Eur J Med Chem. 2014 Sep 12;84:584-94. doi: 10.1016/j.ejmech.2014.07.063.

Synthesis, screening and docking of small heterocycles as Glycogen Phosphorylase inhibitors

 

Schweiker SS,a,b Loughlin WA,a,c Lohning AS,b Petersson MJ,a Jenkins IDa

aEskitis Institute for Drug Discovery, Nathan Campus, Griffith University, Brisbane, QLD, 4111, AUSTRALIA

b Faculty of Health Sciences and Medicine, Bond University, Gold Coast, QLD, 4229, AUSTRALIA

cSchool of Natural Sciences, Nathan Campus, Griffith University, Brisbane, QLD, 4111, AUSTRALIA

 

Abstract

A series of morpholine substituted amino acids (phenylalanine, leucine, lysine and glutamic acid) was synthesized. A fragment-based screening approach was then used to evaluate a series of small heterocycles, including morpholine, oxazoline, dihydro-1,3-oxazine, tetrahydro-1,3-oxazepine, thiazoline, tetrahydro-1,3-pyrimidine, tetrahydro-1,3-diazepine and hexahydro-1H-benzimidazole, as potential inhibitors of Glycogen Phosphorylase a. Thiazoline 7 displayed an improved potency (IC50 of 25 mM) as compared to heterocycles 1, 5, 9-13 and 19 (IC50 values of 1.1mM to 23.9 mM). A docking study using the crystal structure of human liver Glycogen Phosphorylase, provided insight into the interactions of heterocycles 5, 7, 913 and 19 with Glycogen Phosphorylase.

KEYWORDS: Fragment screening; Glycogen Phosphorylase a; Heterocycle; Molecular docking; Tetrahydropyrimidine; Thiazoline

PMID: 25062009

 

Supplement:

The incidence of diabetes mellitus continues to rise at accelerated rates, with 347 million cases reported in 2008 (1). Current medication for Type 2 diabetes commonly has adverse side effects, which has identified a need for new and better treatments. Control of glycogenolysis (the breakdown of glycogen to glucose) is a promising approach to control of blood glucose levels. A key enzyme in the glycogenolysis pathway is Glycogen Phosphorylase (GP) that catalyzes conversion of glycogen to glucose-1-phosphate (Glc-1-P) that eventually leads to the production of glucose. Studies have confirmed the efficacy of inhibitors of GP on hepatic glycogen balance and blood glucose control, thus making GP a key target (2) for the design of compounds that lower blood glucose levels. The catalytic site of GP has been the primary focus of inhibitor design efforts, which include glucose-based compounds. However, compounds with considerable structural diversity can be considered as GP is an allosteric protein that is modulated by six-ligand binding sites (2). We explored a fragment screening approach of small heterocycles as part of discovery research to identify new potential inhibitors of Glycogen Phosphorylase, targeted towards non-catalytic site ligand binding sites of GP.

We pursued a targeted fragment screening approach, as we did not have access to a large databank/library of compounds, and thus a high-throughput screening approach was not viable. In high-throughput screening, the identification of nanomolar activity is sought through screening tens of thousands of compounds, with typical molecular weights of ~500 Da. Alternatively, in fragment screening, the identification of millimolar activity is sought through the screening of smaller chemical fragments (MW ~ 200Da), which can identify compounds with weaker binding to the biological target. From this approach, a lead compound could be progressed (in a drug discovery pipeline), and developed with structural modification into compound(s) with more potent activity against GP.

We had previously identified (3) that key peptide residues (EKL), within the peptide C-​terminal sequence of human GL that binds to Glycogen Phosphorylase a (GPa)​, PEWPSYLGYEKLGPYY-​NH2, have additional molecular interactions with GPa. We went on to combine selected chemical structural diversity with the philosophy of a fragment screening approach, which allowed us to gain good synthetic access to a small library of twenty-one compounds and obtain an excellent 86% hit rate for GPa inhibition. Our fragment library was comprised of small heterocycles, including morpholines (designed mimetics of the EKL tri-peptide) and small heterocycles typically not reported against GP.

We asked the question whether a shortlist of potential compounds worth further consideration could be formed from the initial hits? Potency of inhibition was used as a primary screen; where nine compounds displayed low activity, nine compounds had an estimated IC50 more potent than 24 mM and one compound with a MW of 200 and estimated IC50 of 2.5 mM was identified. Further assessment of the mM and mM compounds (nine compounds) was guided by Lipinski parameters, calculated ligand efficiency (LE) and ligand-efficiency-dependent lipophilicity (LELP); which all were pivotal in the selection of the first tier and second tier hits (Figure 1).

 

Fig 1 Loughlin

Figure 1. Classification of compounds hits by: GPa inhibition estimated IC50 value, Ligand efficiency (LE) values (~0.30 or higher) and Ligand-Efficiency-Dependent Lipophilicity (LELP) value (in the range -10 to 10).

 

We explored the in silico interaction of top nine compounds with the three dimensional structure of GPa obtained from the previously determined X-ray ligand structure determination (PDB ID: 1L5Q for purine and indole sites; PDB ID: 3DDS for AMP site). We did a virtual comparison of how each of the nine compounds interacted with GP at the three key binding sites; purine, indole and AMP. It was interesting to note that CScore (a measure of ligand-enzyme interaction) values for our compounds were generally as large as the control inhibitor at the purine site. However the site specificity could not be confirmed due to the small differences between the CScore values obtained for our compounds across the three sites. Nonetheless is was interesting to note the differences between the Tier One-Two compounds and Tier Three compounds when docked in the purine binding site.

 

 Fig 2 Loughlin

Figure 2. Compounds 7, 5, 9 and 11 docked in purine binding site of GPa (PDB ID: 1L5Q). MOLCAD-generated protein lipophilic surface showing lipophilic residues in brown and hydrophilic residues in blue.

 

Figure 2 shows Tier One and Two compounds 7, 5, 9 and 11 docked in the purine binding site of GPa. It was observed that a p-stacking interaction of ligand rings was sandwiched between the aromatic rings of GP peptide residues Phe285 and Tyr613. All four compounds shared a similar binding conformation centred between the aromatic rings of peptide residues Phe285 and Tyr613. Whereas in Figure 3, the Tier Three compounds 10, 12, 13 and 19 with poorer GP inhibition, utilised a broader three-dimensional space within the binding pocket than the stronger inhibition group (7, 5, 9 and 11) with only compounds 12 and 13 p-stacking with GP peptide residues Phe285 and Tyr613.

 

 Fig 3 Loughlin

Figure 3. Binding mode of low inhibition compound group (10, 12, 13 and 19) docked in purine binding site of GPa (PDB ID: 1L5Q).

 

The importance of this study is two-fold.  First, it shows how important and effective a multidisciplinary approach is for the identification of lead compounds for a drug discovery pipeline. The multidisciplinary approach we used involved the design and synthesis small molecules, application of a fragment screening approach to identify potential inhibitors of GP, used compound property predictions and gained structural insights by docking of compounds with the three dimensional structure of Glycogen Phosphorylase. This allowed us to identify from a relatively small compound library, compound 7 as a promising lead compound with micromolar potency of inhibition of GP, and as a scaffold for further development of a more specific inhibitor of the purine site of GP.

Second, the study shows that compound structural diversity is important for targeting the non-catalytic sites of GP, purine, indole and AMP. Development of non-glucose based inhibitors continues to open alternative ways to selectively inhibit GP, and thus potentially modified blood glucose levels.

 

References:

  1. Danaei G, Finucane MM, Lu Y, Singh GM, Cowan MJ, Paci orek CJ, Lin JK, Farzadfar F, Khang YH, Stevens GA, Rao M, Ali MK, Riley LM, Robinson CA, Ezzati M 2011 National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2.7 million participants. Lancet 9785:31–40
  2. Oikonomakos NG, Somsak L 2008 Advances in glycogen phosphorylase inhibitor design. Current Opinion in Investigational Drugs 9:379–395
  3. Schweiker SS, Loughlin WA, Brown CL, Pierens GK 2009 Synthesis of new modified truncated peptides and inhibition of glycogen phosphorylase. Journal of Peptide Science 15(6):442-450

 

Acknowledgements: This research was supported by the Eskitis Institute, Griffith University and partially from a grant from the Diabetes Australia Research Trust.

 

wl fig4Contact:

Wendy A. Loughlin, BSc(Hons), Ph. D., FRACI

Professor of Chemistry

School of Natural Sciences (formerly Biomolecular and Physical Sciences)

Griffith University, Nathan Campus Qld, 4111 AUSTRALIA

w.loughlin@griffith.edu.au

 

 

 

 

Multiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier SchönmannMultiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier Schönmann