Research projects in Human Protein Atlas

In addition to the high throughput protein profiling core project, several projects with more specific research objectives are run based on the resources generated within the Human Protein Atlas. Below is a short description of such selected projects.

Cancer biomarkers

Anca Dragomir, Margret Agnarsdottir, Cecilia Lindskog, Fredrik Pontén
Collaborative projects with Per-Henrik Edqvist (UCAN, UU), Karin Jirström (MAS), Patrick Micke (UAS), Johan Botling (UAS), Irina Alafuzoff (UAS), Michael Bergqvist (UAS), Anja Smiths (UAS), Dan Hellberg (Falu lasarett), Lars Holmberg (ROC), Monica Nistér (KI), Jutta Huvila (Turku University), Olli Carpén (Turku University), Irma Fredriksson (KI/KS), Gerrit Meijer (VUMC, Amsterdam), Meike DeWit (VUMC, Amsterdam), Anna Dimberg (UU), Liam Gallagher (UCD, Dublin) and Mathias Uhlén (SciLifeLab).

In several projects the aim is to further analyze the role of proteins identified as potential cancer biomarkers in the screening effort performed within the Human Protein Atlas project. Tumor material from well defined patient cohorts with tumors representing all major forms of human cancer are being collected and assembled into tissue microarrays. In addition to tumor material, clinical data is also collected to create databases allowing for testing and validation of protein expression patterns of importance for diagnostics, prognostics and functional tumor biology studies. There are special emphasis on i) colorectal cancer for the identification of markers that can stratify patients into groups of high or low risk for recurrent disease, ii) breast cancer in young women based on a large national cohort and extensive clinical database of >1000 patients where the focus is to understand why this patient group has such poor prognosis, and iii) on melanoma markers where image analysis algorithms are being explored for a more reproducible and objective assessment of immunohistochemical signals in melanoma samples. Other collaborative biomarker projects include lung cancer, endometrial cancer, high and low grade gliomas, cervical cancer and prostate cancer.

Antibody validation, performance and characterization

Cecilia Lindskog Bergström, Evelina Sjöstedt, Fredrik Pontén

Collaborative projects with Ulf Landegren (UU), Karl Andersson/Lars Gedda (UU) Tobias Sjöblom (UU), Cecilia Williams (University of Houston) and Mathias Uhlén (SciLifeLab).

The objective of this technical development project is to develop assays and strategies that can validate and verify protein expression data obtained using antibodies in cells and tissues, i.e. immunohistochemistry. A number of projects have been initiated aimed at generating more information on selected antibodies used in the Human Protein Atlas project, and to use existing detection techniques or to develop new techniques for antibody-based simultaneous studies of multiple proteins. For instance, methods for simultaneous detection of multiple epitopes on a common antigen are developed. Primarily, enzyme based, light microscopical immunohistochemistry and immunofluorescence is used, but also proximity ligation. In addition, in situ hybridization and immunohistochemistry on consecutive sections are also explored as a tool for antibody validation. These techniques have different advantages and disadvantages regarding sensitivity, histological and intra-cellular resolution, implementation in high-throughput evaluations, etc. For technical antibody validation transient knock-down of expression, real-time measurements of antibody binding as well as simultaneous stainings using multiple antibodies are employed.

Correlation studies on global RNA transcript and protein expression levels in tissues and organs

Cecilia Lindskog Bergström, Evelina Sjöstedt, Fredrik Pontén

Collaborative project with Emma Lundberg (SciLifeLab), Linn Fagerberg (ScilIfeLab), Åsa Sivertsson (KTH), Frida Danielsson (SciLifeLab) and Mathias Uhlén (SciLifeLab).

The enormous amount of IHC data now available on tissues and cell lines within the HPA project enables large-scale comparative studies of RNA and corresponding protein levels on a global level. Methods include cell microarrays and immunohistochemistry combined with image analysis and bioinformatic tools to compare with RNA data from a range of tissues and cell lines analyzed for transcription profiles using RNAseq. Cell lines are suitable samples for comparative analysis of mRNA and protein levels, since each cell line constitutes a homogenous collection of cells and unlike tissue does not represent a range of different phenotypes. Preliminary data from 17 cell lines, analyzed with RNAseq as well as with IHC using approximately 10 000 antibodies, revealed that 18% of the antibodies generated IHC staining patterns correlating with RNA with a Pearson correlation coefficient of >0,5. This antibody-based proteomics approach also aims to elucidate the protein signatures inflicted by in vitro growth conditions and map cell type specific protein signatures to understand how different cell lines can be utilized as model systems to study different tumor types.

Novel diagnostic tools for determining the origin of cancer metastases

Julia Bergman, Fredrik Pontén

Collaborative project with Karin Jirström (MAS) and Mathias Uhlén (SciLifeLab).

The use of antibodies that target proteins that are tissue- or cell type specific are crucial diagnostic tools in clinical pathology where they are used in immunohistochemistry-applications for the characterization of cancer. Such specific diagnostic antibodies can be used to determine from which original tissue the cancer has developed and to sub-classify the tumor type. The vast amount of data in the Human Protein Atlas is screened for cell and tissue specific proteins. Identified candidates are further validated and characterized for sensitivity and selectivity of specific target binding. Selected antibodies are used to analyze the expression pattern in a large TMA (over 900 cases) containing mainly metastases and primary tumor tissue from tumor types where additional diagnostic markers are needed. The aim is to find and define panels of diagnostic markers to be used in clinical pathology.

Protein profiling using highly characterized antibodies towards cancer proteins

Cecilia Lindskog Bergström, Fredrik Pontén

Collaborative project with Gordon Whitely, Stephen Hewitt (both NCI-CPTC program) and Mathias Uhlén (SciLifeLab).

In an effort to generate highly characterized monoclonal antibodies towards proteins suggested to be involved in cancer growth and spread, the NCI initiated the CPTC program to drive the development of a central community core that would help accelerate biomarker discovery and validation, cancer diagnostics development, and therapeutics monitoring. As part of this effort, CPTC antibodies are tested and used for protein profiling using Human Protein Atlas strategies including immunohistochemistry and immunofluorescence.