pH-sensitive nanomedicine of novel tubulin polymerization inhibitor for lung metastatic melanoma

Abstract

Microtubule binding agents such as paclitaxel and vincristine have activity in metastatic melanoma. However, even responsive tumors develop resistance, highlighting the need to investigate new drug molecules. Here, we showed that a new compound, CH-2-102, developed by our group, has high anti-tumor efficacy in human and murine melanoma cells. We confirmed that CH-2-102 robustly suppresses the microtubule polymerization process by directly interacting with the colchicine binding site. Our results unveil that CH-2-102 suppresses microtubule polymerization and subsequently induces G2 phase cell arrest as one of the possible mechanisms. Notably, CH-2-102 maintains its efficacy even in the paclitaxel resistance melanoma cells due to different binding sites and a non-Pgp substrate. We developed a pH-responsive drug-polymer Schiff bases linker for high drug loading into nanoparticles (NPs). Our CH-2-102 conjugated NPs induced tumor regression more effectively than Abraxane^® (Nab-paclitaxel, N-PTX), free drug, and non-sensitive NPs in B16-F10 cell-derived lung metastasis mouse model. Furthermore, our results suggest that the formulation has a high impact on the in vivo efficacy of the drug and warrants further investigation in other cancers, particularly taxane resistant. In conclusion, the microtubule polymerization inhibitor CH-2-102 conjugated pH-responsive NPs induce tumor regression in lung metastasis melanoma mice, suggesting it may be an effective strategy for treating metastatic melanoma.

1. Introduction

Melanoma is an aggressive and deadly skin cancer in which melanocytes have abnormal growth. It accounts for 5.6% of all new cancer cases and 1.2% of all cancer-related deaths in the United States. The highly metastatic nature of melanoma makes it very dangerous if it is not detected and treated at early stages [1]. Current management of early-stage and non-metastasized melanoma involves surgical removal of cancerous cells. For non-resectable patients, both immune- and genetically targeted treatments have revolutionized the field [2–9]. Specifically, treatment with anti-CTLA4 antibody ipilimumab, and anti-PD1 antibody nivolumab in patients with metastatic melanoma has revolutionized the field [10]. Further, a combination of anti-lymphocyte activation gene-3 (LAG-3) antibody relatlimab with nivolumab was approved for metastatic melanoma patients [11,12]. Despite this, a number of melanoma patients eventually develop resistance to immunotherapy and die due to the lack of any alternative treatments [13].

Dacarbazine (DTIC) and Temozolomide (TZM) are the only two FDA-approved chemotherapy for advanced metastatic melanoma treatment [14–16]. Moreover, the second-line therapy for metastatic melanoma includes N-PTX or paclitaxel in combination with either carboplatin, cisplatin, bevacizumab, temozolomide, etc. N-PTX is albumin-bound NPs (NPs) of paclitaxel which belongs to the class of compounds targeting tubulin proteins to inhibit tumor growth. As compared to DTIC, N-PTX showed clinical benefit with significantly better progression-free survival (4.8 months vs 2.5 months), disease control rate (39% vs 27%), and a manageable safety profile in stage IV chemotherapy-naïve patients [17]. However, PTX cannot shake off chemoresistance, which plagues the treatment of melanoma using chemotherapy [18]. Therefore, there is an immediate need for new, potent therapeutic approaches to the patient resistant to the current standards of care.

Tubulin protein is of interest to researchers as microtubules, the polymeric form of tubulin, form the backbone of cell survival with their involvement in nuclei, cell division, intracellular structure (cytoskeleton), intracellular transport, organelle distribution, and even motility involving cilia and flagella where applicable [19]. Tubulin protein, consisting primarily of α- and β-tubulin monomers, can be targeted mainly through the following three binding sites: i) taxane binding domain, ii) vinca binding domain and iii) colchicine binding domain. As the name indicates, the first domain is the target for PTX, docetaxel (DTX), epothilones, etc. The second domain is targeted by analogs of vinca alkaloids like vinblastine, and vincristine, while the last domain binding agents are colchicine and combretastatin. Interestingly, only the colchicine binding domain does not have reported chemoresistance, [20].

Although the actual mechanism of resistance is not clear, taxanes and vinca alkaloids are substrates for several efflux pumps known to decrease their intracellular concentration. These include the ATP-binding cassette (ABC) transporters (P-glycoprotein [Pgp]), multidrug resistance-associated protein 1 (MRP1), and breast cancer resistance protein (BCRP/ABCG2). Furthermore, the modification of target proteins where aberrant overexpression of ß-tubulin isotypes (I, II, III, IV, V) is also known to increase the resistance to taxane [21]. Therefore, development of potent, novel molecules which are not substrates of the efflux transporter, less prone to acquired resistance to target protein, and with little to no toxicity, are urgently required for proper management of metastatic melanoma.

Keeping the aforementioned goal in mind, we have synthesized a series of colchicine binding compounds with computer-aided drug design method. We have already reported on SMART-OH, LY293, ABI-III, QW-296, and CH-3–8, where these compounds show superior potency than taxanes as anticancer agents [22–25]. Furthermore, we utilized different nanoparticulate systems for the delivery of these molecules to various cancer models in mice. Some of these compounds have unfavorable physicochemical properties and the physical encapsulation of these small molecules in polymeric systems is not always possible. Therefore, we even utilized lipid/polymer conjugation to enhance the payload in our delivery system [24].

Polymer conjugation is a viable option wherever possible to increase blood circulation time, bioavailability, stealth protection, tumor accumulation through enhanced permeability and retention (EPR) effect. In this study, we have synthesized a novel tubulin polymerization inhibitor, CH-2-102, a potent analog of QW-296, and conjugated it to poly (ethylene glycol)-block-poly (2-methyl-2-carboxyl-propylene carbonate) (mPEG-b-PCC) through a pH-sensitive linker to get mPEG-b-PCC-g-BA-g-CH-2-102. The biodegradable polymer shows the stealth properties with mPEG on the outside, while the polycarbonate backbone with CH-2-102 forms the inner hydrophobic core. Moreover, we checked the impact of the compound on taxane-resistant cells, metabolic respiration of cells, and interaction with the mitochondria. We evaluated the efficacy of CH-2-102 in vitro and in vivo using a syngeneic metastatic mouse model.

2. Material and methods

2.1. Materials

Melanoma cell lines B16-F10 (murine) and A375 (human) were purchased from the American Type Culture Collection (ATCC). The PTX resistance versions of these cells i.e., B16-F10-TXR and A375-TXR, were generated by repeated treatment with increasing drug concentration. HyClone Dulbecco’s Modified Eagle’s Medium (DMEM) was purchased from Cytiva (Marlborough, MA). Heat inactivated fetal bovine serum (FBS) was obtained from R&D Systems (Minneapolis, MN). The antibiotic-antimycotic mixture and trypsin EDTA were obtained from ThermoFisher Scientific (Waltham, MA). Primary antibodies were obtained from Cell Signaling (Beverly, MA), Abcam (Cambridge, MA), Santa Cruz biotech (Dallas, TX) and while IR dye-labeled secondary antibodies were obtained from Licor (Lincoln, NE). The synthesis reagents 2,2-Bis(hydroxymethyl) propionic acid, benzyl bromide, methoxy poly(ethylene glycol) (mPEG, 5,000 Da) 8- diazabicycloundec-7-ene (DBU), L-lactide, and BOC-glycine were purchased from Sigma-Aldrich (St. Louis, MO). Methylene chloride (DCM), ethyl acetate, and n-hexanes were obtained from Fisher Scientific (Waltham, MA). Merck precoated plates (silica 60 F254, 0.25 mm) were used for thin-layer chromatography (TLC), and bands were observed under ultraviolet (UV) light. ¹H NMR spectra were obtained using a Bruker Advance-III HD 500 MHz or 400 MHz NMR spectrometer, and data were processed by TopSpin 3.5 (Bruker). The difference of chemical shifts is reported in δ value (ppm) compared to the internal standard tetramethylsilane.

2.2. Synthesis of CH-2-102

CH-2-102 (6-aryl-2-benzoyl-pyridine) was synthesized according to the scheme in Fig. 1. Briefly, 2,6-dibromopyridine 1 was reacted with 3,4,5-trimethoxybenzaldehyde in the presence of n-BuLi to afford intermediate 2, Subsequently, 2 was reacted with Dess-Martin periodinane in a solution of dichloromethane to give compound 3 as a major product. The Suzuki coupling reaction of compound 3 with 4-hydroxymethylphenyl boronic acid under tetrakis(triphenylphosphine)palladium [Pd (PPh3)4] condition generated the final compound 4 CH-2-102. The chemical structure of CH-2-102 was confirmed by ¹H NMR and Waters Xevo G2-S qTOF high-resolution mass spectrometry (HRMS). The purity of CH-2-102 (≥95%) was verified by analytical high-pressure liquid chromatography (HPLC) using a BEH C18 column (2.1 mm × 50 mm, 1.7 μm particle size) elution was done with the acetonitrile/water (with 0.1% formic acid) solvent mixture (60:40) at a flow rate of 0.3 mL/min. The aqueous solubility of CH-2-102 was determined by incubating it in water (pH 7.4 or 5.0, n = 3) for 24 h at 37 °C with continual shaking at 125 rpm and centrifuged for 10 min at 5000 rpm. The drug concentrations of the supernatant were determined relative to the peak areas of drug standards (0.2–100 μg/mL) by HPLC at λ_max of 241.5 nm using a C18 column (150 mm × 4.6 mm, 5 μm, Phenomenex, Torrance, CA) and using the same elution media indicated above.

Fig. 1.

Synthesis and characterization of CH-2-102. A) Synthetic scheme and B) ¹H NMR spectra.

2.3. Molecular modeling

Molecular modeling was accomplished using the crystal structure of a closely related analog; DJ-101 in complex with tubulin at a resolution of 2.8 Å (PDB code: 5H7O). For the modeling, Schrödinger Molecular Modeling Suite 2018 (Schrödinger LLC, New York, NY) was used as described previously [25]. Briefly, using the Protein Preparation Wizard workflow, the structures of the protein-ligand complexes were prepared. Next, by using the Receptor Grid Generation, we generated the native ligand tubulin receptor grid. CH-2-102 tubulin inhibitor was built and prepared for docking using the Ligprep module before it was docked into 5H7O. The estimated free energy of binding (kcal/mol) was calculated using the Glide docking score; a lower (more negative) number suggests a more favorable interaction. Maestro interface of the Schrödinger program was used for hydrogen bonding and data analysis.

2.4. In vitro tubulin polymerization assay

A fluorescence-based tubulin polymerization assay kit (Cytoskeleton, Inc. Denver, CO) was used to test CH-2-102’s ability to inhibit tubulin polymerization. Specifically, 5 μL of 10× concentrations of controls and samples were aliquoted into a half-area 96-well plate that had been prewarmed (37 °C) (Corning Costar, Corning, NY). The plate was filled with 50 μL of tubulin mix containing GTP, glycerol, and tubulin protein. The reaction movements were measured using a SpectraMax M5e fluorescent microplate reader at a frequency of 30 s and 90 min. The excitation was at 360 nm, and the fluorescent signal was measured at 450 nm.

2.5. Immunofluorescence confocal microscopy

We performed immunofluorescence experiments to observe CH-2-102’s effects on tubulin polymerization at the subcellular level as previously described [21]. A375 cells were seeded on eight-chambered coverslips (500 cells/per well) and incubated for 16 h. Subsequently, cells were treated for 24 h with DMSO, PTX, colchicine or CH-2-102, rinsed twice with ice-cold PBS, fixed overnight at 4 °C with a 4% formalin solution, and washed thoroughly with PBS to remove formaldehyde. Cells were then permeabilized with an equal ratio mixture of PBS containing 2.0% BSA and 0.1% Triton X-100 incubated for 1.0 h on a benchtop. After washing with PBS, cells were then incubated with anti-α, β-tubulin primary antibody (cell signaling, Danvers, MA) at 4 °C for 16 h. The cells were washed three times with PBS to remove the primary antibody before being incubated for 1 h at RT with the secondary antibody (goat anti-Rabbit IgG-FITC, Santa Cruz Biotechnology, Dallas, TX). The cells were rinsed 3 times with PBS to remove the unbound secondary antibody before being stained with a 4′,6-diamidino-2-phenylindole (DAPI) solution (in 0.1% Triton X-100) and incubated at RT for 10 min. Later, the cells were rinsed in PBS before being imaged using a Zeiss 710 confocal microscope and processed with Zen imaging software (Zeiss, Thornwood, NY).

2.6. Cell viability assay

The viability of melanoma cells in presence of CH-2-102 as a free drug or in NPs was determined by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. Briefly, B16-F10 (2000 cells/well) and A375 cells (2500 cells/well) were seeded in 96 well plates for 16 h, and different concentrations of the drug and the equivalent amount of NP formulation were added for 48 h. Subsequently, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (commonly known as MTT) was added to the cells at the concentration of 0.5 mg/mL and incubated at 37 °C for 4 h. The supernatant was removed after centrifuging the plates at 2000 rpm for 5 min, and the violet-colored formazan crystals were dissolved in 150 μL of DMSO, and the absorbance was measured at 570 and 630 nm. Similarly, cell survival assays on PTX resistant (TXR) B16-F10-TXR and A375-TXR cells were used to evaluate if CH-2-102 maintains its potency in these cells. All the cell viability assays were done in triplicate and the standard deviation was calculated (Mean ± S.D.).

2.7. Spheroid formation

We generated three-dimensional (3D) tumor spheroids to replicate the in vivo tumor-like microenvironment using A375 cells. In a low attachment, U-shaped 96 wells plate 500 cells were seeded and centrifuged for 15 s to collect the cells together around the bottom of the wells. We monitored the formation and growth of the 3D structure every day for 7 days. On the 7th day, spheroids were treated with different concentrations of the drug and the control. The spheroids were then monitored every day until the 7th day in which a live-dead assay (ThermoFisher Scientific, Waltham, MA) staining was performed as instructed by the manufacturer. The spheroids were imaged using a fluorescence microscope (ZEISS, Germany).

2.8. Cell cycle analysis

Cell cycle analysis was performed to understand the effect of CH-2-102 on different stages of melanoma cell growth. A375 and B16-F10 cells were seeded (3 × 10⁵/well) on a 6-well plate overnight and treated with CH-2-102 for 24 h. Afterward, the collected cells were fixed in 70% ice-cold ethanol for 60 min and washed. Approximately 1 × 10⁶ cells were then suspended in 0.5 mL of FxCycle^™ PI/RNase staining solution for 15 min at RT in dark. The percentage of cells in each phase was analyzed using a flow cytometer (BD FACS Calibur, NJ) by performing each treatment in triplicate and reported as mean ± S.D.

2.9. Apoptosis assay

A375 and B16-F10 cells were seeded on 6-well plates at the concentration of 3 × 10⁵/well for 16 h and then treated with CH-2-102 for 24 h. Next, cells were gently trypsinized, collected, and resuspended in 100 μL Annexin V binding buffer. The Annexin V antibody labeled with APC and propidium iodide (PI) solutions (BioLegend, San Diego, CA) were added and incubated at RT in the dark for 15 min. 300 μL of Annexin V binding buffer was added and analyzed by a flow cytometer (Ex- 488, Em-610). Each experiment was performed in triplicate, and data were reported as mean ± S.D.

2.10. Western blot analysis

The effect of CH-2-102 treatment on different protein expressions in A375 and B16-F10 cells was determined by Western blot. Briefly, cells were treated with DMSO, CH-2-102, or colchicine for 48 h, washed with cold PBS twice and lysed with RIPA buffer. Protein sample concentration was analyzed by BCA assay (ThermoFisher Sci.) and normalized. After denaturation, samples were loaded into a 4–15% SDS-PAGE and transferred to a PVDF membrane using the iBlot^™ system (Invitrogen, Carlsbad, CA). Membranes were blocked with Blocking solution (Li-COR, Lincoln, NE) at RT for 45 min before probing with different primary antibodies at 4 °C overnight. The next day, membranes were washed with TBST and incubated with IR dye labeled secondary antibody for 1.0 h at RT in dark. β-actin was used as a protein loading control. Blots were analyzed using Li-COR Odyssey^® infrared imaging system (Li-COR, Lincoln, NE).

2.11. Cell migration studies

To determine the migration ability of cells in the presence of CH-2-102 we performed scratch assays. For the assay, A375 cells (3 × 10⁵) in a 24 plate were incubated till they reached 100% confluence. A scratch was made on the cell monolayer by dragging a 200 μL pipette tip through the center of the well. The process removed the monolayer of cells from the scratched area. The scratched cells were removed by washing twice with the culture media. The remaining cells in the well were then treated with the media containing 12 nM colchicine, 8 and 12 nM CH-2-102, or equivalent DMSO. Images of the scratch were obtained immediately and after 6 and 20 h under the microscope (ZEISS, Germany). Boyden chamber assay was performed to visualize the difference in the migratory behavior of cells. A375 cells (3 × 10⁵/well) in 6 well plates were incubated overnight and then treated with 12 nM colchicine, 8 nM, and 12 nM CH-2-102, or equivalent DMSO for 24 h. Next the cells were harvested and seeded in equal numbers in the Boyden chamber with 100 μL of serum-free media. In the lower chamber, 600 μL of DMEM media with 20% FBS was added and incubated for 6 and 24 h. The cells in the upper chamber were gently removed by scrubbing with moistened cotton tip applicator. The chambers were fixed in 70% alcohol and air-dried for 20 min. The dry chambers were then placed in 0.2% solution of crystal violet for staining and then washed multiple times to remove the free stain. The membranes on the chamber were then imaged under the brightfield inverted microscope (ZEISS, Germany). Each experiment was performed in triplicate, and data were reported as mean ± S.D.

2.12. Colony formation studies

The effect of CH-2-102 on the tumorigenic potential of A375 and B16-F10 cells was determined by clonogenic assay by two different methods. To understand the effect of the drug on long-term exposure, A375 cells (1 × 10³) were treated with 8 or 12 nM CH-2-102, 12 nM colchicine, or DMSO for 14 days. To understand the lasting effect of the short-term drug exposure, A375 cells were treated with 8 nM or 12 nM CH-2-102, 12 nM colchicine, or DMSO for 24 h, and the cells (1 × 10³) from each treated groups were allowed to grow for 14 days in fresh cell culture media without any treatments. The colonies, in both the methods, were washed with PBS, fixed with formalin, and stained with crystal violet (0.2% w/v) dissolved in a 10% ethanol solution. The wells were then imaged, and ImageJ software was used to analyze the percentage area covered by the colonies.

2.13. Pgp ATPase activity assay

Pgp-Glo assay system (Promega, Madison, WI) was used to evaluate the effect of CH-2-102 on Pgp ATPase activity. Briefly, CH-2-102 was incubated at the dose of 10, 100, or 1000 nM with recombinant human Pgp membrane (25 μg) in a white untreated 96-well plate. Pgp-GIo assay buffer was used as the negative control, 200 μM verapamil was used as the positive control of drug-induced Pgp ATPase activity, and 100 μM sodium orthovanadate was used as a selective Pgp ATPase inhibitor. 5 mM Mg ATP was added to the mixture to initiate ATPase activity and was incubated at 37 °C for 40 min. Further, 50 μL of ATP detection reagent was added to initiate luminescence, followed by plate scanning after 20 min of incubation at RT. The luminescence of each well was measured on SpectraMax^® i3x microplate reader (Molecular Device, San Jose, CA). Each experiment was performed in triplicate, and data reported as mean ± S.D.

2.14. Metabolic stability and half-life of CH-2-102 in liver microsomes

The metabolic stability of CH-2-102 was performed by using the liver microsomes (LM). The reaction media containing 1 mg/mL microsomal protein and 40 mM MgCl₂, and phosphate buffer (100 mM, pH 7.4) was used. The reaction media was preincubated with 1 mM nicotinamide adenine dinucleotide phosphate (NADPH) at 37 °C for 5 min. CH-2-102 was spiked at 1 μM final concentration. The negative control was incubated without the addition of NADPH while the positive control was incubated with 1 mM verapamil. 100 μL samples were taken between 0 and 60 min, mixed with 300 μL of methanol and then centrifuged at 12,000 g for 15 min. CH-2-102 concentration in the supernatant was analyzed using LC-MS/MS using 60% acetonitrile as mobile phase and the Agilent Zorbax Sb-C18 column (3.0 × 50 mm 1.8 μm).

2.15. Mitochondrial potential assay

Mitochondrial potential assay was performed in A375 and B16-F10 cells (5 × 10³) in a black clear bottom 96 well plate. The cells were seeded for 24 h and then treated with different concentrations of CH-2-102 for 24 h: 12, 25 and 50 nM for A375 cells, while 25 nM, 50 nM, 100 nM for:; B16-F10 cells. Blank wells were used where there were no cells but only the treatment solution. Similarly, 3 depolarized control wells were reserved for FCCP treatment, which only contained the cells but not the treatment with CH-2-102. Four hours prior to the completion of the test, FCCP was added to the control wells. 30 min prior to the completion of the test warmed JC-1 solution prepared in the cell culture media was added to the wells. The plates were incubated for 30 min and washed with dilution buffer containing CH-2-102. Following the last wash, the wells were read using a microplate reader at Ex 475 ± 20 nm/Em 530 ± 15 nm and 590 ± 17.5 nm. The ratio of aggregate/monomer was plotted against the drug concentration.

2.16. Effect of CH-2-102 on cell metabolism

Oxygen consumption rate (OCR) measurement was used for evaluating mitochondrial function/dysfunction in vitro culture conditions. We treated A375 cells with four different concentrations of CH-2-102 (0, 25, and 50 nM) for 16 h and OCR was determined in response to different treatments using Seahorse XFe96 Analyzer (Agilent, USA). The cells were sequentially treated with the ATP synthase inhibitor oligomycin, the mitochondrial oxidative phosphorylation uncoupler FCCP (carbonylcyanide-4-trifluoromethoxyphenylhydrazone), and a mixture of the electron transport chain inhibitors rotenone and antimycin A during the experiment. The measurements were taken 12 times throughout the experiment to calculate for the OCR, attributed to basal ATP coupled, maximal respiration, and spare capacity

2.17. Polymer synthesis

2.17.1. The pH-sensitive polymer was synthesized in the following three steps

2.17.1.1. Synthesis of mPEG-b-PCC and 4-hydroxy benzaldehyde (BA) pendant group-containing polymer.

mPEG-b-PCC was synthesized as per the reported method [26].

2.17.1.2. Synthesis of glycine containing linker.

The synthesis was carried out as described in the publication with minor modifications. CH-2-102 (1.3 mM) was reacted with Boc protected glycine (1.97 mM) linker by the standard coupling reaction using EDC/4-dimethyl aminopyridine (DMAP) as a coupling agent in anhydrous DCM. The Boc group was removed by treating the product with 4 M HCl in dioxane at RT for 3 h. The crude CH-2-102-glycine was purified by column chromatography and characterized by ¹H NMR and Mass spectrometry.

2.17.1.3. Synthesis of mPEG-b-PCC-g-BA-CH-2-102.

A standard Schiff base reaction was utilized to conjugate the carboxylic acid group of CH-2-102-glycine and aldehyde units of mPEG-b-PCC-g-BA backbone in presence of triethylamine (TEA) and glacial acetic acid (GAA) dissolved in DMSO at RT for 48 h. The pH-sensitive polymer was characterized by ¹HNMR.

2.17.2. Synthesis of non-sensitive polymer

mPEG-b-PCC polymer was synthesized as described above. Dodecanol lipid and CH-2-102 were grafted onto the copolymer by EDC/HOBt conjugation reaction at RT for 24 h. The final product poly(ethylene glycol)-block-poly(2-methyl-2-carboxyl-polycarbonate-graft-dodecanol-graft-CH-2-102) mPEG-b-PCC-g-DC-g-CH-2-102 was characterized by ¹H NMR (Bruker 500, MA)

2.18. Preparation and characterization of nanoparticles

NPs were prepared by thin-film hydration. Briefly, the polymer (5 mg) was dissolved in chloroform in a glass vial and evaporated in a rotary evaporator to form a thin film. 1 mL of PBS was added to the film and vortexed until a clear solution was obtained. The hydrodynamic average diameter and surface charge (zeta potential) of NPs were accessed by dynamic light scattering (DLS) with a Zetasizer Nano ZS (Malvern, Westborough, MA). Drug concentration was determined by HPLC as described above. For drug content determination, 500 μL of NPs were mixed with 500 μL of 0.5 N HCl, vortexed for 15 min, and neutralized with 0.5 N NaOH. The mixture was added to 200 μL aliquot of acetonitrile for 5 min, vortexed for 5 min, centrifuged at 10,000 rpm for 5 min at RT, and the supernatant injected into the HPLC. The drug content in all these experiments was determined relative to the peak area of drug standards (0.48–500 μg/mL) in acetonitrile: Water (1:1). The study was performed with three different formulations and reported as the mean ± S.D. The drug release profile of NPs was determined by using a Float-A-Lyzer^® with 3.5 kDa molecular weight cut-off (Repligen, Waltam, MA). Briefly, 1 mL of NPs were taken in the dialysis membrane, and it was dialyzed against 20 mL of 1% tween 80 containing PBS at pH 7.4 and 5.0. Samples were shaken at 120 rpm for 48 h while maintaining a temperature at 37 °C. A 0.5 mL aliquot was drawn at preset time intervals, and the volume was replenished with fresh medium. The samples were then diluted with 0.5 mL acetonitrile, centrifuged at 5000 rpm and the supernatant was analyzed with HPLC. The drug concentration was calculated relateive to the peak area of drug standards (0.48–500 μg/mL). Three independent formulations were analyzed and reported as the mean ± S.D.

2.19. In vivo studies

All animal experiments were performed in accordance with the protocol approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Nebraska Medical Center (UNMC), Omaha, NE. Lung metastatic mouse model of melanoma was established in 8–10 weeks old female C57BL/6 albino mice by injecting B16-F10 cells (2 × 10⁵) expressing luciferase and green fluorescence protein (GFP) in 100 μL PBS through the tail vein. Based on the bioluminescence signal measured by IVIS instrument (1 × 10⁶ p/s/cm² sr), mice were unbiasedly divided into the following four groups: a) control (CTR), b) Free drug (FD), c) NPs and d) N-PTX. Mice in FD and NP groups were treated with free CH-2-102 in co-solvent and PEG-b-PCC-g-BA-g-CH-2-102, respectively at 5 mg/kg dose intravenously while control (CTR) groups were injected with PBS solution. The co-solvent was prepared by dissolving the drug in a 1:2:2 ratio of N, N-dimethyl acetamide, PEG 600, and sterile water for injection. All the groups were injected with 5 doses with injections on alternate days. Bioluminescent radiance of the tumors was measured a day before injection and every 3rd day using the IVIS Spectrum^® imaging system (PerkinElmer Inc., MA). After the treatment period, mice were sacrificed, and tumor metastasized lungs were harvested and processed for analysis. Three representative tumors containing lungs were collected per group and fixed with 10% buffered formalin for 24 h. Sections of 4.0 μm thickness were prepared from the fixed samples and stained with hematoxylin and eosin (H&E), Ki67, and cleaved Caspase-3 (CC3).

2.20. Statistical analysis

Data represented as mean ± S.D. The statistical comparisons of the experiments were performed by two-tailed Student’s t-test, with p < 0.05 was considered statistically significant. One-way ANOVA was used to compare more than two groups.

3. Results

3.1. Synthesis and characterization of CH-2-102

Synthesis scheme for CH-2-102 is provided in Fig. 1A. CH-2-102 is off-white compound, and its chemical structure and purity were confirmed by ¹H NMR (Fig. 1B). ¹H NMR (400 MHz, Chloroform-d) δ 8.11 (d, J = 7.9 Hz, 2H), 8.00 (dt, J = 11.9, 4.7 Hz, 3H), 7.66 (s, 2H), 7.48 (d, J = 7.8 Hz, 2H), 4.78 (d, J = 5.7 Hz, 2H), 3.98 (s, 3H), 3.89 (s, 6H), 1.76 (t, J = 6.0 Hz, 1H). ¹³C NMR (101 MHz, Chloroform-d) δ 191.70, 155.24, 155.03, 152.60, 142.49, 142.28, 138.01, 137.62, 131.20, 127.32, 126.98, 123.19, 122.26, 109.06, 64.92, 61.02, 56.24. HRMS [C22H22NO5⁺]⁺: calculated 380.1498, found 380.1526. HPLC purity 95.25% (tR = 3.44 min). The solubility of CH-2-102 was found to be 0.97 ± 0.16 μg/mL and 1.14 ± 0.06 μg/mLat pH 7.4 and 5.0 respectively at 37 °C.

3.2. CH-2-102 inhibits tubulin polymerization by binding to the colchicine binding site of the tubulin protein

As CH-2-102 was synthesized as a tubulin polymerization inhibitor, we established the efficacy of the compound with three different experiments.

3.2.1. Molecular modeling confirmed the interaction of CH-2-102 with colchicine binding site of the target tubulin protein

The reference compound used for comparison was the crystal structure of the molecule DJ-101, which shows a complex with tubulin at a resolution of 2.8 Å [27]. The simulated binding of DJ-101 (Orange thin tube) and CH-2-102 (rose ball and stick) in the tubulin crystal structure (PDB code: 5H7O) was similar with the Glide docking score – 11.6 and – 10.7, respectively. The study confirmed that CH-2-102 interacts with the tubulin protein similarly (Fig. 2A). The interaction was strongly stabilized by two hydrogen bonds (blue dashed lines): one formed between the carbonyl of CH-2-102 and β-Asp249, and one contributed by hydroxymethyl from the phenyl group of CH-2-102 and β-Asn256. The study confirmed the similar interaction of CH-2-102 with the tubulin protein as the native ligand.

Fig. 2.

CH-2-102 inhibits tubulin polymerization by binding to the colchicine binding site in tubulin protein. A) The simulated binding of DJ-101 and CH-2-102 in the tubulin crystal structure (PDB code: 5H7O). Superposition of DJ-101 (orange thin tube model; Glide docking score – 11.6) with CH-2-102 (rose ball-and-stick model; Glide docking score – 10.7). B) CH-2-102 inhibited tubulin polymerization in vitro as efficiently as colchicine but much more efficiently than QW-296, while PTX increased tubulin polymerization. C) Confocal microscopic images of A375 cells at 24 h post-incubation with DMSO, 12 nM colchicine, 12 nM PTX, 8 and 12 nM of CH-2-102 show reduced microtubule network scattered in the cytoplasm of the cells. In contrast, PTX treated cells had enhanced microtubule polymerization with thick bundle around the nucleus.

3.2.2. In vitro tubulin polymerization assay confirmed the interaction of CH-2-102 with the tubulin protein and identified its potential antineoplastic use

The inhibitory effect of CH-2-102 on tubulin polymerization was similar to that of colchicine but much better than QW-296 (Fig. 2B) [25]. This fluorescence polymerization assay indicates that CH-2-102 has the same mechanism of tubulin polymerization inhibitor as the earlier generation of analogs.

3.2.3. Immunofluorescence confocal microscopy visualized the effect of different tubulin-acting agents in A375 cells post 24 h incubation with colchicine, PTX, or CH-2-102

Colchicine and CH-2-102 treated cells demonstrated reduced and fragmented microtubule networks scattered in the cytoplasm, which indicates the destabilization of microtubules. In contrast, the microtubule-stabilizing agent PTX showed an enhanced tubulin polymerization and formed thick microtubule bundles encircling the nucleus (Fig. 2C). The effect of CH-2-102 on tubulin polymerization, even at a lower concentration of 8 nM was similar to that of colchicine. The microtubule network in the untreated or DMSO alone treated cells was unaffected. These results were consistent with the tubulin polymerization assay and reaffirm that CH-2-102 acts on tubulin protein to destabilize the polymerization as effectively as colchicine.

3.3. CH-2-102 shows the effective anticancer activity in vitro

We determined the anti-tumor and antimotility activities of CH-2-102 in A375 and B16-F10 melanoma cells and found it significantly more potent than colchicine and QW-296.

3.3.1. CH-2-102 inhibits proliferation of cells in 2D monolayer and 3D spheroid cultures

The 2D monolayer cell proliferation assay was performed on two different cancer cells, A375 cells of human origin and B16-F10 cells of murine origin. There was concentration-dependent cell killing with IC₅₀ value of 13.4 ± 0.6 nM and 18.2 ± 0.95 respectively for A375 and B16-F10 cells (Fig. 3A and B). QW-296, the tubulin inhibitor explored in our lab, had the IC₅₀ value of 40.8 nM and 66.8 nM in A375 and B16-F10-cells, respectively. These results from QW-296 are consistent with our earlier publication [25]. Thus, CH-2-102 showed significantly higher inhibition of proliferation than the earlier generation of a tubulin polymerization inhibitor, QW-296.

Fig. 3.

CH-2-102 effectively inhibits melanoma cell proliferation and tumor spheroids growth. CH-2-102 showed better cell killing efficacy than QW-296, as determined by MTT assay with A) A375 and B) B16-F10 melanoma cell lines. Fluorescence microscope images of the tumor spheroids generated with C) A375 cells and D) B16-F10 cells after 7 days of treatment with IC₅₀ CH-2-102 compared to vehicle control (DMSO). Tumor spheroids were grown for 7 days before the treatment.

The 3D spheroids mimic the features of tumors in the human body which mainly includes more realistic physiological responses, spatial architecture, and drug resistance mechanisms [28]. After 7 days of treatment, we observed that the spheroids of A375 cells treated with 12 nM of CH-2-102 were smaller in comparison to the control (Fig. 3C). Also, the control spheroids had extensive branching from the main body, and branching ability was curbed with CH-2-102 treatment. In contrast to the spread-out spheroids of A375 cells, B16-F10 cells had dense spheroids (Fig. 3D). It is clearly observed that the tumor spheroid size decreases with drug treatment at the end of the 7th day. Similarly, the live-dead staining showed that CH-2-102 treated spheroids had a lot of dead (stained red by ethidium bromide) cells in both A375 and B16-F10 cell lines compared to the controls.

3.3.2. CH-2-102 inhibits cell cycle progression at the G2/M phase and leads to apoptosis

In both A375 and B16-F10 cells, we observed the cell arrest in the G2/M phase in concentration-dependent manner when treated with CH-2-102 for 24 h. In A375 cells, the progression of the cell cycle was arrested significantly at G2/M with 13.03% in control vs 60.01% and 82.5% in 8 nM and 12 nM treated groups respectively (Fig. 4A). In the case of B16-F10 cells, the arrest in the cell cycle was significant in control, 12 and 20 nM concentrations with 7.02%, 18.78%, and 81.97% cells arrested in the G2/M phase, respectively (Fig. 4B). Additionally, protein expression of Cyclin B1 was upregulated with increasing concentration in A375 and B16-F10 cells (Fig. 4E). This regulatory protein Cyclin B1 is involved in mitosis and is predominantly expressed during G2/M phase of the cell cycle and increases in 24 h post G2/M phase arrest [29,30]. This confirms that our drug can inhibit cell division through mitosis. CH-2-102 induced apoptotic cell death in A375 and B16-F10 cells post 24 h incubation with the drug at 8 and 12 nM concentrations in the former and 12 and 20 nM in the later cell lines as determined using Annexin V assay (Fig. 4C, D). In A375 cells, the total apoptotic cell population after treatment with DMSO (control), 8 nM and 12 nM CH-2-102 was 13.9% to 31.9% and 56.8%, respectively. Similarly, in B16-F10 cells, the percentage of the apoptotic cells increased from 16.4% in the control to 36.7% and 60.2%, respectively, with 12 nM and 20 nM CH-2-102 treatment groups (Fig. 4C, D). Cells arrested in the G2 phase would tend to accumulate cyclin B1 protein [31]. Consequently, we found a higher levels of cyclin B1 protein in CH-2-102 treated cells (Fig. 4E, Fig. S2A, B). One characteristic event of apoptotic cell death is the proteolytic cleavage of poly(ADP-ribose) polymerase, which is a nuclear enzyme responsible for DNA repair, genomic stability, and transcriptional regulation [32]. In apoptotic cells, the increased Caspase 3 enzyme results in the cleavage of PARP. Therefore, we determined the expression levels of these proteins by Western blot analysis in A375 and B16-F10 cells post-treatment. With increasing drug concentration, there was an increase in cleaved caspase 3 and a decrease in PARP-1 protein (Fig. 4F, Fig. S2C, D). These observations showcase the mechanism of cell death initiated by inhibition of cell cycle progression post 24 h treatment with CH-2-102.

Fig. 4.

CH-2-102 arrests the cell cycle in the G2/M phase and induces dose-dependent apoptosis in melanoma cells. A) A375 and B) B16-F10 cell cycle analysis post CH-2-102 treatment. The figures also show the change in DNA content with different treatments (0, 8 and 12 nM in A375 cells and 0, 12 and 20 nM in B16-F10 cells). Effect of CH-2-102 treatment on the induction of apoptosis in C) A375 cells and D) B16-F10 cells. E) Western blot showed dose-dependent increase in Cyclin B1 signifying the cell arrest in G2/M phase. Similarly, F) showed the dose-dependent increase in cleaved Caspase-3 and a decrease in PARP in both A375 and B16-F10 cells. Data represented as mean ± S.D. (n = 3). *p < 0.05, **p < 0.01, *** < 0.001 compared to Control, ns-non-significant.

3.3.3. CH-2-102 inhibits melanoma cell migration

Incubation of A375 and B16-F10 cells with CH-2-102 at its IC₅₀ concentrations delayed the wound healing significantly compared to lower concentrations (8 nM and 12 nM respectively) of the drug, equivalent concentration of colchicine and DMSO control till 24 h (Fig. 5A& B). The inability of cancer cells to migrate post-treatment with CH-2-102 is reflected by the transwell migration assay. We observed that the drug at 12 nM significantly decreased the migration of A375 cells across the membrane, while the at 8 nM CH-2-102 and positive control (12 nM colchicine), and DMSO control were not as efficient in inhibiting the migration of the cells as indicated by the crystal violet staining. Image J analysis revealed that % area occupied by the migrated A375 cells decreased from 56.3%, 56.7% and 47.7% to 24.3% in control, colchicine (12 nM), CH-2-102 (8 nM), and CH-2-102 (12 nM), respectively. For B16-F10 cells, migrated cells decreased from 70.2%, 47.5%, 19.49% to 3.75% in control, colchicine, lower concentration, and higher concentration respectively (Fig. 5C& D). In addition, the protein expression revealed that vimentin, a filament protein, usually upregulates during the endothelial mesenchymal transition (EMT) and is downregulated in a concentration-dependent manner with CH-2-102 treatment (Fig. 5E). This shows that our compound is effective in mitigating the migratory behavior of cancerous cells.

Fig. 5.

CH-2-102 inhibits the migratory behavior of melanoma cells. A) A375 cells were grown until confluency, scratched, and treated with CH-2-102 (8 or 12 nM) and colchicine (12 nM) for 24 h. Similar treatment was performed on B) B16-F10 cells as well. The images were taken at 0 and 24 h. Transwell membrane assay for cell migration. C) A375 cells were cultured in the transwell membrane (Boyden chamber), treated with CH-2-102 (8 nM or 12 nM) or colchicine (12 nM), incubated for 24 h and visualized under a brightfield microscope. The images also show the ImageJ software analyzed quantification of the % area covered by migrated cells under the influence of the compounds. D) Similarly, B16-F10 cells were treated with CH-2-102 (12 nM of 20 mM) or colchicine (20 nM) 24 h and imaged, which were quantified with the % area covered by migrated cells under the influence of compound E) Effect of CH-2-102 on cell migration and EMT marker, vimentin, a dose-dependent decrease in the protein concentration was observed in both the cell lines. Data represented as mean ± S.D. (n = 3). *p < 0.05, **p < 0.01, *** < 0.001 compared to Control, ns-non-significant.

3.3.4. CH-2-102 inhibits the colony-forming ability of cancer cells

A375 and B16-F10 cells were treated with CH-2-102 for short-term and long-term to evaluate its effect on the colony forming ability of these cells. For the long-term treatment, CH-2-102 was continuously exposed to the cells for 14 days; IC₅₀ concentration was able to significantly reduce the cell proliferation than colchicine. The lower concentrations of CH-2-102 were more effective than colchicine in both the cell lines (Fig. 6A& C). Percentage area covered by the colonies in control, colchicine, CH-2-102 (8 nM) and CH-2-102 (12 nM) treated groups were 32.1%, 7.64%, 2.06% and 0.47%, respectively in A375 cells, while it was 57.5%, 37.5%, 12.9%, 2.3%, respectively in B16-F10 cells. For the short-term treatment, cells were treated with CH-2-102 for 24 h and then allowed to grow for the next 14 days in untreated media. There was a significant difference in colony population in the untreated control cells and the cells treated with CH-2-102 at its IC₅₀ concentration (Fig. 6B& D). Percentage area covered by the colonies in control, colchicine, CH-2-102 (8 nM) and CH-2-102 (12 nM) treated groups were 52.4%, 33.1%, 6.9% and 5.1%, respectively in A375 cells and 91.2, 46.9%, 22.3%, 8.9% respectively in B16-F10 cells. In each case, colchicine in equivalent IC₅₀ concentration was less effective in controlling the colony formation ability. CH-2-102 showed concentration-dependent inhibition of colony formation. This study indicates that CH-2-102 is effective in inhibiting the unlimited growth of cancerous cells post single-dose without any short-term resistance.

Fig. 6.

CH-2-102 inhibits the colony-forming ability of cancer cells. Colonies formed after the incubation of A375 cells with CH-2-102 at 8 nM, 12 nM and colchicine 12 nM for A) long-term and B) short-term; C) long-term D) short-term treatment of B16-F10 cells with 12 nM, 20 nM, and colchicine 12 nM. For long-term treatment, cells were continuously exposed to the drug for 14 days, while the short-term treatment, the cells were treated with the drug for 24 h followed by growth in fresh media for 2 weeks. The quantification by measuring the surface area occupied by the colonies in ImageJ software is also presented. Data represented as the mean ± S.D. (n = 3). *p < 0.05, **p < 0.01, *** < 0.001 compared to Control, ns- non-significant.

3.3.5. CH-2-102 is effective against PTX resistant cell lines

PTX resistant A375 and B16-F10 cells were generated by incubating these cells with increasing concentrations of PTX over 3 months and were named A375-TXR and B16-F10-TXR. CH-2-102 was effective in inhibiting the proliferation of both non-resistant and PTX resistant A375 and B16-F20 cells, with resistance indices of 1.9 and 2.1 in A375/A375-TXR and B16-F10/B16-F10-TXR cells, respectively. In contrast, PTX showed significantly low potency with IC₅₀ values of 1156.4 nM and 2343.9 nM in A375-TXR and B16-F10-TXR cells, respectively. With a resistance indices of 44.5 and 113.8. In contrast, for non-resistant A375 and B16-F10 cells, PTX had IC₅₀ values of 25.99 nM and 20.6 nM, respectively (Fig.7A,B & Table 1). These results demonstrate that CH-2-102 can maintain the cytotoxicity profile of cancers to their drug-resistant variants as a single agent.

Fig. 7.

CH-2-102 shows activity against taxane-resistant cells; is not a substrate of Pgp; doesn’t have a long half-life; impacts the mitochondrial potential and decreases the Oxygen consumption rate. A) Effect of CH-2-102 on taxane resistant A375-TXR and B16-F10-TXR cells post 48 h treatment using MTT assay. B) Effect of Paclitaxel on A375 and B16-F10 cells post 48 h treatment using MTT assay. C) Effect of CH-2-102 on Pgp ATPase activity. Change in luminescence (ΔRLU) compared to 100 μM sodium orthovanadate treated samples was plotted (mean ± S.D., n = 3). CH-2-102 at three concentrations (10 nM, 100 nM and 1000 nM) showed a non-significant effect compared to NT control on Pgp ATPase activity. D) The concentration vs. time graph for the stability of CH-2-102 in human liver microsomes. The insert of log [concentration] vs. time graph represents the straight line used to calculate the half-life of CH-2-102. E) The impact of CH-2-102 on the mitochondrial potential of A375 and B16-F10 cells under treatment with three different concentrations of low (12 nM and 20 nM), medium (25 nM and 50 nM), and high (50 nM and 100 nM), respectively. F) The graph shows the effect of 3 different concentrations of CH-2-102 on the overall mitochondrial respiration in A375 cells. *p < 0.05, **p < 0.01, *** < 0.001 compared to control, ns non-significant.

Table 1.

IC₅₀ values of CH-2-102 and PTX in A375, B16-F10, A375-TXR and B16-F10-TXR cells with their Resistance indices.

Cells	PTX (nM)	CH-2-102 (nM)
A375	25.99	13.4
A375-TXR	1156.4	27.5
Resistance Index (RI)	44.5	2.1
B16-F10	20.6	18.2
B16-F10-TXR	2343.9	34.6
Resistance Index (RI)	113.8	1.9

3.3.6. CH-2-102 is not the substrate for P-gp

Changes in luminescence (ΔRLU) of the samples post CH-2-102 and control treatment samples were plotted to illustrate the stimulatory or inhibitory effect on the Pgp ATPase activity (Fig. 7C). A well-known Pgp substrate, verapamil, stimulated Pgp ATPase activity and consumed ATP, resulting in a sharp change in luminescence. On the other hand, the luminescence of the three different concentrations of CH-2-102 (10, 100, and 1000 nM) was not significantly different from the vehicle control-treated group, which suggests that CH-2-102 has no effect on Pgp ATPase activity. Thus, indicating that CH-2-102 was not a substrate for Pgp and might be the reason behind its effect on taxane resistant cells.

3.3.7. CH-2-102 is not stable against the human liver microsomes

Our previous studies show that several members acting through the colchicine binding site are prone to rapid hydrolysis by liver enzymes and have low bioavailability [33]. Therefore, we determined the CH-2-102 stability in the presence of human liver microsomes. As shown in Fig. 7D, CH-2-102 is almost completely metabolized towards the end of an hour. The equation obtained from the log concentration vs. time graph embedded in the figure was used to calculate the half-life of CH-2-102 in the presence of liver microsomes which was found to be 27.2 min. The short half-life of the compound warrants a robust delivery system which could increase the circulation time and make it more effective in the management of cancer.

3.3.8. CH-2-102 disrupts the mitochondrial membrane potential (ΔΨm) (MMP)

The ratio of aggregate/monomer of JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimi- dazolylcarbocyanine iodide) provides the estimate of the health of the mitochondria in the cells, with healthy cells maintaining the higher ratio compared to the unhealthy cells. Mesoxalonitrile 4-trifluoromethoxyphenylhydrazone (FCCP) is a positive control and collapsed the MMP with aggregate/monomer ratio of 0.22 and 0.085 in A375 and B16-F10 cells, respectively. CH-2-102, on the other hand gradually decreased the aggregate/monomer ratio to 2.51, 2.26 and 1.29 vs 3.39, 2.69, 0.75 with increasing concentration in A375 (12, 25 and 50 nM) and B16-F10 (20, 50 and 100 nM) cells, respectively (Fig. 7E). The negative control of DMSO treated cells maintain the aggregate/monomer ratio of 4.79 and 4.72 than the treated cells, in A375 and B16-F10 cells respectively.

3.3.9. CH-2-102 decreases the oxygen consumption rate in cancer cells

As seen in the graph (Fig. 7F), there was an apparent difference in mitochondrial respiration parameters between the control and treated groups, while different concentrations in the latter had a similar effect. The treatments (12 nM, 25 nM and 50 nM) induced a significant decrease in both basal (164.4 ± 4.2 pmole/min vs 76.6 ± 9.3, 80.8 ± 11.2, 62.0 ± 17.0 pmole/min) and ATP-linked (135.7 ± 3.4 vs 58.6 ± 8.3, 60.5 ± 7.6, 45.7 ± 14.2 pmole/min) OCRs compared to the control groups. The decrease in maximal respiration (185.2 ± 4.3 pmole/min vs 90.3 ± 11.1, 95.0 ± 10.4, 68.1 ± 23.4 pmole/min and spare capacity (20.8 ± 3.9 pmole/min vs 13.8 ± 3.1, 14.2 ± 1.6, 6.1 ± 11.7 pmole/min) post-treatment represents the interference with the mitochondrial electron transport chain (ETC). In addition, a decrease in spare capacity indicates a decrease in response to ATP demand. The proton leak among different treatment groups was similar, referring to the integrity of the mitochondrial electron transport chain (ETC). Overall, the data indicated that CH-2-102 treatment damaged mitochondrial ETC and reduced cellular responses to ATP demand.

3.4. pH-sensitive nanoparticles are suitable for CH-2-102 delivery

The pH-sensitive polymer was synthesized as illustrated in Fig. 8. The products from each step were confirmed with ¹H NMR. We have outlined the details of the polymeric backbone in our earlier publication [34]. For mPEG-g-PCC-g-BA, we observed two aromatic peaks at δ7–8 and one peaks at δ10 (-CHO) (Fig. S3A). For the BOC (tert-butoxycarbonyl) protection of CH-2-102, we observed a prominent peak at δ 1.4, which was lost after the deprotection reaction to obtain CH-2-102-Glycine HCl (Fig. S3B). The product of this reaction was then utilized to synthesize mPEG-b-PCC-g-CH, which showed the representative peaks from the polymeric backbone as well as CH-2-102 aromatic protons at δ7–8, O-CH₃ protons (δ3.7 and δ3.9) and benzyl protons CH₂ at δ5.2 (Fig. S3C). The nonresponsive polymer, mPEG-b-PCC-g-DC-g-CH-2-102 was also synthesized and characterized by ¹H NMR as described above. The final product showed the characteristic peaks of CH-2-102 (aromatic protons at δ7–8, O-CH₃ protons (δ3.7 and δ3.9) and benzyl protons CH₂ at δ5.2) and dodecanal (-CH₃ at δ 0.9 and CH₂ at δ1.2 (Fig. S4). High drug content of 17.3. ±2.4% (w/w) in the pH sensitive polymeric nanoformulation (NP-S) and 40 ± 2.0% non-sensitive NPs was measured by RP-HPLC. NPs showed unimodal particle distribution with an average particle size of 60 ± 2.5 nm with PDI of 0.211 (Fig. S5A). When the pH of ph-responsive NPs was changed to 5.0 and mixed thoroughly using a shaker for 4 h, the size distribution profile changed significantly. We observed a multimodal distribution with multiple peaks, most likely due to cleavage of sensitive linker and aggregation of particles in lower pH of 5.0 (Fig. S5B). In contrast, non-responsive NPs had an average particle size of 76.7 nm and PDI of 0.226. Similar observations were made during the morphological analysis using Atomic Force Microscopy (AFM), where the NP-S particles at pH 7.4 were uniform in size (Fig. S5C) with an average particle size of 29.0 ± 2.2 nm. However, the reduction in pH to 5.0 changed the size and shape in AFM images to multi-model distribution with a wide distribution range. We could observe the distribution of very small as well as large particles in the AFM, just like DLS with a mean particle size of 39.2 ± 10.1 nm (Fig. S5D). We expected NPs to release the drug at a relatively fast rate in the acidic pH. Therefore, we performed the drug release studies at pH 5.0 and 7.4 for both NP-S and NP-NS. We observed that NP-S released almost 50% of the drug in the first 8 h, followed by a slow release for up to 48 h with 80% of the drug released in the acidic pH while the release at neutral pH was only 20% in 48 h (Fig. S6A). For the NP-S, 15% of the drug was recovered from the dialysis bag. Non-responsive NPs, on the other hand, releases around 6% and 8% in 4 h at pH 5.0 and 7.4 respectively. The drug release from non-responsive NPs only increased slightly until 48 h in both the pH. The polymeric formulations (NP-S, NP-NS, and N-PTX) were then filtered through a sterile filter and utilized for the MTT assay. We observed IC₅₀ values of 11.49 nM, 242.0 nM, and 54.14 nM in A375 cells and 15.91 nM, 359.88 nM, and 61.55 nM, in B16-F10 cells for NP-S, NP-NS and N-PTX formulations, respectively (Fig. S6B).

Fig. 8.

3 step synthesis of pH-sensitive mPEG-b-PCC-g-BA-CH-2-102: Step 1: shows the polymerization of mPEG-b-PCC polymer backbone followed by conjugation of 4-hydrobenzaldehyder onto the backbone. Step 2: shows the modification of hydroxyl group amine with glycine conjugation for again of two carbons. Step 3: synthesis of the pH-sensitive polymer by Schiff’s reaction.

3.5. pH-sensitive nanoparticles effectively deliver CH-2-102 to a syngeneic lung metastatic melanoma mouse model

The preclinical effectiveness of CH-2-102 was evaluated in the lung metastatic syngeneic melanoma model generated by injecting luciferase-expressing B16-F10 cells into NSG mice via the tail vein. The mice with the bioluminescent radiance of 1 × 10⁶ photon/s/cm²/sr were randomly divided into four groups. The control (CTR) group was treated with PBS, while the free drug (FD) group was treated with a co-solvent solution of CH-2-102 at the dose of 5 mg/kg. N-PTX group was treated with the 5 mg/kg dose of N-PTX in 100 μL volume just like all other treatment regimens. The NP-S group was treated with a pH-sensitive formulation at the dose of 5 mg/kg of CH-2-102. The NP-NS group was treated with non-pH-sensitive formulation at the dose of 5 mg/kg of CH-2-102. The bioluminescent images show the spread of tumors in the lungs over time. (Fig. 9A). The top images show the bioluminescence at day 0, a day before the treatment began, while the bottom images represent the bioluminescence at day 11, just before euthanasia.

Fig. 9.

In vivo therapeutic efficacy of CH-2-102 conjugated NPs in syngeneic lung tumor metastasis mice model of melanoma. A) In vivo representative bioluminescent images at day 0 (a day before the treatment began) and day 11 of treatment. Bioluminescent images of mice in each group were taken every 3 days during the treatment (n = 6), B) Representative lungs (front and back) showing tumor nodules of each group excised post euthanasia at the end of the study. C) Radiance intensity plot of all treatment groups measured from day 0 to the end of the study. D) Bodyweight change (%) in mice during treatment in all groups. E) Survival plot of all the animals in each treatment groups. Black arrows indicate the day of injection. Data represented as mean ± S.D. (n = 6). The tumor nodules from all treatment groups were stained with hematoxylin and eosin (H&E), proliferation marker Ki67, and apoptosis marker cleaved Caspase 3. The images were captured at 10× magnification, and the floating zoom box magnified further by 4×. Data represented as the mean ± S.D. (n = 6).

The euthanized mice were dissected and the lungs with tumor nodules were harvested along with other vital organs. The lungs were then pumped with 10% formalin solution/PBS to take the images. The lung images clearly show that the mice in the control group had significantly more nodules compared to those in other treatment groups (Fig. 9B). Due to the metastasized melanoma, the lungs in the control group lost the ability to stretch to hold the liquid and leaked. The observed collapsed lungs in the images are due to the lung losing its property due to metastasis. The front and back (show the heart as well) images clearly show the metastasis in the lung as well as the surrounding tissues. The N-PTX treated group, on the other hand, showed significant variation in metastasis. Some lungs had minimal tumor nodules, while others had more. The FD treated group had a lot of small nodules scattered throughout the lungs but was still better than the NP-NS group where the front of the lungs looked similar to that of the N-PTX treated group, but the back had significantly more nodule. NP-S treated group had small nodules scattered on the lungs and looked significantly cleaner than the other groups. We monitored the change in mouse body weight during treatment. The % of the change in weight was acceptable in all treatment groups (Fig. 9C).

During this study, 2 mice from the control group and 1 mouse from the N-PTX group did not make it till euthanasia.(Fig. 9D). The change in luminescence during the treatment is shown in Fig. 9E, and as observed before the N-PTX treated group had a large variation compared to the other groups. The histological images obtained for different treatment groups showed that the extent of metastasis in the lungs was significantly lower in the NP-S treated group compared to the other groups. H&E staining showed extensive metastasis in the control, N-PTX, and non-reponsive NPs groups compared to the free drug and pH-responsive NPs groups. A similar pattern was observed with Ki-67 protein staining where the active phases of cell proliferation were stained in the control, N-PTX, and non-reponsive NPs groups and decreased drastically with free drug and pH-responsive NPs groups. Cleaved Caspase-3, on the other hand, increases with an increase in apoptosis and can be seen even in very small tumor nodules in the lungs in pH-responsive NPs NP-S treated group while other groups seem to have less expression of cleaved-caspase 3 despite the large proliferative tumor.

4. Discussion

Tubulin binding agents induce inhibition of microtubule dynamics and are the most popular chemotherapy drugs to treat human malignancies [35]. In phase II and III clinical trials, N-PTX compared favorably against DTIC in progression-free survival (PFS), overall survival (OS), and response rate [17,36]. However, taxane drugs are prone to resistance. Alternatively, tubulin binding agents having affinity with colchicine binding site (CBS) such as colchicine, which inhibit tubulin polymerization, are less resistant prone [37], but shows high-grade toxicity. Alternatively, combretastatin A-4 (CA-4) is another CBS agent with less toxicity. However, most of CA-4 molecules have low chemical stability and poor aqueous solubility. Our group has developed several new molecules, including exploring the potential therapeutic benefits as summarized elsewhere in detail [33]. The common structural feature of these analogs is 3,4,5-trimethoxyphenyl (TMP) moiety. Otherwise, structural modifications on other sites such as diaryl-ketone chemotypes, with various linkers such as a phenyl ring (I-387), 4-substituted methoxybenozyl arylthiazoles (SMART H, SMART-100), phenylaminothiazoles, (2-(1H-Indol-5-yl) thiazol-4-yl) 3, 4, 5-trimethoxyphenyl methanone (LY293) arylbenzoylimidazoles [ABI-286], reverse arylbenzoylimidazoles (RABIs), and 1H-imidazo[4,5-c]pyridine (DJ101), that selectively binds to the tubulin at colchicine binding site and stabilizes the tubulin at nanomolar concentration [24]. Most recently, we used 2-(4-hydroxy-1H-indol-3-yl)-1H-imidazol-4-yl)(3,4,5-trimethoxyphenyl)methanone (abbreviated QW-296) bearing a 4-hydroxyl group on the indole ring of ABI-III 2. QW-296 loaded NPs showed promising therapeutic efficiency in lung metastatic melanoma-bearing mice [25].

We observed more than three fold higher potency of our new compound CH-2-102 than QW296 in melanoma cell lines. As shown in Fig. 3A and B, the IC₅₀ of CH-2-102 was 13.4 nM and 18.2 nM compared to 40.8 nM, and 66.8 nM of QW296 in A375 and B16-F10 melanoma cell lines, respectively. Lower dose certainly could reduce the systemic toxicity as seen in other anti-tubulin drugs [38]. As CH-2-102 is a more potent analog of QW-296, we hypothesized that it would circumvent TXR mechanisms [24]. Our modeling data indicate that CH-2-102 binds at the interface of α/β-tubulin heterodimer and occupies the same locale as colchicine. In tubulin polymerization assay, while CH-2-102 and colchicine strongly inhibited tubulin polymerization, PTX showed robust tubulin polymerization at the same concentration. These results suggest that CH-2-102 is a more potent microtubule destabilizer than QW-296, and its mode of action differs from PTX (Fig. 2B). In the cell-based assay, the microtubule network was disrupted by CH-2-102 and displayed weaker fluorescence signals due to depolymerization and elevated free tubulin. PTX-treated cells, on the other hand, displayed dense tubulin structures accompanied by a strong fluorescence signal. This is because CH-2-102 and PTX are opposing groups with tubulin destabilizing and stabilizing effects, respectively (Fig. 2C). Notably, in PTX-resistant cells, CH-2-102 showed remarkable antiproliferative effects (Fig. 7A, B & Table 1). This effect may be explained by its interaction with the colchicine binding site of the tubulin and non-Pgp substrate, resulting a high degree of cellular retention compared to PTX (Fig. 7C). Together, CH-2-102 appears to be a unique microtubule destabilizing agent with anticancer properties that are potentially superior to PTX. Because CH-2-102 is an anti-microtubule drug, and a non-Pgp substrate, we reasoned that it would be most useful in those patients where traditional anti-microtubule therapies have failed.

Microtubules are required not only for cell growth but for migration activity of the cells [39]. Therefore, when we treated with CH-2-102, both the cells showed less migration in wound healing and transwell migration assay (Fig. 5A–E). Further, CH-2-102 strongly demonstrated its ability to inhibit anchorage-dependent colony propagation (Fig. 6A–D) and tumor spheroid growth assay (Fig. 3C, D &S1).

Mutations in the taxane binding site on β-tubulin and overexpression of the Pgp drug efflux pump, especially by cancer stem cells (CSCs), or βIII-tubulin are common mechanisms of their resistance [40,41]. Unlike PTX, CH-2-102 is not a Pgp substrate which is in good agreement with the observation that the melanoma cells derived spheroids remained sensitive to the treatment. It has been reported that microtubule disturber significantly decreases the bioenergetic activity of mitochondria and OCR [42]. This can be due to many possible reasons: mitochondria move along with microtubules and generate energy in the absence of microtubules dynamics, when this movement is decreased, and the energy production decreases as well. Furthermore, microtubules participate in the phosphorylation of mTOR, a central energy regulator, involved in cellular physiology. Our study observed that CH-2-102 treatment significantly decreased both basal and ATP-linked OCRs compared to the control groups (Fig. 7). Further, in the case of dysfunctional mitochondria, proton leak increases, and basal and maximal respiration decrease [42]. Basal respiration was lower in CH-2-102-treated cells than in control cells, demonstrating an increase in dysfunctional mitochondria in the latter cells. The SRC was also low in CH-2-102-treated cells than in controls, indicating that moderate energy persists in mitochondria upon microtubule destabilization. Mitochondria is also involved in the intrinsic apoptosis pathway. As we observed, our compound initiates apoptosis post-treatment, we performed a mitochondrial membrane potential (ΔΨm) assay to confirm the involvement of mitochondria in CH-2-102 induced apoptosis. This assay utilizes a lipophilic cationic dye JC-1, which is aggressively taken up by healthy mitochondria, but poorly in unhealthy mitochondria of the apoptotic cells. One of the early events of apoptosis is the reduction of mitochondrial membrane potential (ΔΨm) [43]. Tubulins are known to reversibly inhibit voltage-dependent anion channels (VDAC) through the insertion of the negatively charged c-terminal tail into the VDAC [44,45]. The tubulin destabilizing agents (e.g., colchicine) increase the free tubulin leading to depolarization of MMPs, while stabilizing agents (e.g., taxanes) lead to hyperpolarization [46]. In conclusion, our data confirms that CH-2-102 induces cell apoptosis via mitochondria.

The microsomal stability assay indicated that CH-2-102 is prone to significant metabolism with a half-life of about 27 min (Fig. 7D). It has been shown that the high vitro intrinsic clearance (CL_int) can be scaled to in vivo CL_int and bioavailability [47]. We hypothesized that CH-2-102 will show poor oral bioavailability due to first pass metabolism. A higher dosing frequency may be needed to maintain the effective concentration, which result in dose-related toxicity. Furthermore, CH-2-102 is a hydrophobic molecule that requires some suitable formulation approach for systemic administration. We selected a nanoparticulate approach because it can provide a longer systemic circulation, sustained drug release, and lower side effects. Several strategies such as physical entrapment into nanoformulations, coordinate/charged-based interactions, and polymeric drug conjugates are being actively investigated to improve the blood circulation of labile drugs [48]. Previously, we physically encapsulated hydrophobic labile drugs such as SF2523 into NPs and conjugated hydrophilic gemcitabine to the polymeric delivery systems for effective delivery to the tumor [34,49]. Despite poor aqueous solubility, CH-2-102 loading into the hydrophobic core of NPs was low with rapid drug release. To increase drug loading and decrease metabolism of CH-2-102 in vivo, we conjugated it to a biodegradable polymeric backbone. Drug conjugated to the carrier has high stability under physiological conditions and could effectively prevent premature release. Previously, we have shown that stimulus-responsive NPs are safe and more effective than their non-sensitive NPs due to their controlled drug release at the tumor site [50]. Rapidly growing tumor cells create the low oxygen and mildly acidic tumor microenvironment by producing an excess of lactic acid and hydrolysis of ATPs [51]. Therefore, for therapeutic concentration, drug release at the tumor site through pH-responsive bioconjugation is an attractive approach. To achieve this, we selected a Schiff base pH-sensitive chemical bond (Fig. 8). The bond used for bioconjugation in this study shows cleavage at pH of 5.5–6.5 but remains stable at pH 7.4 [52,53]. Therefore, we hypothesize that protonation of polymers would happen when pH drops from 7.4, which could lead to drug release at the tumor site [54–56].

Post injection of the formulation, it passively accumulates at the lung tumor site through enhanced permeation and retention (EPR) effect [57]. Since melanoma metastatic tumor has been shown to grow faster adjacent to the lung blood arteries, passive targeting via EPR has the potential to deliver NPs to lung metastases. The EPR phenomenon in melanoma metastatic tumors have been confirmed in the literature. For instance, in one of studies, polymeric micelles loaded with Fenretinide (N-(4-hydroxyphenyl) retinamide, 4-HPR) significantly enhanced drug tumor accumulation over the injection of 4-HPR encapsulated in oil-in-water (O/W) emulsions. Further, 4-HPR loaded micelles delayed tumor growth in B16BL6 mice as compared with PBS [58].

Upon drug accumulation into the tumor site and cellular uptake, drug releases from the NPs greatly influences by the pH changes starting the interstitial fluid (slightly acidic) to endosomes (~ pH 6.0) and lysosomes (pH 4.5–5.0). This gives a plenty of opportunities for the drug to be cleaved off the polymer backbone [59]. Consequently, the drug release from the pH-responsive NPs was relatively fast, with 50% of the drug released in the first few hours followed by the sustained release up to 80% of the cargo (Fig. S6A). The release window of 48 h provides us with the opportunity to design the dosing regimen in mice. After cleavage of Schiff’s linkage, there is an ester bond which is cleaved off quickly in the presence of intracellular esterase, thus making free drug available for binding with tubulin to exert its effect. As the drug forms the hydrophobic core, the formulation is expected to be intact till it reaches the tumor. For further validation of its anticancer potential, CH-2-102 conjugated NPs effectively inhibited melanoma tumor growth and lung metastasis in vivo in mouse models. CH-2-102 NPs decreased the tumor nodule counts and overall survival of mice compared to free drug and even N-PTX. Clinical trials and laboratory studies have demonstrated that N-PTX is an effective treatment for a number of cancer types [60–62]. In our in vivo study (Fig. 9), we observed that the effect of N-PTX varied among the mice. Some had a smaller number of nodules, while others had a lot more. This could be attributed to the dose of N-PTX used in this study. Here, we have used 5 mg/kg/2 days to compared to the same dose of CH-2-102, while the effective dose in the literature is 20–30 mg/kg/day. With this vast difference in dosing, the effectiveness of N-PTX could not be counted upon. Toxicity study was carried out by injecting CH-2-102 at the dose of 5 mg/kg every alternate day for two weeks. The results from the in vivo experiment were also represented in the in vitro cytotoxicity study were ph-responsive NPs was found to be more effective compared to N-PTX and non-responsive NPs (Fig. S6B). CH-2-102 conjugated NPs were effective in controlling the cancerous nodules in the lungs compared to the free drug cosolvent solution and even the commercial formulation due to prolonged circulation and controlled release (Fig. 9). Off-target adverse drug reactions often contribute to the high attrition rate in drug discovery and development. There is a clear correlation between dose/toxicity and weight loss for all chemotherapeutics in preclinical studies [63]. In our study, we did not observe any dose-related toxicity in mice.

In summary, we synthesized and evaluated a novel tubulin inhibitor targeting colchicine-binding site. CH-2-102 depolymerizes microtubules in vitro and disrupts microtubule morphology distinctly from PTX. CH-2-102 demonstrates strong antitumor efficacy and decreases the metastatic potential of two melanoma cells, A375 and B16-F10 cells. Remarkably, CH-2-102 maintains its efficacy even in PTX resistant cells. We successfully synthesized pH-sensitive polymer drug conjugate to avoid metabolic degradation, which self-assembles into NPs. These NPs are stable at neutral pH but show fast drug release upon decrease in acidic pH, mimicking the tumor microenvironment, which has a pH of around 5.0–7.2. Furthermore, CH-2-102 maintains potency and efficacy in syngeneic lung metastasis mouse models without causing apparent toxicity to major organs and possesses a good safety profile. Collectively, this preclinical evaluation strongly supports CH-2-102 for further development as a new generation of tubulin inhibitors targeting the colchicine-binding site.