Stephen Michnick talks to James Mitchell Crow about communication at the genomic level.

Stephen Michnick

Stephen Michnick is professor of biochemistry at the University of Montreal, Canada, and a member of the editorial board of Molecular BioSystems. His research in the area of chemical, structural and genome biology focuses on developing approaches to understand biochemical networks in cells.

 

Who inspired you to become a scientist?

My father - he had an infectious curiosity about things. He took any opportunity for a science lesson and I was brought up interested in science.

What are you working on at the moment?

We're interested in how regulatory and protein communication networks work in living cells. At the genome level, we are looking at protein complexes and how these are affected by physical or chemical perturbations. Assembly, disassembly or relocalisation of protein complexes reflects regulatory processes that underlie cellular adaptations to perturbations, including changing gene expression, metabolic pathway flux or cell shape or movement caused by reorganisation of the cytoskeleton. There are some simple mechanistic pictures of these adaptive responses but it is not entirely clear how these processes are orchestrated.  We are trying to systematise the process of discovering how, when, where and what drives the assembly of protein complexes that underlie these adaptations.

How does your protein-fragment complementation assay (PCA) work?

PCA is a strategy to detect protein-protein interactions in biological systems. It takes advantage of nature's invention called protein folding. Since a linear polypeptide has all the chemical information necessary to fold into a three- dimensional structure, in principle, if you split the polypeptide into two pieces and throw them together in solution they might fold up, but generally that doesn't happen.

PCA is based on recreating the unimolecular process of folding by genetically fusing two proteins that interact with the two fragments. If these proteins interact and bring the fragments together, you recreate the unimolecular folding conditions and the two fragments spontaneously fold into the unique three-dimensional structure that the intact polypeptide would have formed. If the physical interaction of the two proteins fused to the fragments is absolutely necessary for folding to occur and the protein that folds catalyses a detectable reaction, then you have a reporter for the protein-protein interaction. We and others have developed a number of different PCAs that allow us to easily address how protein complexes assemble in living cells.

How do you know that it's not the two fragments that you've attached drawing it together, and not the proteins that you're looking at?

The practical answer is to do control experiments because, in principle, that could happen. Once you create a fusion of a protein it's like you've created a monster - you don't know what the consequences are going to be! From our study on the whole genome scale, we have determined how often this happens and the answer is not very often.

The philosophical answer is what's the difference between binding and folding? They are similar things and we don't really know much about how folding happens at all.  So, to talk about attraction between fragments carries little meaning for me. I can see two molecules, whether they are small organic molecules or even proteins or nucleic acids, coming together as a linear process, banging together until they find the correct orientations of complementary surfaces to form a complex. But it's hard to imagine that with folding, as I don't see how a simple collision between two unfolded peptides would lead to some interaction. You could imagine that the fragments are already folded or partially folded and they're docking together. But we don't picture it that way and physical evidence suggests that they remain unfolded until they're brought together.

What will be the next big thing in your field?

An exciting area is the development of biosensors to detect rapid changes in enzyme activities, particularly those mediated by post-translational modifications. We look at physical interactions and how they change but we can't pick up these other events. I'm looking forward to developments of biosensors that will enable us to detect these chemical transformations.

I am interested in the link between what I call 'emergency responses,' how a cell responds immediately to a sudden change in environment, versus adaptive responses, those that occur in the gene expression program. For example, let's say you starve the cell for an essential nutrient. The cell has the enzymes to make this nutrient, but they are expressed at too low a level. The cell can't sit around and wait for the expression of the genes for these enzymes to be turned up. Maybe the existing enzymes are modified to increase their activity by chemosensory signalling pathways. However, we don't know how the chemosensory signal causes the reprogramming of the biochemical pathway that controls the synthesis of a nutrient and how this is done so quickly. Development of assays that detect these processes would help us to figure this out.

What's the secret to becoming a successful scientist?

To contribute to the biological sciences it helps if you develop specific interests in biological questions; you can't impose your views as a chemist on what are important biological questions but your training can help you to devise novel approaches to solve important problems. If you find a way to do something that nobody else has been able to do then you'll find it extraordinarily satisfying.

Will drug discovery ever become a more predictable science?

Everything we do makes it a little more predictable - but in incremental steps. We are already good at picking targets and developing inhibitors or screening for compounds with desired effects, like killing cancer cells. We are not yet able to predict unintended affects of molecules on cells or humans. The problem, as always, is one of biological complexity. Even if we developed a magic technique that allowed us to identify all targets of a compound, we may find that they interact with hundreds of different things; we have to delve deeper into the biology. If we get better at predicting biological responses to small molecules our search for drugs may be more successful.

If you weren't a scientist, what would you be?

I can't imagine living in the 21st century and not being a scientist, but I've thought of writing science fiction in the future. I've always admired scientists who are science fiction writers. My favourite is Gregory Benford because he takes real theories and turns them into dramatisations of things that could happen - like time travel and black holes coming out of cyclotrons.