Proteome analysis of formalin-fixed paraffin-embedded colorectal adenomas reveals the heterogeneous nature of traditional serrated adenomas compared to other colorectal adenomas

Proteomic analysis of the urothelial cancer landscape

Urothelial bladder cancer (UC) has a wide tumor biological spectrum with challenging prognostic stratification and relevant therapy-associated morbidity. Most molecular classifications relate only indirectly to the therapeutically relevant protein level. We improve the pre-analytics of clinical samples for proteome analyses and characterize a cohort of 434 samples with 242 tumors and 192 paired normal mucosae covering the full range of UC. We evaluate sample-wise tumor specificity and rank biomarkers by target relevance. We identify robust proteomic subtypes with prognostic information independent from histopathological groups. In silico drug prediction suggests efficacy of several compounds hitherto not in clinical use. Both in silico and in vitro data indicate predictive value of the proteomic clusters for these drugs. We underline that proteomics is relevant for personalized oncology and provide abundance and tumor specificity data for a large part of the UC proteome ( www.cancerproteins.org ).

Integrated meta-analysis of colorectal cancer public proteomic datasets for biomarker discovery and validation

The cancer biomarker field has been an object of thorough investigation in the last decades. Despite this, colorectal cancer (CRC) heterogeneity makes it challenging to identify and validate effective prognostic biomarkers for patient classification according to outcome and treatment response. Although a massive amount of proteomics data has been deposited in public data repositories, this rich source of information is vastly underused. Here, we attempted to reuse public proteomics datasets with two main objectives: i) to generate hypotheses (detection of biomarkers) for their posterior/downstream validation, and (ii) to validate, using an orthogonal approach, a previously described biomarker panel. Twelve CRC public proteomics datasets (mostly from the PRIDE database) were re-analysed and integrated to create a landscape of protein expression. Samples from both solid and liquid biopsies were included in the reanalysis. Integrating this data with survival annotation data, we have validated in silico a six-gene signature for CRC classification at the protein level, and identified five new blood-detectable biomarkers (CD14, PPIA, MRC2, PRDX1, and TXNDC5) associated with CRC prognosis. The prognostic value of these blood-derived proteins was confirmed using additional public datasets, supporting their potential clinical value. As a conclusion, this proof-of-the-concept study demonstrates the value of re-using public proteomics datasets as the basis to create a useful resource for biomarker discovery and validation. The protein expression data has been made available in the public resource Expression Atlas.

Proteomic Profiling of Colorectal Adenomas Identifies a Predictive Risk Signature for Development of Metachronous Advanced Colorectal Neoplasia

BACKGROUND & AIMS

Colonic adenomatous polyps, or adenomas, are frequent precancerous lesions and the origin of most cases of colorectal adenocarcinoma. However, we know from epidemiologic studies that although most colorectal cancers (CRCs) originate from adenomas, only a small fraction of adenomas (3%-5%) ever progress to cancer. At present, there are no molecular markers to guide follow-up surveillance programs.

METHODS

We profiled, by mass spectrometry-based proteomics combined with machine learning analysis, a selected cohort of formalin-fixed, paraffin-embedded high-grade (HG) adenomas with long clinical follow-up, collected as part of the Danish national screening program. We grouped subjects in the cohort according to their subsequent history of findings: a nonmetachronous advanced neoplasia group (G0), with no new HG adenomas or CRCs up to 10 years after polypectomy, and a metachronous advanced neoplasia group (G1) where individuals developed a new HG adenoma or CRC within 5 years of diagnosis.

RESULTS

We generated a proteome dataset from 98 selected HG adenoma samples, including 20 technical replicates, of which 45 samples belonged to the nonmetachronous advanced neoplasia group and 53 to the metachronous advanced neoplasia group. The clear distinction of these 2 groups seen in a uniform manifold approximation and projection plot indicated that the information contained within the abundance of the ∼5000 proteins was sufficient to predict the future occurrence of HG adenomas or development of CRC.

CONCLUSIONS

We performed an in-depth analysis of quantitative proteomic data from 98 resected adenoma samples using various novel algorithms and statistical packages and found that their proteome can predict development of metachronous advanced lesions and progression several years in advance.

Normics: Proteomic Normalization by Variance and Data-Inherent Correlation Structure

Several algorithms for the normalization of proteomic data are currently available, each based on a priori assumptions. Among these is the extent to which differential expression (DE) can be present in the dataset. This factor is usually unknown in explorative biomarker screens. Simultaneously, the increasing depth of proteomic analyses often requires the selection of subsets with a high probability of being DE to obtain meaningful results in downstream bioinformatical analyses. Based on the relationship of technical variation and (true) biological DE of an unknown share of proteins, we propose the “Normics” algorithm: Proteins are ranked based on their expression level–corrected variance and the mean correlation with all other proteins. The latter serves as a novel indicator of the non-DE likelihood of a protein in a given dataset. Subsequent normalization is based on a subset of non-DE proteins only. No a priori information such as batch, clinical, or replicate group is necessary. Simulation data demonstrated robust and superior performance across a wide range of stochastically chosen parameters. Five publicly available spike-in and biologically variant datasets were reliably and quantitively accurately normalized by Normics with improved performance compared to standard variance stabilization as well as median, quantile, and LOESS normalizations. In complex biological datasets Normics correctly determined proteins as being DE that had been cross-validated by an independent transcriptome analysis of the same samples. In both complex datasets Normics identified the most DE proteins. We demonstrate that combining variance analysis and data-inherent correlation structure to identify non-DE proteins improves data normalization. Standard normalization algorithms can be consolidated against high shares of (one-sided) biological regulation. The statistical power of downstream analyses can be increased by focusing on Normics-selected subsets of high DE likelihood.

•

Normics is a tool for the normalization of proteomic data based on existing algorithms.

•

Specifically addresses data with high shares of differential expression.

•

Combines variance and data-inherent correlation structure.

•

Provides a ranking of differential expression likelihood.

•

Enables normalization based on the most stable proteins.

Serrated Colorectal Lesions: An Up-to-Date Review from Histological Pattern to Molecular Pathogenesis

Until 2010, colorectal serrated lesions were generally considered as harmless lesions and reported as hyperplastic polyps (HPs) by pathologists and gastroenterologists. However, recent evidence showed that they may bear the potential to develop into colorectal carcinoma (CRC). Therefore, the World Health Organization (WHO) classification has identified four categories of serrated lesions: hyperplastic polyps (HPs), sessile serrated lesions (SSLs), traditional serrated adenoma (TSAs) and unclassified serrated adenomas. SSLs with dysplasia and TSAs are the most common precursors of CRC. CRCs arising from serrated lesions originate via two different molecular pathways, namely sporadic microsatellite instability (MSI) and the CpG island methylator phenotype (CIMP), the latter being considered as the major mechanism that drives the serrated pathway towards CRC. Unlike CRCs arising through the adenoma-carcinoma pathway, APC-inactivating mutations are rarely shown in the serrated neoplasia pathway.

Evolving pathologic concepts of serrated lesions of the colorectum

Here, we provide an up-to-date review of the histopathology and molecular pathology of serrated colorectal lesions. First, we introduce the updated contents of the 2019 World Health Organization classification for serrated lesions. The sessile serrated lesion (SSL) is a new diagnostic terminology that replaces sessile serrated adenoma and sessile serrated polyp. The diagnostic criteria for SSL were revised to require only one unequivocal distorted serrated crypt, which is sufficient for diagnosis. Unclassified serrated adenomas have been included as a new category of serrated lesions. Second, we review ongoing issues concerning the morphology of serrated lesions. Minor morphologic variants with distinct molecular features were recently defined, including serrated tubulovillous adenoma, mucin-rich variant of traditional serrated adenoma (TSA), and superficially serrated adenoma. In addition to intestinal dysplasia and serrated dysplasia, minimal deviation dysplasia and not otherwise specified dysplasia were newly suggested as dysplasia subtypes of SSLs. Third, we summarize the molecular features of serrated lesions. The critical determinant of CpG island methylation development in SSLs is patient age. Interestingly, there may be ethnic differences in BRAF/KRAS mutation frequencies in SSLs. The molecular pathogenesis of TSAs is divided into KRAS and BRAF mutation pathways. SSLs with MLH1 methylation can progress into favorable prognostic microsatellite instability-positive (MSI+)/CpG island methylator phenotype-positive (CIMP+) carcinomas, whereas MLH1-unmethylated SSLs and BRAF-mutated TSAs can be precursors of poor-prognostic MSI-/CIMP+ carcinomas. Finally, based on our recent data, we propose an algorithm for stratifying risk subgroups of non-dysplastic SSLs.