p53-Independent Effects of Set7/9 Lysine Methyltransferase on Metabolism of Non-Small Cell Lung Cancer Cells

Set7/9 is a lysine-specific methyltransferase, which regulates the functioning of both the histone and non-histone substrates, thereby significantly affecting the global gene expression landscape. Using microarray expression profiling, we have identified several key master regulators of metabolic networks, including c-Myc, that were affected by Set7/9 status. Consistent with this observation, c-Myc transcriptional targets—genes encoding the glycolytic enzymes hexokinase (HK2), aldolase (ALDOB), and lactate dehydrogenase (LDHA)—were upregulated upon Set7/9 knockdown (Set7/9KD). Importantly, we showed the short hairpin RNA (shRNA)-mediated attenuation of Set7/9 augmented c-Myc, GLUT1, HK2, ALDOA, and LDHA expression in non-small cell lung cancer (NSCLC) cell lines, not only at the transcriptional but also at the protein level. In line with this observation, Set7/9KD significantly augmented the membrane mitochondrial potential (MMP), glycolysis, respiration, and the proliferation rate of NSCLC cells. Importantly, all these effects of Set7/9 on cell metabolism were p53-independent. Bioinformatic analysis has shown a synergistic impact of Set7/9 together with either GLUT1, HIF1A, HK2, or LDHA on the survival of lung cancer patients. Based on these evidence, we hypothesize that Set7/9 can be an important regulator of energy metabolism in NSCLC.


INTRODUCTION
Lysine methylation plays an important role in global transcription regulation. Depending on the location of the target lysine in histone tails, this post-translational modification can either promote or repress transcription by affecting the architecture of chromatin. In addition, lysine methylation, by competing with other lysine-specific modifications (e.g., acetylation, ubiquitinylation, and SUMOylation), can affect the protein stability and hence its functioning (1). Set7/9 (alternative name SETD7) is a SET [Su(var)-3-9, Enhancer-of-Zeste, Trithorax] domain-containing protein that utilizes both histone and non-histone proteins as substrates. Initially, Set7/9 was shown to specifically monomethylate lysine 4 of histone 3 (H3K4me1), which is a positive mark for transcriptional activation (2,3).
Set7/9 is shown to be involved in various signaling pathways associated with several diseases, including cancer (8). We and other researchers have demonstrated the involvement of this lysine methyltransferase in the progression of various malignancies, including lung cancer (16)(17)(18)(19). Lung cancer is one of the most frequent and dangerous groups of malignancies. It tops the list of cancer-related deaths in men and ranks second in women worldwide (20). The 5-year survival rate with all types of lung cancer ranges between only 17% and 21%. Importantly, the role of Set7/9 in lung cancer is rather contradictive, since it was shown to have both oncogenic (16,18) and tumor-suppressive properties (21).
Using RNA-seq analysis, Keating with co-authors (22) have revealed a correlation between the gene expression profile in Set7/9 knocked down cells and the status of several transcription factors known to be the targets of Set7/9, including p53, NFkB, c-Jun, c-Fos, GATA2, ERa, and STAT3. This study linked the cellular status of Set7/9 with global changes in gene expression profiles of various cell lines.
To further elucidate the role of Set7/9 in tumorigenesis, we assessed the effect of Set7/9 ablation on global gene expression in various cancer cell lines. We have identified c-Myc and HIF1A as Set7/9-dependent genes. Moreover, we have shown that NSCLC cells with an attenuated expression of Set7/9 displayed increased mitochondrial membrane potential (MMP), glycolysis, and respiration rates. In line with these observations, Set7/9deficient cells possessed elevated proliferation rates. Bioinformatic analysis revealed the synergistic effect between the Set7/9 and either HIF1A, HK2, GLUT1, or LDHA expression levels on the survival outcome of lung cancer patients.

Cell Cultures and Stable Cell Lines Establishment
All cell lines were purchased from American Type Culture Collection (ATCC) (USA) and genotyped by the shared research facility "Vertebrate cell culture collection," Institute of Cytology, Russian Academy of Science, St Petersburg, Russia. Cells were incubated in either Dulbecco's modified Eagle's medium (DMEM) (U2OS osteosarcoma cells) or Roswell Park Memorial Institute (RPMI) (H1299, A549, and H1975) media, supplemented with 10% fetal bovine serum (FBS) (Gibco, USA), and 50 µg/ml gentamicin.

Reactive Oxygen Species Detection Assay
The total reactive oxygen species (ROS) production was analyzed using the 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA) substrate (Thermo Fisher Scientific, USA) in a final concentration of 50 µM. For detection, superoxide anions (O − 2 ) dihydroethidium (DHE) (Thermo Fisher Scientific, USA) in a final concentration of 5 µM was used. The next day after seeding, cells were treated with H2DCFDA or dihydroethidium (DHE) for 30 min at 37°С in a CO 2 incubator. Following the incubation, cells were washed in PBS, detached with trypsin, and analyzed by flow cytometry (FC) (CytoFlex, Beckman Coulter, USA). Values of the median were used for calculation. Results were represented as the mean ± SD of three experiments.

Proliferation Assay
The proliferation analysis was performed using the xCELLigence technology (ACEA Biosciences, USA) according to the manufacturer's instructions. A total of 20,000 cells were planted into each well of an ACEA E-plate 16 in triplicate. The cell index was registered every 10 min for 10-40 h. Results were represented as the mean ± SEM of three experiments.

Cell Cycle Analysis
Flow cytometry analysis of the cell cycle was carried out. A total of 50,000 H1299 and H1975 cells, with Set7/9 KD or scramble, and A549 control and Set7/9 KO, were planted in triplicates. Two days after seeding, cells were harvested, washed once with PBS, and fixed in 70% ethanol at −20°C for 1 h. Then, 30 min staining of DNA content was carried out by using 50 mg/ml of PI (Invitrogen, USA) and 1 mg/ml RNase A (Thermo Fisher Scientific, USA). Samples were analyzed by a CytoFLEX (Beckman Coulter, USA) flow cytometer. Results were processed by CytoExpert software (Beckman Coulter, USA).

Analysis of Glycolysis and Respiration
The SeaHorse energy profiling of lung cancer cell lines with either Set7/9 KD or scramble was carried out as described in (24) with small modifications. Briefly, 17,000 cells were seeded in SeaHorse 24-well plates a day prior to analysis. The SeaHorse Energy Phenotype test was used in accordance with the manufacturer's recommendations. Stressor mix consisted of carbonyl cyanide p-(trifluoromethoxy)-phenyl-hydrazone (FCCP) and oligomycin (Agilent Technologies, USA) to achieve the final concentrations of 2 and 4 µM, respectively. Results are represented as the mean ± SEM. For further analysis of the impact of Set7/9 on glycolysis in detail, GlycoStress kit (Agilent Technologies, USA) was used according to the manufacturer's instructions. The following concentrations were used in the glycolysis stress experiments: 10 mM of glucose, 4 µM of oligomycin, and 50 mM of 2-DG. Results are represented as the mean ± SEM.

Microarray
The microarray gene expression study was carried out using Human Gene Expression 4x44K Microarrays and a Low Input Quick Amp Labeling Kit (Agilent Technologies, USA) according to the manufacturer's instructions. RNA quality was assessed using a 2100 Bioanalyzer (Agilent Technologies, USA). One hundred nanograms of each RNA sample were used for cDNA synthesis and were simultaneously labeled with Cy-3. After purification, cDNA samples were hybridized with oligonucleotide probes on microarray slides for 18 h. The next day, the slides were washed and scanned. Data were analyzed by GeneSpring GX11.5 software.

Bioinformatic Analysis of Lung Cancer Patients' Survival Rates
The single and synergistic effect of SETD7 (Set7/9), SLC2A1 (GLUT1), HIF1A, HK2, and LDHA expression levels on the survival rate of lung cancer patients was determined by using Syntarget software as described in (25).

Statistical Analysis
Data are represented as mean ± standard deviation (SD) or standard error of the mean (SEM) of at least three replicates. Statistical significance was analyzed using Student's t-test. p < 0.05 was considered significant. p < 0.05 was denoted as * and p < 0.01 as **.

RESULTS
Set7/9 Knockdown Upregulates c-Myc, HIF1A, and Genes of Glycolytic Enzymes in U2OS Human Osteosarcoma Cells Set7/9 lysine methyltransferase is a well-known regulator of transcription (8,10). Previously, we have shown that Tet-inducible Set7/9 knockdown in U2OS human osteosarcoma cells augmented the sensitivity of cells to genotoxic stress by upregulating the transcription of MDM2 (26).
To expand our observations on the role of Set7/9 in transcriptional regulation and to obtain the knowledge on global gene expression affected by Set7/9, we have carried out a microarray gene expression analysis of the same U2OS cells with Tet-inducible Set7/9 knockdown. We have treated U2OS with tetracycline for 2 days, followed by mRNA extraction, cDNA synthesis, labeling, and hybridization with oligonucleotide probes on microarray slides of the Illumina Human Gene Expression 4x44K Microarray Kit. The list of genes affected by Set7/9 KD is given in the Supplementary Table 1. The gene ontology analysis of differentially regulated genes in Set7/9 KD versus Set7/9 wt (U2OS pSuperior) cells revealed several functional pathways, among which the metabolic one was significantly represented. We focused on this pathway, since cancer metabolism is considered to be one of the hallmarks of cancer (27).
We have verified the microarray data by independent analysis using gene-specific real-time PCR. Indeed, Set7/9 knockdown augmented the mRNA levels of both c-Myc and HIF1A and their transcriptional targets-GLUT1, HK2, ALDOA, and LDHA ( Figure 1C).   Figure S1). The results shown in Figure 2A demonstrate that, in general, the suppression of Set7/9 in NSCLC cells augmented the expression of energy metabolism factors similar to that observed in USOS osteosarcoma cells. Western blots were also consistent with RT-PCR results, i.e., attenuated levels of Set7/9 resulted in the increased production of the respective proteins, albeit to a different extent. Notably, the effect of Set7/9 was more pronounced in p53-negative cells (H1299) compared to A549 (p53-positive cells) or H1975 cell (p53 mut) (Figure 2A and Supplementary Figure S1). Furthermore, we found that the protein level of HIF1A was not significantly affected by Set7/9 in H1299 and A549 cells but was altered in H1975 cells with mutant p53 (Figure 2 and Supplementary Figure S1, respectively).

Set7/9 Knockdown Upregulates
Mitochondrial Membrane Potential, Glycolysis, and Respiration c-Myc and HIF1A are the two well-known master regulators of metabolic networks including energy metabolism, deregulations of which are recognized now as one of the "hallmarks of cancer." Enhanced glycolysis in malignant cells is associated with tumor progression, migration, invasion, and resistance against chemoand radiotherapy. Since our results suggested that ablation of Set7/9 caused an increase in c-Myc, GLUT1, and several glycolytic enzymes both at the mRNA and protein levels, we decided to examine whether Set7/9 affected the MMP, glycolysis, and respiration. First, we stained H1299 and A549 cell lines with Set7/9 KD or Set7/9 KO, respectively, with the TMRE agent to measure the levels of MMP intensity. The respective parental control cell lines with wild-type Set7/9 were also used in the experiment. Cell lines with Set7/9 KD or KO increased the MMP level in the range of 10-60% depending on NSCLC cell line ( Figures 3A, B and Supplementary Figure S2A). Importantly, the most pronounced effect was observed in p53-negative H1299 cells.
To confirm that the effect of Set7/9 ablation on MMP was not dependent on the p53 status, we used H1299 cell lines with Tetinducible expression of wild-type p53 (p53wt) or mutant p53 R273H (p53mut) proteins, and control cells with knocked down Set7/9. We showed that in all three H1299-derived cell lines (control, p53wt, and p53mut), suppression of Set7/9 led to an increase in MMP levels (Supplementary Figure S3).  In general, malignant cells have 1.5-2 times higher MMP levels than their non-malignant counterparts, which is the consequence of an elevated energy metabolism. The level of MMP enhancement is associated with the degree of malignancy of neoplastic cells, including enhanced resistance to aggressive conditions and elevation of their metastatic potential (28).
Next, we used the SeaHorse energy profiling technology to study the impact of Set7/9 KD on glycolysis and respiration. Using the SeaHorse Energy Phenotype kit, we showed that Set7/9 KO upregulated only ECAR (glycolysis) in p53-positive A549 cells, whereas p53-negative H1299 cells and p53-mutant H1975 cells with Set7/9 knockdown displayed elevated levels of both glycolysis and respiration compared to control cells ( Figures 4A,  B and Supplementary Figure S2C).
Using the SeaHorse GlycoStress kit, we have further analyzed the impact of Set7/9 on several parameters of glycolysis ( Supplementary Figures S4-S6). The attenuation of Set7/9 expression upregulated both glycolysis and respiration in H1299 and H1975 cell lines (Supplementary Figures S4, S5). On the contrary, only glycolysis was augmented in A549 cells upon Set7/9 attenuation, while respiration did not change significantly (Supplementary Figure S6). These results correlate with the data obtained using the SeaHorse Energy Phenotype kit ( Figure 4B).
A previous study by Shen et al. showed that Set7/9 acts as a negative regulator of b-catenin in response to ROS (15). Importantly, b-catenin affects c-Myc expression, which in turn regulates the expression of glycolytic genes. Thus, it was plausible that Set7/9 mediated its effects on these genes through the bcatenin/c-Myc axis. To test this hypothesis, we assessed the effect of Set7/9 suppression on b-catenin levels in the NSCLC cell lines and observed no correlation between b-catenin levels and the status of Set7/9 (Supplementary Figure S7A).
Furthermore, we have tested the intracellular concentration of ROS in these cell lines. Using DHE staining, we have shown that the level of superoxide anions (O2−) was higher in all three Set7/9-deficient cell lines compared to control lines (Supplementary Figure S7B). However, the total amount of ROS species was either the same or lower in Set7/9 KD or Set7/ 9KO cells (Supplementary Figure S7C) compared to control cells. Since the intracellular superoxide anions content correlates with the activity of mitochondria (29,30), this result further confirms the effect of Set7/9 suppression on mitochondrial activity.

Set7/9 KD Enhances the Proliferation of NSCLC Cell Lines
The increased glycolysis is usually associated with the enhanced proliferation (31) and aggressiveness of tumors. Thus, we decided to study the effect of Set7/9 attenuation on the proliferation of the NSCLC cell lines. To address this question, we used the xCelligence platform. As shown in Figures 5A, B and Supplementary Figure S8A, all three NSCLC cell lines with Set7/9 KD/KO displayed augmented proliferation rates from 1.2to 2-fold, compared to their respective controls. These results are in agreement with our data on the stimulatory effect of Set7/9 KD on glycolysis and respiration. These results prompted us to assess the effect of Set7/9 on the cell cycle. To this end, we analyzed the cell cycle distribution of H1299, A549, and H1975 with different Set7/9 status. We observed a significant increase in S-phase in Set7/9 KD/KO cells in all three cell lines investigated in comparison to control cells ( Figures 5C, D and Supplementary Figure S8B). These results, taken together with the results on proliferation obtained by xCelligence, strongly suggest that either the attenuation or ablation of Set7/9 upregulates proliferation rates of H1299, A549, and H1975 cells.
Expression Levels of Set7/9, HIF1A, GLUT1, HK2, and LDHA Correlate With Survival Rates of Lung Cancer Patients Increased rates of glycolysis in tumor cells are inversely associated with patient survival (32). We sought to investigate whether there are correlations between the expression levels of glycolytic genes HIF1A, GLUT1, HK2, and LDHA, in lung cancer patients and the rates of the patients' survival. We carried out a bioinformatic analysis of several cancer samples databases using an algorithm previously described in SynTarget software (25).
Kaplan-Meier plots ( Figure 6A) demonstrate that the high levels of HK2, SLC2A1 (GLUT1), LDHA, and HIF1a expression were associated with poor survival of lung cancer patients. Moreover, low expression of Set7/9 together with high expression of each of aforementioned genes is also associated with poor outcome. On the contrary, high levels of Set7/9 expression in conjunction with low expression of HK2, SLC2A1 (GLUT1), LDHA, and HIF1A are associated with increased survival of patients ( Figure 6B).
These observations reinforce our hypothesis that Set7/9 may have an impact on survival of cancer patients via regulating the energy metabolism of NSCLC.
In the present manuscript, we have provided evidence that the knockdown of Set7/9 methyltransferase in human NSCLC cell models upregulates a number of glycolytic enzymes (hexokinase, aldolase, and lactate dehydrogenase) and their key transcriptional activators-oncogenes c-Myc and HIF1A-at both transcriptional and protein levels. Enhanced glycolysis has been a well-known feature of malignant cells since Otto Warburg's pioneering works were published in 1925 (35). The so-called "Warburg effect" is part of metabolic reprogramming, which is considered now as one of the "hallmarks of cancer" (36). There are at least three main reasons why malignant cells benefit from glycolysis (37)(38)(39). First of all, glycolysis fuels biosynthetic anabolic pathways of rapidly proliferating cells by diverting glucose flux towards pentosephosphate pathways and one-carbon metabolism. The latter are highly required for the synthesis of DNA, lipids, S-adenosyl methionine (the main donor of methyl groups), glutathione (an important factor of red-ox homeostasis), etc. Second, glycolysis helps neoplastic cells to fine-tune their interaction with the microenvironment by manipulating cancer-associated fibroblasts (CAFs). Finally, glycolysis provides extracellular acidification that protects malignant cells from the attack of immune cells, which are not effective at low pH.
c-Myc and HIF1A are master regulators of metabolic networks. They upregulate and coordinate glycolysis, respiration, one-carbon metabolism, and the metabolism of glutamine and lipids upon tumorigenesis (28,40,41). It is important to note that HIF1A and c-Myc are both known to regulate the expression of tumor-specific isoforms of glycolytic proteins such as GLUT1, HK2, PKM, and LDHA, thereby diminishing the Warburg effect in cancer cells (42)(43)(44). Moreover, both c-Myc and HIF1A play known roles in the upregulation of metabolic networks and the proliferation of NSCLC (23,39,45,46).
Previously, two research groups have shown that Set7/9mediated methylation of HIF1A and HIF2A leads to their destabilization and negatively regulates their transcriptional activity (47). In contrast, our present results suggest that Set7/9 affects HIF1A expression only at the level of mRNA in H1299 and A549 cells and only modestly on the protein level in H1975. Given that these three cell lines have different p53 statuses, it is unlikely that the effect of Set7/9, or lack thereof, was a p53dependent effect. It is possible that the upregulation of HIF1A observed on the level of mRNA in the absence of Set7/9 could also be detected on the protein level should those cells be treated with hypoxia. Future experiments under hypoxic conditions should clarify this possibility. Nevertheless, in the present study, we have directly shown that the ablation of Set7/9 in several NSCLC cell lines upregulated glycolysis and respiration even in the absence of hypoxia (i.e., in normoxic conditions)  Figure S2). These effects were likely due to the increased expression of the key glycolytic genes. Importantly, HIF1A and c-Myc share common target genes such as HK2, ENO1, and LDHA, whose products are involved in glycolysis. In fact, most genes coding for glycolytic enzymes have in their promoter regions consensus binding sequences for both HIF1A and c-Myc (48). Thus, it is plausible that Set7/9 controls glycolysis indirectly, by affecting c-Myc expression. In support of this assumption is the fact that A549 and H1299 cells with suppressed Set7/9 displayed an increased expression of c-Myc both at the transcriptional and protein levels.
In the present study, we demonstrated that the ablation of Set7/9 augmented the levels of HIF1A, c-Myc, and several genes coding for glycolytic enzymes. This points to Set7/9 as a potential tumor suppressor. Along with this notion, studies from several groups including ours also suggest that Set7/9 plays antiproliferative and tumor-protective roles in various cancers (7,15,56,57), including NSCLC (21,58).
In a number of works, Set7/9-mediated tumor suppressive effects have been achieved through p53 stabilization and activation. Indeed, we and others showed that Set7/9 is able to directly methylate the p53 protein, which leads to p53 stabilization and nuclear translocation and enhances the expression of p53 target genes (11,59). Additionally, it was demonstrated that Set7/9 contributes to p53 stability via inactivation of SIRT1 by its methylation. Thus, the Set7/9-mediated methylation of SIRT1 prevented p53 deacetylation and hence increased p53 transactivation under genotoxic stress (60). Here, we demonstrate that Set7/9 acts as a regulator of glycolysis in NSCLC cells regardless of the p53 status. These results contribute to a better understanding of the p53-independent role of Set7/9 in tumorigenesis.
However, several papers on the role of Set7/9 in intestinal regeneration argue against the tumor-suppressive role of Set7/9. Apparently, being part of a YAP/AXIN1/b-catenin complex Set7/9 facilitates Wnt-induced nuclear accumulation of bcatenin, a known oncogenic protein (61).
Thus, the role of Set7/9 in tumorigenesis is likely to be cell tissue dependent. Future studies should discern the molecular mechanisms involving Set7/9 that affect tumor growth and metastases. Our results presented in this study indicate that Set7/9 is a potentially important regulator of metabolic networking in NSCLC.

DATA AVAILABILITY STATEMENT
All datasets presented in this study are included in the article/ Supplementary Material. The proceeded microarray data was published in the Mendeley Datasets Repository (doi: 10.17632/ hkgfsz9yhn.1 https://data.mendeley.com/datasets/hkgfsz9yhn/1).

AUTHOR CONTRIBUTIONS
AD, OS, and NB designed experiments and wrote the manuscript. AD and OS carried out majority of the experiments. OF and AZ participated in MMP assays. AP and EB participated in the establishment of cell lines and performed ROS analysis. AK participated in experiments using SeaHorse. LL performed microarray analysis. All authors contributed to the article and approved the submitted version.

FUNDING
The authors acknowledge the support from RSF grant no. 19-75-10059. The shared research facility "Vertebrate cell culture collection" is supported by the Ministry of Science and Higher Education of the Russian Federation #075-15-2021-683.

ACKNOWLEDGMENTS
We are grateful to A. Kostareva (Almazov National Medical Research Center, Russia) for providing opportunity to use SeaHorse XFe24. We thank Alyona Kizenko for the bioinformatic analysis of the microarray data and Vasilisa Aksenova for the preparation of mRNA samples. We also acknowledge Dr P. Muller, University of Leicester, UK, for providing H1299 cell lines with tetracycline-inducible p53 expression.