Interns Zainab, Onyie, Pranati, and Maria interviewed Dr. Eliza Peterson, PhD, a senior research scientist in the Baliga Lab at ISB this summer. It was great talking with a professional with such a fascinating background and areas of expertise. Below is the interview transcript:
What was your journey to becoming a Senior Research Scientist in the Baliga Lab at ISB?
I joined ISB eight years ago, as a Postdoctoral Fellow. Prior to that, I have an undergraduate degree in microbiology from the University of Michigan. While I was at the University of Michigan, I did some undergraduate research with a professor that was studying limnology, which is the study of freshwater, and so we were really looking at the microbial impact of freshwater and water chemistry. I really enjoyed that, and so that research led to me applying to graduate schools in the environmental microbiology area. So I then took a year off between undergraduate and graduate school. I worked at Michigan State University, doing research in freshwater chemistry, doing a lot of field sampling, and then analyzing those samples for dissolved organic carbon and different trace elements as well. I did that for eight months and then I took off and I traveled to Australia and New Zealand for five months. When I came back, I went to MIT (Massachusetts Institute of Technology) for my Masters Degree. I was in the civil environmental engineering side of the program. Again I was studying fresh water, and was specifically looking at the microbial impact of how this polluted watershed was being restored by the use of various microbes. I enjoyed that, but I realized that I missed more of the medical impact that my research could have.
After I finished my masters, I went to work at Pfizer (Cambridge, MA) the pharmaceutical company, and I did early drug discovery. I was finding small molecules that inhibit particular enzymes that were of interest to the company. So I did a lot of biochemistry and mass spectrometry in those assays. This led me to another job at the pharmaceutical company Amgen. I spent three years at these jobs before applying and going back to graduate school for my PhD. I ended up at the University of North Carolina, at the biochemistry and biophysics program, where I did a mix of my interest in microbiology while also building on my experience in early drug discovery. I was in a pharmaceutical sciences lab where we were studying enzymes that are involved in homologous recombination which can lead to the development of antibiotic resistance. So we were trying to find small molecule inhibitors of these enzymes that could be taken along with an antibiotic so that you could kill the bug but also prevent the development of resistance. We were able to identify some small molecules that could delay the development of resistance by inhibiting these enzymes that are involved in homologous recombination.
After I finished my PhD, I was really looking to build on my drug discovery experience. I really enjoyed that aspect of science where I can make an impact that could be translated to patients. But maybe doing it in a more computational kind of way, instead of brunt assays and high-throughput screens, is there a way you can use computational biology and modeling to prioritize drugs and drug targets and accelerate drug discovery. That’s how I ended up at the Institute for Systems Biology and the Baliga Lab. And just as I was joining the Baliga Lab, there was this grant that the Baliga Lab got in collaboration with Seattle Children’s that was studying regulatory networks of mycobacterium tuberculosis. My first study of mycobacterium tuberculosis (TB) had not happened until I joined the Baliga Lab, and I’ve been studying it ever since because I really enjoy it. TB is a very difficult pathogen, in my PhD, I did a lot of antibiotic drug discovery, and TB is in this whole separate category because there are many different complexities to it. A lot of the drugs that can inhibit many other bacterial pathogens don’t work on TB, so it was always one of these difficult organisms that I was hesitant to work on, but now, I am fascinated and interested by TB. TB is a huge public health burden and you can really feel like you’re making a difference. I joined as a post-doctoral fellow and I’ve since been promoted to a research scientist and then a senior research scientist, so that’s how I ended up where I am now.
What made you most inspired to pursue computational biology as a career?
I think the challenge of computational biology as well as the way it brings you closer to a hypothesis without having to do a lot of extra work was inspiring. Computational biology can be very useful in this way of making hypotheses and predictions for experimentation.
(Maria) I’m currently working on a project studying differential gene expression in Mycobacteria tuberculosis and would be interested in hearing about your research and work on network models, genome-wide assays, and phenotype data.
Right now, I am submitting a manuscript of a more advanced gene regulatory network model of mycobacterium TB. I published one in 2014, and this is what is called an ‘ensemble model,’ so we developed an EGRIN (Environment and Gene Regulatory Influence Network) model on mycobacterium TB, and published that, so we are now working on an EGRIN 2.0 by doing this ensembling method where you run multiple EGRIN models over subsets of the gene expression data, trying to form a consensus of statistically reoccurring regulatory interactions across this ensemble of EGRIN networks. It has more statistical rigor and an added element of finding gene regulatory elements which are the DNA sequences that transcription factors bind with, it can find those genome wide. These elements allow us to find all the possible transcription factor binding sites and include those into consensus motifs so that we can find these recurring binding sites of the genome. Using that model, you can get conditional regulation, and because you are training on subsets of the data, it allows you to find conditional relations in small subsets of the data, so ones that could be masked when you run EGRIN on all of the data.
And so we found a regulatory mechanism that is specifically active in low pH during infection (the host cell that MTB enters once inhaled into the lung and goes into the macrophages) along with a lot of other stressors within the macrophage. The bacteria is able to resist killing by the macrophage lysosome, which lowers the macrophage’s pH, so the bacteria is able to survive in these really low, acidic conditions. We found one of the mechanisms that MTB is able to resist and survive within low pH conditions. So we used CRISPR gene silencing to knock down expression of a transcription factor that was predicted to be regulating this regulatory mechanism and validated the regulatory targets found from the EGRIN model. That is some of the phenotype data (phenotype from CRISPR knock down and the RNA sequencing from the CRISPR knock down transcription factors). Some other phenotype data we have is from incubating the cells with a fluorescent alanine analog which is incorporated into the peptidoglycan (part of the cell wall) so that you are able to visualize where the bacteria is synthesizing peptidoglycan. We were able to then see that in the CRISPR knock down strains there were really elongated cells that had multiple septum, so there was definitely a clear phenotype that the bacteria was unable to divide in the knock down in comparison to the wild type. Some of the regulatory targets of the transcription factors are involved in peptidoglycan cleavage which enables the bacteria to separate at the septum when it’s dividing. Another phenotype that we had with the knockdown was when we then tested it in the presence of various antibiotics (two frontline drugs used to test TB). We found that it was much more susceptible to those drugs when you knock down the transcription factor compared to wild type. That is an interesting example of genome wide studies and phenotypic data that we look at.
(Pranati) I am also working on studying DGE in MTB with Maria and Dr. Vivek Srinivas. After reading your paper PerSort Facilitates Characterization and Elimination of Persister Subpopulation in Mycobacteria, I am interested to know more about what sparked your interest in PerSort? Also how was some of your research and collaboration for this paper impacted by the Covid-19 pandemic?
PerSort is very valuable in terms of being able to isolate these persister cells (cells that are not resistant to antibiotics as they do not have mutations and are in a phenotypic state that allows them to tolerate drugs) that exist as a subpopulation. Antibiotics do not completely clear the infection in this case, which is a huge problem in TB. This is a major reason why it takes 6-9 months to treat TB, and that’s with taking four drugs everyday for 6-9 months, and a lot of people aren’t even cured by then. If you have drug resistant TB it can take up to 2-3 years to treat TB. This persister subpopulation is contributing to drug susceptibility. It’s been very difficult to study these persister cells because they are just a small part of the population. So any population wide studies done the persister cells are masked by the bigger population. So it is really important to be able to isolate the persister cells and then study them. And that is what PerSort enables us to do. We are able to identify these translationally dormant cells, which we showed to be persister cells, and then do follow up characterization of them. That way you can better understand what their phenotypic state is, what are the mechanisms for their initial formation, so we can find drugs that target and kill the persister cells and/or prevent them from forming in the first place. That was the motivation behind why PerSort was so needed.
And then the second question was how was the research impacted by Covid. So fortunately, I believe that paper was submitted before Covid so I think we were doing revisions when things started to slow down. So I would say that study was not affected as much. You know once you submit a paper, it gets reviewed, and the reviewers might ask for further experiments. And so that one did ask for further experiments, but I think those were accomplished before the shutdown. But while PerSort wasn’t necessarily affected, you know because of the shutdown a lot of our laboratory work was affected. But fortunately, it was only temporary. You know a couple months was when we weren’t really able to do anything, but in the same sense, while our experimental work was kind of put on hold, it really did give us some opportunity to get some manuscript writing going, and doing some analyses that had really been put on the back burner. So I’d say we were able to keep going, and keep progress going for the most part.
Can you tell us about some of your work on drug discovery and its relation to network models and systems biology as a whole?
Yea, the example I was talking about before, in terms of this transcription factor that we identified from our EGRIN-2 analysis, to me, has been one of the most recent, interesting drug targets that could potentially be followed up. And so we’ve been talking to people at the Melinda Gates Foundation about the attractiveness of this transcription factor as a drug target. So obviously, we’d need to then find a small molecule inhibitor of this transcription factor. The fact that it can potentiate Isoniazid and Rifampicin is very interesting. If you were able to find a drug that inhibits this transcription factor and use that drug along with Rifampicin or Isoniazid, you’d be able to use a lower concentration of Isoniazid or Rifampicin which can then reduce a lot of the side effects that come from taking a high dose of drug. So, I think the attractiveness of our approach of using systems biology to find new drug targets that could potentiate existing drugs.
Link to ISB Profile: Eliza Peterson, PhD · Institute for Systems Biology (isbscience.org)
Link to Baliga Lab: Baliga Lab | The Baliga Lab (isbscience.org)