Small-molecule PROTAC mediates targeted protein degradation to treat STAT3-dependent epithelial cancer

Abstract

The aberrant activation of STAT3 is associated with the etiology and progression in a variety of malignant epithelial-derived tumors, including head and neck squamous cell carcinoma (HNSCC) and colorectal cancer (CRC). Due to the lack of an enzymatic catalytic site or a ligand-binding pocket, there are no small-molecule inhibitors directly targeting STAT3 that have been approved for clinical translation. Emerging proteolysis targeting chimeric (PROTAC) technology-based approach represents a potential strategy to overcome the limitations of conventional inhibitors and inhibit activation of STAT3 and downstream genes. In this study, the heterobifunctional small-molecule-based PROTACs are successfully prepared from toosendanin (TSN), with 1 portion binding to STAT3 and the other portion binding to an E3 ubiquitin ligase. The optimized lead PROTAC (TSM-1) exhibits superior selectivity, potency, and robust antitumor effects in STAT3-dependent HNSCC and CRC - especially in clinically relevant patient-derived xenografts (PDX) and patient-derived organoids (PDO). The following mechanistic investigation identifies the reduced expression of critical downstream STAT3 effectors, through which TSM-1 promotes cell cycle arrest and apoptosis in tumor cells. These findings provide the first demonstration to our knowledge of a successful PROTAC-targeting strategy in STAT3-dependent epithelial cancer.

Keywords: Colorectal cancer; Drug therapy; Head and neck cancer; Oncology; Therapeutics.

Figure 1. Expression of STAT3 and its clinical significance.

(A and D) STAT3 expression was significantly increased in patients with HNSCC (GSE30784 database). (B and E) STAT3 expression was increased in patients with CRC according to the GSE21815 database. (C, F, and G) Increased expression of STAT3 protein was observed in biopsy samples from patients with HNSCC and patients with CRC (n = 3 patients). ***P < 0.001 when compared with the control group. P values are from 2-tailed paired t tests (D–G).

Figure 2. Design, synthesis, and screening of TSMs derivatives.

(A) Computer simulation of the combination between TSN and STAT3 (PDB, 1BG1), where a hydrogen bond was formed between TSN and Glu638 or Ser613 on STAT3. (B) Computer simulation of appropriate site for tethering TSN to lenalidomide (4CI2) for the design of TSM-1. (C) Synthetic route of preparing TSM-1. (D–G) TSM-1 significantly increased the thermal stability of STAT3 in CETSA assays in CAL33 (52°C and 55°C) (E and F) and HCT116 (50°C) cells (D and G) (n = 3 replicates). (H) MST analysis of TSN binding to STAT3 (K_d = 296 nM). (I) MST analysis of TSM-1 binding to STAT3 (K_d = 308 nM). (J) CAL33 cells proliferation was detected using CCK-8 assays after treatment with TSN or TSM-1 for 48 hours (n = 3 replicates). Among them, the degrader TSM-1 exhibited pronounced antitumor effects. *P < 0.05, ** P < 0.01, and *** P < 0.001 when compared with the control group. P values are from 2-way ANOVA with Tukey’s multiple-comparison test (F, G, and J).

Figure 3. TSM-1 inhibited tumor cell proliferation and viability.

(A and B) HNSCC (A) and CRC (B) cell proliferation were detected using CCK-8 assays after treatment with TSM-1 for 48 hours (n = 3 replicates). (C) Protein expression of STAT3 in HNSCC and CRC cells was determined by Western blot. (D) STAT3 expression when knocking down of STAT3 through 3 siRNAs (siRNA 1729, 1878, and 1272). PC represented positive control knocking down GAPDH; NC represented negative control. (E and F) STAT3 knockdown through siRNA 1878 significantly alleviated the inhibition effect of TSM-1 on CAL33 (E) and HCT116 cells (F) (n = 3 replicates). (G) CAL33 and HCT116 cells were stained with a live/dead cell viability/cytotoxicity kit after treatment with TSM-1 for 48 hours. Scale bar: 200 μm. (n = 3 replicates). The green fluorescence was used to mark the living tumor cells, and the red fluorescence was used to mark the dead tumor cells. ***P < 0.001 when compared with the control group. P values are from 2-way ANOVA with Tukey’s multiple-comparison test (A and B).

Figure 4. TSM-1 potently and selectively degraded STAT3 in multiple cell-based systems.

(A) At low concentrations, TSM-1 treatment (36 hours) reduced STAT3 protein expression. (C and D) Analysis revealed that TSM-1 had minimum effects on other STAT family members (n = 3 replicates). (B) TSM-1 induced STAT3 degradation in a time-dependent manner in CAL33 and HCT116 cells. (E–G) Quantitative results of the relative protein levels of STAT3 (E and F) (n = 3 replicates), and the representative microscopic photographs of the CAL33 and HCT116 cells (G). Scale bar: 10 μm (n = 3 replicates). *P < 0.05, **P < 0.01, and ***P < 0.001 when compared with the control group. P values are from 2-way ANOVA with Tukey’s multiple-comparison test (C and D) or ordinary 1-way ANOVA with Dunnett’s multiple-comparison test (E and F).

Figure 5. TSM-1 elicited cell cycle arrest and apoptosis.

(A and C) Cell cycle arrest was detected by flow cytometry after treatment with TSM-1 for 36 hours (n = 3 replicates). (B and D) After exposure to TSM-1 for 48 hours, the extent of apoptosis was monitored by flow cytometry (n = 3 replicates). (E–H) TSM-1 treatment (36 hours) reduced levels of STAT3, p-STAT3, cyclin D, c-Myc, PD-L1, and BCL-XL proteins in CAL33 (E and F) and HCT116 (G and H) cells (n = 3 replicates). *P < 0.05, **P < 0.01, and ***P < 0.001 when compared with the control group. P values are from 2-way ANOVA with Tukey’s multiple-comparison test (A and C) or ordinary 1-way ANOVA with Dunnett’s multiple-comparison test (B, D, F, and H).

Figure 6. Mechanistic studies for TSM-1 induced STAT3 degradation.

(A, B, G, and H) TSM-1 treatment (1 μM, 36 hours) reduced STAT3 protein levels when washed out drugs and replaced with fresh medium for different time points in CAL33 cells (A and G), as well as HCT116 cells (B and H), when additional incubation with fresh medium for 24 hours (n = 3 replicates). (C, D, I, and J) Degradation of STAT3 was blocked in CAL33 (C and I) and HCT116 (D and J) cells when MG132 (10 μM, 1 hour) was added following 36-hour incubation with TSM-1 (n = 3 replicates). (E, F, K, and L) Addition of lenalidomide (15 μM) reversed TSM-1 induced downregulation of STAT3 levels in CAL33 (E and K) and HCT116 (F and L) cells (n = 3 replicates). (M and N) Immunoprecipitation of ubiquitin when treatment with TSM-1 (1 μM) in 293T cells. IP: STAT3, IB: ubiquitin, and the input samples. (O) High-resolution microscopy of GFP-fluorescence images at different time points when treated with 1 μM TSM-1 through GE DeltaVision OMX SR. Formation of ternary complexes in 293T cells after treatment with 1 μM TSM-1 for 36 hours was reversed by pretreatment with 10 μM TSN or 15 μM lenalidomide through High Content Analysis System (n = 3 replicates). *P < 0.05, **P < 0.01, and ***P < 0.001 when compared with the control group. P values are from ordinary 1-way ANOVA with Dunnett’s multiple-comparison test (G and H) or 2-tailed unpaired t test (I–L)

Figure 7. TSM-1 decreased STAT3 protein levels and inhibited tumor growth in vivo.

(A, E, and I) The treatment regimen diagrams (n = 5 mice). (B, C, F, G, J, and K) Both tumor volume (B, F, and J) and body weight (C, G, and K) were monitored every 2 days. (A, D, E, H, I, and L) When mice were sacrificed, the tumors were photographed (A, E, and I), and tumor weight was recorded (D, H, and L). (M–R) Western blot analyses show that TSM-1 treatment inhibited STAT3 and its downstream signaling pathway–related target gene expression in vivo. *P < 0.05, **P < 0.01, and ***P < 0.001 when compared with the control group. P values are from ordinary one-way ANOVA with Dunnett’s multiple-comparison test (D, H, and L) or 2-way ANOVA with Tukey’s multiple-comparison test (B, F, J, P, Q, and R).

Figure 8. TSM-1 enhanced anti–PD-L1 immune checkpoint blockade.

(A) The pattern diagram. For the combinatorial immunotherapy, anti–PD-L1 antibody (12.5 μg per animal) was i.p. injected on days 1, 3, and 5 in addition to the vein injection of 2 mg/kg TSM-1. (B) When mice were sacrificed, the tumors were photographed (n = 5 mice). (C and D) Both tumor volume (C) and body weight (D) were monitored every day (n = 5 mice). (E) When mice were sacrificed, tumor weight was recorded (n = 5 mice). (F) TSM-1 led to decreased STAT3 and PD-L1 expression compared with control group. (G and H) Statistical analysis results are shown (n = 3 mice). *P < 0.01 and ***P < 0.001 versus the control group. P values are from ordinary 1-way ANOVA with Dunnett’s multiple-comparison test (E, G, and H) or 2-way ANOVA with Tukey’s multiple-comparison test (C).

Figure 9. TSM-1 suppressed organoid formation and survival.

(A) The pattern diagram. (B–D) Concentration-dependent effects of TSM-1 on organoid survival, based on organoid sizes (%) (n = 3 replicates). (E) Pictures of patient-derived organoids stained with Live/Dead fluorescent dye after TSM-1 treatment for 6 days (n = 3 replicates). Scale bars: 500 μm.

Figure 10. TSM-1 inhibited HNSCC tumor growth in PDX models.

(A) The treatment regimen diagrams. (B–G) Both tumor volume (C and F) and body weight (B and E) were monitored every 2 days (n = 5 mice in B, and n = 7 mice in E); when mice were sacrificed, the tumors were photographed (B and E), and tumor weight was recorded (D and G). (H and I) Western blot analyses showed that TSM-1 treatment inhibited STAT3 and its downstream signaling pathway–related target gene expression in PDX SCC342 model. (J) TSM-1 treatment led to significantly increased necrotic lesion, TUNEL⁺ cells, and decreased STAT3 expression in PDX models. *P < 0.05, **P < 0.01, and ***P < 0.001 when compared with the control group. P values are from 2-way ANOVA (C and F), 2-tailed unpaired t test (D and G), or 2-way ANOVA with Tukey’s multiple-comparisons test (I).