Machine-Learning-Guided Peptide Drug Discovery: Development of GLP-1 Receptor Agonists with Improved Drug Properties.

Peptide-based drug discovery has surged with the development of peptide hormone-derived analogs for the treatment of diabetes and obesity. Machine learning (ML)-enabled quantitative structure-activity relationship (QSAR) approaches have shown great promise in small molecule drug discovery but have been less successful in peptide drug discovery due to limited data availability. We have developed a peptide drug discovery platform called streaMLine, enabling rigorous design, synthesis, screening, and ML-driven analysis of large peptide libraries. Using streaMLine, this study systematically explored secretin as a peptide backbone to generate potent, selective, and long-acting GLP-1R agonists with improved physicochemical properties. We synthesized and screened a total of 2688 peptides and applied ML-guided QSAR to identify multiple options for designing stable and potent GLP-1R agonists. One candidate, GUB021794, was profiled in vivo (S.C., 10 nmol/kg QD) and showed potent body weight loss in diet-induced obese mice and a half-life compatible with once-weekly dosing.

Molecular Engineering of Insulin Icodec, the First Acylated Insulin Analog for Once-Weekly Administration in Humans.

Here, we describe the molecular engineering of insulin icodec to achieve a plasma half-life of 196 h in humans, suitable for once-weekly subcutaneously administration. Insulin icodec is based on re-engineering of the ultra-long oral basal insulin OI338 with a plasma half-life of 70 h in humans. This systematic re-engineering was accomplished by (1) further increasing the albumin binding by changing the fatty diacid from a 1,18-octadecanedioic acid (C18) to a 1,20-icosanedioic acid (C20) and (2) further reducing the insulin receptor affinity by the B16Tyr → His substitution. Insulin icodec was selected by screening for long intravenous plasma half-life in dogs while ensuring glucose-lowering potency following subcutaneous administration in rats. The ensuing structure-activity relationship resulted in insulin icodec. In phase-2 clinical trial, once-weekly insulin icodec provided safe and efficacious glycemic control comparable to once-daily insulin glargine in type 2 diabetes patients. The structure-activity relationship study leading to insulin icodec is presented here.

Food intake rather than blood glucose levels affects the pharmacokinetic profile of insulin aspart in pigs.

In humans, food intake and glucose infusion have been reported to increase subcutaneous blood flow. Since local blood flow influences the rate of insulin absorption from the subcutaneous tissue, we hypothesised that an increase in blood glucose levels-occurring as the result of glucose infusion or food intake-could modulate the pharmacokinetic properties of subcutaneously administered insulin. The pharmacokinetic profile of insulin aspart was assessed in 29 domestic pigs that were examined in a fed and fasted state or included in hyperinsulinaemic clamp studies of 4 vs. 10 mmol/L glucose prior to subcutaneous (30 nmol) or intravenous (0.1 nmol/kg) insulin administration. Results showed that food intake compared to fasting accelerated absorption and decreased clearance of insulin aspart (P < 0.05). Furthermore, higher c-peptide but also glucagon levels were observed in fed compared to fasted pigs (P < 0.05). The pharmacokinetic profile of insulin aspart did not differ between pigs clamped at 4 vs. 10 mmol/L glucose. Hence, food intake rather than blood glucose levels within normal range modulates the pharmacokinetic properties of insulin aspart upon subcutaneous and intravenous administration in pigs.

Engineering of Orally Available, Ultralong-Acting Insulin Analogues: Discovery of OI338 and OI320.

Recently, the first basal oral insulin (OI338) was shown to provide similar treatment outcomes to insulin glargine in a phase 2a clinical trial. Here, we report the engineering of a novel class of basal oral insulin analogues of which OI338, 10, in this publication, was successfully tested in the phase 2a clinical trial. We found that the introduction of two insulin substitutions, A14E and B25H, was needed to provide increased stability toward proteolysis. Ultralong pharmacokinetic profiles were obtained by attaching an albumin-binding side chain derived from octadecanedioic (C18) or icosanedioic acid (C20) to the lysine in position B29. Crucial for obtaining the ultralong PK profile was also a significant reduction of insulin receptor affinity. Oral bioavailability in dogs indicated that C18-based analogues were superior to C20-based analogues. These studies led to the identification of the two clinical candidates OI338 and OI320 (10 and 24, respectively).

Molecular engineering of safe and efficacious oral basal insulin.

Recently, the clinical proof of concept for the first ultra-long oral insulin was reported, showing efficacy and safety similar to subcutaneously administered insulin glargine. Here, we report the molecular engineering as well as biological and pharmacological properties of these insulin analogues. Molecules were designed to have ultra-long pharmacokinetic profile to minimize variability in plasma exposure. Elimination plasma half-life of ~20 h in dogs and ~70 h in man is achieved by a strong albumin binding, and by lowering the insulin receptor affinity 500-fold to slow down receptor mediated clearance. These insulin analogues still stimulate efficient glucose disposal in rats, pigs and dogs during constant intravenous infusion and euglycemic clamp conditions. The albumin binding facilitates initial high plasma exposure with a concomitant delay in distribution to peripheral tissues. This slow appearance in the periphery mediates an early transient hepato-centric insulin action and blunts hypoglycaemia in dogs in response to overdosing.

The counterregulatory response to hypoglycaemia in the pig.

The domestic pig is commonly used as animal model in the pharmaceutical development of new therapeutics for treatment of diabetes. Since a formal definition of hypoglycaemia only exists in humans, the purpose of this study was to assess the counterregulatory response in the domestic pig at glucose levels known to induce symptoms of hypoglycaemia in humans. Six pigs were included in hyperinsulinaemic glucose clamps with plasma glucose targets of 2, 3 and 5 mmol/L in a cross-over design, and the associated glucose counterregulatory response was assessed by measuring glucose kinetics and levels of glucagon, c-peptide, catecholamines, cortisol and growth hormone. Results showed that the 2 and 3 vs 5 mmol/L clamps significantly decreased and increased the secretion of c-peptide and glucagon, respectively (P < .05). This finding was associated with increased rate of glucose appearance (Ra ) and decreased rate of glucose disappearance (Rd ) (P < .001). No marked differences in the catecholamine, growth hormone or cortisol response were observed. Consequently, like humans, pigs respond to hypoglycaemia by decreasing the pancreatic output of insulin while increasing that of glucagon, with increased glucose mobilization and decreased glucose disposal as a result. The hypoglycaemic clamps did not result in a marked secretion of the other counterregulatory hormones.

Intraintestinal and Parenteral Administration of an Insulin Analogue Leads to Comparable Activation of Signaling Downstream of the Insulin Receptor in the Small Intestine.

BACKGROUND: Oral delivery of insulin was recently demonstrated to have therapeutic relevance in patients with diabetes. Insulin receptors are expressed in the gastrointestinal tract and can be activated by insulin in the bloodstream, but it is not known if the large amount of insulin in the intestinal lumen required for sufficient oral delivery will induce a different effect. The aim of this study was to compare the acute effect in the intestine of insulin administered in the intestinal lumen with that of insulin administered by a parenteral route. METHOD: Intraintestinal (ii) injection in the mid-jejunum of anaesthetized rats with insulin analogue 106 (I106), formulated with the absorption-enhancer sodium caprate, was used as an animal model of oral insulin administration. As control treatment, rats were treated with I106 by iv infusion according to algorithms which precisely mimicked the pharmacokinetic and pharmacodynamic properties of ii administered I106. Several fold more I106 was administered by ii injection than by iv infusion. Phosphorylated Akt (Ser473) was used as indicator of insulin-stimulated acute effects in the intestine. RESULTS: Treatment with I106 resulted in activation of Akt in the intestine, with no significant difference between the effects of ii or iv administration. CONCLUSION: The results from this rat model of orally administered insulin indicate that the unabsorbed insulin in the intestinal lumen after oral administration will not result in an enhanced acute effect in the intestine.

Subcutaneous Administration of Insulin is Associated With Regional Differences in Injection Depot Variability and Kinetics in The Rat.

BACKGROUND: In humans, subcutaneous administration of insulin in the abdominal region or arm is associated with a faster absorption compared to the thigh or buttocks. We hypothesised that this is partly caused by differences in injection depot structure and kinetics and that the variability in insulin exposure differs between injection sites. MATERIAL AND METHODS: Regional effects on insulin pharmacokinetics were evaluated in a series of studies in Sprague Dawley rats dosed subcutaneously with insulin aspart in the neck or flank. Injection depots were visualised using µCT after subcutaneous dosing with insulin aspart mixed with the contrast agent iomeprol, and insulin exposure was determined between the scans by Luminescent Oxygen Channeling Immunoassay. RESULTS: Insulin absorption was significantly delayed by subcutaneous dosing in the flank compared to the neck region (p<0.01 or less). This delay was associated with smaller depots, as measured by reduced depot volume and surface area (p<0.001). Furthermore, the delayed absorption correlated with a slower depot disappearance (p<0.001). Regional differences in depot variability were not reflected by similar differences in pharmacokinetic variability. CONCLUSION: Structure and kinetics of subcutaneous injection depots-as detected by µCT scans-predict insulin exposure and may thus contribute to the regional differences in insulin pharmacokinetics. The present methodology is applicable for visualisation of insulin injection depots in vivo. Our results did however not support a link between the variability in depot size and insulin pharmacokinetics.

Dual treatment with a fixed ratio of glucagon and insulin increases the therapeutic window of insulin in diabetic rats.

The current available insulin therapies decrease blood glucose but are associated with the risk of developing hypoglycemia. Glucagon is a counter regulatory hormone and we hypothesize that a fixed ratio of insulin and a long-acting glucagon-analogue can reduce the risk of hypoglycemia. To define an appropriate ratio we tested two fixed glucagon doses (3.5 and 10 nmol/kg) in combination with increasing doses of insulin in diabetic rats. We observed a plateau in blood glucose at 15.2 mmol/L with 10 nmol/kg of the glucagon-analogue. The mechanism behind this plateau, protecting against hypoglycemia, was investigated by measuring the glucose output, cAMP production, and hormone binding in primary rat hepatocytes. While glucose output could contribute to the observed plateau in blood glucose, cAMP response or hormone binding did not explain the observation. Though such plateau indicated decreased risk of hypoglycemia a full normalization of blood glucose was still needed. Based on the data obtained with 3.5 nmol/kg of the glucagon-analogue, a 1:23 (glucagon-analogue:insulin) ratio was chosen and a dose-response was performed in diabetic rats. At low doses (≤20 nmol/kg), insulin and the 1:23 ratio showed similar efficacy of lowering blood glucose. Interestingly, the insulin-dose resulting in hypoglycemia was increased from 40 nmol/kg insulin alone to 160 nmol/kg insulin in the 1:23 ratio. Analysis of the liver glycogen content at the end of the experiment showed that the highest dose in the 1:23 ratio almost emptied the liver from glycogen. Thus, liver glycogen is essential for the protective effect of glucagon in hypoglycemia.

Sustained effect of glucagon on body weight and blood glucose: Assessed by continuous glucose monitoring in diabetic rats.

Insulin is a vital part of diabetes treatment, whereas glucagon is primarily used to treat insulin-induced hypoglycemia. However, glucagon is suggested to have a central role in the regulation of body weight, which would be beneficial for diabetic patients. Since the glucagon effect on blood glucose is known to be transient, it is relevant to investigate the pharmacodynamics of glucagon after repeated dosing. In the present study, we used telemetry to continuously measure blood glucose in streptozotocin induced diabetic Sprague-Dawley rats. This allowed for a more detailed analysis of glucose regulation compared to intermittent blood sampling. In particular, we evaluated the blood glucose-lowering effect of different insulin doses alone, and in combination with a long acting glucagon analog (LAG). We showed how the effect of the LAG accumulated and persisted over time. Furthermore, we found that addition of the LAG decreased body weight without affecting food intake. In a subsequent study, we focused on the glucagon effect on body weight and food intake during equal glycemic control. In order to obtain comparable maximum blood glucose lowering effect to insulin alone, the insulin dose had to be increased four times in combination with 1 nmol/kg of the LAG. In this set-up the LAG prevented further increase in body weight despite the four times higher insulin-dose. However, the body composition was changed. The insulin group increased both lean and fat mass, whereas the group receiving four times insulin in combination with the LAG only significantly increased the fat mass. No differences were observed in food intake, suggesting a direct effect on energy expenditure by glucagon. Surprisingly, we observed decreased levels of FGF21 in plasma compared to insulin treatment alone. With the combination of insulin and the LAG the blood glucose-lowering effect of insulin was prolonged, which could potentially be beneficial in diabetes treatment.

Rodent model choice has major impact on variability of standard preclinical readouts associated with diabetes and obesity research.

Laboratory rodents are available as either genetically defined inbred strains or genetically undefined outbred stocks. As outbred rodents are generally thought to display a higher level of phenotypic variation compared to inbred strains, it has been argued that experimental studies should preferentially be performed by using inbred rodents. However, very few studies with adequate sample sizes have in fact compared phenotypic variation between inbred strains and outbred stocks of rodents and moreover, these studies have not reached consistent conclusions. The aim of the present study was to compare the phenotypic variation in commonly used experimental readouts within obesity and diabetes research, for four of the most frequently used mouse strains: inbred C57BL/6 and BALB/c and outbred NMRI and CD-1 mice. The variation for all readouts was examined by calculating the coefficient of variation (CV), i.e., the relative variation, including a 95% confidence interval for the CV. We observed that for the majority of the selected readouts, inbred and outbred mice showed comparable phenotypic variation. The observed variation appeared highly influenced by strain choice and type of readout, which suggests that these collectively would serve as more predictive of the phenotypic variation than the more general classification of mice as inbred or outbred based on genetic heterogeneity.

Construction of Insulin 18-mer Nanoassemblies Driven by Coordination to Iron(II) and Zinc(II) Ions at Distinct Sites.

Controlled self-assembly (SA) of proteins offers the possibility to tune their properties or to create new materials. Herein, we present the synthesis of a modified human insulin (HI) with two distinct metal-ion binding sites, one native, the other abiotic, enabling hierarchical SA through coordination with two different metal ions. Selective attachment of an abiotic 2,2'-bipyridine (bipy) ligand to HI, yielding HI-bipy, enabled Zn(II)-binding hexamers to SA into trimers of hexamers, [[HI-bipy]6]3, driven by octahedral coordination to a Fe(II)  ion. The structures were studied in solution by small-angle X-ray scattering and on surfaces with AFM. The abiotic metal ligand had a higher affinity for Fe(II) than Zn(II)  ions, enabling control of the hexamer formation with Zn(II) and the formation of trimers of hexamers with Fe(II)  ions. This precise control of protein SA to give oligomers of oligomers provides nanoscale structures with potential applications in nanomedicine.

Oral insulin (human, murine, or porcine) does not prevent diabetes in the non-obese diabetic mouse.

Studies have shown oral insulin prevents type 1 diabetes (T1D) in mouse models, however human trials were inconclusive. We tested the ability of different insulins to prevent T1D in non-obese diabetic mice. Mice received oral insulin or PBS twice weekly and disease was monitored. Contrary to previous studies, no insulin tested showed significant ability to prevent T1D, nor did testing of linked suppression in a delayed type hypersensitivity model have reproducible effect. To investigate delivery of antigen within the GI tract, blue dye was fed to mice. Dye traveled 5-8 cm from stomach to small intestine within 10s, suggesting orally administered antigen may not get digested in the stomach in mice. Insulin incubated with jejunum extracts was instantly digested. Thus, in humans large doses of insulin may be required to achieve tolerance as antigen may be more vulnerable to digestion in the stomach even before reaching the small intestine.

Long-chain PUFA in granulocytes, mononuclear cells, and RBC in patients with cystic fibrosis: relation to liver disease.

BACKGROUND AND AIM: Patients with cystic fibrosis (CF) have low levels of n-3 long-chain polyunsaturated fatty acids (n-3 LCPUFA) in plasma or red blood cells (RBC), as also seen in other chronic and acute liver diseases. The differences may be more pronounced in CF transmembrane conductance regulator protein (CFTR)-regulated tissues such as granulocytes, monocytes, and lymphocytes. The aim of the present study was to investigate whether patients with CF-related liver disease have lower n-3 LCPUFA level than patients with CF without liver disease. METHODS: Twenty patients with known CF-related liver disease were matched with 20 CF patients without. Blood samples were analysed for liver biochemistry and haematology. Granulocytes, mononuclear cells, and RBC were separated by density gradient centrifugation, and fatty acid composition was measured by gas chromatography. Hepatic ultrasound was scored according to Williams et al. Hepatic transit time (HTT) was measured with the ultrasound contrast agent SonoVue. RESULTS: No significant differences were seen in either n-6 or n-3 LCPUFAs in any cell line when the 2 groups were compared. In a multiple regression analysis including HTT, age, Pseudomonas aeruginosa infection, diabetes mellitus, treatment with ursodeoxycholic acid, forced expiratory volume in 1 second (% of predicted value), and Williams' ultrasound scoring scale, only n-3 LCPUFA docosahexaenoic acid in mononuclear cell membranes was positively associated with HTT (P = 0.02). The arachidonic acid/docosahexaenoic acid ratio within the mononuclear cells was negatively associated with both HTT (P = 0.003) and Williams' ultrasound scoring scale (P = 0.03). For RBC-LCPUFAs, no significant associations were seen. CONCLUSIONS: These findings indicate that in patients with CF, the degree of liver disease was negatively associated with LCPUFA n-3 levels in CFTR-expressing white blood cells but unrelated to those levels in CFTR-negative RBC.

The protease inhibitors ritonavir and saquinavir influence lipid metabolism: a pig model for the rapid evaluation of new drugs.

BACKGROUND: Studies of the effects of antiretroviral drugs on lipid metabolism are limited by the availability of suitable models. We have thus developed an animal model utilising Göttingen mini-pigs. The normal lipid metabolism of mini-pigs closely reflects that of humans and they are expected to have similar reactions to antiretroviral drugs. METHODS: The pigs were treated orally with high doses of the protease inhibitors ritonavir and saquinavir for 4 weeks. The model allows repeated concomitant biopsies from liver, muscle, adipose tissue and plasma samples. RESULTS: The study showed a general decrease in polyunsaturated fatty acids; changes in both saturated and monounsaturated fatty acids were also apparent after antiretroviral treatment. The changes were observed after 4 weeks of treatment. At 4 weeks post-treatment, the levels of all fatty acids were lower compared with pretreatment levels, suggesting a prolonged effect of the antiretroviral drug treatment lasting beyond the 4 week post-treatment observation period. CONCLUSIONS: The Göttingen mini-pig model is a promising animal model for rapid screening of the metabolic effects induced by antiretroviral drugs.

Postprandial lipid responses of butter blend containing fish oil in a single-meal study in humans.

The postprandial effects of a butter product containing fish oil were investigated in a single-meal, randomized crossover study with a commercial butter product as the control. Twelve healthy males consumed two test meals with (13)C-labelled cholesterol (45 mg) and either an interesterified butter blend with fish oil (352 mg n-3 long-chain PUFA (LCPUFA)) or the commercial butter blend. Blood samples were collected after the meals and in the fasting condition on the test day and the following morning, and were analysed for cholesterol absorption, plasma lipid profile and fatty acid composition. No significant difference in the postprandial plasma fatty acid composition was observed between the groups, neither difference in cholesterol absorption, plasma cholesterol or the cholesterol contents of plasma lipoproteins. The incorporation of fish oil in the butter resulted in a significant lower concentration of triacylglycerols in the plasma 2 h after the meal in comparison with the commercial butter blend (p = 0.02); there was, however, no significant difference 24 h after the meal. In conclusion, fish oil-enriched butter blend provides a source to increase the intake of n-3 LCPUFA in the population, but has no acute effect on cholesterol absorption and plasma cholesterol concentration in human.

Butter blend containing fish oil improves the level of n-3 fatty acids in biological tissues of hamster.

Many studies have shown beneficial effects of long chain n-3 polyunsaturated fatty acids (PUFA) on human health. Regardless of the positive effects of n-3 PUFA, the intake of these fatty acids remains low. An approach to increase the intake of n-3 PUFA in the population is to incorporate fish oil into food. In the present study, fish oil was incorporated into butter blends by enzymatic interesterification. The aim of the study was to investigate the effects of this butter product in comparison with a commercial butter blend and a product produced by interesterification but without fish oil. Golden Syrian hamsters received hamster feed blended with one of the three butter products. After 6 weeks of feeding, the fatty acid compositions of plasma, erythrocytes, liver, brain, and visceral fat were determined. The intake of butter product with fish oil resulted in a higher level of n-3 PUFA in plasma, erythrocytes, and liver. The incorporation of n-3 PUFA was significantly higher in phospholipids than in triacylglycerols. The results suggest that enriching butter blends with small amounts of fish oil can be used as an alternative method for improving the level of n-3 PUFA in biological tissues.

The form of dietary conjugated linoleic acid does not influence plasma and liver triacylglycerol concentrations in Syrian golden hamsters.

Several studies have shown that conjugated linoleic acid (CLA) supplementation can improve the plasma lipid profile and thereby probably decrease the risk for development of atherosclerosis. The aim of the present study was to compare the effects on plasma and organ lipids of different dietary forms of CLA: triacylglycerol (TAG), diacylglycerol (DAG), monoacylglycerol (MAG), and fatty acid ethyl esters (FAEEs). DAG-, MAG-, and FAEE-CLA were produced by enzymatic interesterifications and all supplements were composed of a 1:1 mixture of the 2 major CLA isomers: cis-9, trans-11 and trans-10, cis-12. Male Syrian Golden hamsters were fed mildly atherogenic diets (10 g butter/100 g, 0.1 g cholesterol/100 g) supplemented with 0.5 g CLA/100 g or without CLA (control) for 8 wk. Liver weights were greater in the TAG- and FAEE-CLA groups than in the control group. In general, the form of CLA did not differentially affect plasma or liver cholesterol or plasma lipoprotein cholesterol concentrations, but only the TAG-CLA group had a higher final plasma TAG concentration than the control group. Both CLA isomers were incorporated into plasma, livers, and spleens. The results of the present study suggest that the form in which CLA is supplemented in the diet does not affect hamster plasma and liver TAG concentrations. The TAG-CLA form, a frequently used form of supplemental CLA, increases plasma TAG concentrations. If similar effects occur in humans, supplemental TAG-CLA cannot be considered to be beneficial given the relation between plasma TAG and the development of atherosclerosis.

The metabolism of structured triacylglycerols.

The triacylglycerol (TAG) structure in addition to the overall fatty acid profile is of importance when considering the nutritional effect of a dietary fat. This review aims at summarizing our current knowledge of the digestion, absorption, uptake, and transport of structured TAGs, with particular emphasis on the following aspects: gastric emptying, specificity of pancreatic lipase, lymphatic transport and clearance of chylomicrons, effects of lipid structure on tissue lipid compositions and the fecal loss of fats. So an overview will be provided for how the structure and fatty acid composition of TAGs affect their absorption and the distribution of the fatty acids in the body following digestion and absorption.