Introduction
Lymphomas are a very prevalent form of cancer that can be classified into Hodgkin and non-Hodgkin lymphoma subtypes (HL and NHL, respectively). Diffuse large B-cell lymphoma (DLBCL) is the most common NHL subtype globally, accounting for 30–40% of overall NHL cases [
1]. DLBCL is a highly heterogeneous and aggressive disease that can exhibit highly varied outcomes in affected patients. Treatment of DLBCL patients with rituximab and cyclophosphamide-doxorubicin-vincristine-prednisone chemotherapy (R-CHOP) has led to rising long-term patient survival and a 50–75% 5-year survival rate for patients with this disease [
2]. However, roughly 40% of patients ultimately suffer from relapsed or refractory disease [
3]. The genetic and molecular etiology of DLBCL is also highly complex and some studies have focused on the identification of genetic drivers and their functional roles in DLBCL to determine novel therapeutic targets and/or diagnostic or prognostic biomarkers to promote the disease’s diagnosis and treatment [
4,
5].
High-throughput and next-generation sequencing (NGS) technologies have enabled many biomarker discovery studies in DLBCL samples to date [
6,
7]. Findings from the majority of these studies, however, have yet to be validated or implemented in clinical settings. Many of these prior studies have also been limited by factors such as tissue heterogeneity, small sample sizes, and single-cohort approaches, resulting in inconsistent results [
8]. One approach to overcoming these limitations is the re-analysis of multiple independent previously published NGS datasets in order to more reliably identify potential cancer-specific biomarkers [
9,
10].
Herein, we sought to identify novel DLBCL diagnostic, prognostic, or therapeutic biomarkers via an integrative bioinformatics approach. Through this strategy, we identified phospholipase A2 group VII (PLA2G7) as a novel biomarker of interest associated with this cancer type. PLA2G7 has previously been reported to control a range of oncogenic processes through mechanisms associated with metabolic alterations [
11]. PLA2G7 has also been identified as a prognostic biomarker in patients with prostate cancer (PCa) [
12] and melanoma [
13]. Together, our data suggest that PLA2G7 may function as a driver of the proliferation, migration, and survival of tumor cells and a regulator of immune cell infiltration within the DLBCL tumor microenvironment. PLA2G7 may therefore represent a viable therapeutic target in DLBCL patients.
Materials and methods
Ethics statement and clinical specimens
The study was approved by the Ethics Committee of Fujian Medical University Union Hospital. All experiments were performed according to the relevant regulations and written informed consent was obtained from patients. A total of 18 DLBCL tissues, 11 lymphadenitis tissues and 53 DLBCL peripheral blood samples were obtained from Fujian Medical University Union Hospital. Tissue samples were frozen in liquid nitrogen until further analysis. DLBCL cases only were confirmed by pathological examination of lymph node biopsy or lymphadenectomy according to the 2017 WHO Classification of Lymphoid Neoplasms. Patients samples used for this study had not received chemotherapy or radiotherapy before surgery.
DEG identification
The GSE32018 and GSE56315 datasets from the Gene Expression Omnibus (GEO) database (
http://www.ncbi.nlm.nih.gov/geo/) were downloaded. Raw CEL files were background corrected, subjected to z-score transformation and quantile normalization, and subjected to further analysis. The GSE32018 dataset [
14] contained 22 DLBCL tumor samples and 7 normal lymph node tissue samples and was prepared using the GPL6480 platform. The GSE56315 dataset [
15] contained 55 DLBCL tissue samples and 33 normal tonsil tissues and was prepared using the GPL570 platform.
In addition, gene expression profiles and clinical data pertaining to 48 DLBCL patients were downloaded from The Cancer Genome Atlas (TCGA) database (
https://portal.gdc.cancer.gov/). The ESTIMATE algorithm was used to calculate stromal and immune scores [
16].
Genes that were differentially expressed between DLBCL and control tissues were identified with the R ‘limma’ package using the following cutoff criteria: log2 FoldChange (FC) > 1 and adjusted
P-value < 0.05. The RobustRankAggreg (RRA) R package was then used to integrate the DEGs from these two datasets. RRA utilizes a probabilistic model to aggregate and monitor the genes that are ranked consistently better than expected under the null hypothesis of uncorrelated inputs, and defines a significance score for each gene [
17]. Cutoff criteria: score < 0.05 was applied.
Functional enrichment analysis
Identified DEGs were subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses [
18‐
20] using the DAVID database (
https://david.ncifcrf.gov/summary.jsp), with
P < 0.05 as the threshold of statistical significance.
Weighted gene co-expression network analysis (WGCNA)
Co-expression analyses were conducted using the WGCNA R package using TCGA data corresponding to 48 DLBCL patients for whom detailed survival data were available. A weighted adjacency matrix was constructed by determining Pearson’s correlation coefficients for individual pairs of genes. An appropriate soft power threshold was then selected to yield standardized scale-free networks, after which the resultant matrix was subjected to transformation into a topological overlap matrix (TOM). In addition, topological overlap dissimilarity (1-TOM) was utilized as input for hierarchical clustering analyses. Gene dendrogram and module identification were constructed with a dynamic tree cut. Any modules with a size < 30 were then deleted. Modules that had a dissimilarity of < 0.25 were merged, after which relationships between clinical traits and module eigengenes were assessed.
Cell culture and transfection
Human DB and SU-DHL-2 DLBCL cells were obtained from Procell Life Science and Technology and were grown in RPMI-1640 (Invitrogen, CA, USA) containing 10% fetal bovine serum (Gibco, CA, USA) at 37 °C in a 5% CO2 humidified incubator. Cells were transfected with Negative Control (NC) and PLA2G7 siRNAs using Lipofectamine 3000 (Invitrogen). The siRNA sequences were as follows: PLA2G7-si1: CCUGUUGCCCAUAUGAAAUTT,
AUGGUUAAUGUUUGCAGGCAT; PLA2G7-si2: CCUGGAUGCAUGGAUGUUUTT,
AAACAUCCAUGCAUCCAGGGC.
Quantitative real-time PCR
Trizol (Invitrogen) was utilized to extract RNA from cells, after which a cDNA synthesis kit (Thermo Fisher Scientific, USA) was used based on provided directions to prepared cDNA. The expression levels were detected by QRT-PCR analysis with FastStart Universal SYBR-Green Master (Roche), an ABI7500 sequence detector (Applied Biosystems, Foster City, CA, USA) and calculated by 2 − ΔΔCt method. β-actin expression was used for normalization purposes, and primers used in this study were as follows: PLA2G7, Forward (F): 5′-GAACACACTGGCTTATGGGC-3′, Reverse (R): 5′-GAGATGCCAGGTCAATGCCA-3′.
For colony formation assays, 500 cells were added per well in methylcellulose-based media. Plates were then cultured for 2 weeks, after which methanol was used to fix colonies which were subsequently stained with 0.5% crystal violet. All colonies containing > 50 cells were then counted via microscopy.
For migration assays, 24-well transwell chambers (8 μm aperture, BD Biosciences) were utilized. DLBCL cells were suspended in FBS-free media and were added to the upper well of these chambers, whereas cells in media supplemented with 10% FBS were added into the lower chamber. Following a 48 h incubation, cells that had not migrated to the lower chamber were removed with a cotton swab. Migrated cells were then fixed and stained using 0.5% crystal violet and imaged via microscopy. In total, five fields of view per sample were analyzed at random. ImageJ was utilized to quantify the number of migrated cells per well.
Apoptosis assay
A PE Annexin V Apoptosis detection kit (BD Pharmingen, USA) was utilized at 48 h post-transfection to analyze cells. Briefly, cells were stained using Annexin V-PE and 7-AAD for 15 min at room temperature. A BD Accuri C6 flow cytometer was then used to analyze samples, after which FlowJo was used for data analysis.
CCK-8 assay
Cells were seeded into 96-well plates and incubated in different concentrations of darapladib for 72 h, then CCK-8 dye solution was added to each well and incubated at 37 °C for 2 h. The absorbance was measured at 450 nm using a microplate reader (BioTek, USA).
Statistical analysis
ROC (receiver operating characteristic) curve was conducted to analyze the effectiveness of target gene expressions between tumor and healthy samples. Area under the ROC curve (AUC) values were calculated using SPSS 20.0 (SPSS Inc., IL, U.S.A.) and were used to assess the sensitivity and specificity of individual DEGs as diagnostic biomarkers between DLBCL and normal tissues.
Patients were separated into two groups based upon PLA2G7 expression levels, after which two-tailed chi-squared tests were utilized to compare clinicopathological characteristics between the PLA2G7-high and –low patient groups. In addition, comparisons of PLA2G7 expression levels were made via Student’s t-tests and one-way ANOVAs using GraphPad Prism 8. Kaplan-Meier survival curves were drawn to show the relationship between expression of PLA2G7 and overall survival (OS) of patients, which was tested by the log-rank test. P < 0.05 was the significance threshold for this analysis.
Discussion
DLBCL is a common and highly heterogeneous form of NHL associated with high morbidity and mortality rates [
26]. While roughly half of DLBCL cases can be cured via standard chemotherapy, treated relapsed or refractory DLBCL remains challenging [
27]. It is thus essential that prognostic and diagnostic biomarkers of DLBCL be identified in order to guide patient detection and treatment. Herein, we therefore sought to identify potential prognostic and diagnostic biomarkers of DLBCL. In total, we identified 30 and 38 genes that were significantly up- and down-regulated, respectively, in DLBCL samples from two GEO datasets. Functional enrichment analyses suggested that these DEGs were associated with angiogenesis, cell proliferation, the immune response, and other key tumor progression-related pathways. Of these 68 DEGs, five were identified as promising diagnostic biomarkers of DLBCL through ROC curve analyses.
We next utilized the TCGA database in order to conduct WGCNA analyses exploring the relationship between gene expression profiles and DLBCL patient prognosis. Through this approach, we identified a key gene module that was related to patient outcomes. PLA2G7 was a hub gene within this module, and also exhibited favorable diagnostic utility in the above ROC curve analyses. Prior research indicates that PLA2G7 is a tumor-associated macrophage-derived factor that plays a key role in the regulation of tumor cell migration. In nasopharyngeal carcinoma (NPC) cells, there is evidence that PLA2G7 can enhance tumor cell migration and survival [
28], and similar findings have also been observed in PCa cells [
11]. Further research has led to the identification of this gene as a biomarker of primary and metastatic PCa, and a viable therapeutic target in patients with ERG positive PCa [
29]. In the GTex and TCGA databases, PLA2G7 was found to be upregulated in 17 different tumor types. Consistent with prior studies of PCa and NPC, we also determined that PLA2G7 promoted DLBCL cell proliferation and migration while suppressing the apoptotic death of these cells.
The tumor microenvironment includes macrophages, dendritic cells, T helper cells, T cytotoxic cells, and reactive B lymphocytes. Shain et al. [
30] previously demonstrated that B cell tumor interactions with the local TME can influence tumor cell behavior by controlling the oncogenesis’s growth and progression. Among these components, tumor-associated macrophages (TAM) were found to play a major role. The data generated in the present study further suggest that PLA2G7 expression may be associated with DLBCL tumor stromal and immune scores. Kua et al. [
31,
32] reported that the CIBERSORT algorithm was used to analyze the DLBCL immune cell infiltration in the TME. As such, we explored the association between PLA2G7 expression and TIICs, revealing that patients expressing high levels of this gene also exhibited increased monocyte infiltration. Monocytes develop into macrophages in tumors, and are associated with poor prognosis in DLBCL patients due to IL-34 production [
33]. It is reported that the OS and PFS of DLBCL patients with high expression of TAMs are often poor, and there is a positive correlation with the peripheral absolute monocyte count (AMC) [
34]. Therefore, AMC is a useful prognostic marker that reflects the status of the tumor microenvironment (TME) in DLBCL. Increased tumor-infiltrating monocyte numbers may thus be responsible for the relationship between high PLA2G7 expression and poor DLBCL patient outcomes. Together, these data, therefore, offer insight into potential immunotherapeutic treatment strategies for this cancer type.
Notably, our study focused on the identification of PLA2G7 as a novel biomarker for DLBCL. However, it had some limitations. First, the sample size obtained from TCGA to validate the GEO data sets was not large, the data in TCGA about DLBCL lacked normal samples as controls and complete survival information was lacking in some cases. Future studies are required and should include a large sample size to validate such observations. Also, our study focused on bioinformatics methods and in vitro experiments to screen candidate genes for DLBCL and identified PLA2G7 as a novel biomarker, the functional role of PLA2G7 could be explored further to determine tumor cell migration using in vivo studies and a detailed mechanistic approach. Future studies are required and should include these approaches.
In summary, we herein identified PLA2G7 as a biomarker that is upregulated in DLBCL and that is related to the enhancement of DLBCL cell proliferation, invasion, and tumor microenvironmental composition. These findings suggest that PLA2G7 may not only be a diagnostic and prognostic biomarker of DLBCL, but also a viable therapeutic target for the improvement of patient outcomes.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.