A Potent Small-Molecule Inhibitor of the MDM2-p53 Interaction (MI-888) Achieved Complete and Durable Tumor Regression in Mice

Abstract

We previously reported the discovery of a class of spirooxindoles as potent and selective small-molecule inhibitors of the MDM2-p53 interaction (MDM2 inhibitors). We report herein our efforts to improve their pharmacokinetic properties and in vivo antitumor activity. Our efforts led to the identification of 9 (MI-888) as a potent MDM2 inhibitor (K_i = 0.44 nM) with a superior pharmacokinetic profile and enhanced in vivo efficacy. Compound 9 is capable of achieving rapid, complete, and durable tumor regression in two types of xenograft models of human cancer with oral administration and represents the most potent and efficacious MDM2 inhibitor reported to date.

INTRODUCTION

Tumor suppressor protein p53 is a transcription factor and plays a key role in regulation of cell cycle, apoptosis, DNA repair, senescence, and angiogenesis, among other cellular processes.^1–3 In approximately half of human cancers, TP53, the gene encoded p53 protein, is deleted or mutated, which renders loss of the tumor suppressor function of p53.⁴ In the other half of human cancers, p53 retains wild-type status but its tumor suppressor function is effectively inhibited by its primary endogenous inhibitor, the murine double minute 2 protein (MDM2, or HDM2 in humans).⁵ MDM2 inhibits the function of p53 by multiple mechanisms, but all are mediated by their direct protein-protein interaction.⁵ The co-crystal structure of MDM2 in complex with p53 has shown that their binding involves primarily three key residues in p53 (Phe19, Trp23, and Leu26) and a well-defined, surface binding pocket in MDM2,⁶ suggested the possibility to design non-peptide, small-molecule inhibitors to block the MDM2-p53 protein-protein interaction (hereafter called MDM2 inhibitors). Such small-molecule inhibitors may have a promising therapeutic potential for the treatment of human cancer by reactivation of the powerful tumor suppressor function of p53.^7–10 Vassilev and his colleagues from Hoffmann-La Roche were the first to report potent, selective, and orally bioavailable, small-molecule MDM2 inhibitors, as exemplified by Nutlin-3.⁷ In the last few years, there has been intense research interest in the design of new small-molecule MDM2 inhibitors, from both academia and industry.^{8, 11–21}

Our group has previously reported structure-based design of a series of spirooxindoles as a new class of potent and selective MDM2 inhibitors.^{8, 11–13, 21} Two of such compounds, 1 (MI-219)⁸ and 2,²¹ bind to MDM2 with low nanomolar affinities (Table 1) and are orally active in pre-clinical animal models of human cancer. While both compounds inhibited tumor growth in the SJSA-1 xenograft model in mice, high oral doses (200–300 mg/kg) were needed. Furthermore, these MDM2 inhibitors failed to achieve complete tumor regression.^{8, 21} Collectively, these data suggest that the pharmacokinetic (PK) properties and in vivo efficacy for 1 and 2 need to be further improved for clinical development.

Table 1.

MDM2 binding affinity and cell growth inhibition data for 1–10

Compounds	Binding IC50 (nM)	Ki (nM)	IC50 (µM) SJSA-1 Cell Line	IC50 (µM) RS4;11 Cell Line
1	44.6±2.7	10.5±2.0	0.89 ±0.21	0.54±0.14
2	9.8±2.7	0.86±0.11	0.25±0.04	0.10 ±0.01
3	7.1±1.5	0.61±0.10	0.24±0.05	0.12±0.01
4	8.2±2.3	1.1±0.3	0.41±0.04	0.30 ±0.06
5	8.8±1.4	0.94±0.07	0.32±0.06	0.23 ±0.04
6	8.4±1.1	0.62±0.05	0.20±0.02	0.10 ±0.02
7	16.3±2.1	2.4±0.5	0.35±0.06	0.21 ±0.05
8	11.1±2.1	0.97±0.3	0.60±0.16	0.43 ±0.07
9	6.8±0.9	0.44±0.22	0.08±0.02	0.06 ±0.01
10	7.6±0.9	0.62±0.19	0.25±0.04	0.10±0.01

Since 2 has a higher binding affinity to MDM2 and more potent cellular activity in cancer cells than 1,²¹ we have focused on our modifications on 2 with the goal to improve its PK properties and in vivo antitumor activity. Our efforts have now yielded a highly potent MDM2 inhibitor 9 (MI-888)²¹ with improved oral PK properties and in vivo antitumor activity. In fact, 9 was capable of achieving complete and long-lasting tumor regression in animal models of human cancer upon oral administration.

RESULTS AND DISCUSSION

Both 1 and 2 have a flexible 1,2-diol side chain, which was shown to enhance their binding affinities to MDM2 by several times and play a role in modulating their oral pharmacokinetic properties in our previous studies.^{11, 21} Our metabolic studies of 2 revealed that the major metabolic softspots are located in the 1,2-diol side chain (data not shown). We hypothesized that the overall oral PK profile of 2 can be improved by conformationally constraining the 1,2-diol side chain, thus reducing the number of rotatable bonds in the molecule, and by further improving the metabolic stability.

Although the side chain in 1 and 2 contains two hydroxyl groups, we retained only one hydroxyl group in 3–7 containing a conformationally constrained side chain for consideration of synthetic feasibility (Figure 2). Their binding affinities to MDM2 were determined using our optimized fluorescence-polarization (FP) binding assay²¹ and the results are summarized in Table 1.

Figure 2.

Chemical structures of compounds with constrained side chains.

The binding data showed that 3–6 with a cyclic alcohol side chain bind to MDM2 with high affinities (K_i = 0.61–1.1 nM). These compounds are as potent as 2 but 10 times more potent than 1. However, 7 with a tert-butanol side chain is 3-times less potent than 2 in binding to MDM2.

We tested 1–7 for their ability to inhibit cell growth in SJSA-1 osteosarcoma and RS4;11 acute leukemia cell lines, which have wild-type p53 and are sensitive to MDM2 inhibitors. The results are summarized in Table 1. Compounds 3 and 6 have IC₅₀ values of 0.24 µM in the SJSA-1 cell line and 0.12 µM in the RS4;11 cell line and are as potent as 2. Compounds 4, 5, and 7 are slightly less potent than 2, 3, and 6 in these two cancer cell lines.

Compounds 3–7 were evaluated for their metabolic stability in mouse, rat, and human microsomes, in direct comparison with 1 and 2 (Table 2). In mouse microsomes, 1, 2, 3, 6, and 7 have similar T_1/2 values (T_1/2 > 60 min) and are stable. However, 4 has modest stability (T_1/2 = 20.6 min) and 5 is very unstable (T_1/2 = 1.5 min). In rat microsomes, compounds 1, 2, 3, and 6 have a shorter T_1/2 than that in mouse microsomes, but 7 is quite stable (T_1/2 >60 min), whereas 5 is very unstable (T_1/2 = 3.8 min). In human microsomes, the trend is very similar to that in mouse microsomes and 1, 2, 3, 6, and 7 have good microsomal stability. The excellent microsomal stability of 7 suggested that blocking the oxidation of the hydroxyl group in the side chain can yield compounds with excellent microsomal stability in all three species.

Table 2.

Stability of MDM2 inhibitors in mice, rat, and human microsomes

Compounds	T1/2 (minutes)
	Mouse microsome	Rat microsome	Human microsome
1	>60	20.2	>60
2	>60	16.3	>60
3	>60	19.1	54.1
4	20.6	22.1	30.8
5	1.5	3.8	4.0
6	>60	21.1	55.1
7	>60	>60	>60
8	>60	>60	>60
9	>60	>60	>60
10	>60	>60	53.3

The good stability profile for 7 in mouse, rat, and human microsomes suggested that trisubstituted-alcohol-containing side chain may block or retard the potential oxidation by cytochrome P450 enzymes. To test this hypothesis, we installed a methyl group on the hydroxyl-substituted carbon of 3 and 4, which resulted in 8 and 9 both tethering a trisubstituted-alcohol-containing side chain. We also replaced the hydroxyl group with a methylsulfonamide group in 3, which yielded 10. Microsomal stability evaluation confirmed that 8–10 are very stable in mouse, rat, and human microsomes (Table 2). Furthermore, our binding data (Table 1) showed that 8–10 have essentially the same high binding affinity to MDM2 as 2. In the cell growth inhibition assay, 8 is 2–4 times less potent than 2, whereas 9 is two times more potent than 2, and 10 has the same potency as 2 in both cell lines (Table 1). Therefore, 9 is the most potent MDM2 inhibitor among these compounds based upon their potencies in inhibition of cell growth in these two cancer cell lines.

Based upon the high binding affinity, potent cellular activity, and excellent microsomal stability, we selected 9 for evaluation of its oral PK properties in rats, in comparison to 2. Our PK studies showed that 9 has an improved PK profile in rats over 2 (Figure 3 and Table 3). The cMax and AUC values for 9 are 2.6 and 7-fold higher than those for 2, respectively, at the same dose and these two compounds have a similar T_1/2 value (3.4 and 3.2 h for 2 and 9, respectively).

Figure 3.

Oral pharmacokinetics of 2 and 9 in Sprague Dawley rats.

Table 3.

Summary of oral pharmacokinetic data for 2 and 9 in Sprague Dawley rats.

	Dose (mg/kg)	Route	Cmax (µg/L)	AUC (0→t) (µg/L*h)	T1/2 (h)
2	25	Oral	675 ± 274	2138 ± 775	3.4 ± 0.5
9	25	Oral	1726 ±100	14874 ± 2491	3.2 ± 0.5

Next, we examined p53 activation by 9 in the SJSA-1 osteosarcoma cell line by Western blot analysis and included 2 and Nutlin-3a⁷ as control compounds (Figure 4). All three compounds effectively and dose-dependently activate p53, as evidenced by the accumulation of the p53 protein, as well as MDM2 and p21, two p53 targeted gene products. In addition, treatment of three compounds dose-dependently diminished full-length poly (ADP-ribose) polymerase (PARP) at 24 h, which indicated dose-dependent apoptosis induction. Treatment of 9 activated p53 at concentrations as low as 30 nM, as evident by clear accumulation of p21 and p53. In comparison, 9 is at least 3 times more potent than 2 in p53 activation and is 10 times more potent than Nutlin-3a.

Figure 4.

Western blot analysis of p53 activation and PARP cleavage in the SJSA-1 cancer cell line by compounds 2, 9 and Nutlin-3a. Cells were treated with different concentrations of the compounds for 24 h and western blot analysis was performed using specific antibodies. FL-PARP, full-length PARP; Cl-PARP, cleaved PARP; Pro-Cas3, pro-caspase-3; Cl-Cas3, cleaved caspase-3.

Based upon the molecular mechanism of action, the cellular activity of a specific MDM2 inhibitor should be dependent upon the activation of wild-type p53. To firmly assess the cellular specificity of 9, we employed the HCT-116 p53^+/+ colon cell line and its isogenic p53 knock-out (HCT-116 p53^−/−) cell line. Compound 9 potently inhibited cell growth in the HCT-116 p53^+/+ colon cell line with an IC₅₀ value of 92 nM and had an IC₅₀ value >10 µM in the HCT-116 p53^−/− cell line, thus displaying >100-fold selectivity for the HCT-116 cell line with wild-type p53 over its isogenic p53 deleted cell line.

Encouraged by its promising in vitro activity and excellent oral PK profile, we evaluated 9 for its antitumor activity in the SJSA-1 osteosarcoma and RS4;11 acute leukemia xenograft models in SCID mice.

In the efficacy experiment using the SJSA-1 xenograft model, SJSA-1 tumors were grown to approximately 100 mm³ and 9 was administered via oral gavage to mice, daily for 2 weeks at 10, 30, and 100 mg/kg (Figure 5). While 9 had no significant activity at 10 mg/kg, it effectively inhibited tumor growth at 30 mg/kg. Impressively, 9 at 100 mg/kg achieved rapid and complete tumor regression. After 5 days of daily dosing, the average tumor volume was decreased by >70% and after 10 days of dosing, all the mice (8 out of 8 mice) treated with 9 had undetectable tumor. The complete tumor regression was durable; all the mice remained tumor free 60 days after the last dose. There was no significant difference in animal weight between the vehicle control group of mice and the three groups of mice treated with 9. Furthermore, there was minimal weight loss and no sign of toxicity observed in mice treated with 9 at all doses during the entire experiment. Collectively, these data showed that 9 was well tolerated in mice at all the doses tested.

Figure 5.

Antitumor activity of 9 in the SJSA-1 osteosarcoma tumor xenograft model in mice. (A) Tumor volume; (B) Animal weight change.

To gain an insight into the in vivo mechanism of action of 9, we tested its ability and kinetics in activation of p53 and induction of apoptosis in the SJSA-1 xenograft tissues (Figure 6). Mice bearing SJSA-1 xenograft tumors were given a single oral dose of 9 at 100 mg/kg. Mice were then sacrificed at different time points and tumors were harvested for Western blot analysis. Our Western blot data showed that 9 induced robust upregulation of p53, as well as p21 and MDM2 proteins at 3 h and 6 h time-points, indicative of strong p53 activation in the tumor tissues. The levels of p53, p21, and MDM2 proteins were significantly diminished at the 24 h time-point, suggesting that p53 activation was transient in tumor tissues. Interestingly, cleavage of PARP and caspase-3 was minimal at the 1, 3, and 6 h-time points but became very clear at the 24 h time-point, indicating that while p53 was activated by 9 in tumor tissues quickly, apoptosis induction occurred at later time-points.

Figure 6.

Activation of p53 and cleavage of caspase-3 and PARP in SJSA-1 tumor tissues by 9. Mice bearing SJSA-1 tumors were dosed with a single oral dose of 9 at 100 mg/kg and tumor tissues were harvested at different time points for Western blot analysis using specific antibodies. FL-PARP, full-length PARP; Cl-PARP, cleaved PARP; Pro-Cas3, pro-caspase-3; Cl-Cas3, cleaved caspase-3.

In the efficacy experiment using the RS4;11 acute leukemia xenograft model, when tumors grew to an average volume of 150 mm³, compound 9 was administered at 100 mg/kg and 200 mg/kg via oral gavage daily for 3 weeks (Figure 7). At both 100 mg/kg and 200 mg/kg, 9 achieved rapid and complete tumor regression. After 4 days of dosing, tumor volume was decreased by 61% and 80% with the treatment at 100 mg/kg and 200 mg/kg of 9, respectively. Complete tumor regression was observed in all the animals after 7 days of dosing with 9 at 200 mg/kg. The complete tumor regression was also durable; 26 days after the last dose, all mice treated with 9 at 200 mg/kg still remained tumor free. There was no significant weight loss or other signs of toxicity for all the mice treated with 9 during the entire experiment.

Figure 7.

Antitumor activity of 9 in the RS4;11 acute lymphoblastic leukemia xenograft model: (A) Tumor volume; (B) Animal weight change.

CHEMISTRY

The synthesis of 3 to 10 is shown in Scheme 4. The synthesis of 9 has been reported in a previous communication.²¹

Compound 12 was synthesized using previously reported methods.¹¹⁻¹³, ²¹ Most amines required for the synthesis of intermediate 13 were commercially available with the exception of 8, 9, and 10. The intermediates 14 were synthesized following previously reported procedures. ^{11−13, 21} Compound 14 was isomerized in MeCN-H₂O or MeOH for 3–4 days affording final compounds in 50–60% yield as determined by analytical HPLC or UPLC.

The trans-N-(3-aminocyclobutanyl)methyl sulfonamide 17 used for the synthesis of 10, was synthesized from commercially available material 15 (Scheme 6). Compound 15 was converted into mesylate 16 in 84% yield and the N-Boc group was cleaved in TFA-CH₂Cl₂ solution, affording 17. The amine 17 was neutralized using 2N NaOH aqueous solution prior to use.

CONCLUSION

In the present study, we designed and synthesized a set of new and potent MDM2 inhibitors with the goal to improve the oral pharmacokinetic profile over 2 by constraining the conformation of the 1,2-diol moiety and blocking the metabolic oxidative site. Among these new compounds, 9 is a highly potent MDM2 inhibitor and has an excellent oral pharmacokinetic profile. Significantly, 9 is capable of achieving complete and durable tumor regression in both SJSA-1 osteosarcoma and RS4;11 acute lymphoblastic leukemia tumor xenograft models. Our in vitro and in vivo data show that 9 is a promising MDM2 inhibitor and warrants extensive evaluation as a potential clinical development candidate for the treatment of human cancer.

EXPERIMENTAL SECTION

General Chemistry

All reactions were conducted in a round-bottom flask equipped with a Teflon-coated magnet stirring bar. Experiments involving moisture and/or air sensitive components were performed under a nitrogen atmosphere. Commercial reagents and anhydrous solvents were used without further purification. The crude reaction product was purified by flash column chromatography packed with silica gel. Further purification was performed on a semipreparative HPLC (Waters Delta 600) with a SunFire C18 reverse phase column (19 mm × 150 mm). The mobile phase was a gradient flow of solvent A (water, 0.1% of TFA) and solvent B (CH₃CN, 0.1% of TFA or MeOH, 0.1% of TFA) at a flow rate of 10 mL/min. Proton nuclear magnetic resonance (¹H NMR) and carbon nuclear magnetic resonance (¹³C NMR) spectroscopy were performed in Bruker Advance 300 NMR spectrometers. Low resolution ESI mass spectrum analysis was performed on a Thermo-Scientific LCQ Fleet mass spectrometer. The analytical HPLC model was a Waters 2795 Separation (UV detection at 230 nm and 254 nm wave lengths) using the reverse phase column (SunFire, C18-5 µm, 4.6 × 150 mm). The analytical UPLC model was Waters Acquity H class (UV detection at 230 nm and 254 nm wave lengths). The reverse phase column used was the Acquity UPLC^® BEH (C18-1.7 µm, 2.1 × 50 mm). All final compounds were purified to ≥95% purity as determined by analytical UPLC or HPLC analysis. The optical rotation was measured on an Autopol^® III automatic polarimeter (Rudolph Research Analytical). The specific optical rotation was calculated by application of the following formula: [α]_D^T = α / (l × c), where α is the observed angular rotation, l is the length of the cell in decimeters (l = 1 dm), c is the concentration in grams per milliliter (g/mL).

Synthesis of 3 from the corresponding intermediate 14 (Scheme 1, R = trans-3-hydroxycyclobutyl) through isomerization: 14 (60 mg) obtained from CAN oxidation was placed in a round bottom flask equipped with magnetic stirring bar. MeOH (5 mL) was added to dissolve 14, followed by water (4 mL) and saturated NaHCO₃ solution (0.5 mL) to set pH at 8–9. This solution was stirred at room temperature for 3–4 days and 3 constituted 52% of all four isomers as determined by UPLC analysis. Final purification of 3 was performed on a semi-preparative or preparative reverse phase HPLC. The isomerization and purification of 4–8 and 10 followed the same isomerization procedure.

Scheme 1. General synthetic route for final compoundsa.

^aReagents and conditions: (a) Free amine, 1.0 to 2.0 equiv, THF, rt, 12 h; (b) Ammonium cerium(IV) nitrate (CAN) 2.1 equiv, MeCN: H₂O = 1:1, rt, 1 h; (c) NaHCO₃, MeOH-H₂O = 1:1, 3–4 days.

(2'S,3R,4'S,5'R)-6-Chloro-4'-(3-chloro-2-fluorophenyl)-N-(trans-3-hydroxycyclobutyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (3)

¹H NMR (300 MHz, MeOD-d4): 8.77 (d, J = 5.78 Hz, 1H, NH), 7.64-7.52 (m, 2H), 7.41 (t, J = 6.85 HZ, 1H), 7.23-7.13 (m, 1H), 7.14 (dd, J = 8.11, 1.82 Hz, 1H), 6.79 (d, J = 1.77 Hz, 1H), 5.21 (d, J = 11.34 Hz, 1H), 4.58 (d, J = 11.34 Hz, 1H), 4.50 (dd, J = 8.22, 1.94 Hz, 1H), 4.34-4.22 (m, 1H), 4.20-4.06 (m, 1H), 2.30-2.06 (m, 3H), 1.94-1.80 (m, 2H), 1.15 (dd, J = 15.13, 1.70 Hz, 1H), 0.87 (s, 9H). ¹³C NMR (75 MHz, MeOD-d4): 177.87, 167.38, 157.87 (d, J_C–F = 245.04 Hz), 145.18, 137.29, 132.68, 128.70, 126.79, 126.63 (d, J_C–F = 4.91 Hz), 124.22, 123.47, 122.63 (d, J_C–F = 18.68 Hz), 121.64 (d, J_C–F =13.39 Hz), 112.17, 65.20, 64.67, 64.27, 62.88, 43.46, 42.97, 40.40, 39.83, 31.01, 29.59. ESI-MS calculated for C₂₇H₃₁³⁵Cl₂FN₃O₃ [M+H]⁺: 534.17, Found: 534.14.

To a round-bottom flask, 12 (2.9 g, 4.50 mmol), cis-3-aminocyclobutanol (0.40 g, 4.56 mmol), and THF (30 mL) were added and the solution was stirred at room temperature for 12 h. THF was removed and the crude material was purified on flash column chromatography and the desired ring-opening adduct was isolated in 2.13 g (2.93 mmol). The ring-opening adduct and MeCN (20 mL) were placed in a round-bottom flask. CAN (3.37 g, 6.15 mmol) and water (20 mL) were subsequently added and the mixture was stirred at room temperature for 1 h before quening with NaHCO₃ saturate solution. The aqueous layer was extracted with ethyl acetate (50 mL × 3) and the combined organic layers were washed with water and brine and dried over anhydrous Na₂SO₄. The solvent was removed and the crude material was purified on flash column chromatography. The corresponding intermediate 14 was isolated in 624 mg (1.17 mmol, 26% yield over two steps). 4 constituted 56% of all four isomers upon isomerization of 14 (Scheme 1, R = cis-3-hydroxycyclobutyl) for 3 days.

(2'S,3R,4'S,5'R)-6-Chloro-4'-(3-chloro-2-fluorophenyl)-N-(cis-3-hydroxycyclobutyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (4)

¹H NMR (300 MHz, MeOD-d4): 8.77 (d, J = 6.95 Hz, 1H), 7.62-7.51 (m, 2H), 7.37 (d, J = 7.00 Hz, 1H), 7.20-7.12 (m, 1H), 7.12 (dd, J = 8.02, 1.74 Hz, 1H), 6.78 (d, J = 1.68 Hz, 1H), 5.23 (d, J = 11.33 Hz, 1H), 4.60 (d, J = 11.33 Hz, 1H), 4.50 (dd, J = 8.24, 1.49 Hz, 1H), 3.98-3.84 (m, 1H), 3.84-3.70 (m, 1H), 2.64 (ddd, J = 12.94, 12.71, 6.64 Hz, 1H), 2.50 (ddd, J = 12.46, 12.50, 6.55 Hz, 1H), 1.90 (dd, J = 15.46, 8.33 Hz, 1H), 1.80 (dd, J = 19.42, 8.81 Hz, 1H), 1.53 (dd, J = 19.42, 8.81 Hz, 1H), 1.13 (dd, J = 15.46, 1.46 Hz, 1H), 0.86 (s, 9H). ¹³C NMR (75 MHz, MeOD-d4): 177.78, 166.96, 157.86 (d, J_C–F = 249.82 Hz), 145.19, 137.24, 132.58, 128.67, 126.79, 126.58 (d, J_C–F = 4.98 Hz), 124.15, 123.47, 122.55 (d, J_C–F = 18.94 Hz), 121.58 (d, J_C–F = 12.97 Hz), 112.15, 64.55, 64.26, 62.72, 61.22, 43.37, 41.44, 41.36, 38.05, 30.97, 29.56. ESI-MS calculated for C₂₇H₃₁³⁵Cl₂FN₃O₃ [M+H]⁺: 534.17, Found: 534.28. [α]_D²⁵ = 16.7° (MeOH, c = 0.0160 g/mL) To a round-bottom flask, 12 (3.2 g, 4.95 mmol), (1R,3S)-3-aminocyclopentanol (0.5 g, 4.95 mmol), and THF (25 mL) were added and the solution was stirred at room temperature for 12 h. THF was removed and the crude material was purified on flash column chromatography and the desired ring-opening adduct was isolated in 2.68 g (3.46 mmol). The ring-opening adduct and MeCN (20 mL) were placed in a round-bottom flask. CAN (3.98 g, 7.27 mmol) and water (20 mL) were subsequently added and the mixture was stirred at room temperature for 1 h before quening with NaHCO₃ saturate solution. The aqueous layer was extracted with ethyl acetate (50 mL × 3) and the combined organic layers were washed with water and brine and dried over anhydrous Na₂SO₄. The solvent was removed and the crude material was purified on flash column chromatography. The corresponding intermediate 14 was isolated in 811 mg (1.48 mmol, 30% yield over two steps). 5 constituted 64% of all four isomers upon isomerization of 14 (Scheme 1, R = (1S,3R)-3-hydroxycyclopentyl) for 3 days.

(2'S,3R,4'S,5'R)-6-Chloro-4'-(3-chloro-2-fluorophenyl)-N-((1S,3R)-3-hydroxycyclopentyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (5)

¹H NMR (300 MHz, MeOD-d4): 8.24 (d, J = 7.50 Hz, 1H, NH), 7.63-7.52 (m, 2H), 7.37 (t, J = 6.96 Hz, 1H), 7.19-7.11 (m, 1H), 7.12 (dd, J = 8.10, 1.78 Hz, 1H), 6.78 (d, J = 1.70 Hz, 1H), 5.28 (d, J = 11.52 Hz, 1H), 4.60 (d, J = 11.52 Hz, 1H),4.51 (dd, J = 8.35, 1.65 Hz, 1H), 4.26-4.08 (m, 2H), 2.20-2.06 (m, 1H), 1.91 (dd, J = 15.44, 8.35 Hz, 1H), 1.84-1.60 (m, 2H), 1.56-1.40 (m, 2H), 1.38-1.24 (m, 1H), 1.14 (dd, J = 15.37, 1.48 Hz, 1H), 0.87 (s, 9H) ¹³C NMR (75 MHz, MeOD-d4): 177.78, 166.74, 157.91 (d, J_C–F = 249.49 Hz), 145.18, 137.23, 132.26, 128.67, 126.74, 126.67 (d, J_C–F = 4.93 Hz), 124.15, 123.50, 122.62 (d, J_C–F = 19.14 Hz), 121.61 (d, J_C–F = 12.95 Hz), 112.16, 73.04, 64.43, 64.30, 62.82, 51.66, 43.41, 42.00, 34.65, 31.79, 30.99, 29.56 ESI-MS calculated for C₂₈H₃₃³⁵Cl₂FN₃O₃ [M+H]⁺: 548.19, Found: 548.20; [α]_D²⁵ = 18.7° (MeOH, c = 0.0174 g/mL) The same procedure for the synthesis of 5 was followed using (1S,3S)-3-aminocyclopentanol as the reactant. The CAN oxidation furnished the corresponding intermediate 14 in 32% over 2 steps. 6 constituted 70% of all four isomers upon isomerization of 14 (Scheme 1, R = (1S,3S)-3-hydroxycyclopentyl) for 3 days.

(2'S,3R,4'S,5'R)-6-Chloro-4'-(3-chloro-2-fluorophenyl)-N-((1S,3S)-3-hydroxycyclopentyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (6)

¹H NMR (300 MHz, MeOD-d4): 8.47 (d, J = 7.23 Hz, 1H, NH), 7.68-7.54 (m, 2H), 7.41 (t, J = 6.94 Hz, 1H), 7.24-7.14 (m, 1H), 7.16 (dd, J = 8.13, 1.82 Hz, 1H), 6.81 (d, J = 1.74 Hz, 1H), 5.23 (d, J = 11.38 Hz, 1H), 4.61 (d, J = 11.38 Hz, 1H), 4.54 (dd, J = 8.22, 1.64 Hz, 1H), 4.42-4.28 (m, 1H), 4.28-4.19 (m, 1H), 2.04-1.86 (m, 3H), 1.80-1.60 (m, 2H), 1.56-1.42 (m, 1H), 1.17 (dd, J = 15.34, 1.39 Hz, 1H), 1.14-1.00 (m, 1H), 0.89 (s, 9H). ¹³C NMR (75 MHz, MeOD-d4): 177.81, 167.10, 157.86 (d, J_C–F = 248.99 Hz), 145.17, 137.22, 132.57, 128.75, 126.79, 126.61 (D, J_C–F = 4.71 Hz), 124.18, 123.54, 122.54 (d, J_C–F = 18.86 Hz), 121.63 (d, J_C–F = 12.93 Hz), 112.16, 72.88, 64.59, 64.13, 62.89, 51.52, 51.41, 43.44, 42.29, 34.30, 31.61, 30.99, 29.57. ESI-MS calculated for C₂₈H₃₃³⁵Cl₂FN₃O₃ [M+H]⁺: 548.19, Found: 548.32. [α]_D²⁵ = 22.9° (MeOH, c = 0.0112 g/mL) To a round-bottom flask, 12 (3.0 g, 4.67 mmol), 1-amino-2-methylpropan-2-ol (0.41 g, 4.60 mmol), and THF (25 mL) were added and the solution was stirred at room temperature for 12 h. THF was removed and the crude material was purified on flash column chromatography and the desired ring-opening adduct was isolated in 2.19 g (2.99 mmol). The ring-opening adduct and MeCN (10 mL) were placed in a round-bottom flask. CAN (3.44 g, 6.28 mmol) and water (10 mL) were subsequently added and the mixture was stirred at room temperature for 1 h before quening with NaHCO₃ saturate solution. The aqueous layer was extracted with ethyl acetate (50 mL × 3) and the combined organic layers were washed with water and brine and dried over anhydrous Na₂SO₄. The solvent was removed and the crude material was purified on flash column chromatography. The corresponding intermediate 14 was isolated in 690 mg (1.29 mmol, 28% over 2 steps). 7 constitute 62% of all four isomers upon isomerization of 14 (Scheme 1, R = 2-hydroxy-2-methyl-propyl) for 3 days.

(2'S,3R,4'S,5'R)-6-Chloro-4'-(3-chloro-2-fluorophenyl)-N-(2-hydroxy-2-methylpropyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (7)

¹H NMR (300 MHz, MeOD-d4): 7.65-7.56 (m, 2H), 7.35 (t, J = 7.04 Hz, 1H), 7.20-7.14 (m, 1H), 7.12 (dd, J = 8.02, 1.66 Hz, 1H), 6.78 (d, J = 1.64 Hz, 1H), 5.42 (d, J = 11.51 Hz, 1H), 4.64 (d, J = 11.51 Hz, 1H), 4.54 (dd, J = 8.22, 1.27 Hz, 1H), 3.40-3.30 (m, 1H), 2.98 (d, J = 13.44 Hz, 1H), 1.93 (dd, J = 15.42, 8.38 Hz, 1H), 1.14 (dd, J = 15.45, 1.18 Hz, 1H), 0.97 (s, 3H), 0.87 (s, 9H), 0.83 (s, 3H). ¹³C NMR (75 MHz, MeOD-d4): 177.70, 168.15, 157.90 (d, J_C–F = 249.83 Hz), 145.20, 137.21, 132.57, 128.83, 126.83, 126.68 (d, J_C–F = 4.69 Hz), 124.13, 123.51, 122.66 (d, J_C–F = 18.96 Hz), 121.63 (d, J_C–F = 13.25 Hz), 112.16, 75.58, 71.16, 64.48, 64.43, 62.85, 51.51, 43.45, 31.01, 29.58, 27.41, 26.65. ESI-MS calculated for C₂₇H₃₃³⁵Cl₂FN₃O₃ [M+H]⁺: 536.19, Found: 536.44. [α]_D²⁵ = 24.0° (MeOH, c = 0.0186 g/mL)

To a round-bottom flask, 12 (643 mg, 1.0 mmol), trans-3-(tert-butyldimethylsilyloxy)-3-methylcyclobutanamine²¹ (0.10 g, 0.46 mmol), and THF (10 mL) were added and the solution was stirred at room temperature for 12 h. THF was removed and the crude material was purified on flash column chromatography and the desired ring-opening adduct was isolated in 0.36 g (0.41 mmol). The ring-opening adduct and MeCN (5 mL) were placed in a round-bottom flask. CAN (450 mg, 0.82 mmol) and water (5 mL) were subsequently added and the mixture was stirred at room temperature for 1 h before quening with NaHCO₃ saturate solution. The aqueous layer was extracted with ethyl acetate (50 mL × 3) and the combined organic layers were washed with water and brine and dried over anhydrous Na₂SO₄. The solvent was removed and the crude material was purified on flash column chromatography. The corresponding intermediate 14 was isolated in 62 mg (0.11 mmol, 24% yield over two steps) and (2'S,3R,4'S,5'R)-N-((1r,3R)-3-((tert-butyldimethylsilyl)oxy)-3-methylcyclobutyl)-6-chloro-4'-(3-chloro-2-fluorophenyl)-1'-((1R,2S)-2-hydroxy-1,2-diphenylethyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (chiral-auxiliary-cleaved product with the hydroxy group protected by TBS, 53 mg, 0.08 mmol, 17% yield over two steps). 8 constituted 64% of all four isomers upon isomerization of 14 (Scheme 1, R = trans-3-hydroxy-3-methylcyclobutyl) for 3 days.

(2'S,3R,4'S,5'R)-6-Chloro-4'-(3-chloro-2-fluorophenyl)-N-(trans-3-hydroxy-3-methylcyclobutyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (8)

¹H NMR (300 MHz, MeOD-d4): 8.68 (d, J = 5.95 Hz, 1H, NH), 7.64-7.52 (m, 2H), 7.41 (t, J = 7.56 Hz, 1H), 7.24-7.14 (m, 1H), 7.14 (dd, J = 8.06, 1.64 Hz, 1H), 6.79 (d, J = 1.76 Hz, 1H), 5.19 (d, J = 11.28 Hz, 1H), 4.58 (d, J = 11.28 Hz, 1H), 4.50 (dd, J = 8.06, 1.65 Hz, 1H), 4.37-4.20 (m, 1H), 2.44-2.22 (m, 2H), 1.96-1.82 (m, 2H), 1.62 (dd, J = 11.69, 5.79 Hz, 1H), 1.15 (dd, J = 15.43, 1.66 Hz, 1H), 1.09 (s, 3H), 0.87 (s, 9H). ¹³C NMR (75 MHz, MeOD-d4): 177.84, 167.28, 157.90 (d, J_C–F = 249.25 Hz), 145.17, 137.28, 132.66, 128.70, 126.79, 126.68 (d, J_C–F = 4.96 Hz), 124.20, 123.41, 122.65 (d, J_C–F = 18.90 Hz), 121.63 (d, J_C–F = 12.76 Hz), 112.16, 70.79, 64.65, 64.25, 62.84, 44.80, 43.73, 43.39, 40.89, 30.97, 29.56, 29.43. ESI-MS calculated for [M+H]⁺: C₂₈H₃₃³⁵Cl₂FN₃O₃ [M+H]⁺ 548.19, Found: 548.56.

To a round-bottom flask, 12 (1.286 g, 2.0 mmol), 17 (1.8 mmol based on 16), and THF (15 mL) were added and the solution was stirred at room temperature for 12 h. THF was removed and the crude material was purified on flash column chromatography and the desired ring-opening adduct was isolated in 0.776 g (0.96 mmol). The ring-opening adduct and MeCN (10 mL) were placed in a round-bottom flask. CAN (1.053 g, 1.84 mmol) and water (10 mL) were subsequently added and the mixture was stirred at room temperature for 1 h before quening with NaHCO₃ saturate solution. The aqueous layer was extracted with ethyl acetate (50 mL × 3) and the combined organic layers were washed with water and brine and dried over anhydrous Na₂SO₄. The solvent was removed and the crude material was purified on flash column chromatography. The corresponding intermediate 14 was isolated in 626 mg (1.023 mmol, 56% yield over two steps). 10 constituted 62% of all four isomers upon isomerization of 14 (Scheme 1, R = trans-3-(methylsulfonamido)cyclobutyl) for 3 days.

(2'S,3R,4'S,5'R)-6-Chloro-4'-(3-chloro-2-fluorophenyl)-N-(trans-3-(methylsulfonamido) cyclobutyl)-2'-neopentyl-2-oxospiro[indoline-3,3'-pyrrolidine]-5'-carboxamide (10)

¹H NMR (300 MHz, MeOD-d4): 8.96 (d, J = 6.37 Hz, 1H, NH), 7.64-7.54 (m, 2H), 7.36 (t, J = 6.96 Hz, 1H), 7.20-7.16 (m, 1H), 7.11 (dd, J = 8.13, 1.79 Hz, 1H), 6.78 (d, J = 1.75 Hz, 1H), 5.32 (d, J = 11.51 Hz, 1H), 4.63 (d, J = 11.51 Hz, 1H), 4.52 (d, J = 8.47, 1.49 Hz, 1H), 4.28-4.14 (m, 1H), 3.83 (quintet, J = 3.83 Hz, 1H), 2.84 (s, 3H), 2.40-2.30 (m, 2H), 2.30-2.20 (m, 1H), 2.06-1.96 (m, 1H), 1.93 (dd, J = 15.47, 8.42 Hz, 1H), 1.13 (dd, J = 15.47, 1.23 Hz, 1H), 0.86 (s, 9H). ¹³C NMR (75 MHz, MeOD-d4): 177.73, 167.29, 157.80 (d, J_C–F = 249.64 Hz), 145.13, 137.16, 132.63, 128.69, 126.75, 126.61 (d, J_C–F = 4.84 Hz), 124.13, 123.52, 122.51 (d, J_C–F = 18.77 Hz), 121.55 (d, J_C–F = 13.01 Hz), 112.15, 64.51, 64.11, 62.70, 46.60, 43.43, 40.93, 38.47, 38.11, 30.97, 29.54. ESI-MS calculated for C₂₈H₃₄³⁵Cl₂FN₄O₄S [M+H]⁺: 611.17, Found: 611.42. [α]_D²⁵ = 13.8° (MeOH, c = 0.0208 g/mL)

N-(trans-3-Aminocyclobutyl)methanesulfonamide (17)

To a round-bottom flask, 15 (400 mg, 2.15 mmol, 1.0 equiv), Et₃N (1 mL, 7.5 mmol, 3.5 equiv) and THF (25 mL) were added. The reaction mixture was cooled in an ice-water bath and methanesulfonyl chloride (MsCl, 492 mg, 4.3 mmol, 2.0 equiv) was added dropwise via a syringe. The reaction was stirred at room temperature for 12 h before quenching with water. The aqueous layer was extracted with ethyl acetate (70 mL ×2) and the combined organic layers were washed with brine (50 mL) and dried over anhydrous Na₂SO₄. The solvents were removed and the crude product was purified on flash column chromatography. 16 was obtained as a white solid (0.475 g, 84%) 16: ¹H NMR (MeOD-d4, 300 MHz): 4.10-3.92 (m, 2H), 2.87 (s, 3H), 2.30 (t, J = 6.71 Hz, 4H), 1.43 (s, 9H) ¹³C (MeOD-d4, 75 M Hz): 158.07, 80.52, 46.61, 43.58, 40.90, 38.90, 28.91; ESI-MS: calculated for C₁₀H₂₀N₂NaO₄S [M+Na]⁺ = 287.10; obtained: 287.16

To a round-bottom flask, 16 (0.475 g, 1.8 mmol), Et₃Si-H (0.1 mL) and CH₂Cl₂ (15 mL) were added. TFA (4 mL) was added via a syringe and the reaction was stirred at room temperature for 1 h. The reaction mixture was concentrated to remove solvent and TFA. MeOH (15 mL) was added and the pH was adjusted to be 10–11 by NaOH (2 N aqueous solution). MeOH and water were then removed affording crude free amine 17, which was directly used for the next step without further purification.

FP-based Protein Binding Assay

The binding affinity of MDM2 inhibitors was determined by an optimized, sensitive, and quantitative fluorescence polarization-based (FP-based) binding assay using a recombinant human His-tagged MDM2 protein (residues 1–118) and a FAM tagged p53-based peptide as the fluorescent probe. The design of the fluorescent probe was based upon a previously reported high affinity p53-based peptidomimetic compound (5-FAM-βAla-βAla-Phe-Met-Aib-pTyr-(6-Cl-LTrp)-Glu-Ac3c-Leu-Asn-NH₂).²² This tagged peptide was named as PMDM6-F. The K_d value of PMDM6-F to the MDM2 protein was determined to be 1.4 ± 0.3 nM by monitoring the total fluorescence polarization of mixtures composed with the fluorescent probe at a fixed concentration and the MDM2 protein with increasing concentrations up to full saturation. Fluorescence polarization values were measured using the Infinite M-1000 plate reader (Tecan U.S., Research Triangle Park, NC) in Microfluor 2 96-well, black, round-bottom plates (Thermo Scientific). In the saturation experiments, 1 nM of PMDM6-F and increasing concentrations of proteins were added to each well to a final volume of 125 µl in the assay buffer (100 mM potassium phosphate, pH 7.5, 100 µg/ml bovine γ-globulin, 0.02% sodium azide [Invitrogen], with 0.01% Triton X-100 and 4% DMSO). Plates were mixed and incubated at room temperature for 15–30 minutes with gentle shaking to assure equilibrium. The polarization values in millipolarization units (mP) were measured at an excitation wavelength of 485 nm and an emission wavelength of 530 nm. Equilibrium dissociation constants (K_d) were then calculated by fitting the sigmoidal dose-dependent FP increases as a function of protein concentrations using Graphpad Prism 5.0 software (Graphpad Software, San Diego, CA). K_i values of tested compounds were determined in a dose-dependent competitive binding experiment. Mixtures of 5 µl of the tested compound in different concentrations in DMSO and 120 µl of preincubated protein/fluorescent probe complex with fixed concentrations in the assay buffer (100 mM potassium phosphate, pH 7.5, 100 µg/ml bovine γ-globulin, 0.02% sodium azide, with 0.01% Triton X-100) were added into assay plates and incubated at room temperature with gentle shaking. The incubation time was precisely controlled at 15 min to minimize the influence of the initial isomerization to the binding affinities. Final concentrations of the protein and fluorescent probe in the competitive assays were 10 nM and 1 nM, respectively, and final DMSO concentration is 4%. Negative controls containing protein/fluorescent probe complex only (equivalent to 0% inhibition), and positive controls containing free fluorescent probe only (equivalent to 100% inhibition), were included in each assay plate. FP values were measured as described above. IC₅₀ values were determined by nonlinear regression fitting of the sigmoidal dose-dependent FP decreases as a function of total compound concentrations using Graphpad Prism 5.0 software (Graphpad Software, San Diego, CA). K_i values of tested compounds to the MDM2 protein were calculated using the measured IC₅₀ values, the K_d value of the fluorescent probe to the protein, and the concentrations of the protein and fluorescent probe in the competitive assays.²³

Cell Growth Inhibition Assay

The SJSA-1 osteosarcoma cell line and RS4;11 leukemia cell line were purchased from the American Type Culture Collection (Manassas, VA). HCT-116 p53^+/+ colon cell line (HCT-116 p53^+/+) and its isogenic p53 knock-out (HCT-116 p53^−/−) cell line are kind gifts from Dr. Bert Vogelstein, Johns Hopkins University at Baltimore. All cell lines were grown in PRMI 1640 medium (Gibco, life Technologies, Grand Island, NY) supplemented with 10% FBS (Invitrogen, Carlsbad, CA) in a humidified tissue culture incubator at 37 °C, 5% CO₂. For cell growth inhibition assay, cells were seeded in 96-well, flat bottom cell culture plates at a density of (3–4) × 10³ cells/well and allowed to grow overnight, then incubated with a compound at different concentrations. The effect of cell growth inhibition of a compound was determined with a lactate dehydrogenase-based WST-8 assay (WST-8; Dojindo Molecular Technologies Inc., Gaithersburg, MD). For RS4;11 cell line, WST-8 solution was directly added to culture medium in each well to a final concentration of 10% upon the end of treatment. For SJSA-1, HCT-116 p53^+/+ and HCT-116 p53^–/–, culture medium in each well was discharged and fresh RPMI 1640 Medium with no Phenol Red (Gibco, life Technologies, Grand Island, NY) containing 10% WST-8 solution was added to each well in the presence of 5% FBS upon the end of treatment, and then the plates were incubated at 37 °C for 2–3 h. The absorbance of the samples was measured at 450 nm using a TECAN ULTRA reader. The effect of each treatment was calculated by the equation: [absorbance value of treatment - absorbance value of (medium with FBS + WST solution only)]/[absorbance value of untreated control - absorbance value of (medium with FBS + WST solution only)]. The IC₅₀ value of a compound is the concentration that inhibits cell growth by 50% over the control.

Pharmacodynamics (PD) Study

For PD studies, the SJSA-1 xenograft tumor model was used. To develop xenograft tumors, 5 × 10⁶ SJSA-1 cancer cells with 50% Matrigel were injected s.c. on the dorsal side of SCID mice from Charles River, one tumor per mouse. When xenograft tumors reached a mean of ∼100 mm³ (70–130 mm³), two mice were treated with vehicle control (10% PEG400: 3% Cremophor: 87% PBS), 8 mice were treated with a single dose of 9 at 100 mg/kg via oral gavage. Mice treated with vehicle control were sacrificed at 6 h and tumor tissue was harvested for Western blot analysis. Two mice treated with 9 were sacrificed at 1, 3, 6, and 24 h (two mice at each time point) and tumor tissue was harvested.

Western Blotting

For western blot analysis, tumor cells or tumor tissues were lyzed in ice-cold RIPA buffer: 20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% sodium deoxycholate, 2.5 mM sodium pyrophosphate, 1 mM b-glycerophosphate, 1 mM sodium orthovanadate and 1µg/ml leupeptin. The expressions of indicated proteins in the whole cell lysates were detected by western blot analysis using the following antibodies: anti-p53 (OP43; Calbiochem), anti-MDM2 (sc-965; Santa Cruz), anti-p21 (556431; BD Biosciences), anti-PARP (9542, Cell Signaling Technology), anti-caspase-3 (AAP-113, Stressgen Bioreagents) and HRP-conjugated anti-GAPDH (sc-5778; Santa Cruz).

In vivo Efficacy Experiments

For in vivo efficacy experiment using the SJSA-1 tumor xenograft model, 5 × 10⁶ SJSA-1 cancer cells with 50% Matrigel were injected s.c. on the dorsal side of SCID mice from Charles River, one tumor per mouse. When xenograft tumors reached a mean of ∼100 mm³ (70–130 mm³), mice were randomized into groups of 8 mice. Vehicle Control (10% PEG400: 3% Cremophor: 87% PBS) was administered orally (p.o.) once per day for 14 days. Compound 9 was administered orally once per day for 14 days at 10 mg/kg, 30 mg/kg and 100 mg/kg. Tumor sizes and animal weights were measured three times per week. Data are represented as mean tumor volumes. Tumor volume (mm³) = (A×B²)/2 where A and B are the tumor length and width (in mm), respectively. Statistical analyses were done by two-way ANOVA and unpaired two-tailed t test, using Prism (version 4.0, GraphPad). P < 0.05 was considered statistically significant.

For in vivo efficacy studies using the RS4;11 tumor xenograft model, the same protocol was used with following modifications: RS4;11 tumor cells were used; dose schedule: 100 mg/kg and 200 mg/kg.

Figure 1.

Two of our previously reported potent and orally active MDM2 inhibitors

Scheme 2. Synthesis of trans N-(3-aminocyclobutanyl)methyl sulfonamide 17a.

^aReagents and conditions: (a). MsCl, Et₃N, THF, 0 °C to rt, 84%; (b). TFA, CH₂Cl₂, rt, 1 h, (c). NaOH, MeOH-H₂O

ABBREVIATIONS

AUC

area under the curve

CAN

ceric ammonium nitrate

C_max

the maximum compound concentration from oral dosing

Cl

clearance

CYP

cytochrome P450

CH₂Cl₂

dichloromethane

EtOAc

ethyl acetate

H₂O

water

LC/MS

liquid chromatography–mass spectrometry

MDM2

murine double minute 2

MeCN

acetonitrile

NaOH

sodium hydroxide

PD

pharmacodynamics

PK

pharmacokinetic

p.o.

oral administration

qD

once a day dosing

SCID

severe combined immunodeficient

T_1/2

half life

TFA

trifluoroacetic acid

T_max

the time at which C_max is reached;