AT A GLANCE:
NAME: Michel Desjardins
POSITION: Associate professor of pathology and cell biology, University of Montreal, since 1994.
Consultant for Caprion Pharmaceuticals.
Post-doc in cell biology, European Molecular Biology Laboratory, 1991-94.
PhD in biology, University of Montreal, 1991.
BS in biology, University of Quebec at Montreal, 1986.
How did you get involved with proteomics?
My interest already [started] during my post-doc at EMBL, when we were working on the basic idea that the function of a system in the cell — either an organelle or a pathway — is directed by the proteins that are present. At that time there were really only a handful of known proteins that were present on the phagosome, [which I was studying]. So we started the systematic characterization of phagosome composition. This led to a first publication, in [Journal of Biological Chemistry] in 1994, which used pre-proteomics approaches — 2D gels, Edman sequencing, co-migration, etc. We identified maybe two to three dozen proteins this way. What changed with what we now call proteomics is really the scale at which we can do this.
What has made our [present] work possible was meeting the mass spectrometrist Jérome Garin at a meeting in Paris in 1998. He said, ‘I’m a mass spectrometrist, and I would like to help you identify all the proteins present on the phagosome.’ This really changed what we’ve been doing in the lab since, and the speed at which we were able to do it.
We started with [Garin] using a 2D gel approach with MALDI TOF. Towards the end, he was one of the first to have a Q-TOF from Micromass. We were sending samples from Montreal. I was quite skeptical about what he was telling me was possible, because I realized in the previous years how long it took to identify just a few proteins and according to him we could identify everything in a much shorter time period. So when we did the first samples, it was something like 300 spots that had been cut from the gel, and I knew some of the spots — what they were — based on our previous work, and I didn’t tell him that I knew and I said, ‘you should analyze these first.’ I received an e-mail the following day, and he had identified the 10 spots or so that I had indicated and he was of course right for each of them. This was when I realized that we were facing a totally new ball game, and this was when we really started to get excited about proteomics. It changed our conception of the way were seeing the organelle that we were studying. It facilitated the prediction of new function because we knew, for example, that a protein is present on lipid rafts, which are at the cell surface — so could it be that lipid rafts are present on phagosomes? We investigated this using traditional cell biology and biochemical approaches, and we were right. And so very rapidly it was as if [we] were looking in a crystal ball — that the identifications were leading us to literally new biology.
Do you have your own mass spec facility now?
For two years in Montreal there has been a center called the Montreal Proteomics Network, based at McGill University. There, we currently have three mass specs and three more are coming. We have at the moment two Q-TOFs, one MALDI TOF, and an FT-MS is coming soon. The head of the facility is John Bergeron, a well-known cell biologist in Montreal. The facility is mainly dedicated to a project that is shared by four groups, called the Cell Map, which is the characterization of every organelle in the eukaryotic cell. John Bergeron is in charge of that.
Tell me about the Cell Map project.
We had success with phagosomes, and then Bergeron used a similar approach to study the Golgi apparatus. The idea is that instead of putting the whole cell in the mass spec, which doesn’t give us any information on the location of the identified protein, we decided to separate cells into their functional entities, which are organelles. And then we analyzed the organelles systematically using mass spec techniques and different machines and we tried to identify as comprehensively as possible the proteins that are present on each organelle. And then we can look at these virtual cells in different conditions — look at the pathology and see how the proteins are changing. So it’s really based on the fact that cells use organelles to perform their function and segregate proteins among organelles. We call it organelle proteomics.
Tell me about the Cell and Nature papers that you published.
The Cell paper was a major contribution in which an obvious system was revealed to us. Then we investigated the system and showed that it was indeed present – this was the publication in Nature.
So for the Cell paper, we isolated phagosomes to a high degree of purity. Then we identified proteins, and we started to get excited when we had identified about 150 proteins and we realized that most of these proteins made sense for an organelle that was aimed at killing and degrading pathogens. But there was also something that any cell biologist would have called contamination — the presence of several proteins from the ER. When we published that paper in JBC in 1994 we mentioned that the ER is a contaminant of the preparation. But with the mass spec, and the larger scale, we realized it was the only contaminant — and with cell biology there is no such thing as a specific contamination. We then used a cell biology and biochemical approach to show what was really a surprise for any biologist, which is that the ER was able to fuse directly with the plasma membrane to form the phagosome. This changed the way we see phagocytosis and how it’s described in textbooks.
So how did you get from your findings in Cell to those in the Nature paper?
Then we realized that if you do a search in the literature and you put these two compartments together, they are both needed and involved in a stepwise way in a very important process for immunity called cross-presentation. One of the dogmas of immunology is that we have two pathways used for antigen presentation that are segregated from one another. One is a pathway for the presentation of exogenous peptides, including bacteria involved in phagocytosis. When you internalize the bacteria in a phagosome, you kill it, degrade it, and generate peptides that are then loaded on MHC Class II molecules in the phagosome, and this immune complex reaches the surface and activates T cells which then further activate B cells and produce antibodies. The other aspect of the immune presentation is made by MHC Class I, and this is normally used to present your own proteins that are present in the cytoplasm, for example after viral infection. This pathway of presentation involves the degradation of proteins in a sub-organelle in the cytoplasm called the proteosome, then the peptides generated in the proteosome degradation are translocated in the ER lumen, loaded on MHC Class I in the ER, and then they later reach the cell surface.
So you have two pathways: Class I and Class II. For class I, the loading takes place in the ER, and the other [part] takes place in endosomes or phagosomes. But somehow peptides that were in a class II presentation organelle — the phagosome — reached the ER where the loading for class I occurred. People thought you had to go from the phagosome to the cytoplasm to be degraded in the proteosome, and then in the ER to load Class I, and they called this process cross-presentation. But when we did further proteomics analysis of our phagosomes, we realized that proteins involved in all of the steps of cross-presentation were present in phagosomes. We showed that all the proteins needed for this process assembled on the phagosome, making it a self-sufficient organelle for cross-presentation.
An interesting outcome of that is that if you look at what enters into the phagosome for cross-presentation, these are large, inert particles — either bacteria or parasites. The cell has another way to capture proteins from the outside — endocytosis. This process is used for the entry of some viruses, or the capture of tumor proteins, and this entryway for endocytosis also leads to a cross-presentation process. But entry by phagocytosis is 10,000 times more efficient for cross-presentation. So the idea is to force the entry of viral or tumor proteins by phagocytosis — for example by attaching them to a bead, or inert particle — and then direct the proteins into a pathway that is 10,000 times more efficient, which means that potentially we could generate antigen-presenting cells that are more efficient to activate cytotoxic T-cells against virally infected cells or cancer cells.
We’re now involved in looking at whether the cross-presentation pathway associated with phagocytosis could be used for the development of vaccines against viral infection or cancer. This is a new tangency that the lab is taking. We want to verify that forcing the entry of the viral or cancer proteins into another machine would trigger a more efficient immune response.
So is looking at whole cells meaningless?
I think looking at whole cells is too complex for a mass spec. Proteins could be present in a cell and not be functional because they’re not in the right place. You would still detect them with the mass spec and think they had a function. But if you use organelles and realize that the protein is no longer in that location, then you know this is where it is active. If you use cells, you lose this dynamic aspect of moving one protein from one place to the other, which is key for the function.