This year, science celebrates mapping 90% of the human proteome, an endeavor that has been achieved by the Human Proteome Project (HPP). Technology Networks explored the journey to this successful feat by speaking with Chris Overall, professor and Canada research chair in proteinase proteomics and systems biology at the University of British Columbia, and chair of the C-HPP component of the HPP.
Molly Campbell: In your editorial piece, you talk about how HPP "minds the gap". Can you talk to us about why there is a 10% missing protein gap, and how the HPP looks to close the gap in the future?
Chris Overall (CO): Missing proteins may be rarely expressed proteins in rare cells, tissues and fetal/childhood development stage/time, or expressed in too low amounts that challenge sampling and analysis; some proteins are chemically or structurally not amenable to current mass spectrometry or require the most recent infrastructure to be analysed that is not generally available to most proteomics labs due to lack of funding directed to proteomics.
MC: How does it feel as a scientist to be involved in a project that contributes so largely to our understanding of human life?
CO: Incredibly exciting! This is all non-funded work in my lab, and in most labs that participated in defining the human proteome, but I feel it is so important a scientific and medical pursuit that we do it unfunded, in our own time. Genomics cannot provide all the answers or diagnostics for diseases lacking a genetic basis. Critically, genomics cannot provide information on disease activity and on-target drug activity. Only proteomics can do so.
For instance, proinflammatory cytokines and chemokines are precisely regulated in expression over time and cell/tissue location. Most chemokines (molecular beacons to control white blood cell migration and activation, are modified by PTMs or precise proteolytic processing that may remove one or a handful of amino acids and this occurs during the disease phase to activate, then inactivate chemokines and even to switch cell surface receptors to lead to a totally different signal in the target cells. Cleavage can also generate antagonists that activity prevent new signals reaching the cell. But all these proteoforms are generated from exactly the same gene and mRNA. Hence, knowing that information alone does not provide the temporal information on relative and absolute levels of these chemokine proteoforms.
Deciphering this by proteomics using the features of these proteoforms as biomarkers provides accurate and timely information and the active disease activity status the patient is experiencing. This is vital information to inform diagnosis, treatments and patient management. Thus, proteomics really holds the key to devising new accurate diagnostics for personalised medicine. These analytical approached can then be translated as the simpler ELISA and other tests suitable for deployment in hospitals, diagnostic labs and even bed side. I feel privileged to be involved in such a worthwhile endeavour and this drives me and my lab members. However, our work would progress so much faster and more accurately if funding was more available for essential infrastructure.
MC: How do you expect the data gathered through the HPP to impact the future of modern medicine?
CO: Proteins are the fabric of life. Our genomes just provide the instructions on how our proteins are weaved together to form this amazing complex tapestry, that defines our individuality. Our individuality in turn, defines how prone we may be to different diseases, aging and other stressors. Like clothing, our fabric is ever changing and adapting to the conditions when they change – just like we need different clothes in the different seasons or during the day versus the night. Some genes allow us to make such changes quickly, other genes adapt slower. In this case, "slow" is a method of control as consistency is needed to enable a stable platform for our bodies. It is to understand this pattern, and how it changes, that proteomics is so integral to understanding health and disease and for personalised medicine.
In addition, and separately, machine learning of the normal and pathological cellular expression and distribution of disease relevant protein proteoforms biopsy sample processing by histology will become more accurate and faster, providing more nuanced information that can be key to accurate, early and appropriate medical decision making.
Professor Chris Overall was speaking to Molly Campbell, Science Writer, Technology Networks.