Phosducin-like 3 is a novel prognostic and onco-immunological biomarker in glioma: A multi-omics analysis with experimental verification

Abstract

Malignant glioma is the most frequent primary tumor of the central nervous system. PDCL3 is a member of the phosducin-like protein family, and its imbalance has been shown to be associated with several human diseases. However, the underlying role of PDCL3 in human malignant cancers, especially in malignant gliomas, is unclear. In this study, we combined public database analysis and experimental verification to explore the differential expression, prognostic value and potential functions and mechanisms of PDCL3. The results revealed that PDCL3 is upregulated in multiple cancers and acts as a potential prognostic biomarker of glioma. Mechanistically, PDCL3 expression is associated with epigenetic modifications and genetic mutations. PDCL3 may directly interact with the chaperonin-containing TCP1 complex, regulating cell malignancy, cell communication and the extracellular matrix. More importantly, the association of PDCL3 with the infiltration of immune cells, immunomodulatory genes, immune checkpoints, cancer stemness and angiogenesis suggested that PDCL3 may regulate the glioma immune landscape. Furthermore, PDCL3 interference also decreased the proliferation, invasion and migration of glioma cells. In conclusion, PDCL3 is a novel oncogene and can be adopted as a biomarker with value in assisting clinical diagnosis, predicting patient outcomes and assessing the immune landscape of the tumor microenvironment in glioma.

Keywords: PDCL3; glioma; immune landscape; prognostic biomarker; tumor microenvironment.

Figure 1

PDCL3 is upregulated in multiple cancers and predicts poor prognosis. (A) The expression of PDCL3 across cancers from the TIMER database. (B) The expression of PDCL3 across cancers based on TCGA and GTEx databases. (C) Forest plots showing the correlation between PDCL3 expression and overall survival across cancers. (D) Forest plots showing the correlation between PDCL3 expression and progression-free survival across cancers. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001.

Figure 2

Upregulated PDCL3 is a prognostic biomarker in glioma. (A–D) Stratification analysis of PDCL3 expression in four different datasets. (E–H) Kaplan-Meier curves showing the OS of the PDCL3 high and low groups in four different datasets. (I) PDCL3 mRNA expression in the cohort composed of 94 glioma patients. (J) Kaplan-Meier curves of this 94-patient cohort. (K) Representative pictures of IHC staining in the glioma tissue microarray cohort (scale bars of upper = 500 µm, scale bars of lower = 100 µm). (L) PDCL3 protein expression in the glioma tissue microarray (***p<0.001 and ns, no significance). (M) Kaplan-Meier curves of this microarray cohort.

Figure 3

Construction and evaluation of the nomogram. (A, B) Univariate and multivariate Cox regression analyses in the TCGA cohort. (C) Sankey diagram showing the overall prognostic trend and living status of the inner relationship. (D) Nomogram based on age, WHO grade, IDH status, 1p/19q codeletion and PDCL3 expression. (E) Calibration curves showed the concordance between predicted and observed 1-, 3-, and 5-year OS. (F) ROC curve analyses of the nomogram in predicting 1-, 3-, and 5-year OS.

Figure 4

Functional analysis of PDCL3 in glioma. (A) Volcano plot showing DEGs between the PDCL3 high and low groups in the TCGA-GBMLGG dataset. (B) Histogram showing the top 10 significant terms of BP, MF, CC and KEGG enrichment analysis. (C, D) Ridge plot showing the top 10 pathways of GSEA enrichment analysis, including KEGG and Reactome pathways. (E) The protein-protein interaction network of PROS1 was constructed using GeneMANIA. (F) The immunoprecipitation results were observed using silver staining. (G) Mass spectrometry identified the top 15 differentially expressed proteins between the PDCL3-IP group and the IgG-IP group.

Figure 5

Correlation of PDCL3 with immune cell infiltration in the glioma microenvironment. (A) The relationship between PDCL3 expression and immune cell infiltration across cancers was evaluated using the quanTIseq algorithm (*p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001; glioma cohort is shown in red squares). (B) The proportion of 10 immune cells in the PDCL3 high and low subgroups. (C) Differences in immune cell infiltration between the PDCL3 high and low subgroups in glioma were verified using TIMER. (D) Comparison of the immune score, stromal score and ESTIMATE score between the PDCL3 high and low subgroups. (E) The expression of 5 immune cell markers was detected in 12 glioma clinical specimens using IHC. The percentage of positive area was used to evaluate the degree of immune cell infiltration. (F) Representative IHC staining of five immune cell markers in the PDCL3 high and low groups (20X, scale bars = 100 µm).

Figure 6

Association between PDCL3 expression and immunomodulators and immune checkpoints. (A) Heatmap showing the correlation between PDCL3 expression and 150 common immunomodulators across cancers (*p<0.05; glioma cohort is shown in red squares). (B) Scatter diagram showing the correlation between PDCL3 expression and the top 5 relevant immunomodulators. (C) Heatmap showing the correlation between PDCL3 expression and 60 common immune checkpoint genes across cancers (*p<0.05; glioma cohort is shown in red squares). (D) Box plot showing the expression distribution of the top 10 relevant immune checkpoint genes (***p<0.001). (E–G) The correlation between PDCL3 and the 7 most important known immune checkpoints was visualized using chord plots in three independent datasets.

Figure 7

PDCL3 is associated with cancer stemness and angiogenesis. (A–F) Six different stemness indexes, including DNAss, EREG-METHss, DMPss, ENHss, RNAss and EREG.EXPss were calculated with PDCL3 expression in pan-cancer. (G–I) Three different algorithms, including xCELL, MCPcounter and EPIC, were used to evaluate the infiltration of endothelial cells in the PDCL3 high and low subgroups. (J) Immunofluorescence showed the distribution relationship of PDCL3, CD31 and Nestin in 3 GBM clinical specimens (60X, scale bars = 50 µm). (K) The expression of CD31 and α-SMA was detected in 12 glioma clinical specimens using IHC. The microvasculature per field was used to evaluate the relationship between PDCL3 expression and angiogenesis. (L) Representative IHC staining of CD31 and α-SMA in the PDCL3 high and low groups (20X, scale bars = 100 µm).

Figure 8

Experimental validation of PDCL3 function in glioma cell lines. (A, B) PDCL3 interference efficiency was detected using Western blotting and qRT-PCR (**p<0.01). (C) Heatmap revealing the 327 DEGs identified by transcriptome sequencing analysis. (D) Bubble diagram showing the results of GO and KEGG pathway enrichment. (E) Representative pictures of EdU, transwell and wound healing assays in U251 and U87-MG cells. (F–K) Statistical results of EdU, transwell and wound healing assays in U251 and U87-MG(*p<0.05, **p<0.01, ***p<0.001).

Figure 9

Verification of PDCL3 function and expression at the single-cell level. (A) Functional analysis of PDCL3 in glioma using the CancerSEA database. (B) The expression of PDCL3 in different cell types from 17 independent single-cell transcriptome sequencing datasets. (C, D) Two representative datasets, GSE163108_10X and GSE148842, show the expression and distribution of PDCL3, CD276, IL10RB, HLA-A, CXCL10 and TGFBR1.