Development of the First Small-Molecule Inhibitor Targeting Oncostatin M for Treatment of Breast Cancer

Abstract

Oncostatin M (OSM) is a proinflammatory cytokine implicated in inflammatory diseases and multiple cancers, especially breast cancer. To date, no federally approved anti-OSM therapeutics exist. We computationally screened ∼1.65 million compounds to identify small-molecule inhibitors (SMIs) of the OSM, and candidates were validated in human breast cancer models. We identified a tetrasubstituted furan (SMI-10) that inhibited OSM signaling, and optimization generated SMI-10B (K_D = 12.9 μM) and SMI-10B13 (K_D = 6.6 μM). SMI-10B13 strongly inhibited OSM-mediated STAT3 phosphorylation in T47D and MCF-7 cell lines (IC₅₀= 136 and 164 nM, respectively). Fluorescence quenching, NMR, and surface plasmon resonance assays were used to characterize SMI/OSM interactions and identify a number of analogs with low-micromolar affinity for OSM. In a human breast cancer mouse model, SMI-10B13 reduced tumor growth (p < 0.001). Kaplan-Meier analysis showed improved survival in SMI-10B13-treated mice (p = 0.04), highlighting its potential as the first anti-OSM therapeutic to inhibit breast cancer progression and extend survival.

1

OSM structure and signaling. (A) In order to form a complete signaling complex, OSM first binds to gp130, which then allows for the recruitment of LIFRβ or OSMRβ. Afterward, several signaling cascades are activatedmainly the JAK/STAT3, PI3K/AKT, MAPK/ERK, and JNK pathways (not shown). (B) OSM is a four α-helical bundle protein with two distinct binding pockets for ligand–receptor interaction. Site II is responsible for gp130 binding and Site III is utilized for interaction with OSMRβ.

2

High-throughput virtual screening identifies three novel compounds capable of interacting with site III of OSM. (A) Virtual screening analysis of over 300,000 compounds identified 16 top molecules to interact with crucial amino within Site III of OSM (SMI-10 displayed in (A). (B) Autodock Vina-computationally predicted pose of parent SMI-10 docked to the crystal structure of OSM (PDB: 1EVS), with literature amino acids that have a role in OSMRβ binding shown. (C) MDA-MB-231 human breast cancer cells were treated with OSM (10 ng/mL) and/or SMI (10 μM) for 30 min, analyzed via pSTAT3 ELISA. SMI-8, SMI-10, and SMI-11 were identified as effective inhibitors of OSM. (D) MDA-MB-231 and (E) T47D human breast cancer cell lines were treated with SMI-8, SMI-10, and SMI-11 (10 μM) and OSM (10 ng/mL) for 30 min, and immunoblots were performed for pSTAT3, pAKT, pJNK, pERK, and β-actin. (F) Structure of parent SMI-10. Data are presented as mean ± SD relative to + OSM treatment, analyzed by one-way ANOVA with Tukey’s posttest *p < 0.05, **p < 0.01.

1. General Synthetic Route to SMIs-10A–10J

2. Full Synthesis of SMI 10A–J Analogs

3

First-generation SMI-10 analogs inhibit OSM-induced pSTAT3 via ELISA. (A) Human T47D and (B) MDA-MB-231 cells were treated with SMI-10A-J analogs (10 μM) and OSM (10 ng/mL) for 30 min, and cell lysates were analyzed via ELISA. In vitro analysis identified several compounds capable of inhibiting OSM-induced pSTAT3 via ELISA, including 10B, 10C,10D, 10E, 10F, and 10G. SMI-8 is used as a positive control. Data are expressed as mean ± SD and assessed relative to + OSM treatment by one-way ANOVA with Tukey’s posttest *p < 0.05, **p < 0.01.

3. Synthesis of S MI 10B Analogs

4

10B analogs with increased inhibition of OSM-mediated pSTAT3 levels. Human T47D breast cancer cells were treated with OSM (10 ng/mL) and SMI-10B analogs (10 μM) for 30 min, and cell lysates were evaluated for pSTAT3 induction via ELISA. Analysis of 10B analogs suggests that further modification of 10B creates more effective inhibitors of OSM, specifically SMI-10B13. Data are expressed as mean ± SD and assessed relative to + OSM treatment by one-way ANOVA with Tukey’s posttest *p < 0.05, **p < 0.01.

5

SMI-10B and SMI-10B13 NMR CSP assays with ¹⁵N OSM. (A) An overlay of ¹H, ¹⁵N HSQC NMR spectra of 100 μM ¹⁵N-labeled OSM with cross peaks color-ramped from red to blue with increasing SMI-10B concentration, from 20 μM (red) to a final concentration of 260 μM (blue). The no-SMI control is shown in black. (B) Zoomed regions of panel A for Asp87, Gln90, Arg91, and Leu92. (C) An overlay of ¹H, ¹⁵N HSQC NMR spectra of 100 μM ¹⁵N-labeled OSM with cross peaks color-ramped from red to blue with increasing SMI-10B13 concentration, from 0.5 μM (red) to a final concentration of 60 μM (blue). Note that spectra of ¹⁵N-labeled OSM titrated with up to 200 μM SMI-10B13 were collected but showed very little difference compared to 60 μM SMI and were therefore not included in this overlay. The no-SMI control is shown in black. (D) Zoomed regions of panel C for Asp87, Gln90, Arg91, and Leu92.

6

SMI-10B13 inhibits tumor growth in vivo. (A) Athymic nu/nu mice were injected with SMI-10B13 or vehicle control (n = 6) 12 h before injection of MCF-7-luc-OSM-overexpressing cells. Afterward, mice were injected three times weekly for a period of 40 days with either vehicle or SMI-10B13. (B) Once weekly, mice were evaluated for tumor growth via bioluminescence imaging. Mice treated with SMI-10B13 displayed a significant reduction in bioluminescent detection compared to vehicle control. (C) Tumor size diameter (mm) was also measured 3x weekly. At the end point of experiment, mice given SMI-10B13 compared to vehicle control displayed significant reduction in tumor mm size, suggesting SMI-10B13 reduces tumor growth in vivo. Data are expressed as mean + SD and assessed relative to + OSM treatment by unpaired t-test, ***p < 0.001.

7

SMI-10B13 inhibits metastasis in vivo. (A) 12 h prior to cell injection, athymic nu/nu mice were injected with SMI-10B13 (n = 13) or vehicle control (n = 15). The following day, mice were injected with 2 × 10⁶ MCF-7-luc-OSM-overexpressing breast tumor cells in the 4th mammary fat pad and subsequently injected 3x weekly with treatment or vehicle for a period of 50 days. Tumor growth was monitored via caliper measurement, and once weekly, mice were imaged for bioluminescent tumor detection via IVIS imaging. (B) To assess survivability, mice were monitored for weight and tumor size. Mice that reached a clinical end point (weight loss greater than 20% or tumor size greater than 2 cm) were immediately sacrificed, and ex vivo imaging was performed for detection of metastasis. After 50 days, all mice were sacrificed. At the end point of the experiment, mice given SMI-10B13 compared to vehicle control displayed prolonged survival and overall smaller tumors than those treated with vehicle. (C,D) Ex vivo imaging revealed that mice treated with SMI-10B13 displayed a trend toward significance in decreasing lung metastasis intensity compared to mice treated with vehicle control.