PT2385

3-[(1S,2S,3R)-2,3-Difluoro-1-hydroxy-7-methylsulfonyl-indan-4- yl]oxy-5-fluoro-benzonitrile (PT2977), a Hypoxia-Inducible Factor 2# (HIF-2#) Inhibitor for the Treatment of Clear Cell Renal Cell Carcinoma

Rui Xu, Keshi Wang, James P. Rizzi, Heli Huang, Jonas A. Grina, Stephen T. Schlachter, Bin Wang, Paul M. Wehn, Hanbiao Yang, Darryl D. Dixon, Robert M. Czerwinski, Xinlin Du, Emily L. Ged, Guangzhou Han, Huiling Tan, Tai Wong, Shanhai Xie, John A. Josey, and Eli M. Wallace
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00719 • Publication Date (Web): 24 Jun 2019
Downloaded from http://pubs.acs.org on June 26, 2019

Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts. is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

6 3-[(1S,2S,3R)-2,3-Difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-
8
9 5-fluoro-benzonitrile (PT2977), a Hypoxia-Inducible Factor 2 HIF-

13 2 Inhibitor for the Treatment of Clear Cell Renal Cell Carcinoma

17 Rui Xu,*† Keshi Wang,† James P. Rizzi, Heli Huang, Jonas A. Grina, Stephen T. Schlachter,‡
18
19
20 Bin Wang, Paul M. Wehn, Hanbiao Yang, Darryl D. Dixon, Robert M. Czerwinski, Xinlin Du,
21
22 Emily L. Ged, Guangzhou Han, Huiling Tan, Tai Wong, Shanhai Xie, John A. Josey, Eli M.
23
24 Wallace
26
27 Peloton Therapeutics, Inc., 2330 Inwood Road, Suite 226, Dallas, TX 75235

† These authors contribute equally

ABSTRACT

42 The hypoxia-inducible factor 2 (HIF-2) is a key oncogenic driver in clear cell renal cell
44
45 carcinoma (ccRCC). Our first HIF-2 inhibitor PT2385 demonstrated promising proof of
46
47
48 concept clinical activity in heavily pretreated advanced ccRCC patients. However, PT2385 was
49
50 restricted by variable and dose-limited pharmacokinetics resulting from extensive metabolism
51
52
53 of PT2385 to its glucuronide metabolite. Herein we describe the discovery of second-
54
55 generation HIF-2 inhibitor PT2977 with increased potency and improved pharmacokinetic

3 profile achieved by reduction of phase 2 metabolism. Structural modification by changing the
5
6 geminal difluoro group in PT2385 to a vicinal difluoro group resulted in enhanced potency,
7
8
9 decreased lipophilicity and significantly improved pharmacokinetic properties. In a phase 1
10
11 dose-escalation study, the clinical pharmacokinetics for PT2977 supports the hypothesis that
12
13
14 attenuating the rate of glucuronidation would improve exposure and reduce variability in
15
16
17 patients. Early evidence of clinical activity shows promise for PT2977 in the treatment of

35 LogD7.4: 2.2
36 VEGFA EC50: 46 nM
37 Human PPB: 82% bound
38 Free fraction adjusted EC85: 540 ng/mL

INTRODUCTION

LogD7.4: 1.2
VEGFA EC50: 17 nM
Human PPB: 52% bound
Free fraction adjusted EC85: 75 ng/mL

46 Kidney cancer is among the 10 most common cancers in both men and women. Nearly 74,000
47
48 new cases of kidney cancer are estimated to be diagnosed in the United States in 2019 with over
50
51 14,000 deaths.1 Clear cell renal cell carcinoma (ccRCC) is the most common form of kidney
52
53 cancer. Although targeted therapies and immunotherapies have considerably improved the
54
55 prognosis for patients with ccRCC,2 only 12% of patients with advanced or metastatic disease

1
2
3 will survive 5 years.3 There exists a significant medical need for better therapeutic options for
4
5 ccRCC patients with metastatic disease.
7
8
9 In the majority of patients, ccRCC is characterized by inactivation of the tumor suppressor von
10
11
12 Hippel-Lindau (VHL) due to genetic predisposition, somatic mutations or methylation.4,5 The
13
14 VHL protein (pVHL) is a component of an E3 ubiquitin ligase complex that mediates protein
15
16 degradation by the proteasome. A principal role of pVHL is the regulation of the hypoxia-
18
19 inducible factor (HIF) family of transcription factors consisting of HIF-1, HIF-2, and the less
20
21 well characterized HIF-3.6-9 Unlike HIF-1, in adults the expression of HIF-2 is limited to
23
24 endothelial cells, kidney fibroblasts, hepatocytes, intestinal lumen epithelial cells, pancreatic
25
26 interstitial cells, heart myocytes, interstitial cells, and lung type II pneumocytes.10,11 Importantly,
27
28 HIF-2 activity has been demonstrated to be a key oncogenic driver in ccRCC.12-15 In mouse
30
31 ccRCC tumor models, knockdown of HIF-2 expression in pVHL defective cell line blocked
32
33 tumor growth comparable to reintroduction of pVHL. In addition, expression of a stabilized
34
35 variant of HIF-2 was able to overcome the tumor suppressive role of pVHL. HIF-2 has also
37
38 been shown to contribute to tumorigenesis in other malignancies such as glioblastoma
39
40 multiforme where hypoxia is a key feature of the tumor microenvironment and a correlation
41
42
43 between HIF-2 activity and survival has been observed.16,17
44
45
46 The activity of HIF-2 is controlled by oxygen-dependent hypoxia-inducible factor prolyl-
47
48
49 hydroxylases (PHDs).18,19 When oxygen availability is high (normoxia), these enzymes
50
51 hydroxylate specific proline residues on the oxygen-dependent domain of HIF-2, creating a
52
53 substrate recognition site for pVHL. pVHL recognizes the modified form of HIF-2, binds to it
55
56 and targets it for rapid proteasomal degradation. Under hypoxic conditions, there is insufficient
oxygen for the PHDs to catalyze hydroxylation of HIF-2, and pVHL cannot bind unmodified
4
5
6 HIF-2. As a result, HIF-2 accumulates and translocates to the nucleus where it dimerizes with
7
8 the constitutively expressed aryl hydrocarbon receptor nuclear translocator (ARNT, also known
9
10 as HIF-1) to form an active transcription factor complex. This complex recognizes hypoxia-
12
13 responsive elements (HRE) on DNA leading to increased expression of a variety of proteins,
14
15 many of which coordinately regulate angiogenesis, proliferation, migration and immune
16
17 evasion.20-22 In ccRCC, pVHL is defective or absent in the vast majority of patients which leads
19
20 to the accumulation and transcriptional activation of HIF-2 even under normoxic conditions.
21
22 While many gene products involved in the hypoxic response have been explored individually as
23
24 therapeutic targets for cancer, broad inhibition of the pathway through direct targeting of HIF-2α
26
27 offers an exciting opportunity to attack tumors on multiple axes.22
30
31 Despite the inherent difficulties in targeting DNA binding transcription factors, groundbreaking
32
33 research demonstrated that small molecule binding to an inner pocket in HIF-2 could
34
35 allosterically inhibit the protein-protein interaction between HIF-2 and ARNT leading to
37
38 inhibition of transcriptional activity.23-26 In a previous report,27 we disclosed our efforts guided
39
40 by iterative structure-based drug design and rational modification to identify a novel series of
41
42 HIF-2 inhibitors with excellent potency and physical and pharmacologic properties. This work
44
45 led to the discovery of compound 1 (PT2385), the first HIF-2 inhibitor to enter clinical
46
47 development.28-30 Although PT2385 was associated with highly variable drug exposure, anti-
48
49
50 tumor activity was observed in heavily pretreated advanced ccRCC patients, providing HIF-2
51
52 target validation.31 Herein we describe the path from PT2385 to the clinical candidate compound
53
54 2 (PT2977) by optimizing potency and DMPK properties.

BACKGROUND

23 A first-in-human study to evaluate PT2385 for safety, pharmacokinetics (PK) and
25
26 pharmacodynamics (PD) was completed in patients with advanced ccRCC previously treated
27
28
29 with one or more VEGF inhibitors.31 PT2385 was administered orally at twice-per-day (b.i.d.)
30
31 doses of 100 to 1800 mg, according to a standard 3 + 3 dose-escalation design, followed by an
33
34 expansion phase at the recommended phase 2 dose (RP2D). The dose-escalation and expansion
35
36
37 phases enrolled 26 and 25 patients, respectively. PT2385 was well-tolerated in this study, with
38
39 no dose-limiting toxicities observed at any dose level tested. Exposure of PT2385 was variable
40
41
42 and increased up to 800 mg b.i.d. No additional increases in compound circulating levels were
43
44
45 observed up to the highest dose of 1800 mg b.i.d. PT2385 treatment resulted in rapid decreases
46
47 in plasma erythropoietin (EPO), a HIF-2 regulated gene, at all doses, demonstrating target
48
49
50 engagement and biological activity. No further reduction was observed above 800 mg b.i.d.
51
52 Based on safety, PK, and PD data, 800 mg b.i.d. was recommended as the phase 2 dose. PT2385
54
55 demonstrated promising anti-tumor activity in these heavily pretreated patients. Of the 47

1
2
3 evaluable patients, one patient had a complete response (CR), six patients had a partial response
5
6 (PR), and 26 patients had stable disease (SD). Three patients remained in the study for over 3
7
8
9 years. Statistical analyses were performed to determine the relationship between PK
10
11 parameters and anti-tumor activity of PT2385. PT2385 steady-state trough level was found to
12
13
14 be significantly correlated with progression free survival (PFS). Patients with PT2385 trough
15
16 concentrations greater than 500 ng/mL experienced improved PFS (14.3 months) compared to
18
19 patients with PT2385 levels below 500 ng/mL (1.4 months) (Figure 2). PT2385 plasma
20
21
22 concentration of 500 ng/mL is equivalent to the free-fraction adjusted EC85 in the vascular
23
24 endothelial growth factor A (VEGFA) secretion assay in 786-O cells, a VHL-mutant ccRCC cell.
26
27 In the 786-O mouse xenograft model, PT2385 trough plasma concentration of 500 ng/mL was

52 Figure 2. Comparison of progression-free survival (PFS) profiles in patients experiencing
53
54 steady-state exposure trough concentrations < 500 ng/mL (red line) versus those with trough

3 concentrations > 500 ng/mL (black line). Data include all evaluable patients in dose escalation

18 Evaluation of the pharmacokinetic profiles of PT2385 in patients showed that a significant
19
20 proportion were underexposed. Individual time-plasma concentration curves for all 25 patients
21
22 who received a single 800 mg dose (the RP2D) of PT2385 shows that patients could be
23
24 categorized into “normal exposure” and “low exposure” groups (Figure 3). From a total of 25
26
27 patients, 6 patients belonged to the “low exposure” group. These patients had significantly lower
28
29 area under the curve (AUC) and trough plasma concentrations than those in the “normal
30
31 exposure” group, and importantly, achieved little clinical benefit from PT2385 treatment.

3 Figure 3. Concentration-time profile on day 1 for individual patients receiving 800 mg PT2385.
4
5 Each line represents a separate individual who received PT2385. Red lines represent patients
7
8 who had low PT2385 plasma concentrations.
12 The hepatic clearance of PT2385 was predicted to be low to moderate in humans with an

15 extraction ratio (ER or observed clearance/hepatic blood flow) of 22% in human liver
16
17 microsomes (HLM) and 44% in human hepatocytes. In human hepatocyte incubations, the only
18
19 metabolite detectable by HPLC-UV chromatogram was the glucuronide PT2639 (Figure 4A),
20
21 suggesting that the primary metabolic pathway for PT2385 in humans may be glucuronidation.
23
24 The structure of PT2639 was confirmed by isolation of an authentic sample from dog urine
25
26 samples after oral administration of PT2385 and chemical synthesis. In preclinical studies
27
28 PT2639 did not bind to HIF-2 with an IC50 greater than 100 M in the scintillation proximity
30
31 assay (SPA).27 In the Phase 1 clinical trial, the average plasma concentration–time profiles of
32
33 PT2385 and PT2639 following a single oral administration of PT2385 at 800 mg are presented in
34
35
36 Figure 4B. The levels of PT2639 were significantly higher than PT2385. The mean AUC and
37
38 Cmax metabolite/parent ratios were 6.9 and 9.7, respectively. The high levels of PT2639 in
39
40 patients was somewhat surprising based on preclinical data. While circulating levels of PT2639
41
42 were abundant in dogs (AUC ratio of PT2639:PT2385 = 1.8), it was not present in mice and was
44
45 found in significantly lower levels than PT2385 in rats (AUC ratio of PT2639:PT2385 = 0.24).

18 Figure 4. (A) Structure of the glucuronide metabolite 3 (PT2639). (B) Mean plasma
20
21 concentration of PT2385 and PT2639 after single oral administration of 800 mg PT2385.
22
23
24 In the phase 1 dose escalation study, the mean AUC of PT2385 on day 1 increased with
25
26 increasing dose from 100 to 800 mg in a generally dose-proportional manner, while a further
28
29 increase was not achieved when escalating to 1800 mg (Figure 5A). Unlike the parent drug, its
30
31 metabolite PT2639 showed more than proportional increases in exposure with higher doses
32
33 (Figure 5B). For orally administered drugs, aqueous solubility is one important parameter that
35
36 could contribute to inadequate and variable bioavailability. PT2385 is a BCS Class 232
37
38 compound with moderate thermodynamic aqueous solubility at pH 7.4 (15 g/mL) and high
39
40 permeability. In preclinical species, PT2385 exhibited good oral bioavailability in mouse, rat and
42
43 dog.27 These results combined with the fact that parent exposure plateaued at dose levels starting
44
45 at 800 mg, but metabolite exposure continued to increase, suggest that limited solubility of
46
47 PT2385 in gastrointestinal fluids is not the main reason for lack of dose-proportionality at higher
49
50 doses. Instead, extensive, variable metabolism of PT2385 to PT2639 is likely the major factor
51
52 contributing to its suboptimal pharmacokinetic profile.
Figure 5. (A) Dose-response plot showing dose-limiting exposure of PT2385. (B) Dose-response
23
24 plot showing super-proportional exposure of PT2639.
25
26
27 To determine which human uridine 5′-diphospho-glucuronosyltransferases (UDP-
29
30 glucuronosyltransferases, or UGTs) are responsible for direct glucuronidation of PT2385, eleven
31
32 recombinantly expressed UGTs were incubated with 50 M PT2385 and the formation of
33
34 PT2639 was monitored (Figure 6). Significant PT2639 was formed after 60-minute incubation
36
37 with UGT2B17 supersomes, indicating PT2385 is a substrate of UGT2B17. There was none or
38
39 very little PT2639 detected after incubation of PT2385 with the ten other UGTs, suggesting that
40
41 UGT2B17 is the major enzyme responsible for PT2639 formation. UGT2B17 belongs to the
43
44 UGT2 superfamily of UGTs and plays an important role in testosterone metabolism in humans.33
45
46 UGT2B17 is primarily expressed in human intestine where it is found in much greater abundance
47
48 than liver. It was reported that mRNA expression of UGT2B17 in the intestine is about 13-fold
50
51 higher than in the liver,34 and protein abundance in intestine is 4.4-fold higher than in liver.35
52
53 Importantly, among all UGTs the 2B17 isoform has the highest expression in the intestine. It
54
55 constitutes 59% of total UGT isoforms in the intestine but only 3% in the liver.36 UGT2B17 is

3 one of the most frequently deleted genes in humans.37 It also shows high inter-patient variability
4
5 in its protein abundance in both adult human liver microsomes (HLM) and human intestine
7
8 microsomes (HIM).38 Genetic polymorphism, unique abundance in the intestine and high inter-
9
10 individual expression all contribute to variable exposure for orally administered UGT2B17
11
12 metabolized drugs, likely as a result of first pass metabolism.38,39 These data strongly suggest
14
15 that gastrointestinal metabolism of PT2385 to PT2639, mediated by UGT2B17, is likely the
16
17 cause of the significant variability and low drug plasma exposure in the clinic.
Figure 6. Formation of PT2639 by recombinant human UGTs.

45 We also performed studies to evaluate additional potential mechanisms responsible for dose-
46
47 limiting drug exposure of PT2385 observed in patients. In small intestine, efflux transporters
48
49 play important roles in the disposition of glucuronide metabolites. P-gp, BCRP and MRP2 are
51
52 three major efflux transporters expressed in the small intestine.40 The efflux ratio of PT2385 and
53
54 PT2639 across MDCKII-MDR1 and MDCKII-BCRP cells was less than 2 at concentrations up

to 300 M, suggesting that they are not substrates of P-gp or BCRP. PT2385 and PT2639 were
4
5
6 also evaluated in the MRP2 transporter assay in membrane vesicles expressing human MRP2
7
8 protein. There was no MRP2-dependent uptake observed for PT2385. However, the MRP2
9
10 uptake of PT2639 was approximately 32-fold higher in vesicles that contained ATP compared to
11
12 those without suggesting active transport. As the concentration of PT2639 was increased, uptake
14
15 into MRP2-expressing vesicles plateaued suggesting active transport was saturated. The
16
17 maximum velocity (Vmax) and substrate concentration at half maximum velocity (Km) were
18
19 determined to be 1230 pmol/min/mg protein and 97 μM, respectively. The ATP dependent
21
22 activity was abolished by the MRP2 inhibitor benzbromarone. These data support that PT2639 is
23
24 a substrate of MRP2 and at higher concentrations PT2639 can saturate MRP2. We also have
25
26 shown that PT2639 can be converted to PT2385 by bacterial -glucuronidase in vitro.
Figure 7. Proposed UGT2B17, MRP2 and -glucuronidase interplay in the absorption of
Based on these studies and clinical pharmacokinetic data, we propose an absorption model to
4
5 explain the intestinal metabolism of PT2385 that, we believe, is responsible for its suboptimal
7
8 oral exposure in patients (Figure 7). After oral administration, PT2385 is absorbed through the
9
10 enterocytes by passive transcellular diffusion. A small portion of PT2385 is transported to the
11
12 blood as parent, while the rest of the drug is extensively metabolized by UGT2B17 to form
14
15 PT2639. PT2639 can then follow one of two routes; absorption into the blood stream or active
16
17 MRP2-mediated transport to the intestinal lumen. Once in the gastrointestinal lumen, PT2639
18
19 can then be converted by the action of bacterial -glucuronidase to PT2385, which can undergo
21
22 intestinal re-absorption resulting in a re-circulation cycle. Additional data to support this model
23
24 includes PT2639 dog PK. After intravenous (i.v.) and oral gavage (p.o.) administration of
25
26 PT2639 to dogs, significant levels of PT2385 were generated by both routes, with oral dosing
28
29 generating more PT2385 than i.v. dosing.
30
31 This proposed mechanism involves UGT2B17, MRP2 and -glucuronidase functioning
32
33 individually and working in concert. The genetic polymorphism and irregular expression of these
35
36 three components likely all play roles in the observed variable exposure of PT2385. Finally, the
37
38 local enterocyte concentration of PT2639 generated by UGT2B17 after oral doses of PT2385
39
40 greater than 800 mg likely surpassed the K of PT2639 for MRP2, resulting in saturation of the
41
42
43 transporter and thus dose-limited exposure of PT2385 and super-proportional exposure of
44
45 PT2639. Therefore, the preclinical and clinical PT2385 and PT2639 data supports our model,
46
47 and provides a framework for medicinal chemistry optimization to overcome PT2385
48
49 pharmacokinetic liabilities.
51
52
53
54 Compound Design and Lead Optimization

3 Despite the complexity of our proposed model, the pharmacokinetic issues with PT2385 may all
4
5 relate to extensive metabolism of PT2385 to PT2639 by UGT2B17 in the intestine. Thus, our
7
8 primary strategy to enhance HIF-2 inhibition in patients was to reduce formation of the
9
10 glucuronide metabolite while retaining potency, selectivity and pharmacological profile of
11
12 PT2385.
14
15 Glucuronidation is an important pathway of human drug metabolism. Most often glucuronidation
16
17 is the second step of drug metabolism by acting on hydroxylated products generated by oxidative
18
19 metabolism. However, it can also react directly with drugs containing hydroxyl, carboxylic acid,
21
22 amino or thiol functional groups. Glucuronidation is a SN2 reaction catalyzed by the UGT
23
24 enzymes in which the metabolized or parent drug acts as a nucleophile, the secondary alcohol in
25
26 the case of PT2385, and the cofactor UDP--D-glucuronic acid (UDPGA, 4) as the electrophile
28
29 (Figure 8). The rate of glucuronidation is dependent on both nucleophilicity and the steric
30
31 environment of the drug. For O-glucuronidation, deprotonation of the alcohol is considered to be
32
33 a key step.41-43 Alcohols substituted by electron withdrawing groups tend to have lower pKa, and
35
36 thus higher reactivity in the glucuronidation reaction.

3 Previously,27 we described the discovery of a series of indanols as potent and selective inhibitors
4
5 of HIF-2. Protein-inhibitor X-ray crystal structure of PT2385 as shown in Figure 9 revealed
7
8 two key protein-ligand interactions. The hydroxyl group in PT2385 engages a hydrogen bonding
9
10 network with Tyr281, a bound water molecule and His293. Three residues in HIF-2, Met252,
12
13 His293 and Tyr278 must move to accommodate the formation of this hydrogen bonding
14
15 network. These movements lead to highly unfavorable steric clash between Tyr278 in HIF-2
16
17 and Phe446 in ARNT, resulting in HIF-2:ARNT dimer disruption and inhibition of HIF-2
19
20 transcriptional activity. Another key interaction between the ligands and HIF-2 is a putative n
21
22  *Ar type of interaction44 between the phenolic oxygen of Tyr281 and the centroid of the
24
25 benzene portion of the indane ring. Maintaining an electron deficient ring system is critical to
26
27 engage this interaction for optimum ligand binding. The electron-withdrawing methyl sulfone
28
29 group in PT2385 enhances binding affinity by decreasing the electron density of the benzene
31
32 ring.
33
34 Due to the importance of the hydroxyl group to dimer disruption, cellular potency and physical
35
36 properties, our design strategy for the second-generation HIF-2 inhibitors was to maintain the
38
39 hydroxyl group, but significantly reduce its reactivity in glucuronidation and thereby improve
40
41 pharmacokinetic performance. The geminal difluoro group in the indanol series provides a 10 to
42
43 20-fold enhancement of cell potency compared to unsubstituted analogs. We speculated that the
45
46 cis-fluorine atom engages in intramolecular C−F···H−O electrostatic interactions and thus
47
48 reduce the desolvation penalty required for ligand binding.27,45 However, the highly
49
50 electronegative fluorine atoms, through a strong inductive effect, also likely increase the
51
52 hydroxyl moiety’s reactivity in the glucuronidation reaction. In order to compensate for the
likely decreased activity of analogs without the geminal difluoro group, additional optimization
4
5 was required to identify potency enhancing moieties.
Figure 9. X-ray crystal structure of PT2385 (in orange) in complex with the HIF-2 PAS-
29
30 B*/ARNT PAS-B* dimer (PDB:5TBM).27 Tyr281 and His293 are highlighted (magenta).
31
32 Dotted lines indicate hydrogen bond interactions and the red sphere denotes a water molecule Based on our understanding of the key n  *Ar
type of interaction between the phenolic oxygen
38
39 of Tyr281 and the small molecule ligands, we hypothesized that introducing a fluorine group at
40
41 the benzylic position of PT2385 may further decrease the electron density of the phenyl ring
43
44 resulting in a stronger n  *Ar electrostatic interaction and increased potency. A series of
45
46 compounds with fluorine substitution on the indanol ring were prepared and tested in the SPA
47
48 binding assay as well as HIF-2 luciferase assay in 786-O cells. As summarized in Table 1,
50
51 removal of the geminal difluoro group in PT2385 (compound 6) led to 8-fold decrease of
52
53 potency in the biochemical assay and 30-fold decrease of potency in the cellular assay, consistent
54
55 with the previous SAR. One particularly interesting finding that emerged from mono-fluorine

3 substitution on the 2-position of the indanol core is the different potency profiles between
4
5 compounds with cis and trans configurations. The cis-fluorohydrin 7 retained most of the
7
8 potency of PT2385, with EC50 within 2-fold of PT2385 in the luciferase assay. On the contrary,
9
10 the trans-fluorohydrin 8 was more than 10-fold less potent than 7. Mono-fluorination at the
11
12 benzylic position resulted in even more dramatic potency differences between the two
14
15 diastereomers. While the cis-diastereomer 9 was 2-fold more potent than 6 in the luciferase
16
17 assay, introduction of a fluorine atom with trans-configuration at the benzylic position
18
19 completely abrogated binding affinity (compound 10). Combination of the two fluorine atoms at
20
21
22 cis relative configuration to the hydroxyl group resulted in potency improvement in a synergistic
23
24 manner. Compound 2 was about 2-fold more potent than PT2385 in the luciferase assay.
25
26 Reverting the fluorine stereochemistry at the benzylic position again decreased potency.
27
28 Compound 11 was almost 200-fold less potent than 2 in both the binding and cellular assays.
30
31 Weak activity for compounds with trans-fluorine at the benzylic position was also observed in
32
33 compound 12, which has a geminal difluoro group at the benzylic position. Inspection of the X-
34
35 ray crystal structures of these analogs revealed that the benzylic trans-fluorine would probably
37
38 point to His248, resulting in detrimental steric and electronic interactions. Finally, the trifluoro
39
40 compound 13 was the most potent compound in this group. This was predicted by favorable
41
42 electrostatic interactions between the geminal difluoro group with the alcohol, and the additional
43
45 inductive effects on the phenyl ring from the benzylic cis-fluorine atom.
46
To explore the effect of fluorination on the rate of glucuronidation, compounds 1, 2, 6-9 and 13
42
43 were evaluated in enzyme kinetic studies in HIM and UGT2B17 supersomes. The corresponding
44
45 glucuronide metabolites were prepared by chemical synthesis and used as standards for
46
47 quantification. Substrate saturation plots for the formation of glucuronide metabolites from
49
50 parents are shown in Figure 11. Data were fitted using the Michaelis-Menten equation. Enzyme
51
52 kinetics parameters Vmax and Km are summarized in Table 2. For PT2385, enzyme kinetics
53
54 appeared to follow simple hyperbolic Michaelis-Menten kinetics with Vmax values of 49 and 280

1
2
3 pmol/min/mg protein, and Km values of 20 and 11 μM in HIM and recombinant UGT2B17,
4
5 respectively. As expected, compound 13 with the geminal difluoro and benzylic fluoro groups
7
8 had the highest rate of glucuronidation as determined by highest Vmax and lowest Km values in
9
10 both systems. Surprisingly, the compound with the third highest glucuronidation rate was the
11
12 trans-fluorohydrin compound 8. The Vmax of 8 was about 5-times of the corresponding cis-
14
15 diastereomer 7 in HIM and 4-times in UGT2B17 supersomes. In a recent paper,45 the influence
16
17 of fluorination on H-bond acidity was investigated using the conformationally restricted 4-tert-
18
19 butylcyclohexanol model. It was demonstrated that cis-fluorination on the  or  carbon of the
21
22 cyclohexanol can significantly attenuate alcohol H-bond acidity through intramolecular F···HO
23
24 interactions. We reasoned that two geminal fluorine atoms in PT2385 probably play different
25
26 roles in contributing to the reactivity of PT2385 in the glucuronidation reaction. For the cis-
28
29 fluorine, the electron-withdrawing effect and the intramolecular F···HO electrostatic interaction
30
31 cancel out each other, as evident in similar glucuronidation rates between the cis-fluorohydrin 7
32
33 and the nonfluorinated analog 6. In the trans-diastereomer, such an intramolecular hydrogen
34
35 bonding interaction cannot form. As a result, the trans-fluorine significantly increases the acidity
37
38 of the hydroxyl group in compound 8 through an inductive effect. It was also reported that 1,3-
39
40 coaxial fluorohydrin in the 4-tert-butylcyclohexanol ring system forms a favorable F···HO
41
42 interaction. Such a strong impact of 3-cis-fluorination was not observed in compound 9, possibly
44
45 due to a different fluorohydrin dihedral angle in the fused five-membered ring system. In
46
47 compounds 2 and 7, only the cis-fluorine atom that contributes positively to improved potency
48
49 and decreased reactivity is maintained. Comparing to PT2385, 2 and 7 were at least 20-times
51
52 slower in generating glucuronide metabolites in enzyme kinetic studies in HIM and UGT2B17
53
54 supersomes. Based on these in vitro results, we expect that these two compounds will afford

1
2
3 improved oral exposure with less variability in humans. In addition, the cis-benzylic fluorine in
4
5 compound 2 further enhances potency. Based on potency and enzyme kinetic study results,
7
8 compound 2 was selected for additional profiling.
9
Characterization of PT2977
4
5 Compared to PT2385, PT2977 was found to be 2 to 3-fold more potent in the HIF-2 luciferase
7
8 assay and the VEGFA secretion assay. Simply moving the trans- fluorine atom of the geminal
9
10 difluoro group in PT2385 to the benzylic position with cis- relative configuration to the hydroxyl

of PT2977 is about 1 log unit

15 lower than PT2385. In general, replacement of a hydrogen with a fluorine results in slight
16
17 increase in lipophilicity. However, examples have been reported in the literature that show
18
19 fluorination decreases molecular lipophilicity. One extreme example is all-cis 1,2,3,4,5,6-
21
22 hexafluorocyclohexane, a facial polarized ring reported to be one of the most polar aliphatic
23
24 molecules ever measured.46 It was also reported that compounds with fluoro substitution two or
25
26 three carbons away from an oxygen atom tend to have decreased lipophilicity.47,48 The increased
28
29 polarity of PT2977 is likely due to the combination effects of fluorine atoms in close vicinity to
30
31 the alcohol, and the high molecular dipole created by two vicinal electronegative fluorine atoms
32
33 with cis-configuration. The decreased lipophilicity translates to significantly reduced plasma
34
35 protein binding. PT2977 was about 30% less bound than PT2385 in the human plasma protein
37
38 binding assay (52% bound versus 82% bound). Combining the improvement in free fraction and
39
40 cellular potency, the free fraction adjusted EC85 for PT2977 in the VEGFA secretion assay was
41
42 calculated to be 75 ng/mL. Additional profiling of PT2977 showed it to be a low clearance
44
45 compound in microsomes and hepatocytes across species, and that it does not inhibit CYP P450
46
47 enzymes up to 50 μM (the highest concentration evaluated). Safety profiling of PT2977 showed
48
49 low potential for QT prolongation with an estimated IC50 greater than 50 μM in the patch clamp
51
52 hERG channel assay. It also had no appreciable activity at 10 μM in a screening panel that
53
54 included 126 receptors, ion channels, kinases and phosphatases.

23 Pharmacokinetic Profile of PT2977 in Preclinical Species
25
26 Based on its encouraging in vitro profile, PT2977 was evaluated in animal pharmacokinetic
27
28 studies in mice, rats, dogs and monkeys. The PK profile of PT2977 after i.v. administration is
29
30 summarized in Table 4. PT2977 had low plasma clearance (Cl ) in mice, dogs and monkeys, and
31
32
33 moderate clearance in rats. The in vivo hepatic extraction ratios (ER) are consistent with low to
34
35 medium clearance predicted by in vitro stability assays in microsomes and hepatocytes, and
36
37 lower potential to form glucuronide metabolite in the enzyme kinetic study.

16 aPT2977 was delivered as a solution in 20% EtOH, 40% PEG400, 40% water. bPT2977 was delivered as a solution in
17
18 10% dimethylacetamide, 10% EtOH, 40%PEG400, 40% water. cER = extraction ratio: observed clearance / hepatic
19 blood flow.
20
24 Oral administration of PT2977 in mice, rats, dogs and monkeys resulted in good plasma
25
26 exposure. Comparison of mean plasma concentration–time profiles of PT2977 and PT2385 are
28
29 shown in Figure 12, and summary of parent and glucuronide metabolite drug exposure (AUC) is
30
31 provided in Table 5. Interestingly, the oral exposure of PT2977 was only slightly higher than
32
33 PT2385 in rodents, while these two compounds behaved very differently in higher species. The
35
36 dose-normalized AUC of PT2977 was 9- and 20-fold higher than that of PT2385 AUC in dogs
37
38 and monkeys, respectively. This phenomenon turned out to be related to the relative rate of
39
40 glucuronide metabolite formation for each analog in these species. As shown in Table 5, the
41
42 AUC of the PT2977 glucuronide metabolite (PT3317) in dogs was about 30% of the parent,
44
45 while the AUC of PT2639 (PT2385 glucuronide metabolite) was almost 2-fold higher than the
46
47 parent. Similarly, a low amount of circulating metabolite was observed for PT2977 in monkeys,
48
49 with a metabolite/parent ratio of 0.19. Both compounds formed significantly less amounts of
51
52 glucuronide metabolites in rats, suggesting an alternative metabolic pathway in this species. For
53
54 PT2385, dog pharmacokinetics were a better predictor of glucuronide metabolite PT2639

3 formation in humans. Based on these data, we predicted that PT2977 would have a reduced
4
5 propensity for glucuronidation in humans and thus demonstrate a significantly improved
7
8 pharmacokinetic profile over PT2385 in patients.

aPT2385 and PT2977 were delivered as a suspension in 10% EtOH, 30% PEG400, 60% (0.5%
26 methylcellulose, 0.5% Tween 80 (aq)). bPT2385 and PT2977 were delivered as a suspension in 0.5%
27
28 methylcellulose and 0.5% Tween 80 in water.
29
30 PK/PD Relationship and In Vivo Efficacy
31
32 A pharmacokinetic/pharmacodynamic (PK/PD) study was performed in mice bearing
34
35 subcutaneous 786-O ccRCC tumors to correlate plasma drug levels of PT2977 with PD effects in
36
37 the tumors. Six doses of PT2977 at 0.3, 1 and 3 mg/kg or PT2385 at 10 mg/kg b.i.d. were
38
39 administered orally. Plasma and tumor tissue samples were collected 12 hours after the last dose.
41
42 Gene expression analyses in excised tumors by qPCR showed that PT2977 potently and dose-
43
44 dependently reduced mRNA levels of human cyclin D1, a target gene regulated by HIF-2
45
46 (Figure 13A). Maximum PD response was achieved at 1 mg/kg for PT2977 while a much higher
48
49 dose of 10 mg/kg was required for PT2385.28 And to achieve the same PD effect, the steady state
50
51 trough plasma drug concentration of PT2977 was about 1/10 of PT2385 concentration (Figure
52
53 13B). Consistent with the improvement in potency and physical properties, compound PT2977
54
55
56 was about 10-fold more potent than PT2385 in vivo. Compound PT2977 and PT2385 were

3 subsequently evaluated for anti-tumor activity in the 786-O mouse xenograft model.
4
5 Administration of PT2977 at 0.3, 1 and 3 mg/kg all led to rapid regression of established tumors
7
8 (Figure 13C), confirming superior in vivo activity of compound PT2977 compared to PT2385.28
9
Figure 13. (A) PK/PD evaluation of PT2977 and PT2385 in 786-O mouse xenograft (female
33
34 SCID/beige mice, n=3) dosed at 0.3, 1 and 3 mg/kg and 10 mg/kg p.o., b.i.d., respectively.
35
36 Tumors were harvested 12 hours after administration of the final dose and total RNA was
38
39 isolated from the tumor samples and used to make single-stranded cDNA. The resulting cDNA
40
41 was used in quantitative PCR reactions; mRNA levels were normalized to internal control
42
43 cyclophilin B mRNA levels in each sample. (B) Plasma levels of PT2385 and PT2977 at 12
44
45 hours after the last dose in the PK/PD study. (C) In vivo efficacy study of PT2977 and PT2385 in
47
48 786-O xenografts (female SCID/beige mice, n=8).

52 Based on its improved potency, free fraction, pharmacokinetics and metabolite profile,
54
55 compound PT2977 was predicted to have a low clinically active dose and less inter-patient

1
2
3 exposure varability. PT2977 was chosen as our second-generation HIF-2 clinical candidate for
4
5
6 further investigation as a potential new treatment for ccRCC and VHL Disease.

10 PT2977 Phase 1 Clinical Data
11
12 In a phase 1 dose-escalation trial,49,50 patients with advanced solid tumors were treated with
14
15 PT2977 with doses ranging from 20 to 240 mg once a day. The pharmacokinetic profile on day
16
17 15 is summarized in Figure 14. Exposure of PT2977 increased with dose up to 120 mg (Table 6).
18
19 Doses above 120 mg did not provide markedly increased exposure. Target steady-state trough
21
22 concentration of 75 ng/mL was achieved at 20 mg and exceeded with doses at 40 mg and above.
23

44 Figure 14. Mean PT2977 plasma concentration on day 15 versus time following once daily oral
45
46 administration at the dose levels of 20, 40, 80, 120, 160, and 240 mg in patients with solid
48
49 tumors. Dotted line denotes free fraction adjusted EC85 of 75 ng/mL for PT2977 in the VEGFA
50
51 secretion assay in 786-O cells.

22 As shown in Figure 15A, treatment with PT2977 led to rapid and dose-dependent reduction in
23
24 EPO expression. EPO reduction with PT2977 at 120 mg q.d. was comparable to the level of
26
27 reduction in PT2385 high exposure patients (Figure 15B). PT2977 was well tolerated with
28
29 anemia as the most common adverse event consistent with pharmacologic action. Maximum
30
31 tolerated dose (MTD) was not reached in the phase 1 dose-escalation study. Based on PK, PD
33
34 and safety data, RP2D was determined to be 120 mg q.d.
Figure 15. (A) Pharmacodynamic response as assessed by decreases in the HIF-2 target
54
55 erythropoietin (EPO) following once daily oral administration of PT2977 at the dose levels of

3 20, 40, 80, 120, 160 and 240 mg in patients with solid tumors. (B) Decreases in EPO in high
4
5 exposure PT2385 patients with > 500 ng/mL trough plasma concentration versus all PT2977
7
8 patients in the 120 mg cohort.
9
10 Figures 16A shows mean PT2977 and glucuronide metabolite (PT3317) plasma concentrations at
11
12 steady-state versus time. On day 15, the trough concentration (24 h) for PT2977 at 120 mg q.d.
14
15 was 500 ng/mL and the metabolite/parent ratio was 0.32. For PT2385, a 13-fold higher total
16
17 daily dose was required to achieve a trough concentration of 860 ng/mL at 12 h, and the
18
19 metabolite/parent ratio (6.0) was much higher than that of PT2977 (Figure 16B). In addition to
20
21
22 greater potency and free fraction, PT2977 exposure is more consistent with no underexposed
23
24 patients (Figure 17). These PK results are in-line with our prediction based on UGT2B17
25
26 enzyme kinetic data and PK in preclinical species, further supporting our hypothesis that
27
28 glucuronidation plays a critical role in the absorption and metabolism of this series of HIF-2
30
31 inhibitors.

3 Figure 16. (A) Mean PT2977 and the glucuronide metabolite PT3317 plasma concentration on
4
5 day 15 versus time following once daily oral administration at 120 mg. Dotted line denotes free
7
8 fraction adjusted EC85 of 75 ng/mL for PT2977 in the VEGFA secretion assay in 786-O cells.
9
10 (B) Mean PT2385 and the glucuronide metabolite PT2639 plasma concentration on day 15
11
12 versus time following twice daily oral administration at 800 mg. Dotted line denotes free fraction
14
15 adjusted EC85 of 540 ng/mL for PT2385 in the VEGFA secretion assay in 786-O cells. Bottom:
16
17 Day 15 metabolite/parent AUC ratios and trough plasma concentrations for PT2977 and PT2385.

39 Figure 17. AUClast on day 15 for individual patients received 120 mg PT2977 q.d. or 800 mg
41
42 PT2385 b.i.d. Each data point represents a single patient.

CONCLUSIONS
47
48 In this paper we investigated potential causes of inter-individual PK variability and low exposure
50
51 of our first-in-class HIF-2 inhibitor PT2385 in patients. Based on circulating metabolite
52
53 identification and UGT phenotyping, we hypothesized that attenuating the rate of
54
55
56 glucuronidation would improve exposure and reduce variability in patients. Structural

3 modification by moving one of the geminal difluoro atoms in PT2385 to the benzylic position
4
5 resulted in enhanced potency, significantly decreased lipophilicity and drastically reduced
7
8 glucuronidation. Enzyme kinetic studies in HIM and UGT supersomes revealed the impact of
9
10 fluorine substitution and stereochemistry on rate of glucuronidation. To our knowledge, this is
11
12 the first systematic study of the influence of these factors on glucuronidation. The application of
14
15 fluorine in modern medicinal chemistry and drug development has expanded rapidly in recent
16
17 years.51,52 More than 40% of new chemical entities (NCE) approved by FDA in 2018 contain one
18
19 or more fluorine atoms.53 This research represents a case study highlighting the significant
20
21
22 impact of fluorine substitution pattern on intrinsic potency, metabolite stability and
23
24 pharmacokinetics properties.
25
26 Our work culminated in the advancement of PT2977, a novel HIF-2 inhibitor with excellent in
27
28
29 vitro potency, pharmacokinetic profile and in vivo efficacy in mouse tumor models. The clinical
30
31 pharmacokinetics of PT2977 demonstrated consistent and significantly higher exposure than
32
33 PT2385, achieved with much lower dose and once daily dosing. PT2977 also showed
34
35 encouraging outcomes in patients with advanced renal cell carcinoma in an expansion cohort of
37
38 fifty-five patients with ccRCC treated at 120 mg q.d. As of January 1, 2019, 12 patients (22%)
39
40 had a confirmed partial response. Median progression free survival (PFS) was not yet reached in
41
42 the study with a median follow-up of 9 months, and 36% of patients remained on study at the
44
45 time of the data cut-off.50 Data supports continued clinical development with planned PT2977
46
47 monotherapy Phase 3 trial. In addition, an international Phase 2 study in patients with VHL
48
49 Disease is ongoing.

3 Chemical Synthesis
4
5 Preparation of compounds 1, 7 and 8 has been previously reported.27 Syntheses of compounds 2,
7
8 6, 9-11 and 13 are summarized in Scheme 1. Reduction of ketone 14 by Noyori’s asymmetric
9
10 transfer hydrogenation afforded 6 enantioselectively. Preparation of 2, 9-11 and 13 began with
11
12 acetylation of the alcohol in 1, 6 and 7 followed by benzylic bromination with N-
14
15 bromosuccinimide (NBS) catalyzed by 2,2’-azobis(2-methylpropionitrile) (AIBN). Subsequent
16
17 treatment of 16 with Ag2CO3 or AgClO4 in 1,2-dimethoxyethane and water gave alcohols 17.
18
19 Fluorination of the hydroxy groups in 17 followed by deprotection yielded the target compounds.
20
21
22 Mixtures of diastereomers were readily separated by chromatography on silica gel, either at the
23
24 alcohol 17 step or after fluorination.
25
26 Scheme 1. Synthesis of 2, 6, 9-11 and 13a

45 aReagents and conditions: (a) [(R,R)-Ts-DPEN]RuCl(p-cymene), HCO2H, Et3N, CH2Cl2, rt,
46
47 83%; (b) Ac2O, Et3N, DMAP, CH2Cl2, rt, 61-96%; (c) NBS, AIBN, 1,2-dichloroethane or carbon
48
49 tetrachloride, 80 °C; (d) Ag2CO3 or AgClO4, 1,2-dimethoxyethane, water; (e) DAST, CH2Cl2, -
50
51 78 °C → 0 °C; (f) LiOH, THF, water, rt, 23-86%.

1
2
3 The installation of the benzylic geminal difluoro group in 12 was accomplished by fluorination
4
5 of the ketone 19, which was easily obtained by oxidation of the alcohol 17c. Hydrolysis of the
7
8 acetate protecting group led to 12 in excellent overall yield (Scheme 2).
9
10 Scheme 2. Synthesis of 12a
11

21 aReagents and conditions: (a) Dess-Martin periodinane, CH2Cl2, rt, 88%; (b) 4-(tert-butyl)-2,6-
22
23 dimethylphenyl sulfur trifluoride, hydrogen fluoride pyridine (70%), CH2Cl2, rt, 84%; (c) LiOH,
24
25 THF, water, 0 °C, 86%.
27
28 To synthesize the glucuronide metabolites 3 and 20-25, alcohols 1, 2, 6-9 and 13 were treated
29
30 with the glucuronosyl donor, 2,3,4-tri-O-acetyl-α-D-glucuronide methyl ester
31
32 trichloroacetimidate, in dry DCM using boron trifluoride diethyl etherate or trimethylsilyl
33
34
35 trifluoromethanesulfonate as a promoter, produced the corresponding glycosidic coupling
36
37 products stereoselectively. Subsequent removal of the acetyl protecting groups and the methyl
38
39 ester under basic conditions in THF afforded the glucuronides (Scheme 3).
40
41 Scheme 3. General synthesis of glucuronide metabolitesa
43

3 aReagents and conditions: (a) 2,3,4-tri-O-acetyl-α-D-glucuronide methyl ester
4
5 trichloroacetimidate, boron trifluoride diethyl etherate (Method A) or trimethylsilyl
7
8 trifluoromethanesulfonate (Method B), CH2Cl2, rt; (b) LiOH, THF, water.
9
10
11
12 EXPERIMENTAL SECTION
14
15 General Chemistry: All solvents and reagents were used as obtained. 1H and 19F analysis of
16
17 intermediates and exemplified compounds were performed on an Agilent Technologies 400/54
18
19 magnet system (operating at 399.85 MHz or 376.24 MHz). Vnmrj VERSION 3.2 software pulse
20
21
22 sequences were selected from the default experiment set. Chemical shifts are expressed as δ units
23
24 using trimethylsilane (TMS) as the external standard (in NMR description, s = singlet, d =
25
26 doublet, t = triplet, q = quartet, m = multiplet, and br = broad peak).
27
28 High performance liquid chromatography (HPLC) coupled to a mass spectrometer (MS) was
30
31 used to determine the purity of the compounds synthesized. The data confirmed that the target
32
33 compounds had ≥ 95% of purity. The following analytical method was used to determine
34
35 chemical purity of final compounds: Agilent 1200 series high performance liquid
37
38 chromatography (HPLC) system operating in reverse-phase mode coupled to an Agilent 6150
39
40 Quadrapole spectrometer using an ESI source, water with 0.1% formic acid (mobile phase A),
41
42 acetonitrile with 0.1% formic acid (mobile phase B), Agilent ZORBAX Eclipse Plus C18, 1.8
43
44
45 μm, 2.1×50 mm, 40 °C column temperature, 5−95% mobile phase B in 4.0 min, 95% in 2.0 min,
46
47 700 μL/min flow rate, UV absorbance detection at 220 and 254 nm. Analyte ions were detected
48
49 by mass spectrometry in both negative and positive modes (110 – 800 amu scan range, API-ES
50
51 ionization). For some compounds an additional longer method was also used to assay chemical
53
54 purity: Agilent 1200 series high performance liquid chromatography (HPLC) system operating in

1
2
3 reverse-phase mode, water with 0.1% formic acid (mobile phase A), acetonitrile with 0.1%
4
5
6 formic acid (mobile phase B), Phenomenex Kinetex 2.6 μm C18 100 Å, 30×3.0 mm, 40 °C
7
8 column temperature, 5−95% mobile phase B in 12.0 min, 95% in 2.0 min, 800 μL/min flow rate,
9
10 UV absorbance detection at 214 and 254 nm.
12
13 Routine chromatographic purification was performed using Biotage Isolera One automated
14
15 systems running Biotage Isolera One 2.0.6 software (Biotage LLC, Charlotte, NC). Flow rates
16
17 were the default values specified for the particular column in use. Reverse phase chromatography
18
19
20 was performed using elution gradients of water and acetonitrile on KP-C18-HS Flash+ columns
21
22 (Biotage LLC) of various sizes. Normal phase chromatography was performed using elution
23
24 gradients of various solvents (e.g. hexanes, ethyl acetate, methylene chloride, methanol, acetone,
25
26 chloroform, MTBE, etc.). The columns were SNAP Cartridges containing KP-SIL (50 μm
28
29 irregular particles) or SNAP Ultra (25μm spherical particles) of various sizes (Biotage LLC).
30
31
32
33 The protocols and procedures involving the care and use of animals for this study were reviewed
35
36 and approved by the Institutional Animal Care and Use Committee (IACUC) of AAALAC-
37
38 certified animal facilities.
43 All patients provided written informed consent. The study protocol was approved by institutional
44
45 review boards at all participating institutions. The study (NCT02974738) was conducted in
46
47 accordance with good clinical practice and the Declaration of Helsinki.
Details concerning the SPA assay, luciferase assay, VEGFA secretion assay, in vitro microsomal
53
54 stability assay, plasma protein binding assay, permeability assay, CYP inhibition assessment, in

vivo pharmacokinetic experiments, PK/PD study and in vivo efficacy study were described
4
5 previously.27
7
8 3-[(1S,2S,3R)-2,3-Difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-5-fluoro-
9
10 benzonitrile (2). Step A: Preparation of [(1S,2R)-4-(3-cyano-5-fluoro-phenoxy)-2-fluoro-7-
11
12 methylsulfonyl-indan-1-yl] acetate (15a): To a stirred solution of 3-fluoro-5-[(1S,2R)-2-fluoro-1-
14
15 hydroxy-7-methylsulfonyl-indan-4-yl]oxy-benzonitrile (7) (2.00 g, 5.47 mmol) in DCM (27 mL)
16
17 was added 4-(dimethylamino)pyridine (0.200 g, 1.64 mmol) and triethylamine (1.53 mL, 10.9
18
19 mmol). Acetic anhydride (1.00 mL, 10.9 mmol) was added dropwise at 0 °C under nitrogen. The
20
21
22 reaction mixture was stirred at ambient temperature overnight. The reaction mixture was diluted
23
24 with DCM, washed with saturated aqueous NaHCO3 and brine, dried and concentrated. The
25
26 residue was purified by flash chromatography on silica gel (20-40% EtOAc/hexane) to give
27
28 [(1S,2R)-4-(3-cyano-5-fluoro-phenoxy)-2-fluoro-7-methylsulfonyl-indan-1-yl] acetate (1.95 g,
30
31 87%). LCMS ESI (+) m/z 408 (M+H).
32
33 Step B: Preparation of [(1S,2S)-3-bromo-4-(3-cyano-5-fluoro-phenoxy)-2-fluoro-7-
34
35 methylsulfonyl-indan-1-yl] acetate (16a): To a stirred solution of [(1S,2R)-4-(3-cyano-5-fluoro-
37
38 phenoxy)-2-fluoro-7-methylsulfonyl-indan-1-yl] acetate (15a) (1.95 g, 4.79 mmol) in 1,2-
39
40 dichloroethane (24 mL) was added N-bromosuccinimide (NBS) (0.94 g, 5.27 mmol) and 2,2′-
41
42 azobisisobutyronitrile (AIBN) (8 mg, 0.05 mmol). The reaction mixture was heated at 80 °C for
43
44
45 3 hours. After cooling, the reaction mixture was diluted with DCM, washed with saturated
46
47 aqueous NaHCO3 and brine, dried and concentrated. The residue was purified by column
48
49 chromatography on silica gel (20-30% EtOAc/hexane) to give [(1S,2S)-3-bromo-4-(3-cyano-5-
50
51 fluoro-phenoxy)-2-fluoro-7-methylsulfonyl-indan-1-yl] acetate (2.10 g, 90%). LCMS ESI (+)
53
54 m/z 486, 488 (M+H).

3 Step C: Preparation of [(1S,2R,3S)-4-(3-cyano-5-fluoro-phenoxy)-2-fluoro-3-hydroxy-7-
4
5 methylsulfonyl-indan-1-yl] acetate (17a) and [(1S,2R,3R)-4-(3-cyano-5-fluoro-phenoxy)-2-
7
8 fluoro-3-hydroxy-7-methylsulfonyl-indan-1-yl] acetate (17b): To stirred solution of [(1S,2S)-3-
9
10 bromo-4-(3-cyano-5-fluoro-phenoxy)-2-fluoro-7-methylsulfonyl-indan-1-yl] acetate (16a) (2.05
11
12 g, 4.22 mmol) in 1,2-dimethoxyethane (28 mL) and water (0.050 mL) was added silver
14
15 perchlorate hydrate (1.42 g, 6.32 mmol). The reaction mixture was heated at 70 °C for 2 hours.
16
17 After cooling, the reaction mixture was diluted with EtOAc and filtered through Celite. The
18
19 filtrate was washed with water and brine, dried and concentrated. The residue was purified by
20
21
22 flash chromatography on silica gel (20-50%) to give [(1S,2R,3S)-4-(3-cyano-5-fluoro-phenoxy)-
23
24 2-fluoro-3-hydroxy-7-methylsulfonyl-indan-1-yl] acetate (0.416 g, 23%) as the less polar
25
26 product. LCMS ESI (+) m/z 441 (M+NH +). Further elution with 60% EtOAc/hexane gave
27
28 [(1S,2R,3R)-4-(3-cyano-5-fluoro-phenoxy)-2-fluoro-3-hydroxy-7-methylsulfonyl-indan-1-yl]
30
31 acetate (0.580 g, 32 %). LCMS ESI (+) m/z 441 (M+NH +).
32
33 Step D: Preparation of [(1S,2S,3R)-4-(3-cyano-5-fluoro-phenoxy)-2,3-difluoro-7-methylsulfonyl-
34
35 indan-1-yl] acetate (18a): To a stirred solution of [(1S,2R,3S)-4-(3-cyano-5-fluoro-phenoxy)-2-
37
38 fluoro-3-hydroxy-7-methylsulfonyl-indan-1-yl] acetate (17a) (416 mg, 0.980 mmol) in DCM (10
39
40 mL) was added (diethylamino)sulfur trifluoride (DAST) (0.26 mL, 2.0 mmol) at -78 °C under
41
42 nitrogen. The reaction mixture was allowed to warm to 0 °C and stirred for 15 minutes. The
43
44
45 reaction was quenched by saturated aqueous NaHCO3. The mixture was partitioned between
46
47 EtOAc and water. The aqueous layer was extracted with EtOAc. The combined organic layers
48
49 were washed with brine, dried and concentrated. The residue was purified by flash
50
51 chromatography on silica gel (20-40% EtOAc/hexane) to give [(1S,2S,3R)-4-(3-cyano-5-fluoro-
53

phenoxy)-2,3-difluoro-7-methylsulfonyl-indan-1-yl] acetate (310 mg, 74%). LCMS ESI (+) m/z
4
5 426 (M+H).
7
8 Step E: Preparation of 3-[(1S,2S,3R)-2,3-difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-5-
9
10 fluoro-benzonitrile (2): To a stirred solution of [(1S,2S,3R)-4-(3-cyano-5-fluoro-phenoxy)-2,3-
11
12 difluoro-7-methylsulfonyl-indan-1-yl] acetate (18a) (310 mg, 0.730 mmol) in tetrahydrofuran (9
14
15 mL) was added 0.5 N LiOH solution (2.2 mL, 1.1 mmol) at 0 °C under nitrogen. The reaction
16
17 mixture was allowed to warm to ambient temperature and stirred for 4 hours. The reaction was
18
19 partitioned between EtOAc and water. The aqueous layer was extracted with EtOAc. The
20
21
22 combined organic layers were washed with water and brine, dried and concentrated. The residue
23
24 was purified by flash chromatography on silica gel (30-50% EtOAc/hexane) to give 3-
25
26 (1S,2S,3R)-2,3-difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-5-fluoro-benzonitrile (120
27
28
29 mg, 43% yield). LCMS ESI (+) m/z 384 (M+H); 1H NMR (400 MHz, CDCl3): δ 8.13 (d, 1H),
30
31 7.31-7.25 (m, 1H), 7.23-7.19 (m, 1H), 7.14-7.09 (m, 1H), 7.04 (d, 1H), 6.09-5.91 (m, 1H), 5.87-
32
33 5.80 (m, 1H), 5.25-5.05 (m, 1H), 3.32 (s, 3H), 2.95 (d, 1H).
34
35 General Procedure for Preparation of Glucuronide Metabolites (Method A).
37
38 (2S,3S,4S,5R,6S)-6-(((S)-4-(3-Cyano-5-fluorophenoxy)-2,2-difluoro-7-(methylsulfonyl)-
39
40 2,3-dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid (3,
41
42 PT2639). Step A: Preparation of (2S,3S,4S,5R,6R)-2-(methoxycarbonyl)-6-(2,2,2-trichloro-1-
43
44
45 iminoethoxy)tetrahydro-2H-pyran-3,4,5-triyl triacetate: DCM (13 mL) was dried over powdered
46
47 4 Å molecular sieves overnight and then added to 2,3,4-tri-O-acetyl--D-glucuronide methyl
48
49 ester trichloroacetimidate (1.50 g, 3.13 mmol) and 3-[(1S)-2,2-difluoro-1-hydroxy-7-
51
52 methylsulfonyl-indan-4-yl]oxy-5-fluoro-benzonitrile (1) (1.00 g, 2.61 mmol) under nitrogen. The
53
54 reaction mixture was cooled to 4 C. Boron trifluoride diethyl etherate (322 μL, 2.61 mmol) was

3 added dropwise. The reaction mixture was allowed to warm to ambient temperature and stirred
4
5 for 20 hours. The reaction was concentrated in vacuo. The residue was purified by reverse phase
7
8 chromatography with a gradient of 20% to 80% acetonitrile in water to afford ((2S,3S,4S,5R,6R)-
9
10 2-(methoxycarbonyl)-6-(2,2,2-trichloro-1-iminoethoxy)tetrahydro-2H-pyran-3,4,5-triyl triacetate
11
12 (1.50 g, 82% yield). LCMS ESI (–) m/z 698 (M–H).
14
15 Step B: Preparation of (2S,3S,4S,5R,6S)-6-(((S)-4-(3-cyano-5-fluorophenoxy)-2,2-difluoro-7-
16
17 (methylsulfonyl)-2,3-dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-
18
19 carboxylic acid (3): To a stirred solution of (2S,3S,4S,5R,6R)-2-(methoxycarbonyl)-6-(2,2,2-
20
21
22 trichloro-1-iminoethoxy)tetrahydro-2H-pyran-3,4,5-triyl triacetate (1.50 g, 2.14 mmol) in THF
23
24 (20 mL) was added an aqueous solution of lithium hydroxide monohydrate (0.5 M, 21.4 mL,
25
26 10.7 mmol) at 0 C under nitrogen by syringe. The reaction mixture was stirred at 0 C for 2
27
28
29 hours. The reaction was diluted with water and extracted with MTBE. The aqueous layer was
30
31 acidified with 1 N HCl (20 mL) and extracted with EtOAc. The combined organic layers were
32
33 washed with brine, dried over MgSO4, filtered and concentrated in vacuo. The residue was
34
35 purified by reverse phase chromatography with a gradient of 20% to 80% acetonitrile in water to
37
38 afford (2S,3S,4S,5R,6S)-6-(((S)-4-(3-cyano-5-fluorophenoxy)-2,2-difluoro- 7-(methylsulfonyl)-
39
40 2,3-dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid (0.80 g,
41
42 67% yield). HPLC retention time: 1.28 minutes; LCMS ESI (–) m/z 558 (M–H); 1H NMR (400
44
45 MHz, CD3OD): δ 7.96 (d, 1H), 7.46–7.49 (m, 1H), 7.38 (brs, 1H), 7.30 (dt, 1H), 7.17 (d, 1H),
46
47 5.87 (d, 1H), 3.89 (d, 1H), 3.18–3.80 (m, 9H); 19F NMR (376 MHz, CD3OD): δ -103.6 – -
48
49 102.8 (m, 1F), -108.5 (m, 1F), -117.2 – -116.5 (m, 1F).
51
52 Isolation and Characterization of PT2639 from Dog Urine Samples. Six urine samples
53
54 collected following single p.o. administration of 300 mg/kg or 600 mg/kg PT2385 in male beagle

3 dogs were combined. The combined urine was adjusted to pH 1 with 1 N HCl. The urine was
4
5 extracted three times with EtOAc. The combined organic layers were washed with brine, dried
7
8 over Na2SO4, filtered and concentrated in vacuo. The residue was purified by reverse phase
9
10 chromatography with a gradient of 20% to 80% acetonitrile in water to give a beige solid. HPLC
11
12 retention time: 1.28 minutes; LCMS ESI (–) m/z 558 (M–H); 1H NMR (400 MHz, CD3OD): δ
14
15 7.96 (d, 1H), 7.46–7.49 (m, 1H), 7.39 (brs, 1H), 7.30 (dt, 1H), 7.17 (d, 1H), 5.88 (d, 1H), 3.88 (d,
16
17 1H), 3.18–3.70 (m, 9H); 19F NMR (376 MHz, CD3OD): δ -103.6 – -102.8 (m, 1F), -108.5 (m,
18
19 1F), -117.1 – -116.4 (m, 1F). LC-MS and NMR spectroscopic results are in agreement with the
20
21
22 chemically synthesized sample.
23
24 3-Fluoro-5-[(1R)-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-benzonitrile (6). Formic acid
25
26 (1.64 mL, 43.4 mmol) was added slowly to a solution of triethylamine (4.04 mL, 29.0 mmol) in
27
28 dichloromethane (58 mL) at 0 °C under nitrogen. Solid 3-fluoro-5-(7-methylsulfonyl-1-oxo-
30
31 indan-4-yl)oxy-benzonitrile (14)27 (5.00 g, 14.5 mmol) was then added followed by the addition
32
33 of RuCl(p-cymene)[(R,R)-Ts-DPEN] (0.092 g, 0.15 mmol) under nitrogen. The flask was then
34
35 equipped with septa and a limp balloon to relieve the pressure and the reaction mixture was then
37
38 placed directly in a 4 °C refrigerator overnight with intermittent swirling and venting. The
39
40 reaction was allowed to warm to rt and stirred overnight. The reaction mixture was then diluted
41
42 with dichloromethane, washed with saturate aqueous NaHCO and brine, dried and concentrated.
43
44
45 The residue was purified by flash chromatography on silica gel (20-50% EtOAc/hexane) to give
46
47 3-fluoro-5-[(1R)-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-benzonitrile (4.18 g, 83% yield).
48
49 LCMS ESI (-) m/z 392 (M+ HCO -); 1H NMR (400 MHz, CDCl ): δ 7.83 (d, 1H), 7.19-7.16 (m,
50 2 3
51 1H), 7.09-7.07 (m, 1H), 7.01-6.96 (m, 2H), 5.71-5.67 (m, 1H), 3.64 (d, 1H), 3.21 (s, 3H), 3.12-
53
54 3.02 (m, 1H), 2.84-2.75 (m, 1H), 2.52-2.42 (m, 1H), 2.27-2.18 (m, 1H).

3 3-Fluoro-5-[(1R,3S)-3-fluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-benzonitrile (9).
4
5 Step A: Preparation of [(1R)-4-(3-cyano-5-fluoro-phenoxy)-3-hydroxy-7-methylsulfonyl-indan-1-
7
8 yl] acetate (15b): To a stirred solution of 3-fluoro-5-[(1R)-1-hydroxy-7-methylsulfonyl-indan-4-
9
10 yl]oxy-benzonitrile (6) (1.05 g, 3.00 mmol) in DCM (29 mL) was added 4-
11
12 (dimethylamino)pyridine (0.369 g, 3.00 mmol) and triethylamine (0.84 mL, 6.1 mmol). Acetyl
14
15 chloride (0.43 mL, 6.1 mmol) was added dropwise at 0 °C under nitrogen. The reaction mixture
16
17 was stirred at ambient temperature for 2 hours. The reaction mixture was diluted with DCM,
18
19 washed with saturated aqueous NaHCO and brine, dried and concentrated. The residue was
20
21
22 purified by flash chromatography on silica gel (20-50% EtOAc/hexane) to give [(1R)-4-(3-
23
24 cyano-5-fluoro-phenoxy)-7-methylsulfonyl-indan-1-yl] acetate (0.72 g, 61%). LCMS ESI (-) m/z
25
26 434 (M+ HCO -).
27
28 Step B: Preparation of [(1R)-3-bromo-4-(3-cyano-5-fluoro-phenoxy)-7-methylsulfonyl-indan-1-
30
31 yl] acetate (16b): To a stirred solution of [(1R)-4-(3-cyano-5-fluoro-phenoxy)-7-methylsulfonyl-
32
33 indan-1-yl] acetate (15b) (720 mg, 1.85 mmol) in carbon tetrachloride (18 mL) was added N-
34
35 bromosuccinimide (NBS) (362 mg, 2.00 mmol) and 2,2′-azobisisobutyronitrile (AIBN) (3 mg,
37
38 0.02 mmol). The reaction mixture was heated at 80 °C for 2 hours. After cooling, the reaction
39
40 mixture was diluted with DCM, washed with saturated aqueous NaHCO3 and brine, dried and
41
42 concentrated. The residue was purified by column chromatography on silica gel (10-40%
43
44
45 EtOAc/hexanes) to give [(1R)-3-bromo-4-(3-cyano-5-fluoro-phenoxy)-7-methylsulfonyl-indan-
46
47 1-yl] acetate (874 mg, 100%). LCMS ESI (-) m/z: 512, 514 (M+ HCO -).
48
49 Step C: Preparation of [(1R)-4-(3-cyano-5-fluoro-phenoxy)-3-hydroxy-7-methylsulfonyl-indan-
50
51 1-yl] acetate (17c): To a stirred solution of [(1R)-3-bromo-4-(3-cyano-5-fluoro-phenoxy)-7-
53
54 methylsulfonyl-indan-1-yl] acetate (16b) (423 mg, 0.906 mmol) in 1,2-dimethoxyethane (5 mL)

and water (2 mL) was added silver carbonate (374 mg, 1.35 mmol). The reaction mixture was
4
5 stirred at ambient temperature overnight. The mixture was diluted with EtOAc and filtered
7
8 through Celite. The filtrate was washed with water and brine, dried and concentrated. The crude
9
10 was used in the next step without further purification. LCMS ESI (-) m/z 450 (M+ HCO -).
11
12 Step D: Preparation of [(1R,3S)-4-(3-cyano-5-fluoro-phenoxy)-3-fluoro-7-methylsulfonyl-indan-
14
15 1-yl] acetate (18b) and [(1R,3R)-4-(3-cyano-5-fluoro-phenoxy)-3-fluoro-7-methylsulfonyl-indan-
16
17 1-yl] acetate (18c): To a stirred solution of [(1R)-4-(3-cyano-5-fluoro-phenoxy)-3-hydroxy-7-
18
19 methylsulfonyl-indan-1-yl] acetate (17c) (306 mg, 0.750 mmol) in DCM (8 mL) was added
20
21
22 (diethylamino)sulfur trifluoride (DAST) (0.20 mL, 1.5 mmol) at 0 °C under nitrogen. The
23
24 reaction mixture was stirred at 0 °C for 30 minutes. The reaction was quenched by the addition
25
26 of saturated aqueous NaHCO3. The mixture was partitioned between EtOAc and water. The
27
28 aqueous layer was extracted with EtOAc. The combined organic layers were washed with brine,
30
31 dried and concentrated. The residue was purified by column chromatography on silica gel (20-
32
33 40% EtOAc/hexane) to give [(1R,3S)-4-(3-cyano-5-fluoro-phenoxy)-3-fluoro-7-methylsulfonyl-
34
35 indan-1-yl] acetate (144 mg, 47%) and [(1R,3R)-4-(3-cyano-5-fluoro-phenoxy)-3-fluoro-7-
37
38 methylsulfonyl-indan-1-yl] acetate (82 mg, 27%). LCMS ESI (+) m/z 425 (M+NH +).
39
40 Step E: Preparation of 3-fluoro-5-[(1R,3S)-3-fluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-
41
42 benzonitrile (9): To a stirred solution of [(1R,3S)-4-(3-cyano-5-fluoro-phenoxy)-3-fluoro-7-
43
44
45 methylsulfonyl-indan-1-yl] acetate (18b) (144 mg, 0.350 mmol) in tetrahydrofuran (3.5 mL) was
46
47 added 0.5 N LiOH solution (1.41 mL, 0.710 mmol) at 0 °C under nitrogen. The reaction mixture
48
49 was stirred at rt overnight. The reaction was partitioned between EtOAc and water. The aqueous
50
51 layer was extracted with EtOAc. The combined organic layers were washed with water and
53
54 brine, dried and concentrated. The residue was purified by column chromatography on silica gel

(30-60% EtOAc/hexane) to give 3-fluoro-5-[(1R,3S)-3-fluoro-1-hydroxy-7-methylsulfonyl-
4
5 indan-4-yl]oxy-benzonitrile (91 mg, 70% yield). LCMS ESI (+) m/z 383 (M+NH +); 1H NMR
7
8 (400 MHz, CDCl3): δ 8.04-8.01 (m, 1H), 7.25-7.22 (m, 1H), 7.18-7.16 (m, 1H), 7.11-7.06 (m,
9
10 1H), 7.00 (d, 1H), 6.09-5.79 (m, 1H), 5.69-5.61 (m, 1H), 3.54 (d, 1H), 3.23 (s, 3H), 2.94-2.80
11
12 (m, 1H), 2.52-2.41 (m, 1H).
14
15 3-Fluoro-5-[(1R,3R)-3-fluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-benzonitrile (10).
16
17 10 was prepared from 18c by the same procedure described for 9. LCMS ESI (+) m/z 383
18
19 (M+NH +); 1H NMR (400 MHz, CDCl ): δ 8.00 (dd, 1H), 7.25-7.22 (m, 1H), 7.18-7.16 (m, 1H),
20 4 3
21
22 7.09-7.06 (m, 1H), 6.99 (d, 1H), 6.35-6.18 (m, 1H), 5.95-5.88 (m, 1H), 4.02 (d, 1H), 3.21 (s,
23
24 3H), 2.89-2.73 (m, 1H), 2.58-2.43 (m, 1H).
25
26 3-[(1S,2S,3S)-2,3-Difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-5-fluoro-
27
28 benzonitrile (11). Step A: Preparation of [(1S,2S,3S)-4-(3-cyano-5-fluoro-phenoxy)-2,3-
30
31 difluoro-7-methylsulfonyl-indan-1-yl] acetate (18d): To a stirred solution of [(1S,2R,3R)-4-(3-
32
33 cyano-5-fluoro-phenoxy)-2-fluoro-3-hydroxy-7-methylsulfonyl-indan-1-yl] acetate (17b) (28
34
35 mg, 0.070 mmol) in dichloromethane (0.7 mL) was added (diethylamino)sulfur trifluoride
37
38 (DAST) (0.018m L, 0.13 mmol) at 0 °C under nitrogen. The reaction mixture was stirred at 0 °C
39
40 for 30 minutes. The reaction was quenched by sat. aq. NaHCO3. The mixture was partioned
41
42 between EtOAc and water. The reaction was quenched by saturated aqueous NaHCO . The
43
44
45 mixture was partitioned between EtOAc and water. The aqueous layer was extracted with
46
47 EtOAc. The combined organic layers were washed with brine, dried and concentrated. The
48
49 residue was purified by flash chromatography on silica gel (20-40% EtOAc/hexane) to give
50
51 [(1S,2S,3S)-4-(3-cyano-5-fluoro-phenoxy)-2,3-difluoro-7-methylsulfonyl-indan-1-yl] acetate (20
53
54 mg, 71%). LCMS ESI (+) m/z 426 (M+H).

3 Step B: Preparation of 3-[(1S,2S,3S)-2,3-difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-5-
4
5 fluoro-benzonitrile (11): To a stirred solution of [(1S,2S,3S)-4-(3-cyano-5-fluoro-phenoxy)-2,3-
7
8 difluoro-7-methylsulfonyl-indan-1-yl] acetate (18d) (20 mg, 0.047 mmol) in tetrahydrofuran (0.5
9
10 mL) was added 0.5 N LiOH solution (0.19 mL, 0.094 mmol) at 0 °C under nitrogen. The
11
12 reaction mixture was allowed to warm to ambient temperature and stirred for 1 hour. The
14
15 reaction was partitioned between EtOAc and water. The aqueous layer was extracted with
16
17 EtOAc. The combined organic layers were washed with water and brine, dried and concentrated.
18
19 The residue was purified by C18 reversed-phase flash chromatography (20–80%
20
21
22 acetonitrile/water gradient) to give 3-(1S,2S,3S)-2,3-difluoro-1-hydroxy-7-methylsulfonyl-indan-
23
24 4-yl]oxy-5-fluoro-benzonitrile (12 mg, 67% yield). LCMS ESI (+) m/z 384 (M+H); 1H NMR
25
26 (400 MHz, CDCl3): δ 8.04-8.01 (m, 1H), 7.25-7.22 (m, 1H), 7.18-7.16 (m, 1H), 7.11-7.06 (m,
27
28 1H), 7.00 (d, 1H), 6.09-5.79 (m, 1H), 5.69-5.61 (m, 1H), 3.54 (d, 1H), 3.23 (s, 3H), 2.94-2.80
30
31 (m, 1H), 2.52-2.41 (m, 1H).
32
33 3-[(1R)-3,3-Difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-5-fluoro-benzonitrile (12).
34
35 Step A: Preparation of [(1R)-4-(3-cyano-5-fluoro-phenoxy)-7-methylsulfonyl-3-oxo-indan-1-yl]
37
38 acetate (19): To a stirred solution of [(1R,3S)-4-(3-cyano-5-fluoro-phenoxy)-3-hydroxy-7-
39
40 methylsulfonyl-indan-1-yl] acetate (17c) (366 mg, 0.904 mmol) in DCM (9 mL) was added
41
42 Dess-Martin periodinane (574 mg, 1.35 mmol). The reaction mixture was stirred at ambient
43
44
45 temperature for 1 hour. The reaction mixture was partitioned between EtOAc and saturated
46
47 aqueous NaHCO3. The aqueous layer was extracted with EtOAc. The combined organic layers
48
49 were washed with water and brine, dried and concentrated. The crude was purified by flash
50
51 chromatography on silica gel (10-50% EtOAc/hexane) to give [(1R)-4-(3-cyano-5-fluoro-
53
54

3 phenoxy)-7-methylsulfonyl-3-oxo-indan-1-yl] acetate (320 mg, 88%). LCMS ESI (-) m/z 402
4
5 (M-H).
7
8 Step B: Preparation of [(1R)-4-(3-cyano-5-fluoro-phenoxy)-3,3-difluoro-7-methylsulfonyl-indan-
9
10 1-yl] acetate: To a plastic tube containing [(1R)-4-(3-cyano-5-fluoro-phenoxy)-7-
11
12 methylsulfonyl-3-oxo-indan-1-yl] acetate (19) (109 mg, 0.270 mmol) and DCM (1.2 mL) was
14
15 added 4-(tert-butyl)-2,6-dimethylphenyl sulfur trifluoride (115 mg, 0.460 mmol) under
16
17 nitrogen. Hydrogen fluoride pyridine (70%, 0.020 mL, 0.27 mmol) was added, and the mixture
18
19 was stirred at ambient temperature for 4 hours. The solvent was removed under reduced pressure.
20
21
22 The residue was taken up in EtOAc, washed with saturated aqueous NaHCO3 and brine, dried
23
24 and concentrated. The residue was purified by flash chromatography on silica gel (10-50%
25
26 EtOAc/hexane) to give [(1R)-4-(3-cyano-5-fluoro-phenoxy)-3,3-difluoro-7-methylsulfonyl-
27
28 indan-1-yl] acetate (97 mg, 84%). LCMS ESI (+) m/z 426 (M+H).
30
31 Step C: Preparation of [(1R)-3,3-difluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-5-fluoro-
32
33 benzonitrile (12): To a stirred solution of [(1R)-4-(3-cyano-5-fluoro-phenoxy)-3,3-difluoro-7-
34
35 methylsulfonyl-indan-1-yl] acetate (19) (97 mg, 0.23 mmol) in tetrahydrofuran (1.5 mL) was
37
38 added 0.5 N LiOH solution (0.68 mL, 0.34 mmol) at 0 °C under nitrogen. The reaction mixture
39
40 was stirred at 0 °C for 1 hour. The reaction was then partitioned between EtOAc and water. The
41
42 aqueous layer was extracted with EtOAc. The combined organic layers were washed with water
43
44
45 and brine, dried and concentrated. The residue was purified by flash chromatography on silica
46
47 gel (30-70% EtOAc/hexane) to give [(1R)-3,3-difluoro-1-hydroxy-7-methylsulfonyl-indan-4-
48
49 yl]oxy-5-fluoro-benzonitrile (75 mg, 86%). LCMS ESI (-) m/z 428 (M+ HCO -); 1H NMR (400
50
51
52 MHz, CDCl3): δ 8.08 (d, 1H), 7.29-7.23 (m, 1H), 7.19 (brs, 1H), 7.15-7.08 (m, 1H), 7.02 (d, 1H),
53
54 5.78-5.70 (m, 1H), 3.89 (d, 1H), 3.23 (s, 3H), 3.17-3.02 (m, 1H), 2.80-2.64 (m, 1H).

1
2
3 3-Fluoro-5-[(1S,3R)-2,2,3-trifluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-benzonitrile
4
5 (13). Step A: Preparation of [(1S)-4-(3-cyano-5-fluoro-phenoxy)-2,2-difluoro-7-methylsulfonyl-
7
8 indan-1-yl] acetate (15c): To a stirred solution of 3-[(1S)-2,2-difluoro-1-hydroxy-7-
9
10 methylsulfonyl-indan-4-yl]oxy-5-fluoro-benzonitrile (1) (2.00 g, 5.22 mmol) in dichloromethane
11
12 (26 mL) was added 4-(dimethylamino)pyridine (0.19 g, 1.6 mmol) and triethylamine (1.45 mL,
14
15 10.4 mmol). Acetic anhydride (0.99 mL,10.4 mmol) was added dropwise at 0 °C under nitrogen.
16
17 The reaction mixture was stirred at rt overnight. The reaction mixture was diluted with DCM,
18
19 washed with saturated aqueous NaHCO and brine, dried and concentrated. The residue was
20
21
22 purified by flash chromatography on silica gel (20-50% EtOAc/hexanes) to give [(1S)-4-(3-
23
24 cyano-5-fluoro-phenoxy)-2,2-difluoro-7-methylsulfonyl-indan-1-yl] acetate (2.13 g, 96% yield).
25
26 LCMS ESI (+) m/z 426 (M+H).
27
28 Step B: Preparation of [(1S,3S)-4-(3-cyano-5-fluoro-phenoxy)-2,2-difluoro-3-hydroxy-7-
30
31 methylsulfonyl-indan-1-yl] acetate (17d): To a stirred solution of [(1S)-4-(3-cyano-5-fluoro-
32
33 phenoxy)-2,2-difluoro-7-methylsulfonyl-indan-1-yl] acetate (15c) (1.00 g, 2.35 mmol) in DCE
34
35 (24 mL) were added N-bromosuccinimide (NBS) (0.460 g, 2.59 mmol) and 2,2′-
37
38 azobisisobutyronitrile (AIBN) (4 mg, 0.02 mmol). The reaction mixture was heated at 80 C
39
40 overnight. After cooling, the reaction mixture was diluted with DCM, washed with saturated
41
42 aqueous NaHCO3 and brine, dried and concentrated. The crude product was dissolved in 1,2-
44
45 dimethoxyethane (11 mL) and water (0.11 mL). Silver perchlorate hydrate (0.350 g, 1.55 mmol)
46
47 was added. The reaction mixture was heated at 70 C overnight. After cooling, the reaction
48
49 mixture was diluted with EtOAc and filtered through Celite. The filtrate was washed with water
51
52 and brine, dried and concentrated. The residue was purified by flash chromatography on silica
53
54 gel (20-60% EtOAc/hexane) to give [(1S,3S)-4-(3-cyano-5-fluoro-phenoxy)-2,2-difluoro-3-

1
2
3 hydroxy-7-methylsulfonyl-indan-1-yl] acetate (39 mg, 9% yield). LCMS ESI (+) m/z 459
4
5 (M+NH +).
7
8 Step C: Preparation of [(1S,3R)-4-(3-cyano-5-fluoro-phenoxy)-2,2,3-trifluoro-7-methylsulfonyl-
9
10 indan-1-yl] acetate (18e): To a stirred solution of [(1S,3S)-4-(3-cyano-5-fluoro-phenoxy)-2,2-
11
12 difluoro-3-hydroxy-7-methylsulfonyl-indan-1-yl] acetate (17d) (39 mg, 0.090 mmol) in
14
15 dichloromethane (1 mL) was added (diethylamino)sulfur trifluoride (DAST) (0.023 mL, 0.18
16
17 mmol) at -78 °C under nitrogen. The reaction mixture was allowed to warm to 0 °C and stirred
18
19 for 15 minutes. The reaction was quenched by saturated aqueous NaHCO . The mixture was
20
21
22 partitioned between EtOAc and water. The aqueous layer was extracted with EtOAc. The
23
24 combined organic layers were washed with brine, dried and concentrated. The residue was
25
26 purified by flash chromatography on silica gel (20-40% EtOAc/hexanes) to give [(1S,3R)-4-(3-
27
28 cyano-5-fluoro-phenoxy)-2,2,3-trifluoro-7-methylsulfonyl-indan-1-yl] acetate (36 mg, 92%
30
31 yield). LCMS ESI (+) m/z 444 (M+H).
32
33 Step D: Preparation of 3-fluoro-5-[(1S,3R)-2,2,3-trifluoro-1-hydroxy-7-methylsulfonyl-indan-4-
34
35 yl]oxy-benzonitrile (13): To a stirred solution of [(1S,3R)-4-(3-cyano-5-fluoro-phenoxy)-2,2,3-
37
38 trifluoro-7-methylsulfonyl-indan-1-yl] acetate (18e) (19 mg, 0.040 mmol) in tetrahydrofuran (0.3
39
40 mL) was added 0.5 N LiOH solution (0.13 mL, 0.060 mmol) at 0 °C under nitrogen. The
41
42 reaction mixture was allowed to warm to ambient temperature and stirred for 2 hours. The
43
44
45 reaction was partitioned between EtOAc and water. The aqueous layer was extracted with
46
47 EtOAc. The combined organic layers were washed with water and brine, dried and concentrated.
48
49 The residue was purified by flash chromatography on silica gel (20-50% EtOAc/hexanes) to give
50
51 3-fluoro-5-[(1S,3R)-2,2,3-trifluoro-1-hydroxy-7-methylsulfonyl-indan-4-yl]oxy-benzonitrile (4
53
54 mg, 23% yield). LCMS ESI (+) m/z 419 (M+NH +); 1H NMR (400 MHz, CDCl ): δ 8.14-8.11
4 3

(m, 1H), 7.33-7.29 (m, 1H), 7.25-7.23 (m, 1H), 7.16-7.12 (m, 1H), 7.05 (d, 1H), 5.91-5.75 (m,
4
5 1H), 5.71-5.65 (m, 1H), 3.39 (d, 1H), 3.25 (s, 3H).
7
8 General Procedure for Preparation of Glucuronide Metabolites (Method B).
9
10 (2S,3S,4S,5R,6S)-6-(((S)-4-(3-Cyano-5-fluorophenoxy)-2,2-difluoro-7-(methylsulfonyl)-
11
12 2,3-dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid
14
15 (20, PT3317). Step A: Preparation of (2S,3R,4S,5S,6S)-2-(((1S,2S,3R)-4-(3-cyano-5-
16
17 fluorophenoxy)-2,3-difluoro-7-(methylsulfonyl)-2,3-dihydro-1H-inden-1-yl)oxy)-6-
18
19 (methoxycarbonyl)tetrahydro-2H-pyran-3,4,5-triyl triacetate: Dichloromethane (16 mL) was
20
21
22 dried over powdered 4 Å molecular sieves overnight and then added to 2,3,4-tri-O-acetyl-α-D-
23
24 glucuronide methyl ester trichloroacetimidate (1.12 g, 2.35 mmol) and 3-(((1S,2S,3R)-2,3-
25
26 difluoro-1-hydroxy-7-(methylsulfonyl)-2,3-dihydro-1H-inden-4-yl)oxy)-5-fluorobenzonitrile (2)
27
28 (0.60 g, 1.57 mmol) under nitrogen. The reaction mixture was cooled to -78 C. Trimethylsilyl
30
31 trifluoromethanesulfonate (424 μL, 2.35 mmol) was added dropwise. The reaction mixture was
32
33 stirred at -78 C for 30 minutes then the mixture was warmed to ambient temperature and stirred
35
36 for 7 hours. The reaction was quenched by addition of water, diluted with additional
37
38 dichloromethane then washed sequentially with saturated NaHCO3 and brine, dried over MgSO4,
39
40 filtered, and concentrated in vacuo. The residue was purified by chromatography on silica gel
41
42
43 (10-55% ethyl acetate / hexanes gradient) affording (2S,3R,4S,5S,6S)-2-(((1S,2S,3R)-4-(3-cyano-
44
45 5-fluorophenoxy)-2,3-difluoro-7-(methylsulfonyl)-2,3-dihydro-1H-inden-1-yl)oxy)-6-
46
47 (methoxycarbonyl)tetrahydro-2H-pyran-3,4,5-triyl triacetate (0.33 g, 30% yield). LCMS ESI (+)
48
49 m/z 700.2 (M+H).
51
3 Step B: Preparation of (2S,3S,4S,5R,6S)-6-(((1S,2S,3R)-4-(3-cyano-5-fluorophenoxy)-2,3-
4
5 difluoro-7-(methylsulfonyl)-2,3-dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-
7
8 pyran-2-carboxylic acid (20): A stirred solution of (2S,3R,4S,5S,6S)-2-(((1S,2S,3R)-4-(3-cyano-
9
10 5-fluorophenoxy)-2,3-difluoro-7-(methylsulfonyl)-2,3-dihydro-1H-inden-1-yl)oxy)-6-
11
12 (methoxycarbonyl)tetrahydro-2H-pyran-3,4,5-triyl triacetate (0.33 g, 0.47 mmol) dissolved
14
15 in THF (4.7 mL) was cooled to 0 C. To the solution was added an aqueous solution of lithium
16
17 hydroxide monohydrate (1.0 M, 2.36 mL, 2.36 mmol) under nitrogen. The reaction mixture was
18
19 stirred at 0 C for 3 hours. The reaction was diluted with water, adjusted to pH 4-5 with 0.1 N
21
22 aqueous HCl, and extracted twice with ethyl acetate. The combined organic layers were washed
23
24 with brine, dried over MgSO4, filtered and concentrated in vacuo. The residue was purified by
25
26 three sequential C18 reversed-phase flash chromatography runs (20–75% acetonitrile/water
28
29 gradient) affording (2S,3S,4S,5R,6S)-6-(((1S,2S,3R)-4-(3-cyano-5-fluorophenoxy)-2,3-difluoro-
30
31 7-(methylsulfonyl)-2,3-dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-
32
33 carboxylic acid (0.082 g, 31% yield). LCMS ESI (–) m/z 558 (M–H); 1H NMR (400 MHz,
35
36 CD3OD): δ 8.13 (dd, 1H), 7.52-7.49 (ddd, 1H), 7.44-7.42 (m, 1H), 7.36 (dt, 1H), 7.21 (d, 1H),
37
38 6.14 (dd, 1H), 6.03 (dd, 1H), 5.29 (m, 1H), 4.92 (d, 1H), 3.88 (d, 1H), 3.53 (t, 1H), 3.46 (t, 1H),
39
40 3.37 (s, 3H), 3.23-3.19 (m, 1H).
41
42
43 (2S,3S,4S)-6-(((R)-4-(3-Cyano-5-fluorophenoxy)-7-(methylsulfonyl)-2,3-dihydro-1H-inden-
44
45 1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid (21). 21 was prepared from
46
47 6 by the same procedure described for 3. LCMS ESI (–) m/z 522 (M–H); 1H NMR (400 MHz,
48
49 CD3OD): δ 7.87 (dd, 1H), 7.54-7.49 (m, 1H), 7.46-7.42 (m, 1H), 7.34-7.32 (m, 1H), 7.22 (d,
51
52 1H), 6.06 (d, 1H), 4.74 (d, 1H), 4.01 (d, 1H), 3.60 (t, 1H), 3.52 (t, 1H), 3.30 (s, 3H), 3.23-3.19
53
54 (m, 1H), 3.17-3.07 (m, 1H), 2.96-2.86 (m, 1H), 2.54-2.46 (m, 1H), 2.29-2.19 (m, 1H).

(2S,3S,4S)-6-(((1S,2R)-4-(3-Cyano-5-fluorophenoxy)-2-fluoro-7-(methylsulfonyl)-2,3-
4
5 dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid (22).
7
8 22 was prepared from 7 by the same procedure described for 3. LCMS ESI (–) m/z 540 (M–H);
9
10 1H NMR (400 MHz, CD3OD): δ 7.92 (d, 1H), 7.46-7.42 (m, 1H), 7.34-7.31 (m, 1H), 7.27-7.22
11
12 (m, 1H), 7.15 (d, 1H), 6.01-5.98 (m, 1H), 5.48-5.28 (m, 1H), 3.88 (d, 1H), 3.54 (t, 1H), 3.44 (t,
14
15 1H), 3.33 (s, 3H), 3.32-3.19 (m, 3H).
16
17 (2S,3S,4S)-6-(((1S,2S)-4-(3-Cyano-5-fluorophenoxy)-2-fluoro-7-(methylsulfonyl)-2,3-
18
19 dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid (23).
20
21
22 23 was prepared from 8 by the same procedure described for 3. LCMS ESI (–) m/z 540 (M–H);
23
24 1H NMR (400 MHz, acetone-D6): δ 7.93 (d, 1H), 7.59-7.50 (m, 2H), 7.48-7.40 (m, 1H), 7.27 (d,
25
26 1H), 6.03 (d, 1H), 5.64-5.45 (m, 3H), 4.91 (d, 1H), 3.93 (d, 1H), 3.66-3.44 (m, 3H), 3.30 (s, 3H),
27
28 3.28-3.22 (m, 1H), 3.20-3.06 (m, 1H).
30
31 (2S,3S,4S)-6-(((1R,3S)-4-(3-cyano-5-fluorophenoxy)-3-fluoro-7-(methylsulfonyl)-2,3-
32
33 dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid (24).
34
35 24 was prepared from 9 by the same procedure described for 3. LCMS ESI (–) m/z 540 (M–H);
37
38 1H NMR (400 MHz, acetone-D6): δ 8.09 (d, 1H), 7.58 (d, 1H), 7.54-7.51 (m, 1H), 7.47-7.42 (m,
39
40 1H), 7.31 (d, 1H), 6.23-6.19 (m, 1H), 6.10-6.00 (m, 1H), 4.81 (d, 1H), 4.04 (d, 1H), 3.63 (t, 1H),
41
42 3.54 (t, 1H), 3.38 (s, 3H), 3.34-3.27 (m, 1H), 2.80-2.40 (m, 2H).
43
44
45 (2S,3S,4S)-6-(((1S,3R)-4-(3-Cyano-5-fluorophenoxy)-2,2,3-trifluoro-7-(methylsulfonyl)-2,3-
46
47 dihydro-1H-inden-1-yl)oxy)-3,4,5-trihydroxytetrahydro-2H-pyran-2-carboxylic acid (25).
48
49 25 was prepared from 13 by the same procedure described for 20. LCMS ESI (–) m/z 576 (M–
50
51 H); 1H NMR (400 MHz, acetone-D6): δ 8.19 (d, 1H), 7.71-7.64 (m, 2H), 7.63-7.57 (m, 1H), 7.40

3 (d, 1H), 6.14-5.93 (m, 2H), 5.13 (d, 1H), 4.15 (d, 1H), 3.66 (t, 1H), 3.59 (t, 1H), 3.37 (s, 3H),
4
5 3.37-3.28 (m, 1H).
7
8 UGT Phenotyping
9
10
11 The test compound was incubated with a panel of individually-expressed recombinant human
12
13 UGT enzymes (UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9,
14
15 UGT1A10, UGT2B7, UGT2B15 and UGT2B17) expressed in baculovirus-infected insect cell
17
18 membranes (Corning UGT Supersomes). The incubation mixture contained the test compound at
19
20 a final concentration of 50 μM, expressed UGT (0.2 mg protein/ml), Tris–HCl buffer (pH 7.5),
21
22 magnesium chloride (10 mM), alamethicin (25 μg/ml) and UDPGA (2 mM). The mixture
23
24
25 (without the UDPGA cofactor) was pre-incubated at 37°C for 5 min, after which the reaction was
26
27 started by the addition of UDPGA.
28
29
30 Incubations were performed at 37°C. Aliquots (100 μl) were collected at 60 min and quenched
31
32 with two volumes of ice-cold acetonitrile containing an internal standard. The samples were
33
34 centrifuged at 10,000×g at room temperature for 15 min to pellet precipitated proteins. The
36
37 supernatants were then transferred to clean vials containing 200 μl water and analyzed using
38
39 liquid chromatography-tandem mass spectrometry (LC-MS/MS). The glucuronide metabolite
40
41 over internal standard peak area ratio was measured to quantify the glucuronide metabolite
43
44 formation rate.
45
46
47 Enzyme Kinetic Studies in Human Intestine Microsomes (HIM) and UGT2B17 Supersomes
48
49
50 Similar experimental procedure to UGT phenotyping was used for enzyme kinetics studies of
51
52 compounds in human intestinal microsomes and UGT2B17 Supersomes, except that a compound
53
54 concentration range of 2 to 1000 μM was used in the incubations. The absolute glucuronide

1
2
3 metabolite formation rate was measured by LC-MS/MS. The compound concentrations and its
4
5 corresponding glucuronide metabolite formation rates were fitted to the standard Michaelis-
7
8 Menten model to obtain the Vmax and Km. by using Prism (version 6).
9
10
11 Supporting Information
12
13 Additional details concerning human MRP2 mediated transport assay of 3, human MRP2
14
15 transport kinetics study of 3, human P-gp and BCRP transporter studies of 1 and 3, assessment of
17
18 3 in the -glucuronidase hydrolysis assay, hERG profiling of 2 and safety panel screen of 2 are
19
20 available.
21
22 Molecular formula strings with in vitro activity data (CSV)
24
25 AUTHOR INFORMATION
26
27
28 Corresponding Author
29
30
31 *E-mail: [email protected]. Phone: 972-629-4033
32
33
34 Present Addresses:
35
36 ‡ Edgewise Therapeutics, Inc., 3415 Colorado Ave., JSCBB 251E, Boulder, CO 80303
38
39  Gilead Sciences, Inc., 333 lakeside Drive, Foster City, CA 94404
40
41 Notes
42
43 The authors declare no competing financial interests. All authors are employees or former
45
46 employees of Peloton Therapeutics.
47
48
49
50 ABBREVIATIONS USED
51
52
53 ARNT, aryl hydrocarbon receptor nuclear translocator; AUC, area under the concentration-time
54
55 curve; ccRCC, clear cell renal cell carcinoma; CR, complete response; EPO, erythropoietin; HIF,

hypoxia-inducible factor; HIM, human intestine microsomes; HLM, human liver microsomes;
4
5 HRE, hypoxia-responsive element; MTD, maximum tolerated dose; PFS, progression free
7
8 survival; PHD, prolyl-hydroxylase; PR, partial response; pVHL, von Hippel-Lindau protein;
9
10 RP2D, recommended phase 2 dose; SD, stable disease; SPA, scintillation proximity assay;
11
12 UDPGA, UDP--D-glucuronic acid; UGT, uridine 5′-diphospho-glucuronosyltransferase;
14
15 VEGFA, vascular endothelial growth factor A; VHL, von Hippel-Lindau.

REFERENCES

22 (1) American Cancer Society. https://www.cancer.org/cancer/kidney-cancer/about/key-
23
24 statistics.html (accessed April 17, 2019).
25
26 (2) Choueiri, T.K.; Motzer, R.J. Systemic therapy for metastatic renal-cell carcinoma. N. Engl.
28
29 J. Med. 2017, 376, 354-366.
30
31 (3) National Cancer Institute. https://seer.cancer.gov/statfacts/html/kidrp.html (accessed April
32
33 17, 2019).
34
35 (4) Sato, Y.; Yoshizato, T.; Shiraishi, Y.; Maekawa, S.; Okuno, Y.; Kamura, T.; Shimamura,
37
38 T.; Sato-Otsubo, A.; Nagae, G.; Suzuki, H.; Nagata, Y.; Yoshida, K.; Kon, A.; Suzuki, Y.;
39
40 Chiba, K.; Tanaka, H.; Niida, A.; Fujimoto, A.; Tsunoda, T.; Morikawa, T.; Maeda, D.; Kume,
41
42 H.; Sugano, S.; Fukayama, M.; Aburatani, H.; Sanada, M.; Miyano, S.; Homma, Y.; Ogawa, S.
44
45 Integrated molecular analysis of clear-cell renal cell carcinoma. Nat. Genet. 2013, 45, 860-867.
46
47 (5) Kaelin, W.G., Jr. Molecular Biology of Kidney Cancer. In Kidney Cancer: Principles and
48
49 Practice, 2nd ed.; Lara, P.N.; Jonasch, E., Eds.; Springer: Cham, 2015; pp 31-57.
51
52 (6) Shen C.; Kaelin, W.G., Jr. The VHL/HIF axis in clear cell renal cell carcinoma. Semi.
53
54 Cancer Biol. 2013, 23, 18-25.

1
2
3 (7)Wang, G. L.; Jiang, B. H.; Rue, E. A.; Semenza, G. L. Hypoxia-inducible factor 1 is a
4

5
6 basic-helix-loop-helix-PAS heterodimer regulated by cellular O2
7
8 USA 1995, 92, 5510-5514.
9

tension. Proc. Natl. Acad. Sci.

10 (8) Tian, H.; McKnight, S. L.; Russell, D. W. Endothelial PAS domain protein 1 (EPAS1), a
11
12 transcription factor selectively expressed in endothelial cells. Genes Dev. 1997, 11, 72-82.
14
15 (9) Gu, Y. Z.; Moran, S. M.; Hogenesch, J. B.; Wartman, L.; Bradfield, C. A. Molecular
16
17 characterization and chromosomal localization of a third alpha-class hypoxia inducible factor
18
19 subunit, HIF3alpha. Gene Expr. 1998, 7, 205-213.
20
21
22 (10) Wiesener, M.S.; Jurgensen, J.S.; Rosenberger, C.; Scholze, C.K.; Horstrup, J.H.;
23
24 Warnecke, C.; Mandriota, S.; Bechmann, I.; Frei, U.A.; Pugh, C.W.; Ratcliffe, P.J.; Bachmann,
25
26 S.; Maxwell, P.H.; Eckardt, K.U. Widespread hypoxia-inducible expression of HIF-2alpha in
27
28 distinct cell populations of different organs. FASEB J. 2003, 17, 271-273.
30
31 (11) Rosenberger, C.; Mandriota, S.; Jurgensen, J.S.; Wiesener, M.S.; Horstrup, J.H.; Frei, U.;
32
33 Ratcliffe, P.J.; Maxwell, P.H.; Bachmann, S.; Eckardt, K.U. Expression of hypoxia-inducible
34
35 factor-1alpha and -2alpha in hypoxic and ischemic rat kidneys. J. Am. Soc. Nephrol. 2002, 13,
37
38 1721-1732.
39
40 (12) Kondo, K.; Klco, J.; Nakamura, E.; Lechpammer, M.; Kaelin, W.G, Jr. Inhibition of HIF
41
42 is necessary for tumor suppression by the von Hippel-Lindau protein. Cancer Cell
43
44
45 2002, 1, 237-246.
46
47 (13) Maranchie, J.K.; Vasselli, J.R.; Riss, J.; Bonifacino, J.S.; Linehan, W.M.; Klausner, R.D.
48
49 The contribution of VHL substrate binding and HIF1-alpha to the phenotype of VHL loss in
50
51 renal cell carcinoma. Cancer Cell 2002, 1, 247-255.

3 (14) Kondo, K.; Kim, W.Y.; Lechpammer, M.; Kaelin, W.G., Jr. Inhibition of HIF2α is
4
5 sufficient to suppress pVHL-defective tumor growth. PLoS Biol. 2003, 1, 439-444.
7
8 (15) Zimmer, M.; Doucette, D.; Siddiqui, N.; Iliopoulos, O. Inhibition of hypoxia-inducible
9
10 factor is sufficient for growth suppression of VHL-/- tumors. Mol. Cancer Res. 2004, 2, 89-95.
11
12 (16) Mathew, L.; Skuli, N.; Macuj, V.; Lee, S.; Zinn, P.; Sathyan, P.; Imtiyaz, H.; Zhang, Z.;
14
15 Davuluri, R.; Rao, S.; Venneti, S.; Lal, P.; Lathia, J.; Rich, J.; Keith, B.; Minn, A.; Simon, C.
16
17 miR-218 opposes a critical RTK-HIF pathway in mesenchymal glioblastoma. Proc. Natl. Acad.
18
19 Sci. USA 2014, 111, 291-296
20
21
22 (17) Sathornsumette, S.; Cao, Y.; Marcello, J.; Herndon, J.; McLendon, R.; Desjardins, A.;
23
24 Friedman, H.; Dewhirst, M.; Vredenburgh, J.; Rich, J. Tumor angiogenic and hypoxic profiles
25
26 predict radiographic response and survival in malignant astrocytoma patients treated with
27
28 bevacizumab and ironotecan. J. Clin. Oncology 2008, 26, 271-278.
30
31 (18) Jaakkola, P; Mole, D. R.; Tian,Y.-M.; Wilson, M. I.; Gielbert, J.; Gaskell, S. J.; von
32
33 Kriegsheim, A.; Hebestreit, H. F.; Mukherji, M.; Schofield, C. J.; Maxwell, P. H.; Pugh, C. W.;
34
35 Ratcliffe, P. J. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-
37
38 regulated prolyl hydroxylation. Science 2001, 292, 468-472.
39
40 (19) Bruick R. K., McKnight S. L. A conserved family of prolyl-4-hydroxylases that modify
41
42 HIF. Science 2001, 294, 1337-1340.
43
44
45 (20) Kaelin, W. G., Jr.; Ratcliffe, P. J. Oxygen sensing by metazoans: the central role of the HIF
46
47 hydroxylase pathway. Mol. Cell 2008, 30, 393-402.
48
49 (21) Majmundar, A. J.; Wong W. J.; Simon, M. C. Hypoxia-inducible factors and the response
50
51 to hypoxic stress. Mol. Cell 2010, 40, 294-309.

1
2
3 (22) Keith, B.; Johnson, R. S.; Simon, M. C. HIF1α and HIF2α: sibling rivalry in hypoxic
4
5 tumour growth and progression. Nat. Rev. Cancer 2011, 12, 9-22.
7
8 (23) Scheuermann, T. H.; Tomchick, D. R.; Machius, M.; Guo, Y.; Bruick, R. K.; Gardner, K. H.
9
10 Artificial ligand binding within the HIF2 PAS-B domain of the HIF2 transcription factor. Proc.
11
12
13 Natl. Acad. Sci. USA 2009, 106, 450-455.
14
15 (24) Scheuermann, T. H.; Li, Q.; Ma, H. W.; Key, J.; Zhang, L.; Chen, R.; Garcia, J. A.; Naidoo,
16
17
18 J.; Longgood, J.; Frantz, D. E.; Tambar, U. K.; Gardner, K. H.; Bruick, R. K. Allosteric inhibition
19
20 of hypoxia inducible factor-2 with small molecules. Nat. Chem. Biol. 2013, 9, 271-276.
22
23 (25) Rogers, J. L.; Bayeh, L.; Scheuermann, T. H.; Longgood, J.; Key, J.; Naidoo, J.; Melito, L.;
24
25
26 Shokri, C.; Frantz, D. E.; Bruick, R. K.; Gardner, K. H.; MacMillan, J. B.; Tambar, U. K.
27
28 Development of inhibitors of the PAS-B domain of the HIF-2α transcription factor. J. Med.
29
30
31 Chem. 2013, 56, 1739-1747.
32
33 (26) Scheuermann, T. H.; Stroud, D.; Sleet, C. E.; Bayeh, L.; Shokri, C.; Wang, H.; Caldwell,
35
36 C. G.; Longgood, J.; MacMillan, J. B.; Bruick, R. K.; Gardner, K. H.; Tambar U. K. Isoform-
37
38
39 selective and stereoselective inhibition of hypoxia inducible factor-2. J. Med. Chem. 2015, 58,
40
41 5930-5941.
43
44 (27) Wehn, P.W.; Rizzi, J.P.; Dixon, D.D.; Grina, J.A.; Schlachter, S.T.; Wang, B.; Xu, R.;
45
46
47 Yang, H.; Du, X.; Han, G.; Wang, K.; Cao, Z.; Cheng, T.; Czerwinski, R.M.; Goggin, B.S.;
48
49 Huang, H.; Halfmann, M.M.; Maddie, M.A.; Morton, E.L.; Olive, S.R.; Tan, H.; Xie, S.;
50
51
52 Wong, T.; Josey, J.A.; Wallace, E.M. Design and activity of specific hypoxia-inducible factor-
53
54
55 2α (HIF-2α) inhibitors for the treatment of clear cell renal cell carcinoma: Discovery of clinical

1
2
3 candidate (S)-3-((2,2-difluoro-1-hydroxy-7-(methylsulfonyl)-2,3-dihydro-1H-inden-4-
5
6 yl)oxy)-5-fluorobenzonitrile (PT2385). J. Med. Chem. 2018, 61, 9691-9721.
7
8 (28) Wallace, E. M.; Rizzi, J. P.; Han, G.; Wehn, P. M.; Cao, Z.; Du, X.; Cheng, T.; Czerwinski,
10
11 R. M.; Dixon, D. D.; Goggin, B. S.; Grina, J. A.; Halfmann, M. M.; Maddie, M. A.; Olive, S. R.;
12
13 Schlachter, S. T.; Tan, H.; Wang, B.; Wang, K.; Xie, S.; Xu, R.; Yang, H.; Josey, J. A. A small-
14
15 molecule antagonist of HIF2α is efficacious in preclinical models of renal cell carcinoma.
17
18 Cancer Res. 2016, 76, 5491–5500.
19
20 (29) Cho, H.; Du, X.; Rizzi, J. P.; Liberzon, E.; Chakraborty, A. A.; Gao, W.; Carvo, I.;
21
22
23 Signoretti, S.; Bruick, R. K.; Josey, J. A.; Wallace, E. M.; Kaelin, W. G. On-target efficacy of a
24
25 HIF-2α antagonist in preclinical kidney cancer models. Nature 2016, 539, 107-111.
27
28 (30) Chen, W.; Hill, H.; Christie, A.; Kim, M. S.; Holloman, E.; Pavia-Jimenez, A.; Homayoun,
29
30
31 F.; Ma, Y.; Patel, N.; Yell, P.; Hao, G.; Yousuf, Q.; Joyce, A.; Pedrosa, I.; Geiger, H.; Zhang, H.;
32
33 Chang, J.; Gardner, K. H.; Bruick, R. K.; Reeves, C.; Hwang, T. H.; Courtney, K.; Frenkel, E.;
34
35
36 Sun, X.; Zojwalla, N.; Wong, T.; Rizzi, J. P.; Wallace, E. M.; Josey, J. A.; Xie, Y.; Xie, X.-J.;
37
38
39 Kapur, P.; McKay, R. M.; Brugarolas, J. Targeting renal cell carcinoma with a HIF-2 antagonist.
40
41 Nature 2016, 539, 112–117.
42
43
44 (31) Courtney, K. D.; Infante, J. R.; Lam, E. T.; Figlin, R. A.; Rini, B. I.; Brugarolas, J.; Zojwalla,
45
46 N. J.; Wang, K.; Wallace, E.; Josey, J. A.; Choueiri, T. K. Phase I dose escalation trial of PT2385,
48
49 a first-in-class hypoxia inducible factor-2 antagonist in patients with previously treated
50
51
52 advanced clear cell renal cell carcinoma. J. Clin. Oncol. 2018, 36, 867-874.

1
2
3 (32) Amidon, G. L.; Lennern€as, H.; Shah, V. P.; Crison, J. R. A. Theoretical basis for a
5
6 biopharmaceutic drug classification: The correlation of in vitro drug product dissolution and
7
8
9 in vivo bioavailability. Pharm. Res. 1995, 12, 413–420.
10
11 (33) Turgeon, D.; Carrier, J.S.; Chouinard, S.; Bélanger, A. Glucuronidation activity of the
12
13
14 UGT2B17 enzyme toward xenobiotics. Drug Metab. Dispos. 2003, 31, 670–676.
15
16
17 (34) Ohno, S.; Nakajin, S. Determination of mRNA expression of human UDP
18
19 glucuronosyltransferases and application for localization in various human tissues by real-time
20
21
22 reverse transcriptase-polymerase chain reaction. Drug Metab. Dispos. 2009, 37, 32–40.
23
24 (35) Zhang, H.; Basit, A.; Busch, D.; Yabut, K.; Bhatt, D.K.; Drozdzik, M.; Ostrowski, M.; Li,
26
27 A.; Collins, C.; Oswald, S.; Prasad, B. Quantitative characterization of UDP-
28
29
30 glucuronosyltransferase 2B17 in human liver and intestine and its role in testosterone first-
31
32 pass metabolism. Biochem. Pharmacol. 2018, 156, 32-42.
34
35 (36) Sato, Y.; Nagata, M.; Tetsuka, K.; Tamura, K.; Miyashita, A.; Kawamura, A.; Usui, T.
36
37
38 Optimized methods for targeted peptide-based quantification of human uridine 5’-
39
40 diphosphate-glucuronosyltransferases in biological specimens using liquid chromatography-
41
42
43 tandem mass spectrometry. Drug Metab. Dispos. 2014, 42, 885–889.
44
45 (37) McCarroll, S.A.; Hadnott, T.N.; Perry, G.H.; Sabeti, P.C.; Zody, M.C.; Barrett,
47
48 J.C.; Dallaire, S.; Gabriel, S.B.; Lee, C.; Daly, M.J.; Altshuler, D.M. Common deletion
49
50
51 polymorphisms in the human genome. Nat. Genet. 2006, 38, 86-92.
52
53 (38) Bhatt, D.K.; Basit, A.; Zhang, H.; Gaedigk, A.; Lee, S.; Claw, K.G.; Mehrotra, A.M.;
55
56 Chaudhry, A.S.; Pearce, R.E.; Gaedigk, R.; Broeckel, U.; Thornton, T.A.; Nickerson, D.A.;

1
2
3 Schuetz, E.G.; Amory, J.K.; Leeder, J.S.; Prasad, B. Hepatic abundance and activity of
5
6 androgen- and drug-metabolizing enzyme UGT2B17 are associated with genotype, age, and
7
8
9 sex. Drug Metab. Dispos. 2018, 46, 888-896.
10
11 (39) Wang, Y.H.; Trucksis, M.; McElwee, J.J.; Wong, P.H.; Maciolek, C.; Thompson, C.D.;
12
13
14 Prueksaritanont, T.; Garrett, G.C.; Declercq, R.; Vets, E.; Willson, K.J.; Smith, R.C.;
15
16
17 Klappenbach, J.A.; Opiteck, G.J.; Tsou, J.A.; Gibson, C.; Laethem, T.; Panorchan, P.; Iwamoto,
18
19 M.; Shaw, P.M.; Wagner, J.A.; Harrelson, J.C. UGT2B17 genetic polymorphisms dramatically
20
21
22 affect the pharmacokinetics of MK-7246 in healthy subjects in a first-in-human study. Clin.
23
24 Pharmacol. Ther. 2012, 92, 96-102.
26
27 (40) Ge, S.; Tu, Y.; Hu, M. Challenges and opportunities with predicting in vivo phase II
28
29
30 metabolism via glucuronidation from in Vitro Data. Curr. Pharmacol. Rep. 2016, 2, 326–338.
31
32 (41) Peng, J.; Lu, J.; Shen, Q.; Zheng, M.; Luo X.; Zhu, W.; Jiang, H.; Chen, K. In silico site of
33
34
35 metabolism prediction for human UGT-catalyzed reactions. Bioinformatics 2014, 30, 398-405.
36
37 (42) Kerdpin, O.; Mackenzie, P.I.; Bowalgaha, K.; Finel, M.; Miners, J.O. Influence of N-
39
40 terminal domain histidine and proline residues on the substrate selectivities of human UDP-
41
42 glucuronosyltransferase 1A1, 1A6, 1A9, 2B7, and 2B10. Drug Metab. Dispos. 2009, 37, 1948-
43
44 1955.
46
47 (43) Sorich, M.J.; McKinnon, R.A.; Miners, J.O.; Smith, P.A. The importance of local chemical
48
49 structure for chemical metabolism by human uridine 5-diphosphate-glucuronosyltransferase. J.
50
51 Chem. Inf. Model. 2006, 46, 2692-2697.
52
53 (44) Singh, S. K.; Das, A. The n → π* interaction: a rapidly emerging non-covalent interaction.
55
56 Phys. Chem. Chem. Phys. 2015, 17, 9596-9612.

1
2
3 (45) Graton, J.; Wang, Z.; Brossard, A.-M.; Goncalves Monteiro, D.; Le Questel, J.-Y.; Linclau,
4
5 B. An unexpected and significantly lower hydrogen-bond-donating capacity of fluorohydrins
7
8 compared to nonfluorinated alcohols. Angew. Chem. Int. Ed. 2012, 51, 6176-6180.
9
10 (46) Keddie, N.S.; Slawin, A.M.Z.; Lebl, T.; Philp, D.; O’Hagan, D. All-cis 1,2,3,4,5,6-
11
12 hexafluorocyclohexane is a facially polarized cyclohexane. Nature Chemistry 2015, 7, 483–488.
14
15 (47) Böhm, H.J.; Banner, D.; Bendels, S.; Kansy, M.; Kuhn, B.; Müller, K.; Obst-Sander, U.;
16
17 Stahl, M. Fluorine in medicinal chemistry. ChemBioChem 2004, 5, 637−643.
18
19 (48) Zhao, Q.; Manning, J.R.; Sutton, J.; Costales, A.; Sendzik, M.; Shafer, C.M.; Levell, J.R.;
20
21
22 Liu, G.; Caferro, T.; Cho, Y.S.; Palermo, M.; Chenail, G.; Dooley, J.; Villalba, B; Farsidjani, A.;
23
24 Chen, J.; Dodd, S.; Gould, T.; Liang, G.; Slocum, K.; Pu, M.; Firestone, B.; Growney, J.;
25
26 Heimbach, T.; Pagliarini, R. Optimization of 3-pyrimidin-4-yl-oxazolidin-2-ones as orally
27
28 bioavailable and brain penetrant mutant IDH1 inhibitors. ACS Med. Chem. Lett. 2018, 11, 746-
30
31 751.
32
33 (49) Papadopoulos, K.P.; Jonasch, E.; Zojwalla, N.J.; Wang, K.; Bauer, T.M. A first-in-human
34
35 phase 1 dose-escalation trial of the oral HIF-2a inhibitor PT2977 in patients with advanced solid
37
38 tumors. J. Clin. Oncol. 2018, 36, suppl, 2508-2508.
39
40 (50) Choueiri, T.K.; Plimack, E.R.; Bauer, T.M.; Merchan, J.R.; Papadopoulos, K.P.;
41
42 McDermott, D.F.; Michaelson, M.D.; Appleman, L.J.; Zojwalla, N.P.; Jonasch, E. A First-in-
43
44
45 human Phase 1/2 Trial of the Oral HIF-2α Inhibitor PT2977 in Patients with Advanced RCC.
46
47 14th European International Kidney Cancer Symposium, Dubrovnik, Croatia. March 29-30,
51 (51) Gillis,E.P.; Eastman, K.J.; Hill, M.D.; Donnelly, D.J.; Meanwell, N.A. Applications of
54 fluorine in medicinal chemistry. J. Med. Chem. PT2385 2015, 58, 8315−8359.raction adjusted EC85: 75 ng/mL