Bioorg Med Chem Lett. 2018 Sep 1;28(16):2717-2722. doi: 10.1016/j.bmcl.2018.03.010. 

Structure-guided evolution of a 2-phenyl-4-carboxyquinoline chemotype into PPARα selective agonists: New leads for oculovascular conditions.

Dou XZ1, Nath D1, Shin Y2, Ma JX2, Duerfeldt AS3.
1 Department of Chemistry & Biochemistry, Stephenson Life Sciences Research Center, University of Oklahoma, Norman, OK, United States.
2 Department of Physiology, University of Oklahoma Health Sciences Center, Oklahoma City, OK, United States.
3 Department of Chemistry & Biochemistry, Stephenson Life Sciences Research Center, University of Oklahoma, Norman, OK, United States. Electronic address: adam.duerfeldt@ou.edu.

Abstract

Small molecule agonism of PPARα represents a promising new avenue for the development of non-invasive treatments for oculovascular diseases like diabetic retinopathy and age-related macular degeneration. Herein we report initial structure-activity relationships for the newly identified quinoline-based PPARα agonist, Y-0452. Preliminary computational studies led to the hypothesis that carboxylic acid transposition and deconstruction of the Y-0452 quinoline system would enhance ligand-protein interactions and better complement the nature of the binding pocket. A focused subset of analogs was designed, synthesized, and assessed for PPARα agonism. Two key observations arose from this work 1) contrary to other PPARα agonists, incorporation of the fibrate “head-group” decreases PPARα selectivity and instead provides pan-PPAR agonists and 2) computational models reveal a relatively unexploited amphiphilic pocket in PPARα that provides new opportunities for the development of novel agonists. As an example, compound 10 exhibits more potent PPARα agonism (EC50 = ∼6 µM) than Y-0452 (EC50 = ∼50 µM) and manifests >20-fold selectivity for PPARα over the PPARγ and PPARδ isoforms. More detailed biochemical analysis of 10 confirms typical downstream responses of PPARα agonism including PPARα upregulation, induction of target genes, and inhibition of cell migration. Copyright © 2018 Elsevier Ltd. All rights reserved.

KEYWORDS: Age-related macular degeneration; Diabetic retinopathy; PPAR selectivity; PPARα; Structure-based design

PMID: 29628329

 

Supplement:

Diabetic retinopathy (DR) is a serious, sight-threatening complication of diabetes mellitus that affects 40-45% of Americans diagnosed with diabetes.1 Inflammation in the retina plays a key role during the progression of DR,2 and involves pro-angiogenic and inflammatory factors such as interleukin-1β (IL-1β), tumor necrosis factor-alpha (TNF-α), intercellular adhesion molecule-1 (ICAM-1), monocyte chemotactic protein-1 (MCP-1) and vascular endothelial growth factor (VEGF).3-7 These factors induce the breakdown of the blood-retina barrier (BRB) and cause vessel leakage that results in diabetic macular edema (DME), the most common cause of vision loss in diabetic patients.8 Leukocyte adherence to the vasculature and/or leukostasis can cause local ischemia, leading to endothelial cell proliferation, migration and neovascularization (NV).2,9-11 Although treatment options for DR exist, current standard of care suffers from critical challenges: 1) >40% of the DR population are unresponsive to anti-VEGF therapy (the gold-standard); 2) current anti-VEGF treatments are cost-prohibitive for many patients (~$1,000 per dose) and require repetitive (e.g., monthly) intravitreal injections – both of which complicate long-term treatment and/or prevention; 3) secondary approaches like laser photocoagulation create irreversible damage to the retina; and 4) complementary corticosteroid therapy can trigger local/systemic side effects.8,12-14 New therapies that exhibit superiority to or synergism with current approaches are of great value to patients with DME and retinal NV, especially for those that remain refractory to current options.

Studies demonstrate that peroxisome proliferator-activated receptor alpha (PPARα) is expressed at high levels in all types of retinal cells, and that agonism of PPARα with genetic or pharmacological tools not only attenuates VEGF production but also protects against retinal inflammation, mitochondrial dysfunction, vascular leakage, neurodegeneration, and neovascularization (NV) in diabetic animal models. Additionally, it has been shown that diabetes induced PPARα downregulation results in retinal mitochondrial damage and oxidative stress. Thus, PPARα agonism with systemically available small molecules has potential to simultaneously address all major pathological features of DR and high promise to transform the way retinal conditions are treated. The promise of PPARα agonism as a novel strategy for treating DR has been confirmed in two independent, prospective human clinical trials (FIELD and ACCORD), wherein Fenofibrate (Feno, Fig. 1), a clinically approved drug for hyperlipidemia, exhibited robust protective effects against DR and retinal NV in type 2 diabetics.15-16 Since these human trials, it has been determined that the protective effects of Feno on retinal NV and DR are unrelated to its lipid-lowering activity, but rather result from its agonism of PPARα. Feno, however, suffers from poor ocular distribution, low affinity for PPARα, lack of selectivity between PPAR isoforms, and chemotype related dose-limiting toxicities, all of which will limit its use as a DR therapy. All totaled, however, the scientific evidence demonstrates that PPARα agonists with improved potencies and enhanced ocular distribution have high probability to become first-in-class therapeutic options for the treatment of DR is compelling.

 

 

Figure 1. Compounds included in this commentary. Fenofibric acid is the active metabolite of fenofibrate that binds to and agonizes PPARα.

 

Towards identifying new PPARα agonists, Y-0452 (Fig. 1) was identified from a virtual screen as a chemically distinct chemotype predicted to exhibit PPARα agonism.17 This prediction was verified experimentally in vitro and in vivo. Excitingly, Y-0452 exhibits efficacy in DR animal models after systemic (i.p.) administration, providing a new lead for development. As described in the our newest publication, we leveraged in silico studies and structure-based design to develop a novel 4-benzyloxy-benzylamino PPARα agonistic chemotype, which exhibits improved PPARα potency and selectivity over Feno and Y-0452.18 From these studies, compound 10 was identified as the top performer and was assessed in a number of settings to confirm PPARα agonism. In short, compound 10 exhibits improved potency over Y-0452 and >20-fold selectivity for PPARα. Additionally, compound 10 induces typical downstream, responses of PPARα agonism, which includes PPARα upregulation, induction of known target genes (Acadm, Cpt1a, Fabp3, and Slc25a20), and attenuation of cell migration.18 In this study, some additional observations were made that are worth expanding upon.

Fibrate head-group affects potency, level of agonism, and selectivity. Most PPARα agonists contain a carboxylic acid that acts as the pharmacophore and forms a hydrogen bond network with Ser280, Tyr314, His440, and Tyr464. This network is known to play key roles in affinity, selectivity and degree of agonism.19 The 4-benxyloxy-benzylamino chemotype that we developed (Fig. 1) contains a benzoic acid, which provides the carboxylic acid necessary to fulfill the critical hydrogen bonding trigger. Previous structure-activity relationship studies on other PPARα agonistic chemotypes show that incorporation of the fibrate head-group typically results in improved selectivity and potency for PPARα. Interestingly, however, the 4-benzyloxy-benzylamino chemotype fails to follow this trend, as incorporation of the fibrate head-group leads to decreases in potency and selectivity but results instead in a higher level of agonism. This is worth noting, as it suggests that the 4-benzyloxy-benzylamino chemotype interacts with the ligand binding domain differently than other reported chemotypes. This is an important observation, because as a transcription factor, it is well known that the specific downstream effects of PPARα activation are dependent upon ligand induced conformational changes. Thus, it is very possible that the 4-benzyloxy-benzylamino chemotype may display differential downstream effects and/or produce a unique PPARα agonistic profile that differentiates it from the well-studied fibrates.

In hopes of rationalizing why the fibrate head-group installation on this scaffold results in pan-agonism of all isoforms, we conducted in silico experiments to assess potential binding modes to the other PPAR isoforms, PPARδ and PPARγ (Fig. 2). Our models predict that incorporation of the fibrate head-group results in an extra predicted hydrogen bonding interaction with His413 (PPARδ) and His449 (PPARγ), which may strengthen the binding to these two isoforms. Additionally, inclusion of the fibrate head-group is predicted to cause a slight rotation of the pendant benzene ring in the binding pocket, which would allow for a cation-π interaction with Lys331 (PPARδ) and Lys367 (PPARγ). These additional interactions are not observed with the 4-benzyloxy-benzylmino chemotype and thus may be a key factor in the observed pan-agonism for the fibrate head-group containing analogs.

 

 

Figure 2. Compound 22 docked into the LBD of PPARδ (A, PDBID: 3TKM) and PPARγ (B, PDBID: 2VV0). The binding cavity is represented with a surface representation. Predicted hydrogen bonding interactions are represented by yellow dashes and cation-π interactions are represented by magenta dashes.

 

Presence of an amphiphilic pocket. In another docking study, we identified an amphiphilic pocket that we believed might provide an opportunity to enhance potency and selectivity even further (Fig. 3). We proposed that the amphiphilic pocket could be accessed by substituent extension off the meta position of the B-ring through an ether linkage. Two additional derivatives, 26 and 28 (Fig. 1), were evaluated for PPARα agonistic activity and selectivity in a cell-based luciferase assay. The potency of the benzoic acid analog 26 (5.1 μM) is similar to the comparator compound 10 (5.6 μM, Table 1). The fibrate “head-group” containing analog 28, exhibited a ~10-fold improvement in potency (2.1 μM) when compared against 22, but remains a pan-agonist (Table 1).18 While our efforts to exploit this pocket are only preliminary at this point, we still believe this amphiphilic pocket is an opportunity to investigate in the future.

 

 

Figure 3. Predicted binding pose of 28 and the amphiphilic pocket. Binding cavity depicted with a surface representation, PDBID: 2P54.

 

Table 1. Selectivity profile of selective analogues. Data are represented as the EC50 (μM) for the agonism of the corresponding luciferase reporter cell-line (Indigo Biosciences). Dosing was done in triplicate as a single experiment.

 

Although the 4-benxyloxy-benzylamino scaffold, specifically compound 10, represents a promising avenue for further exploration, information regarding the ADME and in vivo efficacy properties of this chemotype are needed. As mentioned previously, the ultimate goal is to develop systemically available treatments for DR and related conditions, and thus viable leads need to be metabolically stable, bioavailable, free of systemic toxicity, and capable of passing the blood-retinal barrier to reach the site of action. These studies are ongoing, and we look forward to communicating the results in due time. Nonetheless, the results presented in this publication provide new SAR insight and highlight a promising new PPARα agonistic chemotype poised for advancement.

 

 

References

  1. National Eye Institute Facts About Diabetic Eye Disease
  2. Joussen, A. M.; Poulaki, V.; Le, M. L.; Koizumi, K.; Esser, C.; Janicki, H.; Schraermeyer, U.; Kociok, N.; Fauser, S.; Kirchhof, B.; Kern, T. S.; Adamis, A. P., FASEB J. 2004, 18, 1450.
  3. Ben-Mahmud, B. M.; Mann, G. E.; Datti, A.; Orlacchio, A.; Kohner, E. M.; Chibber, R., Diabetes 2004, 53.
  4. Makino, N.; Maeda, T.; Sugano, M.; Satoh, S.; Watanabe, R.; Abe, N., J. Diabetes Comp. 2005, 19.
  5. Naldini, A.; Carraro, F., Curr. Drug Targ. 2005, 4.
  6. Yuuki, T.; Kanda, T.; Kimura, Y.; Kotajima, N.; Tamura, J.; Kobayashi, I.; Kishi, S., J. Diabetes Commun. 2001, 15.
  7. Zoppini, G.; Faccini, G.; Muggeo, M.; Zenari, L.; Falezza, G.; Targher, G., J. Clin. Endocrin. Metab. 2001, 186.
  8. Cheung, N.; Wong, I. Y.; Wong, T. Y., Diabetes Care 2014, 37, 900.
  9. Kinoshita, N.; Kakehashi, A.; Inoda, S.; Itou, Y.; Kuroki, M.; Yasu, T.; Kawakami, M.; Kanazawa, Y., Diabetologia 2002, 45.
  10. Miyamoto, K.; Khosrof, S.; Bursell, S. E.; Rohan, R.; Murata, T.; Clermont, A. C.; Aiello, L. E.; Ogura, Y.; Adamis, A. P., Proc. Natl. Acad. Sci. U.S.A. 1999, 96.
  11. Miyamoto, K.; Ogura, Y., Seminars in Ophthalmology. 1999, 14.
  12. Cai, X.; McGinnis, J. F., J. Diabetes Res. 2016, 2016, Article 3789217.
  13. Nentwich, M. M.; Ulbig, M. W., World J. Diabetes 2015, 6, 489.
  14. Ginsberg, H. N.; Elam, M. B.; Lovato, L. C.; Crouse III, J. R.; Leiter, L. A.; Linz, P.; Friedewald, W. T.; Buse, J. B.; Gerstein, H. C.; Probstfield, J.; Grimm, R. H.; Ismail-Beigi, F.; Bigger, J. T.; Goff, D. C.; Cushman, W. C.; Simons-Morton, D. G.; Byington, R. P., N. Engl. J. Med. 2010, 362, 1563.
  15. Keech, A. C.; Mitchell, P.; Summanen, P. A.; O’Day, J.; Davis, T. M. E.; Moffitt, M. S.; Taskinen, M. R.; Simes, R. J.; Tse, D.; Williamson, E.; Merrifield, A.; Laatikainen, L. T.; d’Emden, M. C.; Crimet, D. C.; O’Connell, R. L.; Colman, P. G., Lancet 2007, 370, 1687.
  16. Deng, G.; Moran, E. P.; Cheng, R.; Matlock, G.; Zhou, K.; Moran, D.; Chen, D.; Yu, Q.; Ma, J. X., Invest. Ophthalmol. Vis. Sci. 2017, 58.
  17. Dou, X. Z.; Nath, D.; Shin, Y.; Ma, J. X.; Duerfeldt, A. S., Bioorg. Med. Chem. Lett. 2018, 28, 2717.
  18. Capelli, D.; Cerchia, C.; Montanari, R.; Loiodice, F.; Tortorella, P.; Laghezza, A.; Cervoni, L.; Pochetti, G.; Lavecchia, A., Sci. Rep. 2016, 6, 34792.