Targeting NRAS-Mutant Cancers with the Selective STK19 Kinase Inhibitor Chelidonine

The RAS family of small GTPases (HRAS, NRAS, and KRAS) are binary molecular switches that transition between an active GTP-bound state and an inactive GDP-bound state (1–4). Stimulation of many cell surface receptors activates membrane-bound RAS proteins and its downstream signaling pathways, including RAF–MEK–ERK and PI3K–AKT, which culminate in the promotion of cell growth and suppression of cell death (5, 6). Aberrant RAS activity due to oncogenic mutations is frequently associated with the promotion of tumorigenesis (7, 8); indeed, RAS mutations are present in 20%–30% of all human cancers (8, 9). Melanoma is characterized by gain-of-function hotspot mutations in NRAS at glutamate 61 (Q61; refs. 10, 11), including arginine, lysine, and leucine mutations (Q61R, Q61K, and Q61L), which are present in approximately 30% of melanomas. These mutations result in a constitutively GTP-bound active conformation of NRAS that drives the malignant transformation of melanocytes (10–13). However, pharmacologic targeting of mutant NRAS proteins and the downstream signaling pathways has been challenging. Some drugs with the potential to treat NRAS-mutant cancers have been developed, such as the MEK inhibitor binimetinib, which showed some improvement in progression-free survival of patients with NRAS-mutant melanoma in a phase III trial; however, binimetinib is still in clinical development (14, 15).

Recent work demonstrated that the functionally uncharacterized serine/threonine kinase STK19 is a novel activator of NRAS (16, 17). STK19 phosphorylates NRAS at the evolutionarily conserved residue serine 89 (S89), which enhances binding between NRAS and its effector proteins, activates downstream signaling pathways, and induces malignant transformation of melanocytes (17). Crossing of NRAS Q61R transgenic mice with mice harboring melanocyte-specific expression of STK19 or the gain-of-function mutant STK19 D89N enhances melanoma formation, confirming the ability of STK19 to stimulate NRAS signaling (17, 18). These observations suggest that selective STK19 inhibitors could provide urgently needed therapeutic options to suppress the growth of NRAS-mutant tumors.

Chelidonine is one of the most abundant bioactive isoquinoline alkaloids in extracts of the plant Chelidonium majus, which is also known as the greater celandine (Papaveraceae) and is widely distributed throughout Europe and Asia (19). Crude extracts of Chelidonium majus and purified chelidonine have both been shown to possess antitumor properties, including inhibition of cell proliferation, potentiation of apoptosis, and suppression of cell migration and invasion, in cell lines from such diverse cancers as uveal melanoma, head and neck cancer, gastric carcinoma, liver cancer, and breast cancer (20–24). For example, chelidonine potentiates apoptosis in the HCT116 (KRAS G13D) human colon cancer cell line by inhibiting the NF-κB signaling pathway (25), and it suppresses the migration and invasion of MDA-MB-231 (KRAS G13D) human breast cancer cells by inhibiting formation of the integrin-linked kinase–PINCH–α-parvin complex (26). However, the precise mechanisms of action of chelidonine and its direct targets in cancer cells remain unclear, greatly hindering its translation to the clinic.

In this study, we screened a natural compound library using a phosphorylation assay–based approach and identified chelidonine as a potent and selective inhibitor of STK19. Using biochemical and cellular assays, we show that chelidonine is an ATP-competitive inhibitor of STK19 activity and blocks proliferation and induces apoptosis in a panel of NRAS-mutant cancer cell lines via inhibition of pathways downstream of NRAS, including RAF–MEK–ERK and PI3K–AKT. Similarly, chelidonine impaired cancer cell growth in vivo while having minimal toxicity. Our results suggest that pharmacologic inhibition of STK19 by chelidonine may provide a novel option for targeting NRAS-mutant cancers.

SK-MEL-2, SK-MEL-28, SK-MEL-31, HepG2, Hep3B, NCI-H446, SW-1271, HCT116, and MDA-MB-231 cell lines were purchased from ATCC; WM2032, WM3406, and WM1366 cell lines were purchased from Rockland Immunochemicals; and SNU-719 and SNU-216 cells were purchased from the Korean Cell Line Bank (Seoul, Korea). The mutation status of these cell lines is as below: SK-MEL-2 (Q61R), WM1366 (Q61L), WM2032 (Q61R), WM3406 (Q61K), HepG2 (Q61L), SW-1271 (Q61R), SNU-719 (Q61L), Hep3B [NRAS wild-type (WT)], NCI-H446 (NRAS-WT), SNU-216 (NRAS-WT), HCT116 (G13D), MDA-MB-231 (G13D), SK-MEL-28 (NRAS-WT and BRAFV600E), and SK-MEL-31 (NRAS-WT and BRAFV600E). All cell lines were maintained in DMEM containing 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. Cell lines underwent routine testing for Mycoplasma every 3 months (last confirmed negative date, October 24, 2019). The genetic identity of the cell lines was confirmed by short tandem repeat profiling. The cell lines were used for experiments within 10 passages after thawing.

Twenty-eight tumor samples were collected from patients with NRAS-mutated melanoma after surgical resection at Fudan University Shanghai Cancer Center (Shanghai, China) from January 2011 to May 2017. Written informed consent was obtained from all patients in accordance with institutional guidelines before sample collection. The study was approved by the committees for ethical review of research at Fudan University Shanghai Cancer Center (Shanghai, China).

All animal experiments were conducted in accordance with the guidelines of the NIH for the Care and Use of Laboratory Animals. The study protocol was also approved by the Committee on the Use of Live Animals in Teaching and Research, Fudan University (Shanghai, China). Mice were housed with a time-cycle of 12 hours/12 hours light/dark cycle (6:00 am/pm). Mice were allowed free access to an irradiated diet and sterilized water. The mice were monitored daily for signs related to their health status and distress.

For toxicity profiling of chelidonine, C57BL/6J mice were injected intraperitoneally with vehicle [normal saline containing 5% (w/v) Kolliphor HS 15; Sigma] or chelidonine (10 or 20 mg/kg in vehicle) once daily and body weights were measured daily. After 21 days, the mice were euthanized, and blood and organs were collected. Serum aspartate and alanine aminotransferase (AST and ALT) activity was measured using Assay Kits (Abcam) according to the manufacturer's instructions. The organs were processed by fixing in 4% paraformaldehyde and embedding in paraffin using standard protocols. Tissues were cut into 5-μm thick sections, stained with hematoxylin and eosin (H&E), and observed by light microscopy.

The pharmacokinetic profile of chelidonine was analyzed in mice injected intraperitoneally with chelidonine at 10 mg/kg. Chelidonine concentrations in mouse plasma were measured using an ultra-high performance LC/MS-MS (UHPLC/MS-MS) method established for this study.

In vivo xenograft experiments were performed as described previously (27). Briefly, 2 × 10⁶ SK-MEL-2 (NRAS Q61R), WM1366 (NRAS Q61L), or SK-MEL-28 (NRAS-WT) cells were mixed with Matrigel (1:1) and injected subcutaneously into the left flanks of 8-week-old female nude mice. Tumor size was measured every 3 days with calipers, and tumor volumes were calculated using the following formula: length × width² × 0.5. When the tumor volume reached approximately 200 mm³, mice were injected with vehicle or chelidonine (10 or 20 mg/kg, i.p.) once daily. On the indicated days, the mice were euthanized, and melanoma xenografts were excised, weighed, and processed for further analysis.

The optimal conditions for the 96-well Promega ADP-Glo Kinase Assay (incubation time, and STK19 and ATP concentration) were previously determined according to the manufacturer's protocol (17) and found to be 12.5 nmol/L STK19, 6.36 μmol/L of ATP, and 15 minutes incubation. Individual compounds from the natural compound library (TargetMol) were added to the wells at a final concentration of 10 μmol/L. STK19 kinase activity was quantified on the basis of the luminescence signal detected with a Tecan Infinite M1000 Microplate Reader. The screening results are presented as the percent inhibition of STK19 kinase activity relative to control levels. Compounds exhibiting ≥50% relative inhibition of STK19 activity in the primary screen were selected for secondary evaluation.

Cells were lysed in a buffer containing 50 mmol/L Tris, pH 7.4, 150 mmol/L NaCl, 0.5 mmol/L EGTA, 0.5 mmol/L EDTA, 1 mmol/L phenylmethylsulfonyl fluoride, 1% Triton X-100, 10% glycerol, and complete Protease Inhibitor Cocktail (Roche). The lysates were then homogenized and centrifuged at 14,000 rpm at 4°C for 15 minutes. Protein concentrations were determined using a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Cell lysates were incubated with Pierce Lane Marker Reducing Sample Buffer at 100°C for 10 minutes, and proteins were separated with 8%–16% Mini-PROTEAN TGX Precast Protein Gels (Bio-Rad), and transferred to polyvinylidene difluoride membranes (Bio-Rad). After blocking, the membranes were incubated with specific primary antibodies followed by horseradish peroxidase (HRP)-conjugated secondary antibodies. The antibodies and suppliers were: monoclonal anti-β-actin (AC15), monoclonal anti-Flag M2 (A8592), monoclonal anti-HA (H6533), HRP-conjugated anti-rabbit (A-4914), and HRP-conjugated anti-mouse (A4416) antibodies (all from Sigma-Aldrich); and anti-phosphorylated (p-) MEK1/2 (Ser217/221) (9121), anti-MEK (9122), anti-p-ERK1/2 (Thr202/Tyr204) (9101), anti-ERK1/2 (9102), anti-p-AKT (Ser473) (9271), anti-AKT (9272), anti-cleaved caspase-3 (Asp175) (9661), anti-cleaved caspase-7 (Asp198) (9491), and anti-cleaved poly (ADP-ribose) polymerase (PARP, Asp214) (9541) (all from Cell Signaling Technology). A custom-generated antibody against p-NRAS (Ser89) was obtained from Hangzhou Huaan Biotechnology Co. Ltd.

A human tissue microarray containing 28 melanoma tissues with NRAS mutation were established. Unstained 3-μm thick sections were then prepared from paraffin-embedded tissues. The sections were stained with primary antibodies at 4°C overnight. Staining with the secondary antibody and avidin-biotin peroxidase complex was performed according to the standard protocols provided by the manufacturer (Vector Laboratories). All procedures were performed by two independent assessors and one pathologist, none of whom had any previous knowledge of the clinical outcomes of the cases. An immunoglobulin-negative control was used to rule out nonspecific binding. The primary antibodies used were: anti-STK19 (251814) and anti- phosphorylated (p-) ERK1/2 (Thr202/Tyr204) (138482) antibodies (all from Abcam); and anti-p-MEK1/2 (Ser217/221) (9121) and anti-p-AKT (Ser473) (9271) antibodies (all from Cell Signaling Technology). IHC intensities were assessed by a semiquantitative system according to the immunoreactive score (IRS). The IRS is obtained by multiplying the staining intensity by the percentage of positive cells, resulting in an IRS between 0 and 12. Briefly, the staining intensity (SI) was categorized as 0 (negative), 1 (weak), 2 (intermediate), or 3 (strong), and the percentage of positive cells (PP) was scored as 0 (0% positive), 1 (1%–25%), 2 (26%–50%), 3 (51%–75%), or 4 (76%–100%). The IHC staining IRS = SI × PP. Two senior pathologists performed the scorings independently in a blinded manner.

Total RNA was extracted using a Qiagen RNeasy Kit (Invitrogen) as described previously (28), and cDNA was synthesized with SuperScript II Reverse Transcriptase (Invitrogen). Aliquots of cDNA (30 ng) were amplified by qPCR with TaqManTM Gene Expression Master Mix (Thermo Fisher Scientific). The mRNA levels of the genes of interest (PHLDA1, DUSP4, ETV4, and SPRY2) were normalized against GAPDH mRNA in the same samples. The qPCR data were analyzed using the comparative C_t method. All PCR reactions were performed in triplicate.

The reaction mixture contained recombinant human HA-NRAS protein preloaded with GTP, purified recombinant human STK19-Flag protein in kinase buffer (20 mmol/L MnCl₂, 50 mmol/L HEPES, pH 8.0), 300 μmol/L AMP, and the indicated concentrations of ATP. Samples were incubated for various times at 30°C. Proteins were then immunoprecipitated using anti-HA- or anti-Flag–conjugated beads, separated by SDS-PAGE, transferred to membranes, and subjected to immunoblotting to detect p-NRAS.

Assays were performed as described previously (29). Briefly, melanoma cells were placed in 6-well plates at a density of 2.5 × 10³ cells/well and incubated with the indicated concentrations of chelidonine. After 14 days, colonies were stained with 0.1% crystal violet, visualized by light microscopy, and enumerated.

Cell viability was determined using a CyQUANT NF Cell Proliferation Assay Kit (Invitrogen) according to the manufacturer's protocol. Briefly, cells were plated in 96-well microplates at 500 cells/well and incubated with the indicated concentrations of chelidonine for 4 days. CyQUANT NF dye solution was then added to the wells and fluorescence intensity was measured with a fluorescence microplate reader using a 485/520 nm filter set. Cell viability is presented as the fold-change relative to the initial cell number.

5-Ethynyl-2′-deoxyuridine (EdU) incorporation into DNA was detected using a Click-iT EdU Alexa Fluor 488 Flow Cytometry Assay Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. Briefly, melanoma cells (1 × 10⁶ per sample) were harvested, washed twice in PBS/1% BSA, and fixed in 100 μL Click-iT fixative. After incubation for 15 minutes at room temperature in the dark, the cells were washed twice in 1 × saponin-based permeabilization and wash reagent and incubated with the Click-iT EdU reaction cocktail for 30 minutes. For the staining of cellular DNA, cells were washed once in 1 × saponin-based permeabilization and wash buffer and incubated with DNA staining solution for 15 minutes at room temperature in the dark. Cells were then filtered through a 200-μm mesh and analyzed using a BD FACSCalibur Flow Cytometer (BD Biosciences).

The thermal denaturation of purified recombinant STK19 protein was determined using a Protein Thermal Shift Dye Kit (Thermal Fisher Scientific) as described previously (17). In brief, the purified recombinant protein was diluted to a final concentration of 10 μmol/L in 100 mmol/L of Tris buffer (pH 8.0). Aliquots of 20 μL of the protein sample were mixed with chelidonine (final concentration 100 μmol/L) and a heat gradient (25°C–99°C) was then applied using a QuantStudio 12K Flex Real-Time PCR System. The melting curve was recorded and the melting temperature was determined using the inflection points of the d(RFU)/dT plots.

All quantitative data are presented as the mean ± SD or SEM of at least three independent experiments. Significant differences between groups were assessed using Student t test. Survival analysis was performed using the Kaplan–Meier method and compared using the log-rank test. All analyses were performed using GraphPad Prism 7 or Microsoft Excel 2010. A P value of 0.05 was considered statistically significant.

To identify pharmacologic regulators of STK19 kinase activity, we screened approximately 1,500 natural compounds using a luminescent phosphorylation-based assay with purified recombinant human NRAS protein as the substrate (Fig. 1A). We obtained 20 preliminary hits that inhibited STK19 activity by ≥50% at 10 μmol/L in two independent experiments (Fig. 1B). As a secondary screen, the primary hits were incubated with the human melanoma cell line SK-MEL-2 (NRAS Q61R), and phosphorylation of NRAS at S89 was detected by immunoblotting (30) of cell extracts with an anti-p-NRAS (S89) antibody (Fig. 1C). The specificity of the antibody for S89-phosphorylated NRAS was validated by dot blot analysis using biotinylated peptides (Supplementary Fig. S1A). From these assays, we selected the top candidates, including chelidonine, lycorine, jatrorrhizine, fangchinoline, and daurisoline, and their inhibitory effects on STK19 were confirmed with an in vitro kinase assay. Of note, chelidonine and jatrorrhizine are both benzophenanthridine alkaloids with similar molecular structures (Fig. 1D; Supplementary Fig. S1B). Among the hits, chelidonine was the most potent inhibitor of STK19 kinase activity (IC₅₀ 125.5 ± 19.3 nmol/L) and was selected for in-depth evaluation (Fig. 1E; Supplementary Fig. S1C).

The inhibitory activity of chelidonine against STK19 was further validated using an in vitro kinase assay with purified recombinant human STK19 and NRAS Q61R proteins. These experiments demonstrated that chelidonine inhibited the phosphorylation of NRAS in a concentration- and time-dependent manner (Fig. 1F and G). Moreover, chelidonine appeared to be an ATP-competitive inhibitor of STK19, as indicated by the increase in chelidonine IC₅₀ value as the ATP concentration in the reaction mixture was increased (Fig. 1H). Chelidonine showed a melting temperature (T_m) shift at 6.49°C (Fig. 1I), indicative of a high-affinity interaction between chelidonine and STK19.

We next evaluated the selectivity of chelidonine for STK19 by performing a KINOMEscan assay, which screens a panel of 468 kinases using an in vitro ATP site competition binding assay (31). The KINOMEscan assay scores are reported as the percentage relative to the negative control signal (DMSO), set to 100%. A selectivity score (35%) was defined as the percentage of total kinases (among the 468 kinases) whose kinase activity in the presence of chelidonine was decreased to less than 35% of the control group (Supplementary Fig. S2A; Supplementary Table S1). Among these, six kinases were inhibited to <10% of control activity in the presence of chelidonine: EGFR (L858R and T790M), ERN1, MAP3K13, MAPK9, MATK, and RPS6KA4. To validate these findings, we performed in vitro STK19 kinase assays using the six purified recombinant kinases. As shown in Fig. 1E and Supplementary Fig. S2B, chelidonine was a more potent inhibitor of STK19 compared with EGFR (L858R and T790M), ERN1, MAP3K13, MAPK9, MATK, and RPS6KA4. Because of the relatively similar IC₅₀ values of chelidonine toward STK19, EGFR, and MAPK9, we further overexpressed Flag-tagged STK19, EGFR, or MAPK9 into SK-MEL-2 melanoma cells to explore whether overexpression of STK19, EGFR, or MAPK9 modulates the effects of chelidonine on cell growth. We observed that only the overexpression of STK19 reduced the inhibitory effects of chelidonine on the growth of SK-MEL-2 cells, but not EGFR or MAPK9 (Supplementary Fig. S2C), confirming the specificity of chelidonine toward STK19. Taken together, these results demonstrate that chelidonine is a potent and highly selective ATP-competitive inhibitor of STK19 kinase.

STK19-induced phosphorylation of NRAS S89 enhances NRAS activity and promotes downstream signaling (17). We confirmed STK19 activity in human NRAS-mutant melanoma tissues with IHC staining and observed a positive correlation between STK19 expression and activation of NRAS downstream MAPK and AKT signaling pathways (Supplementary Fig. S3A and S3B). To determine whether chelidonine inhibits signaling downstream of NRAS, we incubated chelidonine with a panel of human melanoma cell lines with various NRAS mutations [SK-MEL-2 (NRAS Q61R), WM2032 (NRAS Q61R), WM3406 (Q61K), and WM1366 (Q61L); refs. 30, 32–34], and examined NRAS pathway activation by immunoblotting. Notably, chelidonine dose dependently reduced the phosphorylation not only of endogenous NRAS S89 but also of MEK, ERK1/2, and AKT in all four NRAS-mutant melanoma cell lines (Fig. 2A). Consistent with these findings, quantitative reverse-transcription PCR (qRT-PCR) analysis indicated that chelidonine also decreased the expression of the ERK transcriptional target genes PHLDA1, DUSP4, ETV4, and SPRY2 (ref. 32; Fig. 2B). Thus, chelidonine effectively inhibits activation of NRAS and its downstream signaling pathways in melanoma cells.

Because oncogenic NRAS plays a critical role in promoting melanoma cell growth and preventing cell death (1), we next evaluated these processes in SK-MEL-2, WM2032, WM3406, and WM1366 melanoma cells after treatment with 5 or 20 μmol/L chelidonine. Indeed, chelidonine substantially inhibited the colony-forming ability (Fig. 3A and B), viability (Fig. 3C), and proliferation (EdU incorporation; Fig. 3D) of the cells. Furthermore, chelidonine induced apoptosis of melanoma cells, as indicated by the appearance of the apoptosis effector proteins cleaved caspase-3, caspase-7, and PARP (Fig. 3E and F) and by the increased activity of caspase enzymes (Fig. 3F). To confirm that these effects of chelidonine were specific for STK19-activated NRAS signaling pathways, we also explored its effect on two KRAS-mutant cancer cells (HCT116 and MDA-MB-231) and two BRAF-mutant, NRAS-WT melanoma cells (SK-MEL-28 and SK-MEL-31). Importantly, chelidonine did not significantly inhibit the growth of either the KRAS-mutant cell lines (Supplementary Fig. S4) or the NRAS-WT melanoma cells (Supplementary Fig. S5A and S5B). Collectively, these data indicate that chelidonine specifically inhibits the proliferation and survival of NRAS-mutant melanoma cells.

Next, we evaluated the clinical potential of chelidonine by determining its toxicity and pharmacokinetic profile in mice. C57BL/6 mice were injected intraperitoneally with 10 mg/kg chelidonine, and blood was collected at various times thereafter. Plasma concentrations of chelidonine were measured using an UHPLC/MS-MS method. The elimination half-life of chelidonine in mouse plasma was 19.78 hours (data not shown), which suggests that high concentrations of drug can be maintained in the plasma by once daily injections. Next, we injected C57BL/6 mice intraperitoneally with 0, (vehicle), 10, or 20 mg/kg chelidonine once daily for 21 days and body weights were measured daily. The mice displayed no overt clinical signs during this time and chelidonine treatment had no significant effects on body weights (Fig. 4A). On day 21, the mice were euthanized, and blood and tissues were collected for blood biochemistry and histopathologic analyses, respectively. Chelidonine had no significant apparent effects on hepatic function, as shown by serum levels of AST and ALT (Fig. 4B), or on tissue integrity, as evaluated by histopathologic analysis of major organs (Fig. 4C). These results suggest that chelidonine has minimal toxicity in vivo.

To confirm the in vivo therapeutic potential of chelidonine for the treatment of melanoma, we injected SK-MEL-2 (NRAS Q61R) melanoma cells subcutaneously into nude mice and allowed tumors to grow to approximately 200 mm³ in volume. Treatment was then initiated by once daily intraperitoneal injections with 0 (vehicle), 10, or 20 mg chelidonine. Chelidonine significantly and dose dependently reduced the volumes and weights of tumor xenografts compared with vehicle treatment (Fig. 5A–C). Tumor-bearing mice treated with chelidonine also survived significantly longer than mice treated with vehicle (Fig. 5D). Xenograft tissues were excised at the end of the experiment and NRAS signaling pathway activation was assessed by immunoblotting of tumor extracts. The results indicated that chelidonine inhibited NRAS signaling in a dose-dependent manner, as illustrated by the reductions in phosphorylated NRAS S89, GTP-bound NRAS, and phosphorylated MEK, ERK1/2, and AKT (Fig. 5E). The inhibitory effects of chelidonine on xenograft tumor growth were also confirmed in another NRAS-mutant WM1366 melanoma cells (Supplementary Fig. S6A–S6C). In contrast, chelidonine did not suppress the growth of SK-MEL-28 (NRAS-WT) melanoma cells (Supplementary Fig. S6D–S6F) confirming the specificity of action of chelidonine observed in vitro. Taken together, these results provide support for the in vivo therapeutic potential of chelidonine by demonstrating its ability to specifically inhibit the growth of NRAS-mutant, but not NRAS-WT, melanoma.

Our in vitro and in vivo studies thus far show that chelidonine is a potent inhibitor of NRAS-mutant melanoma growth, with concomitant downregulation of NRAS, ERK, and AKT signaling. To determine whether chelidonine can inhibit the progression of other types of NRAS-mutant cancers, we examined its effects on the viability of HepG2 (NRAS Q61L), SW-1271 (NRAS Q61R), and SNU-719 (NRAS Q61L) cell lines (Fig. 6A). Consistent with the inhibitory effects of chelidonine on the growth of NRAS-mutant, but not NRAS-WT, melanoma cells and xenografts, the proliferation of these three cell lines was substantially inhibited compared with vehicle (Fig. 6A). To confirm that these effects were mediated via inhibiting NRAS and its downstream signaling pathways, we examined phosphorylation of NRAS, MEK, ERK1/2, and AKT in HepG2, SW-1271, and SNU-719 cells by immunoblotting. Indeed, phosphorylation of each of these signaling proteins was inhibited by chelidonine treatment (Fig. 6B). In contrast, similar experiments with the NRAS-WT counterparts of these cell lines (Hep3B, NCI-H446, and SNU-216) revealed no significant effects of chelidonine on either cell proliferation or activation of NRAS and downstream signaling proteins (Supplementary Fig. S7A and S7B), confirming that chelidonine can inhibit the growth of various cancer cell lines harboring NRAS mutant, while NRAS-WT–expressing cells were unaffected. Taken together, these studies demonstrate that chelidonine is a potent and selective STK19-targeting inhibitor that specifically blocks oncogenic NRAS-driven tumor progression.

KRAS, HRAS, and NRAS (1, 3, 8, 9) were the first identified oncogenes, and it is now well established that about 25% of all human cancers harbor activating mutations in at least one of these proteins (8, 36). In particular, oncogenic KRAS mutations are present in 95% of pancreatic ductal adenocarcinomas and 52% of colorectal adenocarcinomas, with the majority occurring at G12. HRAS mutations (mainly G12 and Q61) are frequently associated with bladder cancer, and 20%–30% of cutaneous melanomas are driven by NRAS Q61 mutations (36–38). Although therapeutic modulation of RAS signaling has been a goal for decades, it has proven difficult to develop strategies to directly inhibit RAS activity. Nevertheless, numerous alternative strategies aimed at exploiting RAS-related vulnerabilities or targeting RAS regulators and effectors have been studied.

RAS is activated at the plasma membrane following its prenylation by farnesyltransferases (39, 40). Several farnesyltransferase inhibitors have been developed to block this step (41); however, they have proven unsuccessful in clinical trials due to the alternative modification of RAS by geranylgeranyl isoprenoid (42). G12C mutations in RAS create a pocket for a potential covalent inhibitor, but this is a relatively minor RAS mutation in cancer (43). Other important breakthroughs in anti-RAS therapies include the small-molecule RAS-mimetic rigosertib and pan-RAS ligands that block RAS binding to effector proteins containing a common RAS-binding domain (44, 45). GTPase-activating proteins and guanine nucleotide exchange factors (GEF) are important regulators of the RAS activation/inactivation cycle (46), and current efforts include therapeutic targeting of RAS/GEF interactions (47, 48). Synthetic lethal strategies have also been employed to identify genes critical for the survival of NRAS-mutant–expressing cancer cells but not those harboring NRAS-WT, and this approach has identified STK33 (49), TBK1 (50), and PREX1 (51) as essential genes. RAS-mutant cancers are highly dependent on upregulated metabolism to maintain their rapid growth (52), and targeting of such metabolic dependence also represents a potentially promising route to therapy. Overall, these new strategies to directly or indirectly target RAS and/or its key regulators and vulnerabilities show great promise for the treatment of RAS-mutant–driven cancers.

Targeting of RAS posttranslational modifications, particularly phosphorylation, is another avenue to the development of anti-RAS therapies. Recent work identified STK19-mediated phosphorylation of NRAS (S89) as a critical mechanism of mutant NRAS activation in melanocytes and their transformation into melanoma (17). STK19 is known to be a top driver gene in this cancer (53), and somatic hotspot mutations in STK19 have been detected in about 5% of melanomas (54) and 10% of skin basal cell carcinomas (55). However, the role of STK19 in the regulation of RAS activity and tumor growth had not previously been appreciated. Here, we observed a positive correlation between STK19 expression and activation of NRAS downstream MAPK and AKT signaling pathways, and pharmacologic inhibition of STK19 suppresses activation of NRAS and the progression of NRAS-mutant–driven cancer both in vitro and in vivo, substantiating the feasibility of STK19 as an anticancer therapeutic target.

Chelidonine has been reported to have broad pharmacologic properties, including antitumor, anti-inflammatory, antimicrobial, and antiviral activities, but its mechanisms of action and molecular functions were poorly understood (21–23, 25, 26). In this study, we demonstrated that chelidonine directly binds to and inhibits STK19, thereby downregulating NRAS and its downstream signaling pathways, leading to inhibition of tumor growth via suppression of proliferation and induction of apoptosis. Chelidonine had good efficacy but minimal toxicity in mice, and it also inhibited the growth of NRAS-mutant cancers of various origins, indicating that this compound could form the basis for new therapies for multiple oncogenic RAS-driven cancers.

The RAF–MEK–MAPK signaling cascade is the key RAS effector pathway for promoting the proliferation and survival of RAS-mutant cancer cells (3). Numerous inhibitors of this pathway have been demonstrated to improve clinical outcomes in patients with various RAS- and RAF-mutant cancers (56–58). However, drug resistance invariably emerges in such cancers, frequently involving an increase in oncogenic RAS/RAF-driver mutations and reactivation of the MEK–MAPK pathway (59–62). In response, various combination therapies have been explored for the treatment of RAS-mutant cancers, such as the RAF inhibitor sorafenib with aspirin (63), the MEK inhibitor trametinib with palbociclib, a CDK4/6 inhibitor (64), and the BRAF/MEK inhibitors dabrafenib/trametinib with magnolol, a natural plant-derived compound (65). These novel combinations have significantly improved the efficacy of RAF–MEK–MAPK pathway inhibitors in RAS-mutant cancers. Considering that STK19 is an activator of NRAS and thus acts upstream of the MEK–MAPK pathway, chelidonine might be a useful additional therapy in combination with MEK inhibitors or other treatments for RAS-mutant cancers. STK19 inhibition warrants further exploration in preclinical and clinical studies.

No potential conflicts of interest were disclosed.

Conception and design: Z. Chen, Z. Meng, P. Wang

Development of methodology: C. Wang, P. Wang

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L. Qian, K. Chen, C. Wang, P. Wang

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L. Qian, K. Chen, C. Wang, P. Wang

Writing, review, and/or revision of the manuscript: L. Qian, K. Chen, C. Wang, P. Wang

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Z. Meng, P. Wang

Study supervision: P. Wang

The authors thank Dr. Shenglin Huang (Fudan University Shanghai Cancer Center, Shanghai, China) for technical support and for valuable advice and discussions. This study was supported by the National Natural Science Foundation of China (81622049 and 81871989), the Shanghai Science and Technology Committee Program (19XD1420900), and the Shanghai Education Commission Program (17SG04).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.