Correlation of the tumor escape phenotype with loss of PRELP expression in melanoma

Background Despite immunotherapies having revolutionized the treatment of advanced cutaneous melanoma, effective and durable responses were only reported in a few patients. A better understanding of the interaction of melanoma cells with the microenvironment, including extracellular matrix (ECM) components, might provide novel therapeutic options. Although the ECM has been linked to several hallmarks of cancer, little information is available regarding the expression and function of the ECM protein purine-arginine-rich and leucine-rich protein (PRELP) in cancer, including melanoma. Methods The structural integrity, expression and function of PRELP, its correlation with the expression of immune modulatory molecules, immune cell infiltration and clinical parameters were determined using standard methods and/or bioinformatics. Results Bioinformatics analysis revealed a heterogeneous, but statistically significant reduced PRELP expression in available datasets of skin cutaneous melanoma when compared to adjacent normal tissues, which was associated with reduced patients’ survival, low expression levels of components of the MHC class I antigen processing machinery (APM) and interferon (IFN)-γ signal transduction pathway, but increased expression of the transforming growth factor (TGF)-β isoform 1 (TFGB1) and TGF-β receptor 1 (TGFBR1). In addition, a high frequency of intra-tumoral T cells directly correlated with the expression of MHC class I and PRELP as well as the T cell attractant CCL5 in melanoma lesions. Marginal to low PRELP expression levels were found in the 47/49 human melanoma cell lines analysis. Transfection of PRELP into melanoma cell lines restored MHC class I surface expression due to transcriptional upregulation of major MHC class I APM and IFN-γ pathway components. In addition, PRELP overexpression is accompanied by high CCL5 secretion levels in cell supernatant, an impaired TGF-β signaling as well as a reduced cell proliferation, migration and invasion of melanoma cells. Conclusions Our findings suggest that PRELP induces the expression of MHC class I and CCL5 in melanoma, which might be involved in an enhanced T cell recruitment and immunogenicity associated with an improved patients’ outcome. Therefore, PRELP might serve as a marker for predicting disease progression and its recovery could revert the tumorigenic phenotype, which represents a novel therapeutic option for melanoma. Supplementary Information The online version contains supplementary material available at 10.1186/s12967-023-04476-x.


Figure S1 :
Figure S1: PRELP expression in neoplastic and non-neoplastic human tissues and cells A. Heterogeneous expression pattern of PRELP across 123 human healthy tissues and subtissues The expression levels of PRELP given as TPM value were analysed using the Omic Horizon Expression database and presented as bar chart.B. Reduced PRELP expression in melanoma cell lines (n=48) using qRT-PCR.The PRELP expression of melanoma cell lines were compared to that of melanocytes.Using cycle of quantification/qualification (Cq) scale, the expression of melanoma cell lines were characterized for their PRELP expression.C. PRELP of tumor skin cutaneous melanoma expression independent of tumor grade, tumor size, gender and age.The data are presented as transcript per million.

Figure S2 :
Figure S2: Correlation plot of HLA-A and PRELP in 133 skin cancer cell lines (GENT2 -Skin cancer cell lines)

Figure S3 :
Figure S3: Reconstitution of PRELP expression in melanoma cells.Overexpression of PRELP in melanoma cell lines was obtained after transfection of a PRELP expression vector in murine and human PRELP low melanoma cells.Transfection with a mock vector served as control.PRELP mRNA expression was determined in PRELP low B16F10 and Buf1088 cells and their PRELP transfectants as described in Material and Methods.qPCR data are shown as bar charts and represent the mean of at least three independent experiments.The statistical significance is presented as * p < 0.05; *** p < 0.001.

Figure S4 :
Figure S4: Distribution of gene mutations of PRELP in melanoma.The Skin Cutaneous Melanoma (TCGA) dataset was analysed for genetic alterations in PRELP.

Figure S5 :
Figure S5: Association between PRELP expression, CD8 + T cell infiltration and activation as well as patients' survival A. Association between infiltration levels of CD8 + T cell subsets and the cumulative overall survival of tumor skin cutaneous melanoma and metastatic melanoma datasets with different PRELP expression levels.In tumor skin cutaneous melanoma, the prognostic signature was built on the immune infiltration of naive, central memory and effector memory CD8 + T cells and the expression of PRELP profiled by xCell algorithm.B. In metastatic melanoma, the prognostic signature built by the immune infiltration of naive, central memory and effector memory CD8 + T cells and the expression of PRELP profiled by xCell algorithm.

Figure
Figure S6: A, B, C. Correlation of PRELP high vs.PRELP low samples with the expression of CCL5 and overall survivalThe TCGA dataset (Tumor Skin Cutaneous Melanoma -TCGA) was analyzed regarding patients' overall survival using the Kaplan-Meier curve for CCL5 expression independent of PRELP expression (A), PRELP high (B) and PRELP low (C) samples.

Table S1 :
Primers used for qPCR analyses.

Table S2 :
Datasets used and number of samples analyzed.

Table S3 :
Correlation of PRELP expression with HLA class I and APM components using two different melanoma datasets.

Table S4 :
Exploring the association between PRELP and immune infiltrates with the clinical outcome in SKCM datasets.The multivariable Cox proportional hazard model was generated by using XCELL algorithm in the TIMER database.Each cell type of the table corresponds to an independent Cox model.The hazard ratio (HR) and p value for the Cox model were displayed.