Metabolic synthetic lethality by targeting NOP56 and mTOR in KRAS-mutant lung cancer

Cancer cell lines used in this study (Table S1) were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured in RPMI-1640 medium or Medium 199 (Cat. #8758 and #4540; Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum/FBS (Cat. #10270–106; Life Technologies, Grand Island, NY, USA) and 1% penicillin/streptomycin solution (Cat. #P0781, Sigma-Aldrich). The cells were authenticated by DNA fingerprinting and confirmed free from mycoplasma contamination (Microsynth, Bern, Switzerland). All inhibitors used in this study were listed in Table S2.

PF139 and PF563 lung cancer cells were established from lung adenocarcinoma malignant pleural effusion and pleural carcinosis specimens of a 67 year-old female patient and a 75 year-old male patient, respectively, at the time of diagnosis prior to any treatment [31]. Authentication was performed by SNP based cell identification (Multiplexion, Heidelberg, Germany).

Lung cancer cells seeded in 96-well plates (2500 cells/well) were dosed 24 h later with different inhibitors for 72 h. Cell viability was determined by PrestoBlue (PB) Cell Viability Reagent (ThermoFisher Scientific) by following the manufacturer’s instructions [14, 31]. The PB reagent was added into media directly (1:10 dilution) and incubated for 30 min-2 h and then the fluorescence was read (excitation 570 nm; emission 600 nm) at recommended time of incubation. The efficacy of drugs on cell growth was normalized to untreated control. Each data point was generated in triplicate and each experiment was done three times (n = 3). Best-fit curve was generated in GraphPad Prism [(log (inhibitor) vs response (−variable slope four parameters)]. Error bars are mean ± SD. The combination index (CI) was calculated by ComboSyn software (ComboSyn Inc., http://​www.​combosyn.​com/​).

Clonogenic assay was done as we described previously [14, 31‐33]. In brief, cells seeded in 6-well plates (3000 cells/well) were dosed 24 h later and continually treated with rapamycin for 7 days (refresh drugs every 3 days), the resulting colonies were stained with crystal violet (0.5% dissolved in 25% methanol).

Lung cancer cells were treated for 72 h with vehicle or rapamycin. After treatment, cells in the supernatant and adherent to plates were collected, washed with PBS and pooled before suspended in binding buffer and stained with the Annexin V Apoptosis Detection Kit -FITC (Cat. #88–8005; Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. Flow cytometry analysis was performed on a BD Biosciences LSRII flow cytometer.

Transient knockdowns were mediated by siRNAs. Cells cultured in triplicate at 50–70% confluency were transfected using SiTran1.0 (TT300001; Origene Technologies, Rockville, MD, USA) according to the manufacturer’s protocol. NOP56 (CAT#: SR307156), EIF4E (CAT#: SR320018), RPS6 (CAT#: SR304160), RAPTOR (CAT#: SR324724), and RICTOR (CAT#: SR326062) were knocked down by specific pooled siRNA duplexes purchased from OriGene Technologies, with control siRNA Duplex as a negative control.

Stable knockdown of NOP56 was achieved via lentiviral delivery of NOP56 Human shRNA Plasmid Kit (SHCLND_006392, MERCK). A scramble shRNA was used as a control. Lentiviral particles were generated and cells infected according to the protocol from Broad Institute. The supernatant containing lentiviruses was collected, filtered through 0.45 μM filters, and stored in aliquots at − 80 °C, or immediately used to infect recipient cells. After infection, cells were selected in puromycin (1.5 μg/ml) and further passaged in culture for functional assays. NOP56 knockout was performed via a CRISPR/Cas9 and non-homology mediated approach using the NOL5A (NOP56) Human Gene Knockout Kit (CAT#: KN411153; OriGene Technologies) according to the manufacturer’s protocol.

Total RNA was isolated and purified using RNeasy Mini Kit (Qiagen, Hilden, Germany). Complementary DNA was synthesized by the High capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA) according to manufacturer’s instructions. Real time PCR was performed in triplicate on a 7500 Fast RealTime PCR System (Applied Biosystems) using TaqMan primer/probes (Applied Biosystems): HSPA5, Hs00607129_gH; ERN1, Hs00980095_m1; EIF2AK3, Hs00984003_m1; ATF4, Hs00909569_g1; DDIT3, Hs00358796_g1, with GAPDH (Hs02786624_g1) and ACTB (Hs01060665_g1) used as endogenous normalization controls.

Cell lysates were prepared and western blot analysis was performed as described [14, 31]. In brief, equal amounts of protein lysates resolved by SDS-PAGE (Cat. #4561033; Bio-Rad Laboratories, Hercules, CA, USA) and transferred onto nitrocellulose membranes (Cat. #170–4158; Bio-Rad). Membranes were then blocked with blocking buffer (Cat. #927–4000; Li-COR Biosciences, Bad Homburg, Germany) for 1 h at room temperature (RT) and incubated with appropriate primary antibodies overnight at 4 °C (Table S3). IRDye 680LT-conjugated goat anti-mouse IgG (Cat. #926–68,020) and IRDye 800CW-conjugated goat anti-rabbit IgG (Cat. #926–32,211) from Li-COR Biosciences were used at 1:5000 dilutions. Finally, signals of membrane-bound secondary antibodies were imaged using the Odyssey Infrared Imaging System (Li-COR Biosciences).

For immunofluorescence, tumor cells grown on poly-lysine-treated coverslides were fixed with 4% paraformaldehyde for 15 min at RT and permeabilized with cold methanol (− 20 °C) for 5 min or with 0.1% Triton X-100/PBS at RT for 15 min before incubated overnight at 4 °C with primary antibodies (Table S3). The cells were incubated for 1 h at RT with Alexa Fluor 647 goat anti-mouse IgG (Cat. #A21236) or Alexa Fluor 488 goat anti-Rabbit IgG (Cat. #A11034) from Invitrogen (Eugene, OR, USA). Nuclei were counterstained by 4′,6-diamidino-2-phenylindole. Images were acquired on a ZEISS Axioplan 2 imaging microscope (Carl Zeiss MicroImaging, Göttingen, Germany) and processed using Adobe Photoshop CS6 v.13 (Adobe Systems, San Jose, CA, USA).

Immunohistochemical study was performed as we described previously [31, 32]. In brief, surgically removed xenograft tumors were formalin-fixed and paraffin-embedded (FFPE). FFPE tumors were sectioned at 4 μm, deparaffinized, rehydrated and subsequently stained with hematoxylin and eosin (H&E) and appropriate antibodies (Table S3) using the automated system BOND RX (Leica Biosystems, Newcastle, UK). Visualization was performed using the Bond Polymer Refine Detection kit (Leica Biosystems) as instructed by the manufacturer. Images were acquired using PANNORAMIC® whole slide scanners, processed using Case Viewer (3DHISTECH Ltd.). The staining intensities of the whole slide (two tumors/group) were quantified by QuPath software.

Mouse studies were conducted in accordance with Institutional Animal Care and Ethical Committee-approved animal guidelines and protocols. All mouse experiments were performed in age- and gender-matched NSG (NOD-scid IL2Rγ^null) as we previously described [31, 32]. Tumor cells in DMEM (H460-shScrambled or H460-shNOP56) 1:1 mixed with BD Matrigel Basement Membrane Matrix (Cat. #356231; Corning, NY, USA) were subcutaneously inoculated in left and right flanks (0.5 × 10⁶/injection). When tumors were palpable, mice were randomly assigned to treatment groups: 1) control; 2) rapamycin (0.1 mg/kg, i.p, 5 days/week) for 5 weeks. Tumors were measured every 3 days, with their size calculated as follows: (length × width²)/2. For survival analysis, the mice were closely monitored on a daily basis, and the size of tumors was measured with a caliper every 4–5 days. Mice were sacrificed when the tumor volume reached 1500 mm³.

To identify synthetic lethal targets in KRAS-mutant cancers, we interrogated functional genomics dataset of CRISPR/Cas9 knockout and RNAi/shRNA knockdown screens from published studies: whole genome RNAi screens in DLD-1 colon cancer cells [12] and in KRAS-mutant lung cancer cells (H2122, H2009, HCC44, H460, H1155) [16], genome-wide CRISPR/Cas9 loss of function screens in KRAS-mutant leukemia cells (PL-21, SKM-1, NB4) [26]. To minimize the effects of cancer lineage and histological subtype, we selected DLD-1 (colon), H460 (large cell lung carcinoma), H2122 (lung adenocarcinoma), SKM-1 (without PML-RARA fusion) and NB4 (PML-RARA) for further analysis, which identified a number of common candidates (n = 21) as KRAS synthetic lethal partners (Fig. 1).

Interrogation of publicly available datasets was performed as we have described [14, 31]. Specifically, transcriptomic data of lung, pancreatic and colon cancer were obtained from the Cancer Genome Atlas (TCGA) (https://​portal.​gdc.​cancer.​gov/​projects/​TCGA). Gene set enrichment analysis (GSEA) was performed by using GSEA software. The transcriptomic dataset (GSE15212) used for GSEA was derived from KRAS-mutant colon cancer cell line (SW480) treated with NOP56-specific siRNAs and downloaded from the Gene Expression Omnibus (GEO) database. For survival analysis, transcriptomic gene expression and corresponding survival data were extracted and analyzed by using the “maxstat”, “survival”, and “survminer” packages in R software (version 3.6.0). Patients were divided into two groups (high_ NOP56 versus low_ NOP56) based on the optimal cutoff value of NOP56 transcripts across all patients to plot the Kaplan–Meier survival curves.

For correlative analysis of NOP56 expression with sensitivity (IC₅₀) to mTOR inhibitors, gene expression data and drug response profiles were downloaded from Cancer Cell Line Encyclopedia (CCLE) and Genomics of Drug Sensitivity in Cancer (GDSC) databases, respectively. Correlation analysis was performed using R software (version 3.6.0).

To identify therapeutic vulnerabilities in KRAS-mutant cancers, we performed integrated analysis of shRNA- and CRISPR-based functional genomics of previously published studies [12, 16, 26]. To minimize lineage-specific effects, we analyzed whole-genome dropout screen dataset in KRAS-mutant lung (H460, H2122), colon (DLD-1), acute promyelocytic leukemia (NB4) and acute myeloid leukemia (SKM-1) cancer cells, which identified 21 common genes whose loss of function is synthetic lethal with mutant KRAS alleles in distinct cancer lineages (Fig. 1A; Table S4). The protein products of these genes fall into several functional categories, with FBL, NOP56, PLK1 and XPO1 as a core set based on their interaction network (Fig. 1B). Remarkably, PLK1 and XPO1 have been reported to be selectively required for KRAS-mutant cancers by counteracting mitotic and nuclear export stress associated with KRAS-induced tumorigenesis [12, 16], and our recent study has implicated PLK1 in metabolic stress response of KRAS-mutant cancers [14]. FBL has also been assigned as a promising target in cancers [34, 35], suggesting the power of functional genomics in identifying oncogene-specific vulnerabilities and the accountability of our analyses. In the present study, we investigated the function of NOP56 in KRAS-mutant cancers.

Our investigations began with NOP56 knockdown using small interfering RNAs (siRNAs), which revealed that downregulation of NOP56 significantly inhibited the proliferation of numerous KRAS-mutant lung (H358, H460, A549, PF563, PF139), pancreatic (MIAPaCa, HPAF-II) and colon (HCT-116, DLD-1) cancer cells, which differ not only in tumor lineages and histological subtypes but also in KRAS mutations, e.g., G12C, G12D, Q61H, etc. (Fig. S1A, B; Table S1). Notably, NOP56 silencing also inhibited NRAS-mutant lung cancer H1299 cells, although the effects on EGFR-mutant (EBC-1) or FGFR1-amplified (H520) lung cancer cells were negligible (Fig. S1A, B). Supporting these observations, KRAS-mutant lung, pancreatic and colon cancer showed significantly higher expression of NOP56 than KRAS-WT tumors (Fig. 1C) and patients with KRAS-mutant lung adenocarcinoma, pancreatic and colon cancer characterized by a higher NOP56 level are associated with significantly shorter survival (Fig. 1D). In contrast, NOP56 expression is not a prognostic marker for KRAS-mutant lung, pancreatic and colon cancers (Fig. S1C). These results indicate a unique function for NOP56 in KRAS-mutant cancers.

To explore NOP56 functions in KRAS-mutant malignancies, we profiled the transcriptomic gene expression data of a previous study [36], whereby NOP56 in KRAS-mutant colon cancer cells (SW480) was silenced by siRNAs. Our analysis revealed that high expression of NOP56 was positively correlated with the gene signature of KRAS signaling (Fig. S1D), in line with the above results (Fig. 1A-D; Fig. S1A, B), and that, importantly, siRNA-mediated NOP56 knockdown led to significant enrichment of the gene sets involved in ROS pathway (consisting of 49 genes upregulated by ROS) and oxidative phosphorylation (a set of 200 genes encoding proteins involved in oxidative phosphorylation), the latter representing a major source of ROS production (Fig. 1E), suggesting a possible role for NOP56 in the suppression of metabolic ROS that is critical for KRAS-induced tumorigenesis [27‐30]. Supporting this notion, NOP56 knockdown (KD) by siRNAs significantly upregulated ROS in H358 and H460 cells, and H₂O₂ treatment, which elevated the already high level of oxidative stress, provoked significantly greater apoptosis in NOP56 KD H358 and H460 cells than the control counterparts (Fig. 1F, G). These results uncover NOP56 as a metabolic dependency in KRAS-mutant cancer by exerting a previously unrecognized role in the surveillance of oxidative stress.

Next, we investigated the mechanism that KRAS-mutant cancer cells utilize to orchestrate cytotoxic ROS upon NOP56 depletion. GSEA of transcriptomic dataset [36] revealed that NOP56 knockdown significantly enriched the genes involved in the unfolded protein response/UPR (a set of 113 genes upregulated during UPR) in KRAS-mutant cancer cells (Fig. 2A), suggesting that tumor cells might engage the UPR to protect from NOP56 KD-induced surge of cytotoxic ROS. To test this possibility, we knocked down NOP56 in KRAS-mutant lung cancer cells (H358, H460) by using short-hairpin RNAs (shRNA) (Fig. S2A, B). In contrast to the results from siRNA-mediated acute depletion, stable expression of two independent shRNAs showed negligible effects on H358 and H460 proliferation (Fig. S2C, D), which may be due to the activation of compensatory mechanisms. Importantly, several UPR genes, in particular ERN1 and HSPA5 encoding the ER stress sensor IRE1α and the chaperon protein BiP, respectively, were markedly upregulated in NOP56-depleted H358 and H460 cells (Fig. 2B). Western blot confirmed the increase of BiP, IRE1α, XBP-1 s (IRE1α effector) and of the master UPR transcription factor HSF-1 (heat shock factor 1) and PDI (protein disulfide isomerase), an important ER chaperone induced during ER stress by carrying out a redox reaction and responsible for the formation of disulfide bonds in proteins (Fig. 2C). Notably, p38 MAPK, a key stress-responsive kinase and an UPR effector [37], was highly phosphorylated (activated) in NOP56-depleted H358 cells (Fig. 2C). Moreover, IRE1α KD blunted p-p38, p-AKT (T308), p-MNK1, p-eIF4E and p-S6 in NOP56-depleted H358 cells, indicating that IRE1α-mediated UPR acts upstream of p38 signaling (Fig. 2D).

Importantly, genetic (siRNA) and pharmacological (4μ8C, an inhibitor of IRE1α) inhibition of IRE1α preferentially impaired NOP56 KD H358 and H460 cells, manifested by significantly greater proliferative inhibition and apoptotic induction in these cells than in control cells (Fig. 2E, F; Fig. S2E, F). Importantly, the increase of IRE1α KD-induced apoptotic cell death was paralleled by ROS upregulation, and addition of NAC, an ROS scavenger largely dampened IRE1α KD-induced apoptosis (Fig. 2F, G), supporting a role for the UPR in response to oxidative stress. Similarly, genetic and pharmacological inhibition (with KRIBB11) of HSF1 suppressed the proliferation and evoked apoptotic cell death to a markedly greater extent in NOP56 KD H358 cells than in control cells (Fig. 2H-J).

The outcome of UPR ranges from adaptation to apoptosis [38] and, as such, perturbations of ER homeostasis in cells with an already high level of ER stress, e.g., treatment with bortezomib and tunicamycin that induce persistent ER stress by targeting the 26S proteasome and the ER chaperone BiP, respectively, evoke programmed cell death [38, 39]. Indeed, NOP56 KD H358 and H460 cells with high basal levels of ROS are highly susceptible to bortezomib and Tunicamycin compared to control cells (Fig. 2K, L).

To identify cellular processes that may present therapeutic vulnerabilities in NOP56 KD cells, we performed synthetic lethal chemical screens with small-molecule drugs (n = 22) that interrogate various oncogenic pathways, with the ER stress inducers (bortezomib and HA15) included as positive controls (Table S2). Our screens showed that, except for bortezomib and HA15, LY294002, AZD5363, and rapamycin, inhibitors of the PI3K/AKT/mTOR pathway, preferentially suppressed the viability of NOP56 KD cells, gauged by their IC₅₀ decrease in NOP56 KD H358 and H460 cells versus control cells (Fig. 3A; Fig. S3A). The greatest change in sensitivity was conferred by rapamycin, which was 0.7 μM and 1.0 μM in H358_shNOP56a and H358_shNOP56b but 12.0 μM in H358_Scr cells, with a selectivity index (IC₅₀ in control cells / IC₅₀ in NOP56 KD cells) of 17- and 12-fold, respectively (Fig. 3A; Fig. S3A). These observations were validated by independent assays, in which NOP56 KD sensitized H358 and H460 cells to PI3K/AKT inhibitors (LY294002, AZD5363), anti-mTOR drugs (rapamycin, everolimus), and ER stress inducers (bortezomib and HA15) (Fig. 3B,C; Fig. S3B,C). Importantly, CRISPR/Cas9-mediated knockout of NOP56 dramatically increased the sensitivity of KRAS-mutant (H358, H460) but not of wild-type (H520, H1703) lung cancer cells to rapamycin (Fig. S3D, E).

Moreover, examining gene expression data of KRAS-mutant cancer cells [36] revealed that NOP56 silencing significantly enriched the mTOR gene signature (Fig. 3D). Mining TCGA and Genomics of Drug Sensitivity in Cancer (GDSC) databases showed that NOP56 expression is negatively correlated with that of mTOR pathway genes in patients with KRAS-mutant lung adenocarcinomas (Fig. 3E) and that NOP56 mRNA levels are a predictive marker of sensitivity (IC₅₀) to rapamycin in KRAS-mutant cancer cell lines but not in KRAS-wild-type cancer cells (Fig. 3F). These data support our in vitro results (Fig. 3A-C; Fig. S3A-C) and further suggest a reciprocal interplay between NOP56 and mTOR signaling.

Indeed, siRNA-mediated NOP56 KD, which slightly increased ribonucleolar proteins (e.g., NOP58, FBL), markedly induced AKT/mTOR (p-AKT, p-mTOR, pS6), translation initiation (p-eIF4E) and the stress-responsive p38 MAPK in H358 cells in a time-dependent manner (Fig. 3G), as did shRNA-mediated stable NOP56 KD in H358 and H460 cells, but not in KRAS-WT lung cancer H1703 cells (Fig. 3H; Fig. S3G). Importantly, NOP56 KD sensitized KRAS-mutant lung (A549), colon (HCT-116, DLD-1, and LS174T), pancreatic (MIAPaCa, HPAF-II) and primary KRAS-mutant lung cancer cells (PF563, PF139) to rapamycin, as well as NRAS-mutant lung cancer H1299 cells but not KRAS-wild-type H2405 (BRAF-mutant), EBC-1 (EGFR-mutant), H1993 (MET amplification) and H520 (FGFR1 amplification) cells (Fig. S3H, I). These results reinforce the notion that NOP56 plays a unique role in KRAS-mutant cancer.

Our results demonstrated that NOP56 and the IRE1α-mediated UPR act upstream of p38 and mTOR signaling (Fig. 2C, D; Fig. 3G), suggesting a signaling cascade from the UPR to mTOR via p38 MAPK. To confirm this, we targeted p38 with the specific inhibitor SB203580, which, as expected, barely affected the upstream IRE1α-dependent UPR (p-IRE1α, XBP-1 s), but strikingly dampened the MNK-eIF4E axis and mTOR signaling (p-AKT, p-S6) in H358 cells (Fig. 3I). Importantly, SB203580 exposure not only decreased activity of the mTOR pathway but also increased PARP expression and promoted PARP cleavage (Fig. 3I), concurrent with substantially elevated cytotoxicity on NOP56 KD H460 and H358 cells compared to that on control cells (Fig. 3J). Together, these results unravel a signaling cascade from the IRE1α-mediated UPR to p38 MAPK and to mTOR signaling in KRAS-mutant lung cancer upon NOP56 suppression.

Our findings that NOP56 KD cells are exposed to higher levels of metabolic ROS and display a greater dependency on UPR-activated mTOR signaling suggest a synthetic lethal vulnerability in KRAS-mutant cancer (Figs. 1, 2, 3). To test this hypothesis, we treated NOP56 KD H358 and control cells with rapamycin, which, as expected, decreased mTOR effectors (e.g., p-S6, p-eIF4E) (Fig. 4A) that were the otherwise adaptively upregulated upon NOP56 depletion (Fig. 3G, H). Strikingly, rapamycin induced PARP cleavage (Cl PARP) in NOP56 KD H358 but not in control cells (Fig. 4A), indicating that concomitant targeting of NOP56 and mTOR caused synthetic lethality. Similar results were observed in NOP56 KD H358 cells that were treated with the AKT inhibitor AZD5363 (Fig. S4A, B). Knockdown of Raptor and Rictor, key components of the mTORC1 and mTORC2, respectively, significantly suppressed the viability of NOP56 KD H358 cells despite to differential extent (Fig. S4C, D). Moreover, whereas individual S6 and eIF4E only partly contributed to the viability of NOP56 KD H358 cells, concomitant inhibition of S6 (siRNA) and eIF4E (siRNA and Briciclib, an eIF4E inhibitor) led to significantly enhanced anti-proliferative effect (Fig. 4B-G; Fig. S4E) and largely recapitulated the impact seen by mTOR inhibition with rapamycin (Fig. 4A, Fig. S4F), gauged by the extent to which apoptotic markers (CI-PARP) were induced in NOP56 KD H358 cells (Fig. 4F). These results interrogate an important role for mTOR to relay the IRE1α-mediated UPR signaling in NOP56-depleted KRAS-mutant lung cancer.

The UPR is a double-edged sword, as its outcome flips from adaption to apoptosis when malfunctional UPR, a condition at which stress stimuli are overwhelming or the UPR signal cannot be properly relayed [38, 39]. We thus assumed that the observed synthetic lethality of NOP56 and mTOR inhibition might be enabled due to malfunctional UPR. Indeed, co-targeting NOP56 and mTOR resulted in synergistic effects that not only increased the expression of IRE1α and p-IRE1α, indicative of hyperactive UPR signals, but also upregulated p-JNK, FOXO3A and BIM, a BH3 only protein and key mediator of apoptotic balance (Fig. 4H). Consistent with their pro-apoptotic roles, this increase of the JNK-FOXO3a-BIM axis was accompanied by PARP cleavage (Cl-PARP) and significantly greater apoptotic cell death in NOP56 KD H358 and H460 cells compared to control cells (Fig. 4H, I). Importantly, IRE1α KD (siRNA) precluded the cytotoxicity of combined NOP56 and mTOR inhibition, evidenced by decreased levels of p-JNK, FOXO3a, BIM, Cl-PARP and of apoptotic populations in NOP56 KD H358 and H460 treated with rapamycin (Fig. 4H, I). Similarly, inhibiting JNK activity by the inhibitor SP600125 dampened the efficacy of rapamycin in NOP56 KD H358 and H460 cells (Fig. 4J). Thus, IRE1α-mediated UPR activates mTOR, which provides a survival signal for NOP56 KD KRAS-mutant cancer cells; conversely, mTOR inhibition leads to overwhelmed UPR and promotes apoptotic cell death by activating the JNK-FOXO3A-BIM axis (Fig. 4K).

Next, we asked if the synthetic lethality of co-targeting NOP56 and mTOR is a result of unresolvable metabolic stress. Indeed, rapamycin sensitivity of NOP56 KD H358 and H460 cells was highly correlated with ROS levels (Fig. 5A, B), and ROS scavenge by NAC significantly compromised the cytotoxicity of rapamycin (Fig. 5B), highlighting a causative link between ROS and rapamycin-induced apoptosis in NOP56 KD cells.

These results uncover a novel homeostatic mechanism of metabolic stress mediated by NOP56 and validate an unexpected synthetic lethality by targeting NOP56 and mTOR that aggregate a metabolic liability in KRAS-mutant lung cancer (Fig. 5C).

Finally, we investigated in vivo efficacy of co-targeting NOP56 and mTOR. In a xenograft model from KRAS-mutant H460 cells, NOP56 knockdown (shNOP56) only mildly inhibited tumor growth compared to control shRNA (shScrambled), as did rapamycin (Fig. 6A). However, the outcome of concomitant targeting of NOP56 and mTOR (shNOP56 plus rapamycin) was superior to that achieved by shNOP56 or rapamycin alone, leading to far more effective and potent suppression of xenograft tumor growth (Fig. 6A, B). Residual tumors (after 3-week treatment) from the combination group (shNOP56 plus rapamycin) were typically tiny and significantly differed from those of the other treatment groups (Fig. 6C). Immunohistochemical (IHC) analysis revealed that the anti-tumor efficacy of combined treatment with shNOP56 and rapamycin was paralleled by marked decrease of mTOR activity (p-AKT, p-mTPOR, p-S6) and increase in apoptosis (Caspase-3) in the residual tumors (Fig. S5A).

Similar results were obtained from H358 xenografts (Fig. 6D, E) and a patient-derived xenograft (PDX) model established from the primary KRAS-mutant PF139 cells (Fig. 6F-I). In both models, NOP56 KD sensitized H358 and PF139 xenograft tumors to rapamycin, leading to potent suppression of tumor growth (Fig. 6D, F-H) and significant improvement of mouse survival (Fig. 6E). IHC of the PF139 residual tumors revealed that NOP56 depletion plus rapamycin strikingly suppressed tumor cell proliferation (Ki-67) and dampened mTOR activity (p-AKT, p-mTOR and p-S6), but increased Caspase-3 cleavage (Fig. 6I), which is consistent with the in vitro results (Figs. 4, 5) and our observations on H460 xenografts (Fig. S5A). Notably, rapamycin showed little beneficial effects in NOP56 KD KRAS wild-type H1703 xenografts (Fig. S5B, C), which mirrors the in vitro data and reinforces the selective activity of co-targeting NOP56/mTOR in KRAS-mutant lung cancer.