Fig 1.
The workflow for urinary protein biomarker discovery.
(A) urine samples were collected from participants and stored at -8000b0c until analysis. (B) urine samples selected for study were pooled, concentrated and depleted for highly-abundant protein removal using an immunodepletion kit, followed by in-solution tryptic digestion and tandem mass spectrometry. Protein identification and database searching were conducted using MaxQuant software (version 1.6.1.0). (C) Biomarker qualification for selected candidates used a custom MRM-MS assay. Data processing was performed using the Skyline and R programs. The statistical analysis software “SIMCA” and “SPSS" were used to perform the univariate/multivariate statistical analysis. (D) Immunohistochemical staining techniques were used to validate potential biomarkers on TMA.
Table 1.
The clinical and demographic information on all subjects for the discovery and qualification groups.
Fig 2.
Venn diagram represents the overlap among protein sample types with 338 identified proteins.
Fig 3.
Biomarker qualification and multivariate analysis.
(A) and (B) PCA and O-PLS-DA score plots of MRM results of urinary candidates that show sample differentiation; normal pathology (green), PDF (blue) and CCA group (red).
Table 2.
The correlation coefficient values of multivariate analysis and the differences of univariate analysis of all qualified urine protein candidates using O-PLS-DA analysis.
Fig 4.
The protein interaction network analysis of 27 significant candidate proteins using the STRING database (http://string-db.org).
A; two main cellular component pathways were identified in this study: lysosome biogenesis (blue circle) and membrane part (red circle). Each node lists the gene name of the candidate according to protein ID from Table 2. The different intensity of lines represents the protein association of confidence.
Fig 5.
Immunohistochemical staining of three candidate proteins.
LAMP1, LAMP2 and CDHR2, was performed on cadaveric donor liver tissues (the first column) and human CCA microtissue arrays which demonstrated low and high expression (the second and third column). The red arrows indicate the positive of LAMP1 and LAMP2 expression at the luminal surface (red arrows).
Table 3.
Clinico-pathological data and urinary candidate expression.