Synthesis and biological evaluation of structurally diverse 6-aryl-3-aroyl-indole analogues as inhibitors of tubulin polymerization

Abstract

The synthesis and evaluation of small-molecule inhibitors of tubulin polymerization remains a promising approach for the development of new therapeutic agents for cancer treatment. The natural products colchicine and combretastatin A-4 (CA4) inspired significant drug discovery campaigns targeting the colchicine site located on the beta-subunit of the tubulin heterodimer, but so far these efforts have not yielded an approved drug for cancer treatment in human patients. Interest in the colchicine site was enhanced by the discovery that a subset of colchicine site agents demonstrated dual functionality as both potent antiproliferative agents and effective vascular disrupting agents (VDAs). Our previous studies led to the discovery and development of a 2-aryl-3-aroyl-indole analogue (OXi8006) that inhibited tubulin polymerization and demonstrated low nM IC₅₀ values against a variety of human cancer cell lines. A water-soluble phosphate prodrug salt (OXi8007), synthesized from OXi8006, displayed promising vascular disrupting activity in mouse models of cancer. To further extend structure-activity relationship correlations, a series of 6-aryl-3-aroyl-indole analogues was synthesized and evaluated for their inhibition of tubulin polymerization and cytotoxicity against human cancer cell lines. Several structurally diverse molecules in this small library were strong inhibitors of tubulin polymerization and of MCF-7 and MDA-MB-231 human breast cancer cells. One of the most promising analogues (KGP591) caused significant G2/M arrest of MDA-MB-231 cells, disrupted microtubule structure and cell morphology in MDA-MB-231 cells, and demonstrated significant inhibition of MDA-MB-231 cell migration in a wound healing (scratch) assay. A phosphate prodrug salt, KGP618, synthesized from its parent phenolic precursor, KGP591, demonstrated significant reduction in bioluminescence signal when evaluated in vivo against an orthotopic model of kidney cancer (RENCA-luc) in BALB/c mice, indicative of VDA efficacy. The most active compounds from this series offer promise as anticancer therapeutic agents.

Keywords: Antiproliferative agents; Indole synthesis; Inhibitors of cell migration; Inhibitors of tubulin polymerization; Molecular docking; Vascular disrupting agents.

Figure 1.

Representative Natural Products [3, 4, 6, 19, 20] and Synthetic Small-Molecule Analogues [18, 21] that Interact with the Tubulin-Microtubule System

Figure 2.

Representative Small-Molecule Inhibitors [–33] of Tubulin Polymerization Previously Synthesized (Pinney Laboratory)

Figure 3.

Structural Motivation for the Design of Hybrid 6-Aryl-3-Aroyl-Indoles Analogues Derived from OXi8006 [32] and Sabizabulin [36, 38]

Figure 4.

Compilation of 6-Aryl-3-Aroyl-Indole Analogues Synthesized in this Study Biological Evaluation

Figure 5:

Scratch Assay for MDA-MB-231 Cells Treated with Compound KGP591. A. Representative images of MDA-MB-231 cells treated with vehicle or KGP591 (50 nM, 100 nM or 200 nM) at time zero and 24, 48, and 72 h after making scratches. Cells were incubated at 37 °C with 5% CO₂. The scratch gap closed after 72 h in control cells while in the 200 nM KGP591 treated cells, the scratch width remained almost unchanged for 72 h. These effects were concentration dependent. B. Wound confluence initially lagged behind the control in 50 nM treated cells, but reached 100% by 72 h as did the control. For 24 h, the 100 nM KGP591 treated cells remained at 0% confluence, but confluence increased slightly at later times. The 200 nM KGP591 treated cells remained near 0% confluence for 72 h. C. The width of the scratch decreased steadily in both control cells and cells treated with 50 nM KGP591. The decrease in wound width was much less in cells dosed with 100 nM KGP591 and remained largely open by 72 h. At 200 nM KGP591, the wound width did not decrease significantly. D. Percent wound closure determined by (area of wound at 72 h/ area of wound at 0 h)x100. The percent wound closure for 100 nM treated cells was greatly decreased compared to control, while there was no significant percent wound closure for 200 nM KGP591 treated cells.

Figure 6.

Microtubule disruption in MDA-MB-231 cells treated with 100 nM KGP591. MDA-MB-231 cells were stained with ViaFluor 488 live cell microtubule stain and Hoechst 33342 for cell nuclei. Diminished microtubule binding was observed at the earliest time point, an effect that increased with time. Profound microtubule disruption was observed in the first 30 min with 100 nM KGP591 treatment. All images were collected with the Biotek Lionheart microscope ELx800 in fluorescence mode with a 40x objective. Scale bar 30 μm. Green: microtubules, Blue: nuclei.

Figure 7.

KGP591 and CA4 induced G2/M cell cycle arrest in MDA-MB-231 cells. (A-C) Representative histograms indicating the difference between (A) vehicle treated control cells and MDA-MB-231 cells treated with (B) 200 nM KGP591 and (C) 5 nM CA4. Propidium iodide-A on the x-axis represents the amount of fluorescence observed at 2N (G1) and 4N (G2/M). (D-E) Bar graphs indicate concentration dependence of KGP591 and percentage of cell events at (D) G1 (p < 0.0001 for all treated samples compared to untreated control) and at (E) G2/M (p < 0.0001 for 5 nM CA4, and 200, 400, and 500 nM treated samples compared to untreated control). The difference between 100 nM KGP591 and untreated control was not significant. Data points in bar graphs represent independent experiments and are presented as mean ± S.E.M. ****represents p < 0.0001.

Figure 8.

KGP591 (Ball and Stick) Docked in the DAMA-Colchicine Binding Site of Tubulin (1SA0).

Figure 9.

A. KGP591 and DAMA-Colchicine (Blue) Co-Crystallized with Tubulin (1SA0). B. KGP588 (Blue), KGP591 (Grey), KGP594 (Yellow), KGP596 (Green) Overlay Showing Similarity in Binding Positions.

Figure 10.

HPLC of KGP618 reaction with alkaline phosphatase; stability of KGP618 in water. (A) HPLC of reaction of KGP618 with 2 U ALP at 37 ^oC at 90 min (blue) and 4 h (red) indicated that reaction was completed in 90 min; (B) HPLC of KGP618 dissolved in water and incubated for 24 h at 37 °C with no enzyme (red) demonstrated the stability of KGP618 (no KGP591 was released).

Figure 11.

Efficacy of KGP618 assessed using BLI. Representative images of acute chronological changes in response to KGP618 in 1,2-propanediol vehicle or vehicle alone observed by BLI at baseline (0 h), and 2.5 h and 24 h after treatment of orthotopic RENCA-luc kidney tumor bearing male BALB/c mice: A. control mouse, and B. KGP618 treated mice. Normalized dynamic BLI curves: C. control mouse (n=1) showed similar signal at each time point: D. KGP618 treated mice (n=6) showed >90% signal reduction at 2.5 h following 150 mg/kg dosage compared to baseline. Some signal recovery was seen at 24 h (n=4).

Figure 12:

H&E Staining for KGP618-Treated RENCA-luc Tumor. A. Whole mount section (bar 2.5 mm) of residual normal kidney tissue at left with RENCA tumor growing at right, showing some regions of necrosis and severe hemorrhage within 2.5 h following dosing, as emphasized in B, the higher magnification area shows massive hemorrhage caused by KGP618 (bar 100 μm).

Scheme 1.

Synthesis of Hybrid 6-Aryl-3-Aroyl-Indole Analogues

Scheme 2.

Synthesis of KGP595 and Prodrug KGP618