Systematic testing of existing drugs and their combinations is an attractive strategy to exploit approved drugs for repurposing and identifying the best actionable treatment options. To expedite the search among many possible drug combinations, we designed a combinatorial CRISPR–Cas9 screen to inhibit druggable targets. Coblockade of the N-methyl-d-aspartate receptor (NMDAR) with targets of first-line kinase inhibitors reduced hepatocellular carcinoma (HCC) cell growth. Clinically, HCC patients with low NMDAR1 expression showed better survival. The clinically approved NMDAR antagonist ifenprodil synergized with sorafenib to induce the unfolded protein response, trigger cell-cycle arrest, downregulate genes associated with WNT signaling and stemness, and reduce self-renewal ability of HCC cells. In multiple HCC patient-derived organoids and human tumor xenograft models, the drug combination, but neither single drug alone, markedly reduced tumor-initiating cancer cell frequency. Because ifenprodil has an established safety history for its use as a vasodilator in humans, our findings support the repurposing of this drug as an adjunct for HCC treatment to improve clinical outcome and reduce tumor recurrence. These results also validate an approach for readily discovering actionable combinations for cancer therapy.

Significance:

Combinatorial CRISPR–Cas9 screening identifies actionable targets for HCC therapy, uncovering the potential of combining the clinically approved drugs ifenprodil and sorafenib as a new effective treatment regimen.

It is becoming increasingly evident that combination therapy can enhance efficacy in treating various cancers (1), including hepatocellular carcinoma (HCC). HCC is the most common type of liver cancer with a very poor prognosis, ranking fourth as the most common cause of death among all cancers globally (2). First-line treatment options for unresectable HCC include sorafenib and lenvatinib, which target multiple kinases including Fms-related tyrosine kinase 4 (FLT4), fibroblast growth factor receptor (FGFR), and platelet-derived growth factor receptor and inhibit tumor growth and angiogenesis (3). However, these first-line treatments only extend overall survival of HCC patients for a few months, and the presence of drug-resistant tumor-initiating cells poses a high chance of tumor relapse (4, 5). There remains a critical need to improve the treatment for this disease.

Prompted by the effectiveness of immune-checkpoint inhibitors in eradicating melanoma (6, 7), clinical trials have been launched to test on therapies that combine first-line drugs with immune-checkpoint inhibitors in HCC patients (8). Combined treatment with immune-checkpoint inhibitor atezolizumab and angiogenesis inhibitor bevacizumab, which targets programmed death-ligand 1 (PD-L1) and vascular endothelial growth factor respectively, was also tested and recently approved for treating unresectable and metastatic HCC (9). Cotreatment with immune-checkpoint inhibitors nivolumab and ipilimumab, which target PD-L1 and cytotoxic T-lymphocyte-associated protein 4, respectively, was approved as a second-line option for advanced HCC patients previously treated with sorafenib (10). So far, the trials on combination therapies are designed based on potential additive effects brought by clinically proven drugs and are largely confined to kinase inhibitors, immune-checkpoint blockers, and angiogenesis inhibitors (8). These drug combinations only modestly increase overall survival of HCC patients and do not effectively prevent tumor relapse.

Systematic screening of therapeutic targets and combinations using high-throughput technologies could broaden the search for and accelerate the discovery of new treatment options. Screening drug arrays may be automated using robotics, by which it requires delicate equipment and is also expensive and offers limited scalability and flexibility in evaluating a vast number of drug combinations. Therapeutic target selection based on genetic evidence could help predict and prioritize candidates for drug discovery and repurposing (11, 12); however, low-throughput classic genetics studies are tedious and time-consuming and much genetic evidence for different types of cancers remains to be discovered. With the development of CRISPR (clustered regularly interspaced short palindromic repeats)-based high-throughput screens (13, 14), pooled characterization of genetic perturbations allows many genes and their pairs to be evaluated affordably using a simple experimental setup. CRISPR screens have been applied in many laboratories to search for novel therapeutic targets for cancers and other diseases. Through screening genes and their combinations from which hits can be translated directly into drug combinations using existing drugs, nonconventional drugs and drug combinations may be discovered and repurposed for treating cancer. The availability of drugs against the identified hits also avoids the need to start the lengthy and costly drug development process, which often has a low success rate (15).

As with other cancers, HCC cells are hypothesized to be heterogeneous and comprise a cancer stem cell (CSC; also known as tumor-initiating cell) subpopulation with the unique ability to self-renew that contributes to drug resistance and tumor recurrence (16, 17). Treatment of bulk cancer cell population within tumors with chemotherapeutic agents and kinase inhibitors was shown to leave behind CSC-enriched cells that re-form tumors (18). We postulated that rendering CSCs susceptible to therapeutic interventions may represent an effective strategy for cancer treatment. Our attempt using Connectivity Map (CMap), a transcriptomics-based drug-repurposing method (19), to map gene expression signatures with a library of FDA-approved drugs, however, did not generate a hypothesis on drug candidates that could be repurposed to revert both CSC signatures (Supplementary Fig. S1). We thus explored whether using a combinatorial targeting approach could enhance efficacy and tackle the problem of recurrence. Using CombiGEM technology (20, 21), we carried out a combinatorial CRISPR–Cas9 screen focusing on a set of druggable targets, of which, their expressions are upregulated in CSC marker-positive HCC cells and/or have been reported to inhibit HCC growth, in order to identify combination therapies with clinically approved drugs for treating HCC. Indeed, our screen discovered two genetic target combinations with strong inhibitory effect on HCC cell growth. These two combinations harbor a common target known as NMDAR1 and its paired targets are two kinases, of which, both their corresponding drug inhibitor is sorafenib. Coadministration of ifenprodil, an NMDAR inhibitor being clinically used as a vasodilator, with sorafenib gave a similar strong inhibitory effect as seen in the screen. This drug combination profoundly suppressed cancer growth and self-renewal in HCC patient-derived organoids and xenograft models. Our overall results reveal ifenprodil as a potential drug candidate to be repurposed for enhancing the treatment efficacy of HCC.

Cell culture

Human HCC cell lines (HepG2 and Hep3B) and HEK293T were purchased from ATCC. The sorafenib-resistant HepG2 cells were established in our previous work (22). Human HCC cell line Huh7 was purchased from the JCRB Cell Bank. HepaRG cells were purchased from Thermo Fisher Scientific. Human HCC cell line MHCC97L was obtained from Liver Cancer Institute, Fudan University. Human liver cell line L02 was obtained from the Institute of Virology, Chinese Academy of Medical Sciences, Beijing, China. RIL-175 cell line is a gift from Dr. Lars Zender (University of Tübingen, Tübingen, Germany). To generate cell lines stably expressing Cas9 protein (i.e., MHCC97L-Cas9 and Hep3B-Cas9), cells were infected with lentiviral pAWp30 (Addgene, 73857; Supplementary Table S1) vector, followed by 2 weeks of selection with Zeocin (200 μg/mL, Thermo Fisher Scientific). To generate the UPR-depleted cells, MHCC97L-Cas9 cells were infected with vectors expressing sgRNAs targeting IRE1-alpha, PERK, and ATF6, and cultured for 6 days prior to downstream assays. Cells were regularly tested for Mycoplasma contamination using PCR detection and were confirmed to be negative.

Organoid culture

HCC patient-derived organoids labeled HCC#23 and HCC#10 were gifts from Meritxell Huch of The Gurdon Institute at the University of Cambridge. HCC organoid lines, named as HCC-HK P1 (or KYM) and HCC-HK P2 (or LKY), were established from resected HCC tissues of HCC patients undergoing hepatectomy at Queen Mary Hospital, Hong Kong SAR. Samples were collected from patients who had not received any previous local or systemic treatment prior to operation. Written informed consent was obtained from all patients before the collection of liver specimens. The studies were conducted in accordance with the ethical guidelines and the protocol for use of human clinical specimens was approved by the Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong West Cluster.

Mouse

All experimental procedures were approved by the Committee of the Use of Live Animals in Teaching and Research at The University of Hong Kong and the Animals (Control of Experiments) Ordinance of Hong Kong, and performed in compliance with the institutional guidelines.

Cell viability assay

To validate the growth inhibition effects brought by the selected sgRNA combinations, MTT assay was applied. On day 7 after infection, 2,000 lentivirus-transduced HCC cells expressing the sgRNA combinations were seeded onto each well of a 96-well plate. Cell viability was then determined every 24 hours by performing MTT assay according to the manufacturer's protocol. Absorbance was detected at 570 and 650 nm (as a reference) by spectrophotometry using the Synergy H1 Microplate Reader (BioTek). To validate the growth inhibition effects brought by the drug(s) in HCC cells, 2,000 HCC cells were seeded onto each well of a 96-well plate, and drugs were added to the cells on the next day. Cell viability was determined by MTT assay after 2 days of drug treatment. To measure the growth inhibition effects brought by the drug(s) in HCC organoids, CellTiter-Glo Luminescent Cell Viability Assay Kit (Promega) was used following the manufacturer's instructions. 1,000 cells were seeded onto each well of a 384-well plate supplemented with Matrigel and complete organoid growth medium, and cultured for 3 days to form organoids. Fresh culture medium together with the 2× concentration of the drug(s) were then added to each well and cultured for another 72 hours. CellTiter-Glo Reagent was added to each well and incubated at room temperature for 30 minutes. Luminescence was detected using the VICTOR3 Multilabel Plate Reader (PerkinElmer).

Drug response study in mice

A total of 5 × 105 MHCC97L cells, and 3.5 × 105 cells (for PDX1) and 1 × 105 cells (for PDX2) dissociated from HCC patient-derived xenograft (PDX), were injected subcutaneously into the left flank of 4- to 5-week-old male BALB/c nude mice and NOD/SCID mice, respectively. 1 × 104 RIL-175 cells were injected into the right median lobe of the livers from 9- to 12-week-old male C57BL/6N mice. Size of subcutaneously xenografted tumors was measured by an external caliper, and tumor volume was estimated by the equation: volume = ½ (length × width2), where the length and width represent the largest and smallest diameters, respectively. Drug administration began when the subcutaneous tumor reached ∼0.05 cm3 in size or at day 5 after injection (for RIL-175 cells), at which point, mice were divided into four groups randomly for daily treatment with DMSO, ifenprodil (20 mg/kg/day, intraperitoneally), sorafenib (28 mg/kg/day, orally), or the combination. Ifenprodil and sorafenib were purchased from Selleckchem (#S4091) and LC Laboratories (#S-8502), respectively. The drug treatment was started after grouping, and lasted for 21 days. Tumor volume and body weight of each mouse were then measured every two days. The estimated tumor volume served as a guide for determining the day when the mice were sacrificed to take actual measurements on the resected tumors. The mice were sacrificed and the tumor tissues were collected at the end of treatment. Part of the tumor tissues was snap-frozen in liquid nitrogen for protein extraction and immunoblot analysis. Part of the tissues was fixed in 4% paraformaldehyde (PFA) at room temperature for histologic staining. The remaining tissues were dissociated into single cells for the limiting dilution spheroid formation assay. To evaluate the tumor-initiating and self-renewal abilities of the drug-treated tumors derived from MHCC97L cells and PDX1 in vivo, limiting dilution and serial transplantation assays were performed. 500, 1,000, 5,000, 10,000 cells dissociated from the residual tumor tissues were injected subcutaneously into either flank of 4- to 5-week-old mice. No drug treatment was further applied. The mice were checked every two days, and tumor incidence and latency were recorded. To evaluate HCC metastases, 4 × 104 RIL-175 cells were injected through tail vein into 6- to 8-week-old male nude mice. Drug administration began at day 5 after injection, and daily treatment with DMSO, ifenprodil (20 mg/kg/day, intraperitoneally), sorafenib (28 mg/kg/day, orally), or the combination was applied for 10 days before resection of lung tumors. At 5 and 12 days after the tail-vein injection, the mice were administered 100 mg/kg D-luciferin via peritoneal injection for bioluminescent imaging (using the PE IVIS Spectrum in vivo imaging system) to monitor the lung metastasis.

Data and code availability

The sequencing data generated during this study are available.

A CRISPR–Cas9 screen identifies combinatorial therapeutic targets that inhibit HCC growth

To identify combinatorial actionable targets for HCC, we adopted the CombiGEM-CRISPR v2.0-based screening strategy (Fig. 1A; refs. 20, 21, 23). In our screening library, we covered a set of genes that are existing or potential drug targets for suppressing HCC growth, as well as those that are upregulated in two liver CSC subsets and have matching drugs being available for targeting them readily. Data mining was performed in search for genes that are upregulated in liver CSCs, using our in-house transcriptome data comparing the sorted CD133+ liver CSCs versus CD133 differentiated subsets (24, 25) and a publicly available transcriptome data set comparing sorted EpCAM+ liver CSCs versus EpCAM differentiated subsets (GSE5975; ref. 26). Based on the above selection criteria, we shortlisted 27 genes to be included in our library (Supplementary Table S2). Three additional genes that were previously identified by our team to be preferentially overexpressed and activated in tumorigenic CD133+ liver CSCs were also included (Supplementary Table S2; refs. 22, 24, 27). We designed three single-guide RNAs (sgRNA) per gene, which have high on-target scores (i.e., at least 0.60 based on the Azimuth 2.0 model; ref. 28), and selected three nontargeting control sgRNAs from the GeCKOv2 library (29) that do not have on-target loci in the human genome for library construction (Supplementary Table S2). We then assembled a high-coverage (98.7%) barcoded pairwise sgRNA library with 93 × 93 sgRNAs (i.e., 8,649 total combinations; Supplementary Fig. S2A).

We conducted a pooled screen to isolate pairs of sgRNAs that modulate growth of MHCC97L HCC cells stably expressing Streptococcus pyogenes Cas9. Barcode abundances between days 7 and 0 (baseline) groups were compared to yield log2 fold changes as a measure of cell growth (see Supplementary Materials and Methods). Based on the genetic screen data, 11 gene combinations had at least two sets of sgRNA combinations with mean log2 ratios of < −0.51 (with at least 30% fewer barcode counts in day 7- versus day 0-cultured cells) and P values of < 0.05 based on results obtained from two biological replicates (Fig. 1B; Supplementary Fig. S2B–S2E; Supplementary Table S3). The resulting hit list contained gene combinations (FLT4 + PDGFRA and FLT4 + FGFR2) that encode the protein targets of the first-line multikinase inhibitors sorafenib and lenvatinib for HCC treatment. The list also included the FGFR3 + ANXA3 combination that can be targeted by using sorafenib plus an anti-ANXA3–neutralizing antibody and was shown to inhibit HCC growth (22). This result highlights the applicability of our screen to isolate anticancer therapeutic targets. Our screen also uncovered new druggable combinations that significantly reduce MHCC97L growth. Among those, NMDAR1 + FLT4 and NMDAR1 + FGFR3 were identified as the top hits to be characterized in this study, because four independent sets of sgRNA combinations targeting those genes gave a mean log2 ratios of < −0.51 at P < 0.05 and their gene products can be inhibited by using same matching drug pairs. Further analysis with The Cancer Genome Atlas (TCGA)–Liver Hepatocellular Carcinoma (LIHC) data set found NMDAR1 expression in HCC tumor to be higher than that in nontumor liver tissues (Fig. 1C) and patients with low NMDAR1 expression showed better overall survival (Fig. 1D), highlighting its potential clinical relevance that awaits further investigation.

Using individual non-pooled cell viability and colony formation assays, we separately validated the growth inhibition effects on MHCC97L cells by the four sgRNA combinations for NMDAR1 + FLT4 or FGFR3 (Fig. 1EH). No or much less inhibition was observed when those genes were genetically ablated individually (Fig. 1EH). The growth inhibition effects brought by the sgRNA combinations were not due to excessive double-strand DNA breaks (30, 31) because simultaneous targeting of two safe harbor loci did not inhibit growth (Supplementary Fig. S3A). The DNA copy numbers of NMDAR1, FLT4, and FGFR3 loci are also not amplified in MHCC97L's genome (32). Similar growth inhibition effects brought by the sgRNA combinations were also observed in Hep3B cells (Supplementary Fig. S4). Because NMDAR1 was upregulated in both EpCAM+ and CD133+ HCC cell subpopulations in our transcriptome analyses described above, we sought to evaluate whether those sgRNA combinations suppressed the self-renewal ability of CSCs that fuels cancer cell growth. Using in vitro limiting dilution spheroid formation assays, we demonstrated that the genetic coablation of NMDAR1 + FLT4 and NMDAR1 + FGFR3 substantially reduced MHCC97L's ability to assemble tumor spheres (Fig. 1I and J; Supplementary Fig. S3B), albeit that the effect brought by NMDAR1 ablation was not further enhanced when combined with FGFR3 ablation.

NMDAR inhibitors synergize with sorafenib and lenvatinib to reduce growth and self-renewal ability of HCC cells

The genetic data led us to evaluate the efficacy of targeting NMDAR1 + FLT4 or FGFR3 for HCC treatment. We examined the growth inhibition effects by treating HCC cells with the matching inhibitor drugs. MK-801 and ifenprodil (IFEN) were used to inhibit NMDARs, whereas infigratinib (INF), erdafitinib (ERA), SAR131675 (SAR), sorafenib (SOR), and lenvatinib (LEN) were used to target FLT4 and/or FGFR3. In line with our genetic data, we found that IFEN when combined with the FGFR inhibitors (INF and ERA) and the selective FLT4 inhibitor (SAR) greatly reduced the growth of MHCC97L cells (Fig. 2AC; Supplementary Fig. S5). IFEN and the broad-spectrum kinase inhibitors SOR/LEN also act synergistically to suppress the growth of MHCC97L cells (Fig. 2D and E; Supplementary Fig. S5). Similar synergy was observed when MHCC97L cells were treated with MK-801 and SOR (Fig. 2F; Supplementary Fig. S5). MK-801 blocks all NMDAR subunits in a nonselective manner, which poses high toxicity and is less tolerated when used as a therapeutic agent (33), whereas IFEN is more selective for NMDAR1/NMDAR2B subunits and has been used as a vasodilator in some countries including Japan and France with known safety history (34). Like when NMDAR1 was genetically ablated, growth inhibition was also observed when NMDAR2B was ablated in SOR-treated MHCC97L cells (Supplementary Fig. S6A). We observed minimal effects on growth inhibition when IFEN was added together with SOR to the NMDAR1- and NMDAR2B-ablated cells (Supplementary Fig. S6B), indicating that the IFEN-induced growth inhibition is likely NMDAR1/NMDAR2B-dependent. We therefore decided to focus on using IFEN in all subsequent work. SOR/LEN were also selected for the further experiments because they are the first-line drugs used in HCC treatment. Similar synergistic effects by IFEN and SOR/LEN were observed for additional HCC cell lines such as Hep3B, Huh7, and HepG2 HCC cells (Fig. 2GI; Supplementary Figs. S5; S7A–S7D), albeit that the synergy observed for treatment with IFEN and SOR in Huh7 cells was relatively weaker. Hepatic L02 and HepaRG cells showed minimal synergistic effects to the IFEN and SOR cotreatment (Supplementary Fig. S8A–S8D). We also observed that cotreatment of IFEN and SOR did not effectively inhibit growth of sorafenib-resistant HepG2 cells and no synergy was detected, suggesting the additional anticancer effect brought by ifenprodil requires sorafenib's activity (Supplementary Fig. S9). This result is consistent with the model that while synergism can provide a steep increment in efficacy when a second drug is added to the treatment regimen, it would also mean that the efficacy to the second drug would likely be decreased if cancer cells develop single-drug resistance to the first drug (35).

To measure the inhibition of HCC cell's self-renewal ability brought by the drug combination, in vitro limiting dilution spheroid formation assays were performed (Fig. 2J). Cotreatment with IFEN and SOR/LEN markedly suppressed the ability of all four tested lines of HCC cells, but not L02, to form tumor spheres (Fig. 2KN; Supplementary Figs. S7D and S8D). Notably, similar levels of suppression were also observed when IFEN and SOR were only treated for two days to the HCC cell cultures prior to plating for spheroid formation assays (Supplementary Fig. S10), suggesting that the drug combination had eliminated the tumor-initiating cell population or abrogated the self-renewal capacity of the cells.

Cotreatment of ifenprodil and sorafenib induces unfold protein response, downregulates WNT signaling, and results in G1 phase cell-cycle arrest in HCC cells

We then explored the mechanism through which the IFEN + SOR combination inhibited HCC cell growth and self-renewal. It was previously reported that a competitive NMDAR antagonist 1l augments the cytotoxic action of sorafenib in murine HCC cells, and it may act by interfering with the lipid signaling pathway, reducing expression of multidrug resistance transporters (MDR), and thereby increasing the accumulation of sorafenib in cancer cells (36). Unexpectedly, we did not observe reduction of MDR transporter expressions in IFEN + SOR-treated MHCC97L cells (Supplementary Fig. S11A). A transcriptome analysis was then performed using RNA-seq to systematically identify differentially expressed genes and their enriched pathways in drug(s)-treated MHCC97L cells as compared with nontreated cells (Fig. 3A and B; Supplementary Fig. S11B–S11D; Supplementary Tables S4–S6). Mapping of the differentially expressed genes to Gene Ontology groups revealed that genes involved in endoplasmic reticulum (ER) stress response are significantly upregulated, whereas many cell-cycle– and DNA replication–related genes are downregulated, in cells treated with IFEN + SOR combination (Fig. 3C).

Upon ER stress, cells activate the unfolded protein response (UPR; ref. 37). We performed Western blot analysis and confirmed that IFEN + SOR treatment activates UPR effectors, including phosphorylated IRE1-alpha, ATF6, and C/EBP homologous protein (CHOP), in MHCC97L cells (Fig. 3D and E). We also confirmed that IFEN + SOR combination treatment greatly increased the expression of p21Cip1, a regulator of cell-cycle progression at G1–S phase, which was accompanied by the downregulation of CDK2 (Fig. 3D) and other cell-cycle– and DNA replication–related proteins including PCNA and TTK, as well as the activation of the apoptosis signaling pathway (Supplementary Fig. S11E–S11G). Similar results were also detected in another HCC cell line HepG2 treated with IFEN + SOR (Supplementary Fig. S12A and S12B). Cell-cycle analysis was then performed, and G1-phase arrest was detected in MHCC97L cells treated with IFEN + SOR, but not with either IFEN or SOR alone (Fig. 3F). To analyze whether IFEN + SOR treatment causes G1-phase arrest through UPR signaling, we created cells that are deficient in UPR sensing. Because UPR is an integrated ER stress–response pathway that is coordinated by three sensor proteins (i.e., IRE1-alpha, PERK, and ATF6), we thus generated cells that are depleted of all the three sensors (Supplementary Fig. S13A and S13B) as previously reported (38). We found that G1-phase arrest and p21Cip1 expression induced by IFEN + SOR treatment was rescued in these cells (Fig. 3F; Supplementary Fig. S13C), indicating that the drug combination-induced G1-phase arrest critically depends on UPR.

Previous studies have shown that UPR induction is associated with the reduction of WNT- and stem cell–related gene expressions and causes the loss of cell stemness properties (39–41). In line with these findings, we detected in our RNA-seq data that WNT signaling–/stemness-related genes (including Lgr5 and Axin2) were significantly downregulated in MHCC97L treated with IFEN + SOR (Supplementary Fig. S11C and S11G; Supplementary Table S4). We further performed the TOPFlash assay to report on the activation of WNT target genes. Our results showed that WNT signaling was markedly downregulated in HCC cells treated with IFEN + SOR, when compared with those treated with vehicle and the single drugs alone (Fig. 3G; Supplementary Fig. S12C). We further confirmed the decreased expressions of WNT target genes including Lgr5, c-Myc, CCND1, CTNNB1, and MMP-7 at the protein level (Fig. 3H; Supplementary Fig. S12A). To directly evaluate whether the loss of cell stemness brought by IFEN + SOR treatment was UPR-dependent, we performed limiting dilution assay on the UPR-depleted MHCC97L cells and found that IFEN + SOR treatment were less effective in inhibiting the self-renewal of these cells (Fig. 3I). The reduced protein expressions of the WNT target genes, including the cell stemness marker Lgr5, brought by the IFEN + SOR treatment were also relieved in the UPR-depleted MHCC97L cells (Fig. 3H). These results show that the drug combination-induced loss of cell stemness requires UPR.

Collectively, our data suggest that IFEN and SOR synergized to induce UPR, thereby suppressing the growth and self-renewal of HCC cells.

Ifenprodil treatment enhances the efficacy of sorafenib in HCC patient-derived organoids and xenograft models

Human primary liver cancer–derived organoid culture models the pathophysiology of a growing tumor in vivo (42). We extended our findings to a panel of organoids derived from different HCC patients to evaluate inhibitions of growth and self-renewal brought by the drug combination. All HCC patient-derived organoids used in this study have been thoroughly characterized, either in-house or in previous studies (43), at both molecular and phenotypic levels, with comparisons made against the original tissue samples (Supplementary Fig. S14; see Supplementary Materials and Methods). Consistent with what we observed in the cell line models, strong synergy between IFEN and SOR on suppressing HCC growth was evident in three out of four different organoids tested (Fig. 4A; Supplementary Fig. S15). Increased expression of phosphorylated IRE1-alpha, CHOP, phosphorylated-eIF2-alpha, and p21Cip1, as well as a decrease in CDK2 and reduced WNT signaling, was detected in organoids treated with the drug combination (Fig. 4B). Furthermore, the drug combination, but not individual drugs alone, profoundly suppressed the ability of those HCC organoids to assemble tumor spheres (Fig. 4C).

Next, we used the MHCC97L cell– and HCC patient–derived xenograft models to determine the efficacy of the drug combination in vivo. In all three xenografts, tumor weight was significantly suppressed in mice treated with the drug combination, but not the single drugs alone (Fig. 5AC; Supplementary Fig. S16). There were signs of increased apoptosis (i.e., more small, swollen cells with hypereosinphilic cytoplasm) in the tumors cotreated with IFEN and SOR (Fig. 5AC), which is consistent with the increase in Annexin V–positive cells observed after the cotreatment in vitro (Supplementary Fig. S11F). None of the mice showed signs of behavior abnormalities or significant changes in body weight (Fig. 5AC). Corroborating with the drug response in vitro, treatment with the drug combination in vivo similarly led to an upregulation of UPR and an increase in p21Cip1 in the MHCC97L cell– and HCC patient–derived xenografts (Supplementary Fig. S17). IHC staining on tissue sections from the resected residual xenografts confirmed the reduced expressions of WNT-/stem cell–related Lgr5 and Axin2 proteins in xenografts of mice treated with the drug combination (Fig. 6A). We further performed the spheroid formation assay on cells harvested from the residual xenografts to evaluate their self-renewal ability. The ability of those cells to assemble into spheroids was profoundly inhibited after the IFEN + SOR treatment, but not after the vehicle, IFEN, or SOR treatment (Fig. 6B). When the residual cells from the PDX xenografts were serially transplanted into secondary mouse recipients, the cells treated with drug combination were less capable of re-forming tumors, whereas the ones treated with only one of the drugs alone had faster and more tumor formation (Fig. 6C). These results indicate the tumor-initiating property of HCC cells is depleted by the IFEN and SOR cotreatment.

The high incidence of recurrence and the limited efficacy of first-line drugs for liver cancer have demanded new therapeutic options. Here we presented a strategy to rapidly discover effective combination therapies and drug-repurposing opportunities by inhibiting multiple druggable targets using a high-throughput functional genomic screen. The screen results guided us to identify the coablation of NMDAR1 and FLT4/FGFR3 inhibits HCC cell growth and self-renewal. We further validated the strong anticancer effects brought by the matching drug pair IFEN and SOR in HCC cell lines, patient-derived organoids, and two xenograft models. This provides compelling evidence that the screening strategy is effective and could be applied as a standard approach in identifying potent therapeutic combinations. Importantly, the presented strategy can be easily adopted by many laboratories and used in studying various cancer cell types and other diseases.

We discovered the NMDAR inhibitor as a novel coblockade agent for enhancing sorafenib's efficacy in treating HCC. Previous studies have reported the role of NMDAR signaling in promoting growth, invasion, and metastasis in several cancer cell types, but the role of NMDAR in HCC remains unclear. High NMDAR2B expression is associated with poor prognosis of pancreatic cancer, breast cancer, ovarian cancer, and glioblastoma (44). Pharmacologic inhibition of NMDAR using MK-801 reduced tumor growth and invasiveness of pancreatic cancer cells (44, 45), whereas genetic knockdown of NMDAR2B reduced breast-to-brain metastasis (46). Based on TCGA LIHC database, we found that high NMDAR1 expression correlates with poor survival rate in HCC patients. Unlike in other cancers, we found that inhibition of NMDAR using MK-801 or IFEN alone did not significantly suppress HCC growth. Notably, unanticipated drug synergy between IFEN and SOR was revealed in our assays, in which inhibitions of both growth and self-renewal ability of HCC cells were greatly enhanced. The addition of IFEN to SOR offers an option to use a lower dose of SOR to reduce its toxicity to normal liver cells that express no or low level of NMDAR, while achieving greater or similar anticancer effects. Being able to lower the treatment dose of SOR could also reduce its discontinuation rate for HCC patients (47). Furthermore, in our experiments, we observed increased cancer stemness in the HCC mouse models after SOR treatment, and addition of IFEN to the treatment reverted such phenotype, suggesting that the combination treatment may be useful for treating HCC cells that gain CSC properties after SOR treatment. The drug combination did not, however, inhibit the metastatic ability of HCC cells in a lung metastasis model in mouse (Supplementary Fig. S18). To evaluate HCC metastases, the RIL-175 cells used for establishing lung metastasis model were genetically modified to harbor p53 knockout and c-Myc overexpression, so that the cells could colonize and invade the lung tissues through tail-vein injection, whereas the RIL-175 cells used for orthotopic injection into the liver do not harbor such modifications. Loss of p53 has been reported to drive metastasis in liver cancer (48, 49). In our study, we showed that the combination treatment could decrease the expression of c-Myc, which is a WNT target gene. However, p53 expression or related signaling was not decreased as shown in our RNA-seq data. We speculate that the drug combination treatment may not affect the activated signaling and metastatic phenotype resulted from p53 loss in RIL-175 cells; therefore, we did not observe any inhibition of lung metastasis. As advanced HCC patients often present with aggressive clinical features such as tumor metastasis to different organs, it will be worthwhile to investigate if the drug combination treatment could inhibit the metastatic ability of HCC using human patient xenografts established from patients with different genetic mutations.

The mechanisms through which IFEN and SOR cotreatment led to a stronger and synergistic suppression of HCC growth and self-renewal were investigated. Reports have shown that NMDAR inhibition could result in different cellular perturbations depending on the cellular context (36, 50, 51), though the underlying mechanisms remain unclear. For example, the reduced expression of MDR transporters observed in murine HCC cells treated with NMDAR antagonist 1l, but not in human HCC cells treated with IFEN, could be due to the difference in cellular regulation in the two species. Another possibility is that 1l bound on its other known targets including AMPA receptors and kainic receptors at a higher affinity to reduce the expression of MDRs, when the drug was used at a high concentration (i.e., 100 μmol/L) to exert its cytotoxic effect with SOR in murine HCC cells (36). In this study, our comparative transcriptomic analysis reveals that UPR was intensified after combining IFEN with SOR in treating MHCC97L cells, and we further confirmed that UPR is important for their growth arrest and self-renewal. Across a panel of HCC cell lines and patient-derived organoids, strong activation of multiple UPR effectors was consistently detected at the protein level when these two drugs were coapplied. In addition, we detected the unexpected synergy between IFEN and SOR in downregulating both CDK2 level and WNT signaling. The depletion of Cdk2 could account for the corresponding cell-cycle arrest at G1 phase being observed in IFEN + SOR-treated cells. In our experiments, we also detected increased apoptosis after cotreatment of IFEN and SOR (Supplementary Fig. S11E and S11F). Further investigation will help delineate the cell death mechanism(s) utilized by the drug combination to result in tumor reduction. On the other hand, WNT signaling is one of the key regulators of cancer stemness (52). The suppressed WNT signaling, including the reduced expression of functional CSC marker protein Lgr5, in IFEN + SOR-treated cells could contribute to the reduced capacity of HCC cells to self-renew and regrow into a tumor.

Clinically, much research has been carried out in search of potent CDK2 blockers and WNT inhibitors for cancer treatment. CDK2 is hypothesized to be dispensable in the cell cycle of normally functioning cells but is important to the growth of cancer cells (53). However, a CDK2 inhibitor is difficult to design given its high similarity to other essential CDK proteins like CDK1 (54), whereas nonselective CDK inhibitors are toxic to most cells (55). On the other hand, many of the WNT inhibitors being developed are still undergoing preclinical or early clinical trials to evaluate their efficacy and safety (56). Our results uncovered IFEN as a safe drug candidate that acts to augment sorafenib's anticancer effects and repress CDK2 and WNT signaling in HCC. In addition to HCC, sorafenib is indicated for treating other cancers including advanced renal cell carcinoma, thyroid cancer, and FLT3-ITD–positive acute myeloid leukemia, in which CDK2 and/or WNT signaling have also been implicated in their tumorigenesis and recurrence (57–60). It will be worthwhile to evaluate the therapeutic potential of combining IFEN and SOR for treating those cancers in follow-up studies.

In translating the genetic targets that were identified in our screen into matching drugs to be readily tested, we took references on the DrugBank (61) to select drugs with known target information and safety profile to maximize the repurposing opportunities. Sorafenib and lenvatinib were used as the matching inhibitors to target FLT4 and FGFR3 because they are the first-line drugs for treating HCC. We chose IFEN to target NMDAR as it is an approved inhibitor drug with known mechanism of action and was tested safe in human. IFEN was originally developed as a drug to treat peripheral circulatory disorders, and there have been ongoing investigations on repurposing IFEN for treating patients with idiopathic pulmonary fibrosis and lung injury associated with COVID-19 infection (NCT04382924). Here we observed strong anticancer effects brought by combined treatment of IFEN and SOR using HCC xenograft models, whereas no toxicity was detected in the drug-treated mice. Using the body surface area normalization method (62) to translate the dose of IFEN used in our xenograft studies in mouse into a human equivalent dose, about 1.6 mg/kg/day of IFEN might be able to result in strong anticancer effect when combined with SOR treatment in human, which is similar to its dose being safe and effective for treating dizziness caused by cerebral infarction sequela in patients. In our animal experiments, we also evaluated the effect brought by the combined treatment of IFEN and SOR on tumor vasculature. We observed reduction of CD31-positive vascular structures after the single-drug treatments of IFEN and SOR in xenografted tumors in mice, whereas the combined treatment did not further enhance the reduction probably because the number of CD31-positive vascular structures was mostly suppressed after sorafenib treatment (Supplementary Fig. S17C). Our result suggests that the drug combination at least retains similar suppressive effect to sorafenib alone on tumor vasculature in the tested animal model. Further studies will be needed to assess any effects on tumor vasculature, as well as clinical toxicity or vasodilatory side effects, for the combined drugs to effectively treat HCC in future trials. For example, blood parameters of toxicity, hepatic function test, and organ histology could be compared.

Our data indicate that IFEN plus SOR could be considered as the initial treatment combination regimen, but not after the cancer cells develop SOR resistance because this drug combination may not effectively inhibit growth of SOR-resistant cells. Based on the model described in Saputra and colleagues (35), nonsynergistic (i.e., antagonistic or additive) drug combinations appear to be more likely than synergistic ones to suppress expansion of resistant subclones and provide a long-term defense against the evolution of therapeutic resistance. Future studies could help in identifying additional or combining other existing drug candidates that may interact with IFEN plus SOR nonsynergistically to help tackle SOR resistance in HCC. In addition, other drug-repurposing candidates, for example those identified by CMap that show connectivity to gene expression signatures from cells cotreated with IFEN and SOR (Supplementary Fig. S19), could be further tested for their efficacy and combinatorial effects with IFEN and/or SOR for improving HCC treatment.

A.S.L. Wong reports grants from The University of Hong Kong internal funds, Hong Kong Research Grants Council (GRF-17102218 GRF-17102218), National Natural Science Foundation of China Excellent Young Scientists Fund (32022089), Croucher Foundation Start-up Allowance, and Innovation and Technology Fund, The Government of Hong Kong Special Administrative Region of the People's Republic of China during the conduct of the study; in addition, F. Xu and A.S.L. Wong have a patent for compositions and methods of using NMDAR inhibitors and kinase inhibitors to treat liver cancer pending. No disclosures were reported by the other authors.

F. Xu: Conceptualization, data curation, formal analysis, validation, investigation, writing–review and editing. M. Tong: Conceptualization, data curation, formal analysis, validation, investigation, writing–review and editing. C.S.W. Tong: Data curation, investigation, and visualization. B.K.C. Chan: Data curation. H.Y. Chu: Formal analysis. T.L. Wong: Formal analysis. J.H.C. Fong: Data curation. M.S.H. Cheung: Data curation. K.H.-M. Mak: Data curation. L. Pardeshi: Formal analysis. Y. Huang: Formal analysis and methodology. K.H. Wong: Formal analysis. G.C.G. Choi: Methodology, writing–review and editing. S. Ma: Conceptualization, resources, formal analysis, supervision, funding acquisition, methodology, writing–review and editing. A.S.L. Wong: Conceptualization, resources, formal analysis, supervision, funding acquisition, methodology, writing–original draft, project administration, writing–review and editing.

The authors thank members of the Wong and Ma labs for helpful discussions. They thank Yuk Kei Wan, Chaya Yuen, and Joe Lam for the technical assistance. The authors thank the Centre for PanorOmic Sciences at LKS Faculty of Medicine, The University of Hong Kong, for providing support on the RNA-seq experiments, flow cytometry analysis, and cell sorting. They thank the Centre for Comparative Medicine Research at LKS Faculty of Medicine, The University of Hong Kong, for providing support to their animal work experiments. The authors also thank the Information Technology Services at The University of Hong Kong for maintaining and providing supports on utilizing the High Performance Computing System to process their RNA-seq data. This work was supported by The University of Hong Kong internal funds, the Hong Kong Research Grants Council (GRF-17102218 to A.S.L. Wong and CRF-C7026-18G to S. Ma), the National Natural Science Foundation of China Excellent Young Scientists Fund (32022089 to A.S.L. Wong), and the Croucher Foundation Start-up Allowance (to A.S.L. Wong). This work was also supported by the Centre for Oncology and Immunology (to A.S.L. Wong) and Laboratory for Synthetic Chemistry and Chemical Biology (to S. Ma) under the Health@InnoHK Program launched by the Innovation and Technology Commission, The Government of Hong Kong Special Administrative Region of the People's Republic of China. This work was partially supported by a funding from the Research Services and Knowledge Transfer Office (MYRG201800017-FHS and MYRG201900099-FHS) and Faculty of Health Sciences (FHS collaboration and innovation grants) of the University of Macau (to K.H. Wong). A.S.L. Wong is a Ming Wai Lau Centre for Reparative Medicine (MWLC) Associate Member.

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.

1.
Al-Lazikani
B
,
Banerji
U
,
Workman
P
. 
Combinatorial drug therapy for cancer in the post-genomic era
.
Nat Biotechnol
2012
;
30
:
679
92
.
2.
Global Burden of Disease Cancer C
,
Fitzmaurice
C
,
Allen
C
,
Barber
RM
,
Barregard
L
,
Bhutta
ZA
, et al
Global, regional, and national cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life-years for 32 cancer groups, 1990 to 2015: a systematic analysis for the Global Burden of Disease Study
.
JAMA Oncol
2017
;
3
:
524
48
.
3.
Yang
JD
,
Hainaut
P
,
Gores
GJ
,
Amadou
A
,
Plymoth
A
,
Roberts
LR
. 
A global view of hepatocellular carcinoma: trends, risk, prevention and management
.
Nat Rev Gastroenterol Hepatol
2019
;
16
:
589
604
.
4.
Llovet
JM
,
Ricci
S
,
Mazzaferro
V
,
Hilgard
P
,
Gane
E
,
Blanc
JF
, et al
Sorafenib in advanced hepatocellular carcinoma
.
N Engl J Med
2008
;
359
:
378
90
.
5.
Kudo
M
,
Finn
RS
,
Qin
S
,
Han
KH
,
Ikeda
K
,
Piscaglia
F
, et al
Lenvatinib versus sorafenib in first-line treatment of patients with unresectable hepatocellular carcinoma: a randomised phase 3 non-inferiority trial
.
Lancet
2018
;
391
:
1163
73
.
6.
Hodi
FS
,
O'Day
SJ
,
McDermott
DF
,
Weber
RW
,
Sosman
JA
,
Haanen
JB
, et al
Improved survival with ipilimumab in patients with metastatic melanoma
.
N Engl J Med
2010
;
363
:
711
23
.
7.
Robert
C
,
Thomas
L
,
Bondarenko
I
,
O'Day
S
,
Weber
J
,
Garbe
C
, et al
Ipilimumab plus dacarbazine for previously untreated metastatic melanoma
.
N Engl J Med
2011
;
364
:
2517
26
.
8.
Greten
TF
,
Sangro
B
. 
Targets for immunotherapy of liver cancer
.
J Hepatol
2017
;
S0168-8278(17)32287-0
.
9.
Finn
RS
,
Qin
S
,
Ikeda
M
,
Galle
PR
,
Ducreux
M
,
Kim
TY
, et al
Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma
.
N Engl J Med
2020
;
382
:
1894
905
.
10.
He
AR
,
Yau
T
,
Hsu
C
,
Kang
Y-K
,
Kim
T-Y
,
Santoro
A
, et al
Nivolumab (NIVO) + ipilimumab (IPI) combination therapy in patients (pts) with advanced hepatocellular carcinoma (aHCC): Subgroup analyses from CheckMate 040
.
J Clin Oncol
2020
;
38
:
512
-.
11.
Plenge
RM
,
Scolnick
EM
,
Altshuler
D
. 
Validating therapeutic targets through human genetics
.
Nat Rev Drug Discov
2013
;
12
:
581
94
.
12.
Nelson
MR
,
Tipney
H
,
Painter
JL
,
Shen
J
,
Nicoletti
P
,
Shen
Y
, et al
The support of human genetic evidence for approved drug indications
.
Nat Genet
2015
;
47
:
856
60
.
13.
Doench
JG
. 
Am I ready for CRISPR? A user's guide to genetic screens
.
Nat Rev Genet
2018
;
19
:
67
80
.
14.
Wong
AS
,
Choi
GC
,
Lu
TK
. 
Deciphering combinatorial genetics
.
Annu Rev Genet
2016
;
50
:
515
38
.
15.
Hay
M
,
Thomas
DW
,
Craighead
JL
,
Economides
C
,
Rosenthal
J
. 
Clinical development success rates for investigational drugs
.
Nat Biotechnol
2014
;
32
:
40
51
.
16.
Kreso
A
,
Dick
JE
. 
Evolution of the cancer stem cell model
.
Cell Stem Cell
2014
;
14
:
275
91
.
17.
Sell
S
,
Leffert
HL
. 
Liver cancer stem cells
.
J Clin Oncol
2008
;
26
:
2800
5
.
18.
Saygin
C
,
Matei
D
,
Majeti
R
,
Reizes
O
,
Lathia
JD
. 
Targeting cancer stemness in the clinic: from hype to hope
.
Cell Stem Cell
2019
;
24
:
25
40
.
19.
Subramanian
A
,
Narayan
R
,
Corsello
SM
,
Peck
DD
,
Natoli
TE
,
Lu
X
, et al
A next generation connectivity map: L1000 platform and the first 1,000,000 profiles
.
Cell
2017
;
171
:
1437
52
.
20.
Wong
AS
,
Choi
GC
,
Cheng
AA
,
Purcell
O
,
Lu
TK
. 
Massively parallel high-order combinatorial genetics in human cells
.
Nat Biotechnol
2015
;
33
:
952
61
.
21.
Wong
AS
,
Choi
GC
,
Cui
CH
,
Pregernig
G
,
Milani
P
,
Adam
M
, et al
Multiplexed barcoded CRISPR-Cas9 screening enabled by CombiGEM
.
Proc Natl Acad Sci U S A
2016
;
113
:
2544
9
.
22.
Tong
M
,
Che
N
,
Zhou
L
,
Luk
ST
,
Kau
PW
,
Chai
S
, et al
Efficacy of annexin A3 blockade in sensitizing hepatocellular carcinoma to sorafenib and regorafenib
.
J Hepatol
2018
;
69
:
826
39
.
23.
Zhou
P
,
Chan
BKC
,
Wan
YK
,
Yuen
CTL
,
Choi
GCG
,
Li
X
, et al
A three-way combinatorial CRISPR screen for analyzing interactions among druggable targets
.
Cell Rep
2020
;
32
:
108020
.
24.
Tong
M
,
Fung
TM
,
Luk
ST
,
Ng
KY
,
Lee
TK
,
Lin
CH
, et al
ANXA3/JNK signaling promotes self-renewal and tumor growth, and its blockade provides a therapeutic target for hepatocellular carcinoma
.
Stem Cell Reports
2015
;
5
:
45
59
.
25.
Ma
S
,
Tang
KH
,
Chan
YP
,
Lee
TK
,
Kwan
PS
,
Castilho
A
, et al
miR-130b Promotes CD133(+) liver tumor-initiating cell growth and self-renewal via tumor protein 53-induced nuclear protein 1
.
Cell Stem Cell
2010
;
7
:
694
707
.
26.
Yamashita
T
,
Ji
J
,
Budhu
A
,
Forgues
M
,
Yang
W
,
Wang
HY
, et al
EpCAM-positive hepatocellular carcinoma cells are tumor-initiating cells with stem/progenitor cell features
.
Gastroenterology
2009
;
136
:
1012
24
.
27.
Ma
S
,
Chan
KW
,
Hu
L
,
Lee
TK
,
Wo
JY
,
Ng
IO
, et al
Identification and characterization of tumorigenic liver cancer stem/progenitor cells
.
Gastroenterology
2007
;
132
:
2542
56
.
28.
Doench
JG
,
Fusi
N
,
Sullender
M
,
Hegde
M
,
Vaimberg
EW
,
Donovan
KF
, et al
Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9
.
Nat Biotechnol
2016
;
34
:
184
91
.
29.
Shalem
O
,
Sanjana
NE
,
Hartenian
E
,
Shi
X
,
Scott
DA
,
Mikkelsen
TS
, et al
Genome-scale CRISPR-Cas9 knockout screening in human cells
.
Science
2014
;
343
:
84
7
.
30.
Aguirre
AJ
,
Meyers
RM
,
Weir
BA
,
Vazquez
F
,
Zhang
CZ
,
Ben-David
U
, et al
Genomic copy number dictates a gene-independent cell response to CRISPR/Cas9 targeting
.
Cancer Discov
2016
;
6
:
914
29
.
31.
Munoz
DM
,
Cassiani
PJ
,
Li
L
,
Billy
E
,
Korn
JM
,
Jones
MD
, et al
CRISPR screens provide a comprehensive assessment of cancer vulnerabilities but generate false-positive hits for highly amplified genomic regions
.
Cancer Discov
2016
;
6
:
900
13
.
32.
Luo
Q
,
Shi
Y
,
Wang
N
,
Lu
Z
,
Han
Z
. 
Genomic features reveal the clonal evolution in human hepatocellular carcinoma cell lines MHCC97 with stepwise metastatic potentials
.
Res Sq
2020
.
doi: 10.21203/rs.3.rs-32352/v1
.
33.
Chen
HS
,
Lipton
SA
. 
The chemical biology of clinically tolerated NMDA receptor antagonists
.
J Neurochem
2006
;
97
:
1611
26
.
34.
Williams
K
. 
Ifenprodil discriminates subtypes of the N-methyl-D-aspartate receptor: selectivity and mechanisms at recombinant heteromeric receptors
.
Mol Pharmacol
1993
;
44
:
851
9
.
35.
Saputra
EC
,
Huang
L
,
Chen
Y
,
Tucker-Kellogg
L
. 
Combination therapy and the evolution of resistance: the theoretical merits of synergism and antagonism in cancer
.
Cancer Res
2018
;
78
:
2419
31
.
36.
Gynther
M
,
Proietti Silvestri
I
,
Hansen
JC
,
Hansen
KB
,
Malm
T
,
Ishchenko
Y
, et al
Augmentation of anticancer drug efficacy in murine hepatocellular carcinoma cells by a peripherally acting competitive N-Methyl-d-aspartate (NMDA) receptor antagonist
.
J Med Chem
2017
;
60
:
9885
904
.
37.
Hetz
C
. 
The unfolded protein response: controlling cell fate decisions under ER stress and beyond
.
Nat Rev Mol Cell Biol
2012
;
13
:
89
102
.
38.
Adamson
B
,
Norman
TM
,
Jost
M
,
Cho
MY
,
Nunez
JK
,
Chen
Y
, et al
A multiplexed single-cell CRISPR screening platform enables systematic dissection of the unfolded protein response
.
Cell
2016
;
167
:
1867
82
.
39.
Okamoto
Y
,
Saito
T
,
Tani
Y
,
Toki
T
,
Hasebe
A
,
Koido
M
, et al
The kinase PERK represses translation of the G-protein-coupled receptor LGR5 and receptor tyrosine kinase ERBB3 during ER stress in cancer cells
.
J Biol Chem
2020
;
295
:
4591
603
.
40.
Spaan
CN
,
Smit
WL
,
van Lidth de Jeude
JF
,
Meijer
BJ
,
Muncan
V
,
van den Brink
GR
, et al
Expression of UPR effector proteins ATF6 and XBP1 reduce colorectal cancer cell proliferation and stemness by activating PERK signaling
.
Cell Death Dis
2019
;
10
:
490
.
41.
Heijmans
J
,
van Lidth de Jeude
JF
,
Koo
BK
,
Rosekrans
SL
,
Wielenga
MC
,
van de Wetering
M
, et al
ER stress causes rapid loss of intestinal epithelial stemness through activation of the unfolded protein response
.
Cell Rep
2013
;
3
:
1128
39
.
42.
Broutier
L
,
Mastrogiovanni
G
,
Verstegen
MM
,
Francies
HE
,
Gavarro
LM
,
Bradshaw
CR
, et al
Human primary liver cancer-derived organoid cultures for disease modeling and drug screening
.
Nat Med
2017
;
23
:
1424
35
.
43.
Huch
M
,
Gehart
H
,
van Boxtel
R
,
Hamer
K
,
Blokzijl
F
,
Verstegen
MM
, et al
Long-term culture of genome-stable bipotent stem cells from adult human liver
.
Cell
2015
;
160
:
299
312
.
44.
Li
L
,
Hanahan
D
. 
Hijacking the neuronal NMDAR signaling circuit to promote tumor growth and invasion
.
Cell
2013
;
153
:
86
100
.
45.
Li
L
,
Zeng
Q
,
Bhutkar
A
,
Galvan
JA
,
Karamitopoulou
E
,
Noordermeer
D
, et al
GKAP acts as a genetic modulator of NMDAR signaling to govern invasive tumor growth
.
Cancer Cell
2018
;
33
:
736
51
.
46.
Zeng
Q
,
Michael
IP
,
Zhang
P
,
Saghafinia
S
,
Knott
G
,
Jiao
W
, et al
Synaptic proximity enables NMDAR signalling to promote brain metastasis
.
Nature
2019
;
573
:
526
31
.
47.
Reiss
KA
,
Yu
S
,
Mamtani
R
,
Mehta
R
,
D'Addeo
K
,
Wileyto
EP
, et al
Starting dose of sorafenib for the treatment of hepatocellular carcinoma: a retrospective, multi-institutional study
.
J Clin Oncol
2017
;
35
:
3575
81
.
48.
Chen
YW
,
Klimstra
DS
,
Mongeau
ME
,
Tatem
JL
,
Boyartchuk
V
,
Lewis
BC
. 
Loss of p53 and Ink4a/Arf cooperate in a cell autonomous fashion to induce metastasis of hepatocellular carcinoma cells
.
Cancer Res
2007
;
67
:
7589
96
.
49.
Luo
YD
,
Fang
L
,
Yu
HQ
,
Zhang
J
,
Lin
XT
,
Liu
XY
, et al
p53 haploinsufficiency and increased mTOR signalling define a subset of aggressive hepatocellular carcinoma
.
J Hepatol
2021
;
74
:
96
108
.
50.
Yamaguchi
F
,
Hirata
Y
,
Akram
H
,
Kamitori
K
,
Dong
Y
,
Sui
L
, et al
FOXO/TXNIP pathway is involved in the suppression of hepatocellular carcinoma growth by glutamate antagonist MK-801
.
BMC Cancer
2013
;
13
:
468
.
51.
Deutsch
SI
,
Tang
AH
,
Burket
JA
,
Benson
AD
. 
NMDA receptors on the surface of cancer cells: target for chemotherapy?
Biomed Pharmacother
2014
;
68
:
493
6
.
52.
Fodde
R
,
Brabletz
T
. 
Wnt/beta-catenin signaling in cancer stemness and malignant behavior
.
Curr Opin Cell Biol
2007
;
19
:
150
8
.
53.
Duensing
A
,
Liu
Y
,
Tseng
M
,
Malumbres
M
,
Barbacid
M
,
Duensing
S
. 
Cyclin-dependent kinase 2 is dispensable for normal centrosome duplication but required for oncogene-induced centrosome overduplication
.
Oncogene
2006
;
25
:
2943
9
.
54.
Wood
DJ
,
Korolchuk
S
,
Tatum
NJ
,
Wang
LZ
,
Endicott
JA
,
Noble
MEM
, et al
Differences in the conformational energy landscape of CDK1 and CDK2 suggest a mechanism for achieving selective CDK inhibition
.
Cell Chem Biol
2019
;
26
:
121
30
.
55.
Asghar
U
,
Witkiewicz
AK
,
Turner
NC
,
Knudsen
ES
. 
The history and future of targeting cyclin-dependent kinases in cancer therapy
.
Nat Rev Drug Discov
2015
;
14
:
130
46
.
56.
Kahn
M
. 
Can we safely target the WNT pathway?
Nat Rev Drug Discov
2014
;
13
:
513
32
.
57.
Jiang
X
,
Mak
PY
,
Mu
H
,
Tao
W
,
Mak
DH
,
Kornblau
S
, et al
Disruption of Wnt/beta-catenin exerts antileukemia activity and synergizes with FLT3 inhibition in FLT3-mutant acute myeloid leukemia
.
Clin Cancer Res
2018
;
24
:
2417
29
.
58.
Wang
X
,
Chen
X
,
Han
W
,
Ruan
A
,
Chen
L
,
Wang
R
, et al
miR-200c targets CDK2 and suppresses tumorigenesis in renal cell carcinoma
.
Mol Cancer Res
2015
;
13
:
1567
77
.
59.
Fendler
A
,
Bauer
D
,
Busch
J
,
Jung
K
,
Wulf-Goldenberg
A
,
Kunz
S
, et al
Inhibiting WNT and NOTCH in renal cancer stem cells and the implications for human patients
.
Nat Commun
2020
;
11
:
929
.
60.
Sastre-Perona
A
,
Santisteban
P
. 
Role of the wnt pathway in thyroid cancer
.
Front Endocrinol
2012
;
3
:
31
.
61.
Wishart
DS
,
Feunang
YD
,
Guo
AC
,
Lo
EJ
,
Marcu
A
,
Grant
JR
, et al
DrugBank 5.0: a major update to the DrugBank database for 2018
.
Nucleic Acids Res
2018
;
46
:
D1074
D82
.
62.
Reagan-Shaw
S
,
Nihal
M
,
Ahmad
N
. 
Dose translation from animal to human studies revisited
.
FASEB J
2008
;
22
:
659
61
.