Fatty acid metabolism-related genes as a novel module biomarker for kidney renal clear cell carcinoma: Bioinformatics modeling with experimental verification

Highlights

• •A computational and experimental approach was applied for key gene prioritization.
• •Novel risk score model and nomogram were developed for KIRC prognosis evaluation.
• •Hub gene ACADM as a novel biomarker for kidney renal clear cell carcinoma.

Abstract

Backgrounds

Lipid metabolism reprogramming is a hallmark of cancer, however, the associations between fatty acid metabolism (FAM) and kidney renal clear cell carcinoma (KIRC) prognosis are still less investigated.

Methods

The gene expression and clinical data of KIRC were obtained from TCGA. Using Cox regression and LASSO regression, a novel prognostic risk score model based on FAM-related genes was constructed, and a nomogram for prediction of overall survival rate of patients with KIRC was proposed. The correlation between risk score and the immune cell infiltration, immune-related function and tumor mutation burden (TMB) were explored. Finally, a hub gene was extracted from the model, and RT-qPCR, Western blot, Immunohistochemical, EdU, Scratch assay and Transwell experiments were conducted to validate and decipher the biomarker role of the hub gene in KIRC theranostics.

Results

In this study, a novel risk score model and a nomogram were constructed based on 20 FAM-related genes to predict the prognosis of KIRC patients with AUC>0.7 at 1-, 3-, and 5-years. Patients in different subgroups showed different phenotypes in immune cell infiltration, immune-related function, TMB, and sensitivity to immunotherapy. In particular, the hub gene in the model, i.e., ACADM, was significantly down-expressed in human KIRC samples, and the knockdown of OCLN promoted proliferation, migration and invasion of KIRC cells in vitro.

Conclusions

In this study, a novel risk score model and a module biomarker based on FAM-related genes were screened for KIRC prognosis. More clinical carcinogenic validations will be performed for future translational applications of the findings.

Introduction

Renal cell carcinoma is the third most common cancer in the urinary system, accounting for 3% of all cancers in women and 5% of all cancers in men [1]. As the most commonly observed subtype of renal cell carcinoma, kidney renal clear cell carcinoma (KIRC) is occurred with high incidence and has now become a serious global public health problem [2]. Currently, KIRC are often detected through health examinations, and surgical resection is the ideal method for treatment of KIRC patients. Unfortunately, a great number of patients are suffered from tumor recurrence after surgery. Despite recent advances in immunotherapy, e.g., immune checkpoint inhibitors (ICIs) targeting programmed death ligand 1/programmed death ligand 1 (PD-1/PD-L1), have delighted new direction for KIRC therapy, only a small proportion of KIRC patients are sensitive enough to ICIs treatment [3]. Hence it is of significance to develop a new model which integrates genetic signature and clinical characteristics to provide help for precision diagnosis and personalized management of patients with KIRC.

Energy metabolism reprogramming is an emerging hallmark of cancer, and accumulating evidence indicates that it plays a crucial role in the rapid growth and proliferation of cancers [4]. Studies have shown that abnormal lipid accumulation in tumor cells, e.g., cholesterol esters, promotes the proliferation and invasion of tumors [5]. In recent years, abnormal fatty acid metabolism has gained increasing attention in biological activities such as cell membrane formation, energy storage and signal transduction in tumorigenesis. It is suggested that fatty acids are indispensable for the emergence and development of cancers [6], [7], [8], [9]. KIRC is a unique subtype of renal cell carcinoma resulting from abnormal accumulation of lipid droplets in the cytoplasm, and lipid accumulation is known to be associated with disease progression [10]. For example, Shen et al. demonstrated that E2F1 promoted the proliferation and metastasis of KIRC by activating SREBP1 dependent fatty acid biosynthesis [11]. Qu et al. demonstrated that inactivation of AMPK-GATA3-ECHS1 pathway induced fatty acid synthesis and promoted tumor growth in KIRC [12]. It could be supposed that FAM pathway may be closely related to KIRC process by promoting KIRC progression or inhibition. At present, the relationship between FAM-related genes and KIRC development are not well investigated, and the single-molecular level identification of genes may limited the systematical understanding of KIRC due to the complexity in disease evolution [13,14]. Therefore, screening FAM-related genes as a panel or module signature to develop a risk score model based on systems biology paradigm is of clinical significance for precision medicine and personalized healthcare of KIRC patients.

In this study, a novel risk score model was constructed by integrating gene expression data and prior knowledge genes associated with FAM to predict the prognosis of KIRC patients. The performance and functions of the model were validated and explored to provide new perspectives for KIRC precision medicine. The study pipeline is shown in Fig. 1.

Fig. 1.

The flowchart of this study.

Materials and methods

Identification of FAM-related genes differentially expressed in KIRC

The mRNA sequencing data derived from 539 KIRC specimens and 72 normal renal tissue specimens and the corresponding clinical information of KIRC patients were downloaded from The Cancer Genome Atlas (TCGA) database. As illustrated in Supplementary Table 1, a total of 530 KIRC patients with complete clinical information were included in this study. A total of 309 known FAM-related genes were collected from Molecular Signature Database v7.2 [15]. The filter for differentially expression (DE) gene identification was set as false discovery rate (FDR) <0.05 and |Log₂FC| >1. The results were clustered using the "Pheatmap" R package.

Construction and evaluation of a risk score model based on FAM-related genes for KIRC prognosis

Univariate Cox regression analysis was performed to evaluate the association between FAM-related DE-genes and KIRC prognosis (P<0.05). LASSO regression analysis with 1500-folds cross-validation was then applied to select robust gene signatures as a module biomarker. The results of LASSO regression analysis were further processed using the "glmnet" R package [16] to develop a risk score model for predicting the overall survival rate of KIRC patients. The equation was defined as follows:

$$\text{Risk}\text{score}=\sum _{i=1}^n(Coefi*ExpGenei)$$

where “Coef” represents the non-zero regression coefficient calculated by LASSO regression analysis, and “ExpGene” represents the gene expression value in the risk score model. Here all samples were divided into low-risk or high-risk group based on the value of median risk score. Kaplan-Meier survival analysis was conducted to compare the difference of overall survival (OS) between different subgroups. ROC curves were drawn using the “Survival” package, “timeROC” package [17], and “SurvMiner” package in R program to evaluate the accuracy of the model, and the independence of prognostic gene characteristics in TCGA_KIRC cohort was evaluated by multivariate analysis. The “rms” package in R was used to develop a nomogram [18] for OS prediction in KIRC. 1 -, 3 -, and 5- year calibration curves and ROC curves were drawn to assess the accuracy of the nomogram.

Functional enrichment analysis of the risk score model

The functional enrichment analysis of FAM-related DE-genes was performed at the GO [19] (gene ontology, including molecular function, cellular component, and biological process) and KEGG [20] (Kyoto Encyclopedia of Genes and Genomes) level using “ClusterProfiler” package [21] in R. The terms with P<0.05 were considered to be statistically significant. “CIBERSORT” package was used to calculate the abundance of 22 immune cell subgroups in all samples [22]. Here perm is set as 1000. Samples with P<0.05 from the results were chosen for further analysis to compare the differences in infiltration of various immune cells between the high and low risk groups. ssGSEA analysis was performed using the “GSEABase” and “GSVA” R packages to obtain scores for immune-related functions, and the scores were calibrated to compare differences in immune-related functions between high- and low-risk groups. TIDE scores of KIRC patients were downloaded from http://tide.dfci.harvard.edu/, and further analyzed the differences of TIDE scores among high- and low-risk groups. Somatic mutation data of KIRC patients were obtained from TCGA database, TMB scores of each patient were calculated, TMB differences between high and low risk groups were compared and survival analysis was performed. In addition, the association between TMB score and prognostic risk score was assessed.

Network-based and literature-guided characterization of hub genes in KIRC pathogenesis

The protein-protein interaction (PPI) among identified genes in the constructed model were investigated using STRING v11.5 [23] online with an interaction score >0.400 (median confidence). The PPI data were then visualized using Cytoscape [24] software (v 3.9.1). The cytoHubba (v 0.1) was applied to prioritize hub genes based on the parameter of degree centrality in the network. Human Protein Atlas [25] (HPA) database was used to validate the expression of hub gene at the protein level, whereas Gene Expression Profiling Interactive Analysis (GEPIA) [26] was used to measure the mRNA-level expression of hub gene between TCGA_KIRC samples and normal kidney tissue samples.

Clinical sample collection

As shown in Supplementary Table 2, cancer tissue and adjacent normal tissues were surgically collected from 8 patients with KIRC at The First Affiliated Hospital of Soochow University from April 2022 to July 2022. Six pairs of frozen samples of KIRC tissues and adjacent normal tissues were used for RT-qPCR and Western blot. The samples were postoperatively pathologically confirmed. None of the patients had anti-tumor therapy prior to operation.

Total RNA isolation and RT-qPCR

TRIzol (Ambion, USA) was used to extract the total RNA from 769P cells and tissues. Then, the HiScript III RT SuperMix for qPCR (+gDNA wiper) (Vazyme, China) was used to perform the reverse transcription reactions and the AceQ qPCR SYBR Green Master Mix (Without ROX) (Vazyme, China) was used to complete qPCR. The specific primer sequences used were listed in Supplementary Table 3.

Immunohistochemistry(IHC)

The paraffin-embedded specimens were cut into 4 µm thick section for immunohistochemical staining. The sections were manually dewaxed with xylene and washed in gradient alcohol and ddH₂O. The sections were then placed in sodium citrate solution for antigen repair under high temperature and high pressure, and washed in ddH₂O and PBS. The samples were incubated with 3% H₂O₂ at room temperature for 15 min and then washed again in PBS. Then 3% BSA was added and the samples were incubated at 37 °C for 30 min. Adding designated antibody (ab92461, Abcam, UK) and incubate overnight at 4 °C to avoid light. IHC was performed by SP Rabbit & Mouse HRP Kit(DAB) (CWBIO, China).

Western blotting

Samples were all lysed in RIPA lysate (Beyotime, China). Protein concentrations were detected using the BCA kit(Beyotime, China). Place 30 μg samples in an SDS-PAGE gel. Separation by gel electrophoresis and then transfer the protein to a PVDF membrane. The membrane was closed with 5% skim milk for 2 h at room temperature and then incubated overnight with the corresponding primary antibody(ab92461, Abcam, UK) at 4 °C. Incubate using horseradish peroxidase-conjugated secondary antibodies and then exposing.

Cell cultures and RNA interference

The human primary clear cell adenocarcinoma cells (769P cells) were maintained in RPMI 1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (Gibco, USA), 100 U/mL of penicillin and 100 U/mL of streptomycin (Beyotime, China). The conditions of cell cultures were 37 °C and 5% CO₂. Inoculate into six-well plates at appropriate density the day before transfection. According to the manufacturer's protocols, siRNA (GenePharma, China) and lipofectamine_2000 transfection reagent (Invitrogen, USA) were used for transfection.

Scratch assay

After 48 h of transfection, 100 μL of Eppen-dorf Tip was used to scratch the cell plate and the cells were washed 2–3 times by PBS. Cells continue to be cultured with 1% fetal bovine serum (Gibco, USA). Observe the changes of cells in each plate at 0 h and 24 h with an inverted microscope.

EdU assay

The EdU assay kit was purchased from Beyotime Company (C0071S, China). After 48 h of transfection, 3  ×  10⁵ cells were added to 12-well plates and cultured for 12 h. Exposed cells to medium containing 50 μM EdU and incubate in an incubator for another 3 h. Cells are fixed with 4% paraformaldehyde for 15 min at room temperature, followed by incubation with 0.3% Triton X-100 for 15 min at room temperature. Incubate with 200 μl Click reaction solution for 30 min at room temperature and then with 1 ml of Hoechst 33,342 at room temperature for 10 min. An Nikon TI2-d-PD inverted microscope (Japan) was used to capture the images.

Transwell assay

The upper chamber was coated with a layer of Matrigel matrix glue (Corning, USA) (matrix glue: Serum-free medium=1:6). After 48 h of transfection, Cells are resuspended with serum-free medium. Then, 5  ×  10⁴ cells were added to upper chamber, and 500 μl full medium were added to lower chamber. After 48 h, fix the cells in the upper chamber with 4% paraformaldehyde for 30 min at room temperature and stain with crystal violet for 20 min. An Nikon TI2-d-PD inverted microscope (Japan) was used to capture the images.

Statistical analysis

Bioinformatics analyses were performed in R4.2.2. Experiments were independently performed at least three times. The statistical analyses were performed using GraphPad Prism 9.0 (GraphPad, San Diego, CA, USA). All data were presented as mean ± SD. The Student's t-test was used to analyze the differences between two groups. One-way or two-way ANOVA analysis were used to analyze differences between three or more groups. P < 0.05 was set as statistically significant.

Results

FAM-related genes identified for KIRC prognosis

In this study, the expression level of FAM-related genes between normal and KIRC tissue samples were compared, and 107 FAM-related DE-genes were identified and summarized in Supplementary Table 4. As described in Supplementary Fig. 1A, 36 genes were up-regulated and 71 genes were down-regulated in KIRC. As shown in Supplementary Fig. 1B, C, functional enrichment analysis indicated that the identified 107 genes were closely involved in FAM-related GO and KEGG terms. Based on univariate Cox regression analysis, as shown in Supplementary Fig. 1D, a total of 46 genes were finally identified to be significantly associated with the prognosis of KIRC patients (P<0.05).

Novel risk score and nomogram for prognosis prediction of KIRC patients

As shown in Fig. 2A, B and Supplementary Table 5 respectively, 20 FAM-related genes were prioritized by LASSO regression analysis as a robust module signature to construct the risk score model for KIRC prognosis. As shown in Fig. 2C, time-dependent ROC at 1, 3, and 5 years showed the accuracy of the model with AUC=0.750, 0.741, 0.787 for predicting OS of KIRC patients, respectively. As shown in Fig. 2D, multivariate Cox regression indicated the risk score as an independent factor for KIRC prognosis (P <0.001, HR:2.638). The samples were further divided into two groups, i.e., the high-risk group and the low-risk group, according to the median value of the risk score, and Kaplan-Meier survival analysis was conducted to compare the OS of patients between high and low risk group. As indicated in Fig. 2E, the OS was significantly reduced in patients in the high-risk group compared with those in the low-risk group. Moreover, as described in Fig. 2F, G, the heatmap and risk curve showed that the mortality of patients in the high-risk group was significantly increased, suggesting the potential of the risk score as a prognostic biomarker for patient survival of KIRC. Fig. 2H showed the heatmap of the twenty optimal FAM-related genes. As shown in Supplementary Fig. 2, The high-risk group exhibited advanced T, N and M stages compared with those of the low-risk group, and clinical stage and pathological grade were more frequent in the high-risk group.

Fig. 2.

Construction of prognostic risk model based on FAM-related genes. (A) LASSO coefficient profiles of the 107 FAM-related genes; (B) Identification of genes for development of prognostic risk score model; (C) Time-dependent ROC curve demonstrated the performance of the model in predicting 1 -, 3 -, and 5-year survival in KIRC; (D) Multivariate Cox regression analysis results; (E) Kaplan-Meier curves of overall survival between patients in the high- and low-risk group; (F and G) Risk score and survival status of KIRC patients in high- and low-risk groups; (H) Heatmaps for the expression of identified genes between high- and low-risk group.

As shown in Fig. 3A, a nomogram based on the risk model and clinical parameters was constructed for survival prediction of KIRC patients. The calibration curves of 1, 3 and 5 years in Fig. 3B indicated the prediction consistency of the nomogram compared with the real observed OS. Moreover, the AUC=0.882 in Fig. 3C also convinced the good predictive power of the nomogram.

Fig. 3.

Nomogram construction and evaluation. (A) The nomogram constructed based on the risk score and clinical characteristic parameters; (B) 1 - year, 3 - year and 5 - year calibration curves of nomogram; (C) Comparison of ROC curves among nomogram, prognostic risk score and clinical pathological characteristics.

Immune-related characteristics between the high- and low-risk groups

In this study, the deconvolution algorithm CIBERSORT was applied to obtain the proportion of 22 immune cells in all samples. As shown in Fig. 4A, plasma cells, T cells CD8, T cells Follicular helper, Tregs, Macrophages M0 were infiltrated significantly in high-risk group, while T cells CD4 memory resting, Macrophages M1, Macrophages M2, Dendritic cells Resting, Mast cells resting were significantly infiltrated in the low-risk group. Analysis of immune-related functions revealed significant differences between high-risk and low-risk groups. As shown in Fig. 4B, the high-risk group activated APC_co_stimulation, APC_co_inhibition, CCR, check-point, activation of cytolysis, HLA, Inflammation-promoting, Parainflammation, T_cell_co_stimulation, T_cell_co_inhibition, Type_Ⅰ_IFN_Reponse and Fig. 4C showed that TIDE score in high-risk group was significantly higher than that in low-risk group.

Fig. 4.

Functional survey of the risk signature in KIRC carcinogenesis. (A) Histogram of immune cell infiltration between high- and low-risk groups; (B) Histogram of immune function between high- and low-risk groups; (C) Violins of TIDE scores between high- and low-risk groups; (D) Differences in TMB between high- and low-risk groups; (E) Scatter plot depicting positive correlation between high- and low-risk scores and TMB; (F) Kaplan-Meier curves of high- and low-TMB group; (G) Kaplan-Meier curves for patients stratified by TMB and prognostic risk score (*: P<0.05; **: P<0.01; ***: P<0.001).

TMB refers to the number of mutated bases per million bases in tumor tissue, which is an emerging biomarker and is increasingly used to predict patient prognosis. As shown in Fig. 4D, the higher TMB was observed to be concentrated in the high-risk score group (p = 0.012). Correlation analysis in Fig. 4E further verified that TMB was significantly positively correlated with prognostic risk score. (R = 0.16, p = 0.0039). As shown in Fig. 4F, Kaplan-Meier analysis demonstrated that the lower TMB was associated with higher survival of KIRC patients (p = 0.001). To further explore the validity of prognostic risk score and TMB consistent prognostic significance, the synergistic effect of these two indicators on prognostic prediction of KIRC patients was validated. As shown in Fig. 4G, the combination of TMB with risk score seemed to have a better prognostic capability for KIRC patients, and patients with high TMB and high risk-scores tended to have the worst prognosis.

ACADM identified and validated as a novel biomarker in KIRC carcinogenesis

Since the development of KIRC is a consequence of disorder at multi-gene level, the PPI network was reconstructed to explore the interaction among genes identified in the risk score model. By ranking the degree of genes in the network, three hub genes, i.e., ACADM, HIBCH and HADH, were extracted as shown in Fig. 5A. In this study ACADM was chosen as the case for further analysis. As shown in Fig. 5B, the mRNA-level expression of ACADM in KIRC specimens was significantly lower than that in normal specimens. Moreover, ACADM protein level was also significantly lower in KIRC tissue compared with normal kidney tissue according to the records in HPA database as shown in Fig. 5C. As illustrated in Fig. 5D, The Kaplan–Meier survival curve indicated that low ACADM expression was associated with poor prognosis. Finally, the expression pattern of ACADM was explored in KIRC tissues. As shown in Fig. 5E, the expression level of mRNA of ACADM notably reduced in tumors (p-value<0.01), and the protein expression of ACADM was also decreased in KIRC tissue samples in Fig. 5F. The immunohistochemistry staining in Fig. 5G further verified the lower expression of ACADM in tumor tissues.

Fig. 5.

The validation and functional survey of hub gene ACADM. (A) Hub genes in the constructed risk score model; (B) The mRNA-level expression of ACADM between KIRC and normal kidney tissue in GEPIA; (C) The protein-level expression of ACADM between KIRC and normal kidney tissue in HPA; (D) Kaplan-Meier survival analysis of ACADM; (E) Expression of ACADM between KIRC and normal tissues measured by RT-qPCR; (F) Immunoblotting analysis of KIRC and normal tissues. The image comes from two different gels; (G) Immunohistochemistry staining of ACADM in KIRC and normal renal tissue (*: P<0.05; **: P<0.01; ***: P<0.001).

In order to further verify the role of ACADM in KIRC, 769P cell lines was selected for follow-up experiments. As shown in Fig. 6A, B, ACADM knockdown was conducted using two different siRNAs. To evaluate the effect of ACADM in cancer cell proliferation, EdU assay was performed. As shown in Fig. 6C, the percentage of EdU-positive cells was increased after ACADM knockdown. Subsequently, scratch assay was performed to assess the effect of knockdown of ACADM on the migration of the 769P As shown in Fig. 6D, compared with the control group, the migration rate of the siRNA groups was dramatically increased at 24 h. Finally, the migrate assay shown in Fig. 6E indicated the percentage of invaded cells was significantly increased after ACADM knockdown at 48 h.

Fig. 6.

Experimental functional verification of ACADM. (A) The mRNA-level expression of ACADM after 36 h siRNA transfection; (B) The protein-level expression of ACADM after 48 h siRNA transfection. The image comes from two different gels; (C) Representative images of the EdU assay in 769P cells with ACADM knockdown; (D) Scratch assay showed the mobility of 769P cells with ACADM knockdown over 24 h. The Image J software calculates the area of wound healing; (E) Transwell assay showed the invasion capacity of 769P cells with ACADM knockdown over 48 h. The Image J software calculates the number of invaded cells at 48 h. All experiments were repeated at least three times, and the data were shown as means ± S.D.(*: P<0.05; **: P<0.01; ***: P<0.001).

Discussion

KIRC is the most common subtype of renal malignancy, with increasing morbidity and mortality over the years [1]. It is urgently needed to identify candidate biomarkers and explore the underlying molecular mechanisms for KIRC prognosis and personalized therapy. Currently, metabolic reprogramming has become a hot topic in cancer research [27,28]. For example, up-regulation of glycolysis metabolism is one of the features of human malignant tumor [29]. With the research progress on metabolic reprogramming, the importance of FAM has gradually been uncovered and it is known that FAM pathway participates in energy production, membrane synthesis and signal transduction in tumor development and progression. In renal tumor, oncogenic mutations could lead to the change in glycolysis, fatty acid biosynthesis, and amino acid metabolism [30,31]. In the past few years, it has been suggested that the regulation of FAM could influence the development and prognosis of KIRC [32,33]. However, the function and clinical significance of FAM-related genes as a module signature for KIRC precision medicine were not well reported.

In this study, the association between FAM-related genes and KIRC were systematically investigated and a risk score model was constructed based on 20 FAM-related genes to predict the prognosis of KIRC patients using integrated univariate Cox regression and LASSO regression analysis. This signature exhibited a greater ability to predict the prognosis of KIRC patients than the traditional TNM stage. Many adverse events, such as lymph node metastases or distant metastases, can also be predicted by assessing a patient's risk score using this model. Multivariate Cox regression indicated that the prognostic risk model could be serve as an independent factor for OS of KIRC patients. Survival analysis showed that patients in the high-risk group had a poor prognosis. In order to better guide clinical decision-making, a nomogram based on the risk model and other clinical parameters was constructed. A growing body of research shows that lipid metabolism is closely related to tumor immunity. Lipids act as intracellular signaling substances that affect the function of a variety of immune cells. KIRC is considered to be highly immunogenic, the tumor immune microenvironment impacts on a portion of KIRC, and some patients may be evaluated to be suitable for immune checkpoint blockade therapy strategies. Based on the above views, we tried to perform immunity analysis. Patients in the high-risk group had a significantly higher TMB, indicating that it is more likely to be detected by the immune system which was consistent with the higher abundance of CD8+T cells, activated CD4+T cells and plasma cells in the high-risk group. It is interesting to note Tregs infiltration was also significantly enriched by high-risk group. Tian et al. demonstrated that Tregs could control a variety of immune cells, such as NK cells and B cells, through a variety of mechanisms, thereby inhibiting tumor immune responses [34]. Although immunotherapy has been applied successfully in KIRC, not all the patients can benefit from this treatment. Therefore, it is indispensable to select appropriate biomarkers to distinguish the patients who are more sensitive to immunotherapy. In our study, patients in the high-risk group showed a higher TIDE score, indicating a greater risk of immune escape and a poorer response to immunotherapy. Therefore, attenuated antitumor immunity in patients at high-risk may be an explanation for their poor prognosis.

Finally, through the PPI network modeling, a hub gene, i.e., ACADM, was identified. ACADM is reported to catalyze the first dehydrogenation of fatty acyl-coA to β -oxidation in mitochondria [35]. Preliminary experimental results from RT-qPCR, Western blot and Immunohistochemistry showed that the expression of ACADM in cancer tissues was significantly lower than that of normal tissues. Although the molecular pathological mechanism of low expression of ACADM in KIRC has not been well decoded yet, many studies proved that the inhibition of fatty acid oxidation was essential in cancer progression. For example, inhibition of ACADM could alter lipid metabolism in hepatoma cells, resulting in elevated levels of triglycerides, phospholipids and cytolipid droplets, ultimately driving liver carcinogenesis. In breast cancer, the downregulation in medium-chain acyl-CoA dehydrogenase activity accelerated cancer progression. In this study, we found that knockdown of ACADM could promote proliferation, migration, and invasion of KIRC. Tumor cells must constantly obtain nutrients from their surroundings to meet their need for rapid proliferation. Lipid metabolism, especially fatty acid β -oxidation, can produce tremendous amounts of energy. The products of fatty acid β -oxidation include NADH and FADH2, which can be directed into the electron transport chain of ATP production. In addition, the inhibition of fatty acid β -oxidation also affects lipid metabolism, resulting in abnormal lipid accumulation. We speculated that ACADM may lead to energy deprivation and abnormal accumulation of lipids in KIRC by inhibiting fatty acid β -oxidation, thereby affecting tumor development and development, which requires further experimental verification.

It should be concerned that several limitations still need to be considered, First, although FAM-related genes were screened and integrated as a module signature for translational KIRC research, the interactions among these genes were not well weighted during model construction, since the alternation in such interactive edges in the network is essential for systems-level decoding of health-disease status transition. Second, the FAM-mediated carcinogenesis between identified genes and KIRC phenotype should be deeply explored and understood. Last but the most important, only expression-level experiments were performed for gene validation in this study, more clinical and pathogenic survey will be conducted in the future work to convince the findings.

Conclusions

In this study, a novel model and a module biomarker based on FAM-related genes was screened for KIRC prognosis using an integrated bioinformatics and experimental approach. More pathological validations using human clinical data will be performed for deciphering the underlying carcinogenesis of these genes during KIRC evolution.

CRediT authorship contribution statement

Zongming Jia: Conceptualization, Methodology, Writing – original draft, Writing – review & editing. Zhenyu Fu: Methodology. Ying Kong: Validation. Chengyu Wang: Validation. Bin Zhou: Methodology. Yuxin Lin: Conceptualization, Funding acquisition, Writing – review & editing. Yuhua Huang: Conceptualization, Supervision, Funding acquisition, Writing – review & editing.

Declaration of Competing Interest

None.