Abstract
Background/Aim: Renal cell carcinoma (RCC) is one of the most commonly diagnosed cancers in the world. Approximately 25-30% of patients identified with initial kidney cancer will have metastasized tumors, thus 5-year survival rates for these patients are poor. Therefore, biomarker research is required to identify and predict molecular signatures in RCC. Materials and Methods: To address this, we used a mass spectrometry (MS)-based proteomics approach to identify proteins related to clear cell RCC (ccRCC) tissues from patients with T1G2, T1G3, T3G2, T3G3, and metastatic RCC (mRCC) stages. Results: We identified and quantified 2,608 and 2,463 proteins, respectively, in ccRCC tissue and identified 1,449 differentially expressed proteins (DEPs). Bioinformatics analysis revealed that serpin family A member 3 (SERPINA3) qualified as biomarker for ccRCC progression. Using indirect enzyme-linked immunosorbent assay (ELISA), immunoblotting, and immunohistochemistry assays it was found that SERPINA3 expression levels in ccRCC tissues were much higher in stages before metastasis. Conclusion: Comparative proteomics analysis of ccRCC tissues provided new evidence of SERPINA3 association with ccRCC progression.
Renal cell carcinoma (RCC) is among the top eight most diagnosed cancers in the United States, of which approximately 79,000 patients are newly diagnosed and 13,920 deaths are recorded (1). Clear cell RCC (ccRCC) disease is the most common histopathological subtype representing approximately 85%-90% of all diagnosed cases (2). During treatment, metastatic RCC (mRCC) is considered a fatal disease. It is accepted that 25%-30% of patients diagnosed with early kidney cancer will have metastasized tumors (3, 4) and their 5-year survival rate is 11.7%, with a very poor prognosis (4, 5). Approximately 30% of patients will experience recurrence despite primary tumor removal, including 10%-15% of T1 patients with clinically removed tumors (4). mRCC patients show very poor responses to chemotherapy and radiotherapy, and the rate of disappointment with these responses is increasing to 25% (6, 7). Common mRCC sites include the lung (40%), bone (30%), lymph node (22%), and liver (20%) (8).
By understanding mRCC clinical and biological properties, therapeutic agents are constantly being developed (9-11). In recent years, VEGF, mTOR, VEGFR/MET/AXL, programmed death-1 (PD-1), and multi-kinase inhibitors have been developed with Food and Drug Administration approval (12, 13). Although diverse therapeutics is used to treat mRCC, new therapeutic targets must be identified to overcome the disease. Omics-based approaches provide incredibly useful information on unknown targets against disease. In recent years, genomics and transcriptomics studies have successfully identified characteristic genes implicated in mRCC (14-16). For example, in a recent study, metastasis-associated prognostic signatures based on single-cell RNA-sequencing in ccRCC were identified (17), whereas levels of ccRCC transcripts using The Cancer Genome Atlas (TCGA) were reported in another article (18). Proteins are important disease targets when exploring new therapies. Even if specific genes are transcribed into RNA and expressed in cancer cells, they may not be translated into functional proteins. Additionally, proteins provide connections between functional pathways in cells and the environment, regardless of changes in RNA levels (19).
In recent decades, mass spectrometry (MS)-based comparative proteomics have been routinely used to detect and quantify thousands of proteins (20). Among proteomics-based protein quantification methods, the isobaric tagging of tryptic peptides has facilitated the simultaneous determination and relative abundance of peptide pairs in complex protein mixtures (21). The stable isotopic labeling of peptides using isobaric reagents [e.g., tandem mass tags (TMTs)] has been used to quantitatively compare proteins across multiple samples (22). By quantitatively profiling thousands of proteins using MS coupled to TMTs, researchers have discovered proteins related to cancer progression or cancer diagnosis biomarkers from patient tissue (23).
Proteomics approaches have been used to identify potential prognostic factors or metabolic pathways in mRCC (24-27). However, cancer is a heterogeneous disease, and it is suggested that the number of proteins implicated in in vivo mechanisms is as high as one million, therefore more information on mRCC progression is required using quantitative proteomic profiling. In this study, using TMT-coupled comparative proteomics, we identified specific differentially expressed proteins (DEPs) in mRCC tissue to clarify novel regulators implicated in kidney cancer metastasis.
Materials and Methods
Clinical tissues from patients with kidney cancer. Study samples were provided by the Kyungpook National University Chilgok Hospital (Patient consent was obtained before acquiring tumor tissues; ethics approval number; KNUCH 2016-05-021-019). Tissue samples were surgically collected from 18 ccRCC patients (previously diagnosed with the disease), with tissue stored at −80°C until required. We divided samples into six group (three samples/stage) according to diagnosed stage: normal kidney, T1G2, T1G3, T3G2, T3G3, and mRCC. The tumor grades used were T stages and Fuhrman grade, the Fuhrman grading system is commonly used to grade ccRCC, and it categorizes nuclear grades from 1 to 4 based on the increasing nuclear size, irregularity, and nucleolar prominence (28). Normal kidney tissue came from distant kidney tissue from patients with ccRCC T1G2 disease. mRCC patients had metastases confirmed before surgery and none had received targeted treatment.
Sample preparation for comparative proteomics analysis. On ice and before lysis, tissue samples were washed in Dulbecco’s phosphate buffered saline to remove blood, then RIPA buffer (Thermo Fisher Scientific, Rockford, IL, USA) and Halt protease inhibitor cocktail (Thermo Fisher Scientific) reagents were added. Tissues were homogenized using a hand-held homogenizer (MIULAB, Hangzhou, Zhejiang, PR China), sonicated five times, centrifuged (12,000 × g for 10 min) at 4°C, and supernatants collected. To prepare an averaged sample for each stage, 200 μg of proteins extracted from individual tissues were pooled per stage. Protein concentrations were determined by bicinchoninic acid assays (Thermo Fisher Scientific). Six sets of reduction samples (600 μg) were prepared by adding 25 mM ammonium bicarbonate (ABC). To reduce and alkylate samples, 15 mM dithiothreitol was added to samples and incubated for 30 min at 56°C, after which samples were mixed with 60 mM IAA in the dark for 30 min at room temperature (RT). After this, trichloroacetic acid (TCA) protein precipitation was performed by slowly adding TCA drop-wise to samples and incubating for 4 h at 4°C. After this, samples were centrifuged (12,000 × g for 7 min) at 4°C and supernatants removed. After washing pellets in 500 μl ice-cold acetone for 5 min, samples were centrifuged (12,000 × g for 7 min) at 4°C, the step repeated, and supernatants removed and dried for 1 min. Before trypsin digestion, 50 mM ABC was added, and samples repeatedly sonicated to re-dissolve debris. Trypsin was then added, the pH measured (pH strips), and at pH 7-8, samples were incubated overnight at 37°C.
After digestion, 10% trifluoroacetic acid (TFA) was added, samples centrifuged (16,000 × g for 15 min) at 4°C and supernatants collected. Then, 20 μl supernatants were taken for peptide quantitation and the remainder were dried in a Speed-Vac (Labconco, Kansas, MO, USA). Purified peptide concentrations were determined using a peptide assay kit (Thermo Fisher Scientific). TMT labeling was conducted according to manufacturer’s (Thermo Fisher Scientific) protocols. Dried peptides were dissolved in 100 mM triethylammonium bicarbonate (Sigma-Aldrich, St Louis, MO, USA), and 41 μl anhydrous acetonitrile (ACN) mixed with a 6-plex TMT labeling reagent (Thermo Fisher Scientific) at RT. TMT labeling reagents (126-131, Thermo Fisher Scientific) were added to samples and incubated for 1 h at RT. Next, 8 μl of 5% hydroxylamine was added to finish reactions and samples incubated for 15 min. Then, equal sample amounts were pooled and dried in a speed-vac.
Peptide fractionation using high-pH reverse-phase (RP) and desalting processes. RP fractionation was performed to remove sample complexities and ensure deep proteome sequencing prior to liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS). A vacuum-dried 100 μg peptide sample was reconstituted in 0.1% TFA and loaded onto a fractionation spin column (Thermo Fisher Scientific). Combined peptide samples were eluted in eight different pH elution solutions and dried in a vacuum drier. Ziptip C18 resin (Millipore, Burlington, MA, USA) was used to remove salt and samples were re-dried and stored at −80°C until analysis by LC-MS/MS.
NanoLC-MS/MS. To perform a global comparative analysis of kidney cancer tissue, fractionated peptides were analyzed using two LS/MS devices: the LTQ Velos-Orbitrap (Thermo Fisher Scientific) and a Q-Exactive Plus Hybrid Quadrupole-Orbitrap (Thermo Fisher Scientific). For the LTQ Velos-Orbitrap with RSLCnano UHPLC (Thermo Fisher Scientific), peptides samples were separated using a two-column system with a nanoViper trap column [75 μm internal diameter (I.D) × 2 mm] and analytical column 75 μm I.D × 50 cm. Peptides were analyzed using a Q-Exactive Orbitrap spectrometer coupled to an online Ultimate 3000 RSLCnano system (Thermo Fisher Scientific), equipped with an Acclaim PepMap 100 C18 HPLC column (75 μm × 2 cm, 3 μm nanoviper) as the loading column and an EASY-Sparay PepMap RSLC C18 column (75 μm × 50 cm, 2 μm; Thermo Fisher Scientific). The linear LC gradient was set to 120 min total running time, with a linear gradient of 5%-28% solvent B (ACN/0.1% formic acid): 0%-28% solvent B (ACN in 0.1%FA) for 110 min, 28%-90% solvent B for 8 min, and 90% solvent B for 12 min at 300 nl/min. High-energy collision dissociation was used to fragment peptide precursor ions. The normalized collision energy was 40%. The injection volume was 2 μl of peptide (1 mg) dissolved in solvent A (water/0.1% FA). The top 10 data-dependent mode switch automatically between MS1 and MS2 acquisition. Full scans were acquired at 300-1,800 m/z and a 1.8 kV voltage in the LTQ Orbitrap system and 350-2,000 m/z and a 2.0 kV voltage in the QE Plus Orbitrap system. Data were collected using MS/MS acquisition (data dependent) and single charged peptides were excluded from fragmentation. Basic Science Research Capacity Enhancement Project, through the Korea Basic Science Institute (National Research Facilities and Equipment Center), performed LTQ Velos-Orbitrap analysis. Raw files from datasets were uploaded into ProteomeXchange via the PRIDE partner repository with the identifier PXD039204.
Database searching. Raw files were loaded into the MaxQuant software package (v.1.5.2.8) and searched based on the Uniprot human database (73,928 sequences and 25,105,724 residues). Reverse decoy database parameters were applied, and the precursor mass tolerance set to 20 ppm. The false discovery rate (FDR=0.01) was used to filter frequently observed contaminants and reversed sequence copies. Enzyme specificity was set as C-terminal to Arg and Lys, also allowing a maximum of two missed cleavages. Fragment ion mass to tolerance was set at 0.02. Carbamidomethyl on Cys residues were set-fixed modifications and acetylation on protein N-terminal and oxidation on Met residues were variable modifications. For protein labeling, TMT 6-plex (N-term) was considered a variable modification. All other MaxQuant parameters were set at default.
Bioinformatics. For comparative proteomics analysis, five comparison groups were used, including normal vs. T1G2, T1G3, T3G2, T3G3, and mRCC. Proteins with calculated fold changes (FCs) >2 or <0.5 were considered DEPs and were analyzed using the UniProt-GOA database (29) and Database for Annotation, Visualization and Integrated Discovery (DAVID) (30, 31). Gene ontology (GO) is a major bioinformatics tool which unifies three annotation categories across species, including, biological process, cellular component, and molecular function (32, 33). The online pathway analysis platform, the Kyoto Encyclopedia of Genes and Genomes (KEGG) (34-36) integrates network information for mapping annotation results. DEPs were also transformed into z-scores and one-way hierarchical clusters clustered (Perseus software platform v. 1.6) into six classification types (A-F). Clusters were visualized using a heat map based on reporter ion values from groups. A volcano plot was expressed between the T3G3 and mRCC group by log10 (p-value) and log2 (fold change), and ratio correlations between technically duplicated analyses were calculated by Pearson’s correlation in the Perseus software platform.
Immunoblotting. Tissues were lysed in RIPA buffer with a Halt protease inhibitor cocktail (Thermo Fisher Scientific) and protein concentrations checked using BCA assay kit (Thermo Fisher Scientific). Then, 20 μg protein was separated by SDS-PAGE using 10% Tris-glycine polyacrylamide gel for 90 min and transferred to polyvinylidene fluoride (PVDF) membranes (Roche, Mannheim, Germany) for 2 h. Membranes were then blocked in 5% bovine serum albumin (BSA) for 1 h at RT and probed with primary antibody for SERPINA3 (Abcam, Cambridge, UK) overnight. The next day, membranes were washed three times in TBST for 10 min. After this, secondary antibody (CST, Danvers, MA, USA) was added and membranes were incubated for 1.5 h. Both antibodies were diluted to 1:1,000 in 5% BSA. Membranes were then washed three times in TBST for 10 min. We used ECL™ prime western blotting detection reagents (Cytiva, Buckinghamshire, UK) to visualize protein bands. Proteins were then exposed and analyzed using iBrignt 1500 instrumentation (Thermo Fisher Scientific).
Indirect enzyme-linked immunosorbent assay (Indirect ELISA). Protein (5 μg) was added to 50 μl coating buffer (filtered 0.2 M sodium bicarbonate, pH 9.4) in a 96-well plate (Corning, NY, USA) and incubated overnight at 4°C. Wells were washed three times in wash buffer (TBS plus 0.05% Tween 20, pH 7.2) every 5 min and 50 μl blocking buffer (3% BSA in TBST) then added to wells and incubated for 6 h. A primary SERPINA3 antibody (Abcam) was diluted (1:250) in blocking buffer and 50 μl added to wells and reacted overnight at 4°C. The next day, the antibody was removed, and wells were incubated with a secondary antibody for 2 h at RT. After washing wells five times, TMB substrate (Thermo Fisher Scientific) was added to wells and the plate incubated for 15-30 min at RT. To stop reactions, stop solution (Thermo Fisher Scientific) was added and the absorbance measured at 450 nm using a microplate reader (TECAN, Männedorf, Switzerland).
Immunohistochemistry (IHC). Tissues used in the experiments were stored at −80°C until required. For IHC, tissues were fixed in 4% formaldehyde and formed into paraffin blocks. Then, sections were cut into 4 um thickness and attached to coated glass slides. After deparaffinization and hydration, sections underwent antigen retrieval and blocking in citrated antigen retrieval buffer and 5% BSA, respectively. The primary antibody (Abcam) was diluted as per manufacturer’s recommendations and used at 4°C for 18 h. A secondary antibody (1:1,000) was applied for 1 h at RT. Sections were dehydrated and cleared using alcohol and xylene and then mounted with DAPI in mount medium.
Statistics. The levels of SERPINA3 were measured using ELISA, and the fold change for each stage was compared to normal values and presented in a bar graph. Statistical significance was determined using Student’s t-test, and a p-value of less than 0.05 was considered statistically significant, denoted by an asterisk (*). More significant values were denoted by **p=0.01 and ***p=0.001.
Results
Comparative proteomics analysis of ccRCC tissue. To identify DEPs related to mRCC progression in RCC, proteins at each disease stage were analyzed using MS and quantitative proteomic analyses (Figure 1A). All normal to mRCC tissue were obtained from the National Biobank of Korea-Kyungpook National University Hospital. Samples were analyzed using high-resolution MS. Peak alignments were performed in MaxQuant 1.6 against a Homo sapiens database (uploaded 2018 December) (FDR <1% and score >40). Overall, we identified and quantified 2,608 and 2,463 proteins, respectively, in RCC tissue (Figure 1B). Normalized data are shown in a box plot and technical coefficients were examined using Pearson correlation coefficient values among biological and technical duplications. R-values were as high as 0.941 which indicated highly correlated replicates.
Mass spectrometry (MS)-based quantitative proteomic profiling of renal cell carcinoma (RCC). (A) Experimental workflow showing 6-plex tandem mass tag labeling of RCC tissue (Control, T1G2, T1G3, T3G2, T3G3, and mRCC) using liquid chromatography mass spectrometry. (B) Overview of DEPs from RCC tissues. (C) Histogram showing DEPs distribution. DEPs: Differentially expressed proteins.
We quantified 1,449 DEPs in RCC tissue by comparing control and different disease stage tissue (Figure 1C). Proteins with a quantitative vs. control ratio of >2.0 were up-regulated and those <0.5 were down-regulated. Consequently, protein distribution at disease stages showed that most DEPs in RCC tissue were increased when compared with controls.
DEPs identification in mRCC. To identify regulatory proteins which progressed from RCC to mRCC, we performed one-way t-tests to identify statistically significant DEPs between T3G3 and mRCC stages and displayed our data in volcano plot (Figure 2). Significant DEPs were filtered using p- values <0.05 and >2-fold (log2 scale ≤−1 and ≥1) change in expression. Accordingly, 147 DEPs were identified, including 95 and 52 up- and down-regulated proteins, respectively, in mRCC when compared with T3G3.
Volcano plot showing DEPs at T3G3 and mRCC stages. Volcano plot showing DEPs compared between T3G3 and mRCC stages. Left- and right-red dots represent significantly down- and up-regulated proteins, respectively. Gray dots show proteins with no significant expression changes. Significant DEPs were filtered using p=0.05 and <2-fold (log2 scale ≤−1 and ≥1) expression changes. DEPs: Differentially expressed proteins.
To investigate DEPs functional differences between mRCC and T3G3 tissue, DEPs functional annotations were performed in DAVID (Figure 3). From KEGG pathway analysis, up-regulated proteins were highly associated with complement and coagulation cascades and also related to cholesterol metabolism categories and regulation of actin cytoskeleton. Down-regulated proteins were related to peroxisome, arrhythmogenic right ventricular cardiomyopathy, and drug metabolism - cytochrome P450. Additionally, we compared significantly regulated proteins in mRCC/T3G3 in GOMF, GOCC, and GOBP categories. Firstly, from GOMF results, up-regulated proteins were associated with “endopeptidase inhibitor activity” and “serine-type endopeptidase inhibitor activity”, while down-regulated proteins were highly associated with “structural constituent of cytoskeleton” and “ATPase binding” processes. From GOCC results, up-regulated proteins were enriched in “blood microparticle” and “extracellular exosome” categories, whereas down-regulated proteins were involved in “spectrin-associated cytoskeleton” and “mitochondrion” categories. Finally, from GOBP results, up-regulated proteins were related to “negative regulation of endopeptidase activity” and “acute-phase response” categories, while down-regulated proteins were associated with “actin filament capping” and “snoRNA localization” processes (Figure 3).
KEGG and GO pathway based-enrichment analysis of DEPs at mRCC/T3G3 stages. (A) Up-regulated and (B) down-regulated proteins at mRCC when compared with the T3G3 stage. The −log Fisher’s exact test was used to represent enrichment indices. GOBP: Gene Ontology Biological Process; CC: cellular component; MF: molecular function; KEGG: Kyoto Encyclopedia of Genes and Genomes; DEPs: differentially expressed proteins.
Verifying DEPs in clinical tissue. From up-regulated proteins, significant DEPs were selected with a >2-fold expression change in volcano plot and Atlas. SERPINA3 was finally selected from significant DEPs; Kaplan-Meier curve showed that ccRCC patients with high SERPINA3 expression levels had poor prognoses when compared with patients with low expression levels (Figure 4A). SERPNA3 was assessed using indirect ELISA and immunoblotting assays to validate comparative proteomics analysis. From indirect ELISA, SERPINA3 was slightly increased at T3G2 but significantly increased in mRCC tissue when compared with normal kidney tissue (Figure 4B). Immunoblotting, which was performed to verify ELISA data, showed similar SERPINA3 expression levels in ccRCC tissue samples (Figure 4C). To correlate SERPINA 3 expression in patient tissue samples, we performed IHC, which showed strong signals and different positive cell numbers as the tumor grade increased. At T1G2, a higher number of positive cells were observed when compared with the normal group; however, at T1G3, stronger SERPINA 3 expression was observed when compared with T1G2. At T3G2, we observed increased positive cell numbers when compared with T1G3. Then, in the mRCC group, the strongest positive reactivity and the highest number of positive cells were observed (Figure 4D).
Validation results in ccRCC tissue – overview. (A) Survival features in kidney renal clear cell carcinoma patients with the database for expression level of SERPINA3 (B) Relative SERPINA3 expression ratios in groups compared with controls by enzyme-linked immunosorbent assay. (C) Immunoblots showing SERPINA3 expression. (D) Immunofluorescence SERPINA3 images in RCC tissue (Control, T1G2, T1G3, T3G2, T3G3, and mRCC). Blue: DAPI; red: SERPINA3. p=0.001.
Discussion
RCC patients can be treated by partial or complete nephrectomy when the disease is not metastatic. Although most common kidney cancers have an asymptomatic clinical course at early stages, 20%-30% of patients will be metastatic at diagnosis (37). mRCC is a highly treatment-resistant malignancy, with surgery and chemotherapy having limited or no effectiveness, thus making a cure difficult (38). Also, more comprehensive strategies, such as gene-expression-array and proteome analysis for new treatment drugs and methods are required given the poor prognosis of advanced RCC. Additionally, an early RCC diagnosis is crucial for metastasis, which means better RCC prognostic techniques are warranted (39, 40).
In this study, we identified and quantified 2,608 and 2,463 proteins, respectively, in ccRCC tissue. When mRCC stage protein levels were compared to T3G3 stage, 147 DEPs were identified from the volcano plot, including 95 and 52 up- and down-regulated proteins, respectively (Figure 2). We also examined highly expressed DEPs distributions among T3G3 and mRCC stages. Among DEPs, some proteins belonged to the serine-protease inhibitor (SERPIN) superfamily, with SERPINA3 having higher expression ratios when compared to other SERPIN proteins. SERPINA3, as a matricellular acute-phase glycoprotein, primarily functions as a protease inhibitor which maintains cellular homeostasis (41-43). Serine proteases and their inhibitors are frequently associated with numerous malignancies and stromal cell functions which promote primary tumor formation. Critically, many different diseases and cancers are linked to SERPINA3 (44, 45).
Using immunoblotting, indirect ELISA, and IHC, we verified that SERPINA3 expression was significantly increased in mRCC tissue (Figure 4B-D). Previous studies also identified associations between SERPINA3 and cancer prognosis; higher disease stages and lower differentiation were linked to SERPINA3 up-regulation in endometrial cancer (46). Clinical parameters, including pathological grade, stage, lymph node metastasis, and vascular invasion, were positively associated with SERPINA3 expression in patients with endometrial cancer (46, 47). SERPINA3 was also positively correlated with a poor prognosis in patients with acute leukemia, breast, and pancreatic cancer, and overall patient survival in gastric cancer patients (48-51). Also, increased SERPINA3 levels were associated with tumor size in HLA-positive cervical carcinoma (52). SERPINA3 is also used as a prognostic and inflammatory marker in many diseases and is crucial in predicting predict survival in different tumors (53-55).
Additionally, SERPINA3 appears to have cancer and compartment-specific biological functions, acting in different cancers as either a tumor promoter or suppressor (43). In recent years, several studies reported SERPINA3 functions in cancer progression. For instance, SERPINA3 was found overexpressed in astroglia/microglia co-cultured glioblastoma stem-like cells and diminished glioma invasion upon silencing, while SERPINA3 also enhanced epithelial-mesenchymal transition (EMT), migration, and invasion in glioblastoma (56, 57). SERPINA3 incidence was also increased in advanced gastric cancer stages when compared with early stages, it contributes to the invasion of the protein (58). Breast cancer tissues showed higher SERPINA3 levels when compared with normal adjacent tissues, while EMT marker expression was increased in triple-negative breast cancer (TNBC) tissue due to SERPINA3 overexpression, showing increased cell migration, invasion, and EMT processes (49). However, SERPINA3 mechanisms of function and effect remain physiologically and pathologically obscure in ccRCC.
In our study, based on a comprehensive proteomics and experimental approach, we successfully showed that SERPINA3 was significantly increased in mRCC compared to stages before metastasis. Further research on SERPINA3 actions is required, but it appears that SERPINA3 is involved in metastatic ccRCC progression as a tumor regulator in kidney and other cancers.
Acknowledgements
This work was supported by a Biomedical Research Institute grant, Kyungpook National University Hospital (2021).
Footnotes
Data and Material Availability
MS raw data are available at ProteomeXchange using the identifier PXD039204.
Conflicts of Interest
The Authors declare no conflicts of interest.
Authors’ Contributions
Conceptualization: Park, J., Lee, S. and Lee, J.N.; Methodology and analysis: Park, J., Lee, E.H., Na, A., Choi, S.Y., and Lee, S.; Resources: Ha, Y., Kwon, T.G. and Lee, J. N.; Data Curation: Park, J., Lee, E.H., Chung, J., and Lee, S.; Writing: Park, J., Lee, E.H., Lee, S. and Lee, J.N.; Funding Acquisition: Lee, J.N.
- Received April 12, 2023.
- Revision received August 5, 2023.
- Accepted August 7, 2023.
- Copyright © 2023, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.