Pioneers of Proteomics

In part seven of the Pioneers of Proteomics video series, Dr. Richard Caprioli discusses imaging mass spectrometry and its application to cancer and personalized medicine as well as the need for continuous innovation and a multidisciplinary approach.

Dr. Caprioli is the Stanley Cohen Professor of Biochemistry and Director of the Mass Spectrometry Research Center at Vanderbilt University School of Medicine. He is also currently Professor in the Departments of Chemistry and Pharmacology at Vanderbilt University. Dr. Caprioli’s lab is focused on the development of new mass spectrometry and data analysis techniques. His laboratory is a pioneer in imaging mass spectrometry (IMS) which enables direct profiling of proteins in sectioned tissue.

On the technique of imaging mass spectrometry (IMS)

Imaging mass spectrometry is a type of analytical procedure which allows you to analyze molecules directly in samples -- in this particular case in tissues. I think an important aspect of what we do is to look at molecules in their native state. It represents a technique where you can analyze biopsies, for example, and other tissues without really cutting them up and dicing them and extracting them and thereby lose a lot of important information that’s present in that tissue in terms of its spatial orientation.

…what you do is you take a biopsy or any piece of tissue -- it doesn’t have to be a biopsy, of course -- and you cut a cryostat section about 12 microns thick. So, roughly it’s the thickness of one cell -- an average mammalian cell is about 10 microns. You put this on a target plate, put the matrix on, and then you put it in the instrument and ablate it with the laser. Molecules then come off, and then you perform the measurement process (to determine the molecular weights).

Viewing where you’re ablating with the laser is very important in this whole process because what it really brings in is the expertise of the pathologists and the histologists. You can take advantage of the things that they’ve learned in their expertise of many, many years, and now you can take the molecular information in those areas and add it to what they’ve learned. So, it brings to them, I think, a brand new type of information.

On the evolution from immunohistochemistry

I think this kind of mass spectrometry, imaging mass spectrometry, in a sense is a natural follow on to things like immunohistochemistry. Those are techniques which are used to look at molecular events, but they’re one step visions of those molecular events. You have to have a special chemical that recognizes a specific protein. This kind of mass spectrometry then is, I think, the next logical step -- at least logical to me -- in looking at suites of molecules. So, in comparison -- well, maybe it’s fair to say that I don’t see that mass spectrometry was ever intended to put immunohistochemistry out of business. I think it adds a whole new dimension, and we’ll have to see how things go. There certainly will be tests which are most easily and very accurately done by immunohistochemical means. But there are certainly going to be others, especially in the area where you look at signatures of disease, where you look at four, five or ten or even more (different molecules) for that matter, and the relative ratio of all of these molecules gives you that signature. And I think those things are going to be difficult to do by immunohistochemistry; so, this kind of platform or something like it is going to be absolutely essential.

On the search for markers in plasma

The search for markers in plasma has obvious advantages and that is it’s a relatively easy tissue to get. One can get a reasonable amount of it from patients and it’s relatively non-invasive. The problem that people are running into -- and, again, many people understood this from the beginning -- is that you’re looking at a molecularly complex medium. And so, I think, an important role of tissue analysis by IMS, by this kind of mass spectrometry, is to discover those molecular events which are closer to the site (of the disease). In fact, if one gets a biopsy, you’re actually at the site of the disease, and so those markers will be relevant to the process you’re looking at, whether it’s stage of disease or whether the treatment of that patient is progressing, and are most likely to be markers of merit. At least this is the way that I would approach it and am approaching it now. The process is to take those markers that are specific for the question being asked about the diseased tissue -- and now do a targeted analysis of the serum (for the markers). And so, one would really love to have -- after having made the discovery of those markers in tissue -- a signature in serum or plasma that reflects that discovery process.

What we envision is a protein database that would have a huge number of entries of these kinds of signatures and proteins so that one day one can take a biopsy and then just do a database scan and essentially ask the question -- have we ever seen this pattern before. I think it’s going to have an immense impact in clinical diagnosis and eventually in patient care.

On the application of IMS to cancer

I think this kind of mass spectrometry -- what we term imaging mass spectrometry -- has really great implications in patient care, in diagnosis and in treatment. So, in the diagnostic process, what we have come to understand over the years, obviously, is the fact that the molecular presentation of the disease gives you a huge amount of information on the disease process, the stage of the disease and perhaps prognostic information of the risk of the patient for developing further stage? So, these molecular events, I think, are critical to understand even at the diagnostic level. The other aspect, I think, which is very exciting is the opportunity to look at individual patient’s treatment with drugs. So that, for example, if a patient gets a drug, one can perhaps assess efficacy by taking maybe a prick of the finger with a little blood smear in order to determine if that patient is reacting positively to the drug -- is that the right level of the drug for that patient and really gets in to this whole idea of individualized medicine. And I think molecular means of assessing patient reaction and treatment are going to be essential if we are to be successful in this process of individualized medicine.

I think the presentation of a variety of phenotypes is common to all cancers. And again as part of that realization -- it’s been going on for years -- that the molecular phenotypes can be very different even in the same nominal kind of cancer. And so, the ability to measure these very accurately is going to help in therapy. Most cancer treatments are effective in just a small percentage of the population. We all expect that, in fact, these drugs may be very highly effective but in a smaller subset of patients. So, the vision is that through these genomic and proteomic measurements that we would be able to assess the molecular state of that disease and to better know which patients would benefit from a given therapy. This is part of a vision that one anticipates working out very well.

On the application of IMS to personalized medicine

I think that the concept of personalized medicine -- if one takes it to the extreme -- is that every patient’s disease is molecularly different. One doesn’t know if that’s true, or even if it is true, whether we could ever measure it to that degree -- but that’s not the issue. In order to begin to make sense of that, then we have to develop molecular means to assess differences among patients. Again, the advantage of imaging mass spectrometry is that one can now make molecular measurements closer to the primary disease -- that you can actually examine different cellular areas as a pathologist would do to analyze different molecular components of the disease. So, for example, in some breast cancers that we’ve been looking at, you can see different molecular patterns in ductal carcinoma in situ, a less aggressive form of the cancer, relative to that of invasive mammary carcinoma. What the pathologist would like to know is what are the relative amounts, how aggressive is this disease, and this information will impact patient treatment and patient care directly. So, I think it’s very important that these molecular technologies be used effectively and that this will aid in the clinical diagnosis of the disease.

On the need for continuous technology innovation

I think that the role of technology in what we’re doing, in what we’re looking at in terms of molecular discovery and disease, is absolutely pivotal. I mean if you look back even over the last five or six years and what we can do today with the technological advances that have occured, it’s almost a quantum leap and has led actually to a paradigm change in how we look at disease, and what we’re doing to try to discover those aspects in disease that are important. So, technology, I think, is critical. It is going to lead the way. I think the things we’re doing today are going to be looked at as perhaps baby steps if we go even ten years down the road. I think it’s a very wonderful vision. We need to keep our focus on technology development certainly as it’s applied to health and disease and I think this will become our door to really great discoveries in the future.

On the need for bioinformatics

When one develops this kind of technology, you go through several phases. The first phase is usually, gee, can we do this? Will this kind of technology do anything useful? The second phase -- when the answer is yes -- is, well, let’s try some pilot projects and you begin to learn critical parameters from these pilot projects. The third phase, which now is a much more mature phase, is to now take those kinds of steps and make them much more broadly applicable. What comes with that is the absolute necessity for databases, for algorithms that interrogate these databases -- and those built in such a way that you can ask questions five years from now of that database that never occurred to you to ask today. Bioinformatics is really a critical part of the whole process. What we put together, is the expertise of the clinician, the pathologist, the analytical scientist and the bioinformatician. And these four disciplines have to come together as an integrated team for us to make any progress. So, it’s absolutely essential to have the right bioinformatics and right biostatistics to do these kinds of studies.

The information, the data, that we get from patient biopsies go through a bioinformatics process, a biostatistical analysis, and then those then go back to the clinicians -- in fact they’re part of that whole process -- for direct use within patient care. So, for example, we’re working on a couple of different tests, which would help the clinicians choose those patients which are most likely to benefit from a particular drug or not. And these are the things that really change the way a clinician would treat a patient -- having some additional information on what the risk of the patient is or at least the likelihood of success of that process for a given patient population.

On the need for a multidisciplinary approach

What we’ve done recently is to put together a team -- a team that consists of pathologists, analytical scientists and clinicians to study disease processes, to look at the tissues as an integrated whole. I think that has come a very long way. There are a few clinicians that have chosen to spend some time in these labs to learn proteomics and what they bring to us is a view and understanding of that outside world in the clinics and what their real problems are. I’ve been amazed at how that has helped us really understand why we’re doing things and perhaps helped us to do things better. Another part of what we wish to do, of course, is to translate this outside of the university laboratories. That becomes a little bit more challenging because you’re dealing with people you don’t deal with every day.

I think I can see that a lot of the departments, pathology departments, for example, or other basic science departments that have a vision to bring basic science into clinical areas -- that this is not a drum one has to beat very hard -- and it’s a wonderful thing to see develop. I think it’s the way we’re going to solve these problems when we bring mathematicians, bioinformaticists, chemists and clinicians together to focus on these problems.

On overcoming the language barrier

It’s a very, very interesting time that we live in right now. In a sense, we have exquisite technologies. We have exquisite biology in our understanding of biology. But in those two areas, what I found is that there’s a chasm that the molecular biologist in their expertise -- and they’ve done wonderful things -- and the instrumentalists in their expertise -- who have done wonderful things -- really don’t often come together to speak the same language. It’s not a lack of willingness to come together - it’s the lack of a common language. It’s an interesting time because it is apparent that chemistry is that common language. And I think as we develop more and more chemistries which address these problems, that satisfy the molecular biologists and satisfy the analytical scientists, I think that bridge is going to be really very effective. So, it’s an exciting time. I think it’s a wonderful time to be in this area, and one comes to work with the feeling that one can make a difference.