The number of newly diagnosed cases of melanoma worldwide in one year reached 300,000 in 2020, which raises a serious concern [1]. NRAS, one of the pivotal melanoma-driven oncogenes, was first discovered in 1984, and its mutations are the second most frequent oncogenic factor in cutaneous melanoma [2], [3]. The NRAS protein is a member of the RAS family, existing in either an active state (guanosine triphosphate (GTP)-bound) or an inactive state (Guanosine diphosphate (GDP)-bound) [4] under physiological conditions. A point mutation in NRAS typically results in substituting glutamine for leucine at position 61 [5] and constitutively disrupts RAS’s GTPase activity by trapping it in an active conformation. Thus, intracellular signaling across a variety of pathways is activated, most notably the mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K) pathways [6], resulting in cell-cycle dysregulation, pro-survival pathways, and cellular proliferation [7]. Therefore, almost 30% of melanomas have NRAS mutations [8] with higher levels of mitosis, worse prognosis, and lower overall survival rates than other subtypes [9], [10]. To date, the treatment for surgically incurable NRAS-mutant melanoma includes immunotherapy with checkpoint inhibitors of programmed cell death protein (anti-PD-1); targeted therapy with pan-RAFi/MEKi combination; traditional cytostatic dacarbazine treatment; etc [11], [12]. However, developing efficient GTP-competitive drugs that directly target RAS proteins is challenging due to the picomolar affinity of RAS to bind to GTP and the lack of distinct drug pockets outside of the nucleotide-binding site [13], [14]. Therefore, the therapeutic success of available drugs for NRAS-mutant melanoma is limited, emphasizing the need to identify new accessible targets [15], [16], [17].

Four guanines form a G-quartet by Hoogsteen hydrogen bonding, and two or more G-quartets form a G-quadruplex (G4) by π-π stacking [18]. Various cellular processes, such as transcription, translation, and telomerase activity, can be regulated by stabilizing the G-quadruplex structure through G4 ligands[19], [20], [21], [22], [23], [24]. Human NRAS proto-oncogene mRNA contains a G-rich sequence (5’-GGGAGGGGCGGGUCUGGG-3’) that may fold into a parallel RNA G4 (rG4). This G-rich sequence is located 14-nt downstream of the 5’cap and 222-nt upstream of the translation start site [25], meaning that the formation of rG4 might regulate the translation of NRAS and NRAS rG4 targeting might be an alternative approach to discover novel NRAS repressors and beneficial for treating NRAS-mutant melanoma [26], [27], [28], [29]. Some G4 ligands can inhibit the translation of the NRAS by stabilizing G4 formation (Fig. S1) [29], [30]. For example, RGB-1 inhibits NRAS expression in MCF-7 cells by stabilizing rG4 without affecting NRAS transcription [29]; our prior investigation also established that quindoline derivatives 4a-10 and 4a-16 exhibited the capabilities to bind and stabilize NRAS rG4 structures, consequently impeding the proliferation of diverse tumor cell lines. [30].

Cryptolepine and quindoline derivatives have been identified as effective DNA G4 (dG4) binders and have been shown to have significant effects against several types of cancer cells [31], [32], [33], [34], [35], [36], [37], [38]. The selective NRAS rG4 ligand mentioned above was gathered by introducing the p-(methylthio)styryl side chain at the 2-position of quindoline scaffold [30] while limiting with some shortcomings such as difficulties in synthesis and poor solubility. Due to the 2’-OH groups in the RNA and their effects on groove and loop widths, ligand side chains interact with rG4s less effectively than dG4s [39]. Also, several findings have revealed that rG4s have a dynamic and transient behavior in vivo [40], [41], making designing rG4s binders meaningful but challenging. Bhattacharya et al., through docking and molecular dynamics simulation studies, found that small molecules with longer side chains have lower binding energy to RNA G4 [42]. Recent research hinted that ligands with long alkyl side chains exhibited high affinities for NRAS rG4 or strong anti-melanoma activities (Fig. S1) [43], [44]. To develop G4 ligands targeting NRAS rG4 and effectively suppressing NRAS-mutant melanoma, we designed a novel series of quindoline derivatives with fork-shaped side chains to evaluate whether longer alkyl side chains provide more favorable hydrophobic interactions with rG4 groove, induce NRAS rG4 formation, and increase cellular uptake. In addition, introducing side chains with fork-shaped branches may be more adaptable to the spatial structure of the NRAS rG4 groove than straight chains. Therefore, we designed and synthesized a series of novel quindoline derivatives (Figure 1). According to the previously reported structure-activity analysis, side chains containing amino and heterocyclic rings were introduced at the 11-position, and fluorine atoms were introduced at the 7-position.

32 quindoline derivatives with fork-shaped side chains were synthesized (Figure 1), and their stabilizing activities with the NRAS rG4 and anti-melanoma activities were evaluated. Introducing a more extensive and flexible fork-shaped side chain at the 2-position could significantly improve the stability of NRAS rG4 and anti-melanoma activity. Moreover, one of the compounds, 10b, was picked out based on these preliminary data to evaluate anti-melanoma activity in vivo and in vitro. All the results indicated that 10b can selectively repress NRAS translation via the interaction with the NRAS rG4 and exhibit good potential in anti-melanoma treatment.