and 9.7 nM for human, dog and rat respectively (Grimshaw et al., 2016). ZY15557 is not only a potent inhibitor of DPP‐4 activity but it also shows prolonged binding to the DPP‐4 enzyme. The observed rate constants for association and dissociation with the DPP‐4 enzyme, that is, K_on and K_off values for ZY15557 (38.6 × 10⁴ and 3.2 × 10⁻⁴ s⁻¹, respectively) are comparable with those of trelagliptin (26.0 × 10⁴ M⁻¹ s⁻¹ and 4 × 10⁻⁴ s⁻¹ respectively) and the half‐life by dissociation assay for ZY15557 (35.5 min) was found to be similar to that of trelagliptin (29 min) (Grimshaw et al., 2016). ZY15557 showed a slower off‐rate similar to that of omarigliptin from the DPP‐4 enzyme‐inhibitor complex, compared with that of sitagliptin. In our work, ZY15557 was incubated with the DPP‐4 enzyme, followed by extensive dialysis of the enzyme‐inhibitor complex. Under these conditions, significant amounts of ZY15557 remained bound to the enzyme even after 24 h, which indicated a prolonged inhibition of DPP‐4 activity (Table 3).

ZY15557 differs from the known DPP‐4 inhibitors, such as sitagliptin, as it has a longer half‐life and duration of action when tested in animal models. Impaired insulin secretion in patients with Type 2 diabetes is often characterized by a decreased first‐phase insulin response, which leads to glucose intolerance and postprandial hyperglycaemia (Lee et al., 2008). ZY15557 treatment decreased the glucose AUC during OGTT in C57 mice and Zucker fatty rats. ZY15557 also dose‐dependently increased plasma active GLP‐1 and insulin levels following OGTT in C57 mice and Wistar rats. We have not measured glucagon in the animals after treatment with ZY15557, but is expected that a DPP‐4 inhibitor, by virtue of its GLP‐1 enhancing actions will also reduce glucagon levels. It will, however, be interesting to understand the changes in glucagon after chronic treatment with ZY15557 in diabetic animals.

There are many long acting DPP‐4 inhibitors in clinical investigation or use, which include omarigliptin and trelagliptin (Sheu et al., 2015; Grimshaw et al., 2016). ZY15557 is chemically novel and different from these compounds. We have observed a beneficial effect in terms of leptin sensitivity and adiponectin secretion after chronic treatment with ZY15557 in mice. ZY15557 reduced hepatic steatosis and improved leptin sensitivity in chronically treated db/db mice (Table 4). In addition, ZY15557 reduced oxidative stress markers in liver of chronically treated db/db mice and improved plasma adiponectin levels. Apart from these efficacy benefits, currently ongoing toxicological and clinical examination of ZY15557 will help to differentiate its profile from the other long term DPP‐4 inhibitors.

ZY15557 showed a long pharmacokinetic half‐life and extended duration of plasma DPP‐4 inhibition in all the animal models. Its terminal half‐life was 44.41, 28.13, 52.89 and 61.16 h in mouse, rat, dog and monkey respectively (Table 5). Oral ZY15557 showed about 50% plasma DPP‐4 inhibition at 7 days post‐dose in dog and monkey. In vitro, ZY15557 showed a threefold slower off‐rate from DPP‐4 enzyme‐inhibitor complex than sitagliptin. Further, following extensive dialysis of the enzyme‐inhibitor complex to remove free drug, it also showed much greater residual inhibitory effect, compared with sitagliptin. The extended duration of inhibition of DPP‐4 shown by ZY15557, in animal models appears to result from a combination of a long plasma half‐life and a slow dissociation of ZY15557 bound to DPP‐4.

Absorption, distribution, metabolism and excretion (ADME) studies are important for drug discovery and development. In fact, the ADME parameters obtained from in vitro and in vivo models aid in the prediction of drug behaviour in patients (Zhang et al., 2012). ZY15557 showed excellent bioavailability and a long plasma half‐life in all the preclinical species tested. Its high bioavailability is consistent with its excellent aqueous solubility (~1 mg·mL⁻¹, across pH range 1 to 6.8), good permeability in Caco‐2 cells (113.3 ± 11.2 nm·s⁻¹; with a low efflux ratio of 1.1), and high metabolic stability in liver microsomes across species (data not shown). The long half‐life of ZY15557 results from a combination of its large volume of distribution and slow plasma clearance. Thus, the volume of distribution was 7.0×, 7.4×, 4.2× and 7.1× total body water in mouse, rat, dog and monkey respectively; the clearance was 13, 37, 10 and 7% of hepatic blood flow, respectively, in the same species. Both volume of distribution and clearance of ZY15557 showed a high allometric correlation across species with body weight, strongly suggesting that these pharmacokinetic features of the compound are likely to be predictable in human (Figure 6). Using different allometric approaches, the predicted human half‐life ranges from 19.97 to 61.53 h. As the exponent for clearance from simple allometry (0.719) exceeds 0.7, the human plasma half‐life estimates of 36.32 (MLP‐corrected approach) and 62 h (brain weight‐corrected approach) are considered the best predictions. Taken together, the pharmacokinetic studies of ZY15557 in animals suggest good prospects of development, with a long half‐life and a sustained effect.

In summary, we have identified ZY15557,as a novel, potent, selective DPP‐4 inhibitor with an extended half‐life and sustained duration of action.

Author contributions

M. R. J. and A.A.J. contributed to the conception of the manuscript, design of experiments, and analysis and interpretation of the data, and writing of the manuscript. A.A.J., S.G.K. and V.J.P. performed the experiments, analysed the data, and wrote the manuscript. R.H.J., H.P. and P.J. performed the experiments and analysed the data. R.C.D. designed the experiments and analysed the data. P.R.P. supervised the work and contributed to the writing of manuscript. All authors have commented on the initial and final drafts of the manuscript and are responsible for approval of the final version of the manuscript in all aspects.

Conflict of interest

The authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organisations engaged with supporting research.

Supporting information

Data S1 Supplementary material for Figure 5. (A) Inhibition of DPP‐4 enzyme activity and plasma concentration of ZY15557 in dogs after single oral administration of ZY15557. (B) Inhibition of DPP‐4 enzyme activity and plasma concentration of ZY15557 in non‐human primates (NHP) after single oral administration of ZY15557.

Jain, M. R. , Joharapurkar, A. A. , Kshirsagar, S. G. , Patel, V. J. , Bahekar, R. H. , Patel, H. V. , Jadav, P. A. , Patel, P. R. , and Desai, R. C. (2017) ZY15557, a novel, long acting inhibitor of dipeptidyl peptidase‐4, for the treatment of Type 2 diabetes mellitus. British Journal of Pharmacology, 174: 2346–2357. doi: 10.1111/bph.13842.