Elsevier

Oral Oncology

Volume 103, April 2020, 104592
Oral Oncology

Increased DSG2 plasmatic levels identified by transcriptomic-based secretome analysis is a potential prognostic biomarker in laryngeal carcinoma

https://doi.org/10.1016/j.oraloncology.2020.104592Get rights and content

Highlights

  • We provide evidence that DSG2 gene overexpression is associated with LSCC metastasis and survival.

  • We confirmed high circulating levels of DSG2 in plasma from LSCC patients.

  • Tumor and plasma DSG2 have the potential to predict disease, survival and/or tumor progression.

Abstract

Objectives

The tumor secretome deconvolution is a promising strategy to identify diagnostic and prognostic biomarkers. Here, transcriptomic-based secretome analysis was performed aiming to discover laryngeal squamous cell carcinomas (LSCC) biomarkers from potentially secreted proteins (PSPs).

Material and Methods

The tumor expression profile (35 LSCC biopsies compared with surrounding normal tissues - SN) revealed 589 overexpressed genes. This gene list was used for secretome analysis based on laryngeal tumors and related secretome databases.

Results

Forty-nine (Laryngeal tumor secretome database) and 50 (Human Protein Atlas and Cancer Secretome Database) PSPs presented an association with worse overall survival. Specifically, DSG2 overexpression was strongly correlated with poor survival and distant metastasis. DSG2 increased expression was confirmed in the LSCC dataset (LSCC = 111; SN = 12) from TCGA. A significant association between shorter survival and DSG2 overexpression was also detected. In an independent cohort of cases, we analyzed and confirmed high protein levels of DSG2 in plasma from LSCC patients.

Conclusion

A set of PSPs including the circulating DSG2, were associated with shorter overall survival in LSCC. DSG2 overexpression was also correlated with distant metastasis. The high plasmatic protein levels of DSG2 suggest its potential to be tested in liquid biopsies and applied as prognostic biomarker of LSCC.

Introduction

Laryngeal cancer is the second most common malignancy among the head and neck squamous cell carcinomas (HNSC) [1]. Approximately 177,000 new cases are diagnosed worldwide every year [2], being laryngeal squamous cell carcinoma (LSCC) the histological type most frequently found [3]. The risk factors associated with LSCC are tobacco and alcohol consumption [4], [5]. Heavy smoker patients have 40-fold higher risk of developing LSCC, while the amount of drinks per day increases LSCC risk [6]. The alcohol usage after the diagnosis increases the mortality risk [7].

The treatment of LSCC patients consists, mainly, at early stages (I-II), in surgery or radiotherapy with high rates of organ preservation and survival [8]. At advanced stages (III-IV; ~40% of the cases), the patients are treated with an aggressive combination of surgery and radiotherapy or radio-chemotherapy and the survival rates are low and remained relatively unchanged for the last fifteen years [9], [10]. The use of molecular biomarkers to diagnose the disease at early stages could improve the prognosis and also could be helpful to predict patient survival or tumor progression.

Proteins or RNAs obtained from large-scale data analysis could be useful to identify biomarkers to selectively apply the most appropriate therapy, and consequently, improve the survival [11], [12]. In addition, decoding the signaling pathways involved in LSCC progression seems to be central to uncover novel potential biomarkers especially in the comparison of differentially expressed genes and proteins in LSCC versus normal laryngeal tissues [1]. Recently, Lapa et al (2019) reported a set of potential biomarkers mapped in 11q13 associated with LSCC prognosis and response to treatment, highlighting potential drug targets [13].

The evaluation of cancer secretome components, which comprises all macromolecules secreted by tumor cells, can lead to the development of new disease biomarkers [14], [15], [16]. Tumor secreted proteins could be easily monitored and detected in liquid biopsies of patients being efficiently applied to the clinical practice. Large-scale data, such as genomic and transcriptomic, can be used as an input for the selection of secretome components, for example, the tumor transcriptome coupled with bioinformatics tools allows the prediction of potentially secreted proteins (PSPs) translated by over-expressed genes [17], [18]. This approach gathers mRNAs and protein data in the same analysis to extract more reliable results. Previously, the circulating proteomic profiling of LSCC patients revealed a panel of 18 proteins, of which 14 were able to differentiate patients according to the lymph node metastasis status [19].

In this study, we used a transcriptome-based secretome analysis to select PSPs by tumor cells from 35 naïve-treatment LSCC patients. We identified a set of putative biomarkers, of which DSG2 showed to be strongly associated with distant metastasis and shorter survival in our internal and external (TCGA) datasets. We further confirmed the potential of DSG2 as a disease biomarker detected in liquid biopsies from an independent set of patients.

Section snippets

Internal data

Tumor tissues (N = 35) were obtained from patients who underwent biopsy or laryngectomy prior to chemoradiation treatment at A.C. Camargo Cancer Center, São Paulo, Brazil. Plasma samples (N = 19) were obtained from an independent group of untreated patients following the same selection criteria. The Human Research Ethics Committee from A.C. Camargo Cancer Center approved the study (Protocols 1608/11 and 1983/14), and all patients provided written informed consent prior to any tissue sample

Results

The transcriptomic analysis revealed 679 over-expressed probes, of which 634 were obtained after filtering (removal of duplicated probes or those non-annotated), and 580 had Ensemble annotation ID (Supplementary Table 1). We compared the 580 over-expressed genes found in our LSCC with 982 secreted proteins of LSCC cells described by Zhang et al (2014) [20]. From this list, we found 49 PSPs in our cases (Figure 2A; Table 2). Interactions among proteins of extracellular matrix (collagens and

Discussion

We used the gene expression profile to identify novel PSPs encoded by over-expressed genes in LSCC tissues. The results improved our knowledge of biomarkers detected in LSCC secretome and could be easily applied in the clinical practice for screening and prognosis prediction.

Two in silico strategies of transcriptome-based secretome analyses applied in LSCC revealed 49 and 50 PSPs. Comparing the 580 over-expressed genes of our internal dataset with a previously published secretome analysis of

Role of the funding source

This study was supported by grants from the National Institute of Science and Technology in Oncogenomics (INCITO - Fundação de Amparo à Pesquisa do Estado de São Paulo - FAPESP 2008/57887–9 and Conselho Nacional de Desenvolvimento Científico e Tecnológico - CNPq 573589/08–9). SS Cury received a fellowship from FAPESP (2017/21223–9) and Danish Government Scholarship under the Cultural Agreements Programme. RML Lopez received a fellowship from CNPq (#153142/2012–0).

Acknowledgements

The results shown here are in part based upon data generated by the TCGA Research Network: http://cancergenome.nih.gov/. We would like to thank the organizations that funded our research: FAPESP, CNPq and the Danish Ministry of Higher Education and Science.

Competing interests

The authors declare no competing interests.

Author Contributions

Conceptualization: SSC and SRR; methodology: SSC, RMLL, FAM and JBH; sample collection: LPK, GBC; histopathological analysis: MACD and CALP; data analysis: SSC, RMLL, FAM.; data validation: SSC and JBH; formal analysis: SSC, RFC, SRR; resources: SRR and LPK.; data curation: SRR, RFC, LPK; writing-original draft preparation: SSC and SRR; writing-review and editing: all authors; supervision: SRR and RFC; project administration: SRR; funding acquisition: SRR and LPK.

Availability of data

This study was performed in part by reanalyzing available gene expression data from Gene Expression Omnibus GSE123986.

References (52)

  • S.T. Mayne et al.

    Alcohol and tobacco use pre- and post-diagnosis and survival in a cohort of patients with early stage cancers of the oral cavity, pharynx and larynx

    Cancer Epidemiol Biomarkers Prev

    (2009)
  • Y. Iizuka et al.

    Recurrence patterns after postoperative radiotherapy for squamous cell carcinoma of the pharynx and larynx

    Acta Otolaryngol

    (2015)
  • C. Zhang et al.

    Organ preservation surgery for patients with T4a laryngeal cancer

    Eur Arch Otorhinolaryngol

    (2018)
  • Cahlon O, Lee N, Le Q, Kaplan M, Colevas D. Cancer of the Larynx. Third. California: Leibel and Phillips Textbook of...
  • C. Bradford et al.

    Biomarkers in advanced larynx cancer

    Laryngoscope

    (2014)
  • M.J. de Miguel-Luken et al.

    A genetic view of laryngeal cancer heterogeneity

    Cell Cycle

    (2016)
  • L. Feng et al.

    Systems biology analysis of the lung cancer-related secretome

    Oncol Rep

    (2018)
  • H. Tjalsma et al.

    Signal peptide-dependent protein transport in Bacillus subtilis: a genome-based survey of the secretome

    Microbiol Mol Biol Rev

    (2000)
  • F.M. Awan et al.

    Identification of circulating biomarker candidates for hepatocellular carcinoma (HCC): An integrated prioritization approach

    PLoS ONE

    (2015)
  • L. Cheng et al.

    SELDI-TOF MS profiling of serum for detection of laryngeal squamous cell carcinoma and the progression to lymph node metastasis

    J Cancer Res Clin Oncol

    (2008)
  • Y. Zhang et al.

    Identification of human tissue kallikrein 6 as a potential marker of laryngeal cancer based on the relevant secretory/releasing protein database

    Dis Markers

    (2014)
  • Peter J. Thul et al.

    A subcellular map of the human proteome

    Science

    (2017)
  • A. Feizi et al.

    Database tool HCSD: the human cancer secretome database

    Database (Oxford)

    (2015)
  • J. Starruß et al.

    Morpheus: a user-friendly modeling environment for multiscale and multicellular systems biology

    Bioinformatics

    (2014)
  • T. Metsalu et al.

    ClustVis: a web tool for visualizing clustering of multivariate data using Principal Component Analysis and heatmap

    Nucleic Acids Res

    (2015)
  • R. Aguirre-gamboa et al.

    SurvExpress: an online biomarker validation tool and database for cancer gene expression data using survival analysis

    PLoS ONE

    (2013)
  • Cited by (0)

    1

    These authors contributed equally.

    View full text