CDMRP Investigator Vignette

Identifying Autoantibody Biomarkers Derived from Cancer-Specific Splicing and Validating Them in Preclinical and Occult Ovarian Cancer Plasma Samples

Martin McIntosh, PhD; Fred Hutchinson Cancer Research Center

My current OCRP funding is actually a continuation of a grant I got initially through the OCRP which was using new technologies to find new biomarkers for predicting ovary cancer development.

The goal is to find proteins that are uniquely derived from a tumor. And that might actually help us detect cancer when it?s smaller and earlier.

One of the problems with finding proteins that are unique to cancer, is that the technologies that allow us to know what proteins are in cancer kind of only allow us to find things that we already know about. But new technology came along recently that not only allowed you to find things that you know are there but also to predict things that shouldn't be there at the same time.

It's derived from new technologies that make human genome sequencing very, very fast and cheap. So, the essential dogma of how biology works is everybody knows what DNA is and we also kind of know what protein is. But in order to make protein, your body first makes RNA. It's kind of like DNA will make RNA and the RNA will then be turned into protein.

So what we're trying to do is instead of sequencing proteins, we sequence the RNA which comes after the genome and before the proteome; we can measure those in a much higher fidelity and actually measure things that we don't know about so we can find proteins that are abnormal in cancer.

And it's kind of like a revolution that people maybe have wanted to do for a long time, but it now may be possible. So, this analysis shows some genes that appear to be new and novel to ovary cancer. And, by new and novel, meaning that nobody has ever seen them before but we found them in ovary cancer. And, what this basically shows is that we think we've discovered a gene that has two variants to them. Each variant on its own is something that occurs in everyone's body, but we're seeing them occur together. And, when they occur together, there's a sequence of amino acids or part of the protein that changes. So, a protein is really assembled from different parts of your genome.

The genes just aren't contiguous. They are assembled by bits. So, I may take a bit from this part and a bit from this part, and a bit from this part, and they are very close together but there are gaps in the middle and we'll stitch them together and say that's the protein. Well, sometimes it will put them together in different ways, almost like, someone will take part one, two, and five; sometimes one, two, five, seven, and nine. Each one of those will be a different isoform from the same gene, so the same gene can make multiple proteins. So, every line you see here is a different map telling us how those proteins are assembled.

So, this shows us there's a variant that occurs in one isoform and a variant that occurs in another isoform that don't ordinarily happen together. And, what this says is we actually seeing in this variant and that variant at the same time. And, this is the measurement of the protein that's derived from that variant. So, here we have one, two, three, four, five examples of five proteins and you can think of the protein as a linear sequence. It's a bunch of letters that go this end, that go from here to here. And, so a lot of things can happen to those proteins. Some of the protein might be deleted because it was never assembled properly. You can have sections that are stuck into it that don't belong there, and you can have a combination of both.

So, here's an example right here, where we see protein, where you see a bright red. It says that sequence is unique to cancer. And the black is a part that is deleted. So, basically, what see here in this gene, we have a protein that should be this number here; 400 amino acids long. The variant we're seeing in cancer may only be about 150 long. But even of that 150, half of that sequence actually is unique to cancer. So that sequence right there may actually be useful as a diagnostic test.

This research I just got funded through the OCRP, I initially sent it through the National Cancer Institute and one of the reviewers had responded that if the work was successful could revolutionize cancer research in general.

The next sentence said however it's risky because it's not guaranteed to work. The things I'm looking for aren't guaranteed to exist and therefore shouldn't be funded. And so I just turned it around and I sent it to the OCRP and I made sure I took that quote from that reviewer and I stuck it on the first page of my grant. And there wasn't a single comment from the OCRP reviewers about the risk of what it?you know, the riskiness of the proposal. They recognized that the hypothesis of the things we're looking for may exist is solid. There's a reasonable chance that it actually is going to exist. And if it does exist it could revolutionize cancer research. And to them that was enough.

So I think my research group is living proof that the OCRP is living up to its mission. A grant that was deemed by NCI to be highly promising and innovative was not funded because it was too risky. And the OCRP didn't seem to bat an eye.