Update: This article was updated to clarify the researchers' sources of funding.
Arizona State University's Joshua LaBaer became interested in using protein arrays to look for markers of disease six years ago, but it was a call for proposals from the National Cancer Institute's Early Detection Research Network that specifically turned his focus to cancer biomarkers and the possibility of creating a diagnostic blood test for patients. "Breast cancer in particular does not have any blood markers to speak of for its detection, and while mammography is a useful tool — and without a doubt has saved thousands of lives — it misses around 25 percent of cancer," LaBaer says. "So if there were blood tests available, we would hope that they would pick up some of these cancers that were missed by mammography." Coupled with mammography, blood tests for cancer markers could also help determine whether women who show an abnormality on a mammogram actually need surgery, he adds.
After spending several years trying to identify autoantibodies associated with breast cancer, LaBaer found that his assumption that he might find biomarkers that would apply equally to all women with breast cancer was faulty, and that gathering large numbers of women with the same molecular subtype of breast cancer for one study would be difficult. So when NCI's Jonine Figueroa contacted LaBaer about a year ago to see if they could combine her collections of breast cancer tissue samples and his expertise on protein arrays, LaBaer jumped at the chance. "She'd already done a huge amount of genomic analysis on these women. She knew all these clinical details about them," he says of Figueroa's sample collection. "This is the best thing that could fall into a biomarker researcher's lap. You couldn't ask for anything better than this. So we started talking about doing something together, and what would make the most sense given this collection of data."
The collaborators decided to focus on a subset of breast cancer — basal tumors — which is particularly tricky to diagnose with mammograms. Many of the women diagnosed with these cancers detected it through other means. ER-negative cancers like these basal tumors often occur in younger women, who tend to have denser breast tissue which makes mammography less successful, Figueroa says. These women, then, could benefit from an additional blood test to pick up biomarkers. "We are specifically looking at autoantibodies as a promising class of blood-based markers. Although individual autoantibodies alone are unlikely to enhance early detection, multiplexed assays for autoantibody panels may achieve the required sensitivity," she adds.
NAPPA at work
What LaBaer contributes to the project is more than just expertise in protein arrays, Figueroa says. Along with his Arizona State colleagues, he developed a technology dubbed NAPPA — Nucleic Acid Programmable Protein Arrays — that allows for the creation of high-density, customized protein arrays for a variety of experiments.
"The idea is you want to display a lot of different proteins to a sample so the autoantibodies in a patient's serum can find any proteins they happen to recognize," LaBaer says. "You want to have every protein and you'd like them to be folded naturally — you'd like every protein to get a fair shake at exposure to the serum." In addition, LaBaer says, it's important the arrays be fairly stable and that they display proteins that won't change or unfold over time.
The NAPPA approach is to take cloned copies of genes and print them on the array. "The cloned copies are cDNA, and are configured in such a way that at the end of the gene, we add an epitope tag at the C-terminus," LaBaer says. "If you get the tag, it means you had to translate the entire protein to get there — you had to go through the whole length of the protein to get to that tag — so anything that's captured by virtue of the tag must have the full-length protein attached to it." Those genes are printed on the array, which can then be stored for months. Once the array is needed, it is floated in expression extract that transcribes and translates the proteins in situ on the glass. That way, LaBaer adds, the proteins are made about an hour before they're used, guaranteeing that they're as fresh as possible. "Between the fact that we have naturally folded protein, we're making the protein only an hour before we're testing it, we're getting good expression of virtually all proteins we've tried, and the levels of proteins from one to the other are pretty normal — all those combined make this a pretty good platform for this type of experiment," he says.
When Figueroa and her team read about NAPPA, they thought it might be a good approach to take in their hunt for biomarkers. "The technology was attractive to us because it uses minimal sample — as little as 0.01 milliliter of plasma or serum — and had good assay reproducibility and reliability based on pilot work, making it particularly attractive for our molecular epidemiology studies," she says. While her group targets specific candidate markers like p53, they also employ agnostic approaches to highlight novel pathways associated with disease. "NAPPA technology and other proteomic platforms are another extension of genomic technologies that might provide us with novel markers for breast cancer," she adds.
The formation of a team
The complementary nature of their skillsets is really what makes this collaboration successful, LaBaer says. "She has this amazing set of samples, we have the platform technology to do the study, and then both groups have strengths in biostatistics," he says. Likewise, Figueroa says it gives her a great advantage to work with LaBaer, whom she calls "one of the foremost experts in proteomics."
However, there are a few challenges the collaborators find themselves having to deal with. For his part, LaBaer says one of the biggest problems is the distance between the two groups. He is in Arizona, while Figueroa is across the country in Washington, DC, so planning requires a little more time, as does coordination for meetings and sharing results. But overall, even if he can't just walk down the hall to meet with Figueroa, LaBaer says these kinds of problems are really small potatoes in the age of modern technology. "It's always nice to do face-to-face meetings, but other than that, [Figueroa's] a really easy person to get along with, and her team is really nice, so that part of it has been pretty simple," LaBaer says.
And there are challenges to doing the work itself. "This is new territory and the biology of autoantibodies is not well understood. We hope to learn how age and lifestyle factors, as well as intra-person and inter-person variability, influence these markers," Figueroa says. "Because basal breast cancers are highly aggressive, it may not be possible to isolate autoantibody markers appropriate for early detection; we may only find markers related to overall tumor aggressiveness." Sample size and power are also issues since basal cancers are so rare, she adds.
LaBaer says the collaborators are looking to wrap up the initial stages of the research in about a year, adding that the work will be continued depending on what the team finds from the initial testing of the breast cancer samples with the NAPPA array, and what questions arise from the work. "Ultimately, our goal is to identify candidate auto-antibody markers that may detect basal breast cancers early or predict survival," Figueroa adds.
Participants: Joshua LaBaer and his team at Arizona State University, and Jonine Figueroa and her team at the National Cancer Institute
Funding: LaBaer is supported by funding from the Early Detection Research Network and Figueroa is supported by intramural National Cancer Institute funds. The team will need about $200,000 for the initial phase of research, and will seek out more funding if the work continues.