Cancer is a heterogeneous group of diseases marked by uncontrolled cell proliferation and metastasis, a major cause of cancer-related mortality [1,2]. Despite intensive efforts to develop novel anticancer agents, drug resistance, often mediated by p-glycoprotein overexpression, remains a major challenge [[3], [4], [5], [6]]. Furthermore, many current treatments are limited by systemic toxicity and adverse effects [7,8], underscoring the urgent need for safer, more selective therapeutics.

Microtubules, composed of α/β-tubulin heterodimers, are vital for cell division and intracellular transport [9,10]. Their dynamic polymerization is essential for mitotic spindle formation and chromosome segregation [[11], [12], [13], [14]]. Notably, microtubules exhibit isotype diversity, with βI, βII, βIII, βIVa, and βV-tubulin isoforms displaying distinct expression patterns across cell types [15]. The βIII-tubulin isotype has emerged as a promising target for cancer therapy [16,17], as it is frequently overexpressed in tumor cells and implicated in drug resistance across various malignancies [18,19]. As a result, small molecules that modulate microtubule assembly represent a promising class of anticancer agents [[20], [21], [22]]. Several FDA-approved microtubule-targeting agents, including paclitaxel [23], vinblastine [24], and eribulin [25] (Fig. 1a), disrupt tubulin dynamics to arrest the cell cycle. While effective, their use is hindered by adverse effects (e.g., neuropathy, neutropenia) [26] and poor bioavailability [27], necessitating intravenous administration and limiting clinical utility.

Colchicine (Fig. 1b), a tricyclic alkaloid first isolated from Colchicum autumnale in 1820, has been used medicinally since ancient times [28]. It inhibits microtubule assembly by binding to tubulin and preventing the conformational change required for α/β-tubulin heterodimerization, leading to metaphase arrest [29,30]. This mechanism highlights its potential in cancer therapy. However, clinical translation of colchicine and its analogs has been limited by significant toxicity. The success of low-dose colchicine in non-cancer indications, such as gout [31], has reignited interest in designing analogs with improved safety and efficacy. This has led to extensive structure–activity relationship (SAR) studies aimed at enhancing potency, isotype selectivity, and therapeutic index. Natural (−)-colchicine is axially chiral, featuring a single stereocenter at C7 [32]. SAR efforts have mainly focused on modifications at C4, C7, and C10 [[33], [34], [35], [36], [37], [38], [39], [40], [41]]. Rings A and C are critical for tubulin binding, with only limited variation tolerated at C4 and C10. Removal of ring A methoxy groups abolishes activity, while the C10 methoxy group can be replaced by small amines or thiols without major loss [42]. Notably, introduction of a C10 N-methylamino group, as in 10-aminocolchicine (1, Fig. 1b), has been shown to further enhance in vitro potency. In contrast, ring B plays a lesser role in tubulin binding, allowing greater flexibility at the C7 acetamide position [43].

The C7 position has thus emerged as a key site for colchicine optimization. For example, Krzywik et al. modified the C7 acetamide group in 10-aminocolchicine (1) with functional groups such as carbamates, sulfonamide triazoles, ureas, and amines [[39], [40], [41]], yielding analogs with improved in vitro potency against cancer cells and, in some cases, enhanced selectivity over healthy cells. Although early SAR efforts largely targeted colchicine's interactions with β-tubulin, more recent studies have explored the possibility of engaging α-tubulin as well. Notably, Marzo-Mas et al. (2016) proposed that linear-chain C7 acetamide colchicine derivatives (Fig. 1b) could interact with a hydrophobic groove in α-tubulin [34]. Supporting this hypothesis, X-ray crystallographic data revealed that analogs with 9–11 methylene unit chains formed novel hydrophobic interactions with α-tubulin, resulting in significantly improved in vitro activity and selectivity compared to colchicine.

Importantly, while these linear-chain derivatives demonstrated that extension from the C7 position can access an additional interaction region on α-tubulin, the interacting motifs were largely limited to nonspecific hydrophobic contacts. This observation suggests that the α-tubulin groove may accommodate diverse interactions beyond simple aliphatic contacts, potentially involving π-stacking, polar, or weak hydrogen-bonding interactions with proximal amino acid residues. We therefore hypothesized that introducing structurally defined pharmacophoric elements at the distal end of the C7 substituent could further enrich these interactions and improve both potency and selectivity.

Building on these insights, we designed a new series of 10-aminocolchicine analogs featuring linear amide linkers with pendant phenoxy moieties at the C7 position (Fig. 1c). The phenoxy group was selected as a conformational and directional element that enables systematic variation of substituents with diverse electronic and steric properties, while maintaining a consistent anchoring geometry toward the α-tubulin interface. By varying both the phenoxy substitution pattern and the linker length, we aimed to fine-tune the spatial reach and orientation of these pharmacophoric groups and to establish a clearer structure–activity relationship (SAR) for α-tubulin engagement.

Previous crystallographic studies indicated that most favorable activity and selectivity were achieved when the total C7 substituent length corresponded to approximately 9-10 carbon units, whereas both shorter and longer chains resulted in a marked reduction in biological performance [34]. Guided by this finding, we designed our analogs to closely mimic this optimal spatial requirement by replacing the distal five to six methylene units of the reported linear chains with a phenoxy moiety, while retaining a proximal 3- or 4-carbon aliphatic linker. A comparative schematic illustrating the carbon count between the previously reported linear-chain derivatives and the present phenoxy-linked analogs is provided in Fig. S1.

Preliminary docking results further supported this design strategy, indicating that 3C-4C linkers position the phenoxy group within reach of the proposed α-tubulin groove (Fig. S1 and Table S1). Detailed analysis of these interactions is discussed in the Results and Discussion section. All compounds were evaluated in vitro for cytotoxicity against HeLa S3, Hep G2, and MCF-7 cancer cell lines, along with the Vero E6 normal cell line to assess selectivity. Immunostaining revealed that two derivatives caused more pronounced microtubule disruption than colchicine. Complementary in silico docking studies were conducted to examine potential interactions with the α/β-tubulin III heterodimer.