Regulation and Therapeutic Targeting of MTHFD2 and EZH2 in KRAS-Mutated Human Pulmonary Adenocarcinoma

Activating KRAS mutations occur in about 30% of pulmonary adenocarcinoma (AC) cases and the discovery of specific inhibitors of G12C-mutated KRAS has considerably improved the prognosis for a subgroup of about 14% of non-small cell lung cancer (NSCLC) patients. However, even in patients with a KRAS G12C mutation, the overall response rate only reaches about 40% and mutations other than G12C still cannot be targeted. Despite the fact that one-carbon metabolism (1CM) and epigenetic regulation are known to be dysregulated by aberrant KRAS activity, we still lack evidence that co-treatment with drugs that regulate these factors might ameliorate response rates and patient prognosis. In this study, we show a direct dependency of Methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) and Enhancer of Zeste Homolog 2 (EZH2) expression on mutationally activated KRAS and their prognostic relevance in KRAS-mutated AC. We show that aberrant KRAS activity generates a vulnerability of AC cancer cell lines to both MTHFD2 and EZH2 inhibitors. Importantly, co-inhibition of both factors was synergistically effective and comparable to KRASG12C inhibition alone, paving the way for their use in a therapeutic approach for NSCLC cancer patients.


Introduction
Lung cancer is one of the most aggressive and deadliest cancer types [1] and causes about 1.4 × 10 6 global deaths every year [2]. Pulmonary adenocarcinoma (AC) is the main subgroup of non-small cell lung cancer, accounting for nearly 40% of all cases [2,3]. On the molecular level, AC can be further subdivided, and activating mutations of KRAS define the largest molecular subgroup (~30%) [4,5]. For a long time, KRAS mutations were considered to be undruggable, but, eventually, the KRAS G12C (KRAS G12C ) specific inhibitor AMG510 (Sotorasib) was FDA-approved for second-line treatment of AC patients. However, mutations other than G12C are still not targetable and the overall response rate for Sotorasib among KRAS G12C cases remains below its expected level, with about 60% of patients evading response [6][7][8][9].
KRAS mutations play a vital role in controlling tumor metabolism, e.g., by stimulating glucose uptake [10][11][12], and an increased need for energy and elevated aerobic glycolysis are closely associated with chemoresistance, tumor progression, and metastasis of malignant tumors [13][14][15][16]. We and others have recently shown that KRAS mutations are connected to an enhanced dependency on one-carbon metabolism (1CM) in non-small cell lung cancer (NSCLC) and colorectal cancer [17][18][19]. One-carbon metabolism includes the methionine and folate cycles and is essential for the maintenance of cellular homeostasis. Integrating the cell's nutritional status 1CM catabolizes different carbon sources to derive one-carbon (methyl) units. In cancer cells, the high proliferation rate requires these one-carbon units for nucleotide synthesis, methylation, and reductive metabolism [20]. Furthermore, we have shown that Methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) is essential for the survival of NSCLC cell lines and is a prognostic factor in AC [18]. MTHFD2 is one of the key enzymes in 1CM and is strongly expressed in embryonic development but it is almost absent in most healthy adult tissues, making it a promising potential therapeutic target for cancer treatment [21]. High levels of MTHFD2 have been associated with tumor recurrence and worse prognosis in multiple solid and liquid malignancies and participate in resistance against gemcitabine and pemetrexed [18,19,[21][22][23][24][25][26][27][28].
Enhancer of Zeste Homolog 2 (EZH2) is a member of the polycomb repressive complex 2 (PRC2), which plays an important role in maintaining cellular identity by regulating the transcription of genes through deposition of the H3K27me3 repressive mark. EZH2 is upregulated in multiple malignancies, including NSCLC [29][30][31][32][33][34][35]; therefore, several attempts have been made to inhibit EZH2 as a clinical treatment approach. The specific EZH2 inhibitors GSK126 and EPZ-6438 have yet not reached clinical stages [36,37]; however, tazemetostat was granted accelerated approval at the beginning of 2020 by USFDA for Follicular Lymphoma and has since been tested in multiple solid tumors [38][39][40]. Recent findings connected EZH2 expression to KRAS mutations and metabolism. Compared to a moderate expression in KRAS wild-type (KRAS WT ) cell lines, expression of EZH2 is increased in cells with an activating KRAS mutation [35]. EZH2 has also been described to facilitate metabolic reprogramming in glioblastomas with a substantial increase in glycolytic metabolism [41].
A better understanding of the metabolic and epigenetic network regulated by aberrant KRAS in AC is therefore of primary importance. In the present study, we investigated the expression of MTHFD2 and EZH2 as dependent on KRAS mutational status in a cohort of primary AC patient samples. Our results highlight a functional connection between mutated KRAS, EZH2, and MTHFD2 and reveal a vulnerability of KRAS-mutated AC cancer cell lines to both MTHFD2 and EZH2 inhibitors. Importantly, our study shows that co-inhibition of both factors was synergistically effective in addition to KRAS G12C inhibition alone, providing evidence that their use in a therapeutic approach for NSCLC cancer patients may increase overall response rate outcomes in KRAS G12C cases.

EZH2 and MTHFD2 Expression Correlate with KRAS Mutation Status and Clinicopathologic Characteristics in AC Patients
KRAS mutations have been connected to metabolism and epigenetic regulation in several cancer types [35,42,43]. Therefore, we investigated a cohort of 109 AC patients for activating KRAS mutations and protein expression of MTHFD2 and EZH2. The clinical characteristics of the patients are summarized in (Table 1 and Supplementary Table S1). All patients underwent surgical tumor resection without prior chemotherapy. Male patients (56%) were slightly more numerous than females and the median age at the time of diagnosis was 67 years (range 34-85 years). Most patients demonstrated a moderately differentiated disease, the frequency of T1-2 stage was 82.6%, and the majority were lymph node-negative (60.2%). The median follow-up time was 23 months and 48% of the patients deceased during the follow-up time. Among 62 patients with a clinical follow-up, we examined the mutational status of exon 2 of the KRAS gene. Eighteen samples were KRAS-mutated (G12C n = 15, G12V n = 3) and 44 were found to be KRAS WT . The MTHFD2 statuses of the investigated samples were published previously [18] and can be found in Supplementary Table S1. Comparing MTHFD2 expression in KRAS MUT and KRAS WT samples, we observed a significant correlation between high expression of MTHFD2 and KRAS MUT tumors (p = 0.0066) (Figure 1a). Furthermore, overall survival (OS) was significantly influenced by MTHFD2 expression in KRAS MUT (p = 0.0178, Figure 1b) but not KRAS WT cases (p = 0.2906, Figure 1c).
EZH2 was strongly expressed in 38.5% of investigated cases and high expression indicated a significantly shorter OS compared to low EZH2 levels (p = 0.0027) (Figure 1d-f). As for MTHFD2, EZH2 high expression was more common in KRAS MUT cases (p = 0.0039, Figure 1g), and patient prognosis was significantly decreased when EZH2 was highly expressed in KRAS MUT patients (p = 0.0419, Figure 1h) but not in KRAS WT cases (p = 0.1126, Figure 1i).
Next, we have shown that strong expression of MTHFD2 was significantly associated both with KRAS mutations and high levels of EZH2 (p = 0.0005, Figure 1j). Patients with strong expression of both EZH2 and MTHFD2 harboring a KRAS mutation had the worst prognosis (p = 0.0128, Figure 1k), whereas no influence on OS was detected based on protein expression of EZH2 and MTHFD2 in KRAS WT and (p = 0.1152, Figure 1l).
In addition, EZH2 expression and KRAS mutation significantly correlated with the occurrence of lymph node metastasis (p = 0.0003 and p = 0.0052, respectively) and poor differentiation grade (p < 0.0001 and p = 0.035) as shown in Table 2.

Expression of MTHFD2 and EZH2 Depends on the Activity of Mutated KRAS in Human Pulmonary Adenocarcinoma Cell Lines
To further investigate the functional interplay between MTHFD2, EZH2, and mutated KRAS, we used two KRAS G12C -mutated (HCC44 and H23) and two KRAS WT (H1993 and HCC78) AC cell culture models. Western blot analyses revealed up to 70% increased levels of MTHFD2 and EZH2 in HCC44 and H23 in comparison to H1993 and HCC78 (Figure 2a and Supplementary Figure S1a

EZH2 Repressive Activity Is Required to Modulate MTHFD2 Expression in KRAS G12C Cell Lines
To characterize the relationship between EZH2, MTHFD2, and KRAS in AC cells, we knocked down EZH2 using two specific siRNAs. Interestingly, next to EZH2, MTHFD2, also, was robustly downregulated only in the KRAS G12C (Figure 3i,j). Altogether, we show that MTHFD2 specifically targets the KRAS G12C background and that, in turn, MTHFD2 expression is subjected to KRAS G12C and EZH2 activity.

Combinational Treatment of KRAS G12C with EZH2 and MTHFD2 Inhibitors
Based on our results that show a strong dependency of MTHFD2 and EHZ2 on activating KRAS mutations, we decided to test whether a combinational treatment may have a synergistic effect on AC viability. We therefore treated HCC44 with increasing concentrations of AMG510 in combination with (I) GSK126 or (II) DS or (III) GSK126 together with DS. Compared to single treatments, all combinations showed an increased and synergistic reduction in cellular survival (Figure 4a-f). In particular, the combination of GSK126 and DS showed a good response that was comparable to KRAS G12C inhibition with relatively low concentrations.  We have previously reported that high MTHFD2 expression induced resistance against Pemetrexed (PTX) [18] and that downregulation of KRAS G12C enhances response to PTX [17]. To test whether PTX resistance in HCC44 can be recovered by KRAS G12C inhibition, we incubated HCC44 and H1993 cells with either AMG510 (2 µM), PTX (20 µM), or in combination. After a single treatment, EZH2 and MTHFD2 protein levels were only partially reduced. However, the combination almost completely reduced EZH2 and MTHFD2 specif-ically in HCC44, meanwhile the expression in H1993 was almost unaffected (Figure 4g). Furthermore, the combination of PTX treatment with GSK126 (2.5 µM) and DS (10 µM) was about 10-fold, and the combination of PTX with AMG510 (2 µM) was 80-fold stronger than the single treatment in HCC44 (Figure 4h). On the contrary, in H1993 cells, no additional effect was observed (Figure 4i).

Discussion
Metastasis and recurrence after tumor resection and adjuvant therapy occur in almost 70% of pulmonary adenocarcinoma cases and keep the five-year survival rate at approximately 15% [44]. The discovery of the small-molecule inhibitor AMG510 directed against G12C-mutated KRAS significantly improved the prognosis of many affected AC patients; however, besides the mutations other than G12C that cannot be targeted, less than 40% of treated patients responded to therapy [6][7][8]. In this study, we show by immunohistochemical and in vitro approaches that KRAS G12C upregulates the epigenetic regulator EZH2 and the 1CM metabolic enzyme MTHFD2, leading to molecular vulnerabilities for combinational treatment approaches.
In our study, we found that expression of both MTHFD2 and EZH2 are increased when KRAS acquires activating mutations in patients with AC. Importantly, only in KRASmutated cases, high expression of both MTHFD2 and EZH2 was predictive of inferior patient prognosis. Aberrant KRAS activity has been described previously as rewiring metabolism and epigenetic regulation [35,42,43]. Riquelme et al. showed that in KRAS G12Cmutant NSCLC, EZH2 expression was preferentially upregulated through the MEK-ERK signaling pathway [35], and a negative prognostic association of EZH2 overexpression has been proposed in several human malignancies, including NSCLC [29][30][31][32][33][34][35]. Indeed, our data not only show a correlation between KRAS G12C and EZH2 expression, but the functional ablation of EZH2 methylase activity also clearly supports its role in AC cancer cell viability. Epigenetic repression of EAF2-HIF1α by EZH2 has been shown to foster metabolic reprogramming in glioblastoma and to promote aerobic glycolysis by upregulating HK2 in prostate cancer [41,45]. In addition, high 1CM activity has been linked to increased tumor aggressiveness and reduced prognosis in several cancer entities, and a dependency of MTHFD2 on KRAS and its prognostic impact was described in AC, colorectal, and pancreatic cancer [18,19,46].
As expected, pharmacological inhibition of KRAS decreased cellular viability and reduced protein levels of MTHFD2 and EZH2 only in KRAS G12C cells. By over-expressing KRAS G12C in the KRASWT cell line H1993 we could validate a direct influence of KRAS activity on metabolism and the regulation of MTHFD2 and EZH2 expression. Specific knockdown or pharmacological inhibition revealed a KRAS G12C -dependent MTHFD2 regulation by EZH2, whereas siRNAs against MTHFD2 had no effect on EZH2 expression. This suggests that KRAS G12C activity is responsible for the upregulation of MTHFD2 subjected to the control of EZH2 methyltransferase activity. The fact that our results show MTHFD2 down-regulation in response to EZH2 inhibition indicates that such regulation most likely occurs indirectly. One obvious mechanistic explanation is that EZH2 may target an MTHFD2 transcriptional repressor. Another possibility to consider is that 3D genome organization analyses indicate that Polycomb-bound loci form insulated and selfinteracting chromatin domains [47] and that the removal of EZH2 activity may induce rewiring of the MTHFD2 gene locus, leading to the preclusion of interaction with its regulatory sequences. Polycomb marks are highly enriched at CpG islands (CGIs), and H3K27me3 distribution is known to be anti-correlated with DNA methylation [48]; in this regard, EZH2-silenced foci near the MTHFD2 gene may acquire DNA methylation (epigenetic switching) [49], triggering the constitutive silencing of the entire locus in response to PRC repression inhibition.
However, the fact that MTHFD2 is transcriptionally regulated by epigenetic fluctuations does not represent the only hypothesis. MTHFD2 is relatively lowly expressed in normal tissues [47] and is upregulated in different cancers, including breast, colorectal, and hepatocellular cancers, where it plays a key role in remodeling the folate metabolism of tumor cells [48]. There have been indications that KRAS regulates MTHFD2 at a transcriptional level [49], but post-translational regulation also seems possible. It has been proposed that MTHFD2 activity and proteasomal degradation are regulated by acetylation [50,51]. Zhang et al. [51] described that the acetylation of MTHFD2 by SIRT4 leads to increased proteasomal degradation in breast cancer. They describe SIRT4 as a guardian of cell metabolism and a sensor of folate availability. Since KRAS and EZH2 are known to increase the metabolic state and the availability of folate, it could be suggested that MTHFD2 is regulated by proteasomal degradation dependent on the metabolic state of the cell. In our study, we aimed to target MTHFD and EZH2 as a therapeutic approach for NSCLC cancer patients and further studies will address the mechanism behind their action. Strikingly, the inhibition of EZH2 or MTHFD2 strongly reduces cellular viability only in KRAS G12C cell lines. The increased response to inhibition of EZH2 and MTHFD2 of the KRAS WT cell line H1993 overexpressing KRAS G12C further suggested that KRAS activity is a main driver of the epigenetic regulation and increased 1CM activity in AC. These findings indicate that the inhibition of EZH2 and MTHFD2 will be mainly effective in KRAS-mutated AC. MTHFD2 and EZH2 inhibition have been proposed as therapeutic options in several solid and lymphatic cancers [19,[36][37][38][39][40]50,51]. Importantly, here we show that the combined inhibition of mutated KRAS and either EZH2 or MTHFD2 exhibit a synergistic effect in KRAS G12C cells, and, surprisingly, the co-inhibition of EZH2 and MTHFD2 had a similar synergistic effect to the combination with AMG510. These results strongly suggest that AC patients with a KRAS mutation other than KRAS G12C might also profit from combined inhibition of MTHFD2 and EZH2.
We previously demonstrated that strong expression of MTHFD2, as seen in KRASmutated cells, leads to resistance of AC cells against treatment with PTX, which is commonly used as first-line therapy for AC patients [18]. While the combination of PTX with AMG510, GSK126, or DS had no additional effect in the KRAS WT cell line H1993, we discovered a synergistic effect when we combined PTX with GSK126 or DS in the KRAS G12C cell line HCC44. Since the combination of PTX with AMG510 led to a similar response to PTX in the KRAS WT cell line, we also suggest PTX as a treatment option in AMG510-treated KRAS G12C AC with a moderate response.
In summary, our findings indicate a causal connection between KRAS mutation status and the expression of the epigenetic regulator EZH2 and the 1CM marker MTHFD2. KRASmutated AC cells are vulnerable to inhibition of EZH2 and MTHFD2, and combinational treatment is as efficient as KRAS G12C inhibition alone, revealing a potential treatment option for non-KRAS G12C -mutated AC. In addition, we suggest PTX as a treatment option in AMG510-treated KRAS G12C AC with moderate response. These findings may lead to a better separation of AC patients and an improved response rate by combinational treatment strategies.

Human Tissue Samples
The NSCLC tissue samples used in this study were collected from the Department of Thoracic and Cardiovascular Surgery, University Medical Center Göttingen after surgical resections. Tissues were formalin-fixed in 4% buffered formaldehyde and paraffin-embedded for diagnostic purposes. The performed experiments were approved by the ethics committee of the University Medical Center Göttingen (#1-2-08, 24-4-20) and all procedures were performed in accordance with the seventh version of the Declaration of Helsinki [52].

Cell Culture
The human AC cell lines HCC44, H23, H1993, and HCC78 were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were maintained in RPMI-1640 supplemented with 1% L-Glutamine, 1% Penicillin-Streptomycin, and 10% fetal bovine serum (Gibco / ThermoFisher, Waltham, MA, USA) at 37°C in a 5% CO 2 humidified environment. The medium was refreshed three times per week and cells were passaged at approximate confluency of 80%.

MTS and ATP Assay
Cell survival analysis was performed using the CellTiter-Glo assay and the MTS assay kit (Promega, Madison, WI, USA), according to the manufacturer's protocol. The CellTiter-Glo assay was used to perform the ATP luminescence assay. Chemiluminescence was measured using a Tecan Reader Infinite 2000 (Tecan, Männedorf, Switzerland).

Cell Transfection with siRNA and Expression Plasmids
All siRNAs in this study were obtained from Qiagen (Qiagen, Hilden, Germany). For the cell transfection, 3 × 10 5 cells were transfected with 30 nM siRNA against MTHFD2 (#1 SI02664928, #2 SI02664921) or 80 nM siRNA against EZH2 (#1 SI02633316, #2 SI02665166) using HiPerFect Transfection Reagent (Qiagen), according to the manufacturer's protocol. Allstars negative siRNA was used as a scrambled control. Plasmid transfection was performed with XtremeGENE HP DNA transfection reagent (Merck, Darmstadt, Germany). In brief, a 100 µL transfection mix containing 4 µL transfection reagent, 2 µg expression vector DNA, and serum-free RPMI-1640 cell culture medium was incubated at room temperature for 15 min and added to 3 × 10 5 cells seeded on a 6-well plate in 2 mL medium. Either a pCMV6-Entry-KRAS G12C vector (Origene Technologies Inc., Rockville, MD, USA) or a pBabe-KRAS WT vector (Addgene, Watertown, MA, USA) was transfected into H1993. Cells expressing de novo KRAS G12C and KRAS WT were selected with G418 and Puromycin at concentrations of 800 µg/mL and 2 µg/mL, respectively, for at least 10 days, and KRAS protein levels were confirmed by Western blotting.

Drug Treatment Assays
Fifteen-hundred cells were seeded on a 96-well plate 24 h before drug treatment. The specific drugs used were AMG510 (MedChemExpress, USA, #HY-114277), GSK126 (Selleckchem, Houston, TX, USA, #S7061), and DS (DS18561882) (MedChemExpress, Monmouth Junction, NJ, USA, HY-130251). Cells were treated with indicated concentrations for indicated time periods. For synergistic effect assays, two inhibitors were combined in a series of nine increasing concentrations. Inhibitor concentrations ranged from 0 to 25 µM for AMG510, 0 to 20 µM for GSK126, and 0 to 100 µM for DS. In corresponding control groups, an equivalent amount of DMSO was added. After 72 h treatment, cell viability was analyzed using MTS, as described above. Drug synergy was evaluated using CompuSyn software (ComboSyn Inc, New York, NY, USA) for the combination index (CI) value calculations. CI values: antagonism (CI > 1.1), additive effect (CI < 1.1 and CI > 0.9), and synergism (CI < 0.9) [57].

Statistical Analysis
Statistical analysis was performed using GraphPad (GraphPad Software LCC, San Diego, CA, USA). Two-group comparisons were performed with Students' t-tests. Halfmaximal inhibitory concentration (IC50) analysis was performed using Pearson's correlation test. Cell growth and resistance comparisons were analyzed using two-way ANOVA. Survival analyses were performed using the Kaplan-Meier method and the log-rank (Cox-Mantel) test. ImageJ was used for the quantification of WB signal intensities [58]. Three biological replicates were performed. The data depicted are the means ± SEMs. A p-value of <0.05 was considered significant (* p < 0.05, ** p < 0.01, *** p < 0.001).
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/metabo12070652/s1, Figure S1, Quantification of EZH2 and MTHFD2 expression in AC cell lines and H1993 cells transfected with KrasG12Cvec or KrasWTvec; Figure S2, Western blot analysis of the four AC cell lines HCC44, H23, H1993, and HCC78 after EZH2 and MTHFD2 knockdown with siRNA; Table S1, Complete patient characteristics.

Institutional Review Board Statement:
The study was conducted according to the Declaration of Helsinki and approved by the Institutional Ethics Committee of the University Medical Center Goettingen (#1-2-08, 24-4-20).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available in article and Supplementary Materials.