#WhyIScience Q&A: A biochemist uses mass spectrometry to find proteins involved in cancer

When Moe Haines first moved from Beirut, Lebanon to the United States for college, his goal was to earn a degree in pharmacy. Haines was the first person in his family to pursue science, and most people he knew with similar interests had chosen a career in medicine. But his first few biology classes in community college exposed him to the possibility of a career in genetics, chemistry, or physics. Haines became particularly interested in protein structure and function and pivoted towards a biochemistry program at Siena College in upstate New York.

After graduating from college, Haines moved closer to his passion for proteomics, working as an analytical development research scientist for a pharmaceutical company in Albany, NY before relocating to Massachusetts. He worked as a drug development research associate at Sanofi and then joined the chemical biology and proteomics group of a small startup in Cambridge. As the startup struggled to find funding, Haines heard about the Proteomics Platform at the Broad Institute of MIT and Harvard. He was excited by the prospect of working in a nationally renowned proteomics lab that houses an impressive 13 mass spectrometers, which enable in-depth and higher throughput analysis of a wide range of chemically modified proteins. 

Now, Haines is a senior research associate in the Proteomics Platform, where he works with Shankha Satpathy, a senior group leader. As part of the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium (CPTAC), Haines and colleagues in the platform investigate proteins involved in cancer and develop new approaches for proteomic analyses of samples from cancer patients at scale. 

We spoke with Haines about his journey from big pharma to the Broad, the value of experimenting with career options, and the untapped potential of proteomics in this #WhyIScience Q&A.

How did you get interested in proteomics?

When I first got to the US, I found a lot of interest in evolution and how science can explain things better than what I was taught back home. That’s how I ended up majoring in biochemistry. And I got into proteomics because I enjoyed thinking about proteins as machines, what they do, and how they operate.

In a traditional family, you’re told to be a doctor — that’s just the way things are done. But I knew that I wouldn’t be happy in medicine. None of my immediate family members have a science background, and it’s challenging to craft your path on your own. But I just really enjoyed the research I was doing, and no longer being in a conservative environment in my home country helped me understand this even more.

How do you think about proteomics and where the field is headed?

Proteomics adds a layer of dimension to the data generated by other teams at the Broad. The human genome consists of about 20,000 genes that code for proteins. Given the dynamic nature of proteins, this translates into more than a million proteins and their modified structures, or proteoforms.

Proteomics also generates so much data, so it’s really analytics-intensive. Searching, storing, and maintaining all that data remain a challenge. 

Compared to fields like genomics, proteomics is in a much earlier stage. We haven’t really tapped into this exponential growth like genomics has. Proteomics is like a puzzle. There is just so much to learn about all the different proteins and their variants, but that’s what I like about this field. 

We’re striving to improve throughput, depth, and reproducibility. Mass spectrometers are now faster and more sensitive, and database searching algorithms are updated frequently, which is improving data confidence and identification rates. This is all coupled with cloud-based resources like Terra at Broad, which allows for scalability for large cohort studies.

How does the Proteomics Platform collaborate with other groups at the Broad?

CPTAC itself is a massive collaboration, not just within the Broad, but with other institutions. We receive a lot of genomic data that we analyze in tandem with our lab’s proteomic data. You see what the DNA is telling you, plot that with protein information, and you end up with a really informative, comprehensive view of what’s happening in cancer cells and tissues.

More generally, the Proteomics Platform helps Broadies who are interested in knowing what the proteins in their experiments are up to. Mass spectrometry enables us to look beyond genetic data, so we can give other scientists a snapshot of each sample or cell of interest and help them infer the protein-level output.

What are you working on within CPTAC?

CPTAC works with a lot of clinical samples. Many of our projects use sample multiplexing to achieve great depth and scalability. Despite the ease of analysis and increased depth and input that multiplexed analysis provides, these methods can be inefficient because of the sample preparation needed to plex your samples and the complications that can arise from multiplexed assays. For this reason, we are switching to label-free analyses for newer projects, called data-independent acquisition or DIA. DIA is emerging as a next generation mass spec approach. Rather than sequencing each multiplexed peptide at a time, we analyze each sample individually by taking advantage of faster instrumentation.

Switching from multiplexed-based proteomics to label-free analysis cuts sample prep time in half and improves the data-generation throughput by almost 4-5 times. We’re talking about running 45-60 samples a day versus 8-12. Proteomics has come a long way, putting us in a better position for clinical studies. 

I first applied this label-free approach to a CPTAC project on formalin-fixed paraffin-embedded tissue (FFPE), which is the standard way tumor tissue is stored after being biopsied or surgically removed. Researchers can easily section these blocks to obtain and study tissue samples and follow disease progression.

There are millions of FFPE blocks out there. Imagine the amount of data that we can get out of that, especially when combined with data about patient outcomes, diagnoses, and medications.

We’ve designed a workflow that circumvents challenges with processing such samples. With this workflow, we can achieve an unprecedented depth of around 10,000 proteins with a turnaround time of almost 4 days from sample preparation to data generation for a 96-well plate of FFPE material. We plan to apply this workflow in our CPTAC study investigating lung adenocarcinoma.

What’s your advice for aspiring scientists?

Experiment with areas that interest you and you’ll narrow down what you’re truly passionate about. For me personally, doing research in undergrad was really helpful. It gave me an opportunity to think for myself in the lab and gave me an idea of what working in research would be like outside of school.

When we say “science”, it doesn’t have to mean chemistry or biology; you might like something else. For example, even though I majored in biochemistry, I didn’t like the biology lectures. I was able to grasp biology better when chemistry got involved. Every person will have to look for these little clues to find what they enjoy in order to navigate their next steps. Go out and experiment. You have nothing to lose.