Michelle Harwood

Reading between the alleles

This interview was created as an assignment in the MMG3001Y (Advanced Human Genetics) course of the MHSc in Medical Genomics program under the guidance of course instructor Dr. Kinjal Desai. By Amy Li and Nasim Azizi.

The possibility of two people carrying the same genetic mutation, where one develops a disease, while the other remains completely healthy, seems impossible – but it is not. Why could that be? For years, scientists emphasized identifying harmful genetic variants, but recent research reveals that simply carrying the mutation does not paint the full picture¹. In fact, what scientists have found is that one version of a mutation in a gene may be expressed more than the other. This phenomenon is known as allele-specific expression (ASE), which acts as a soundboard, with each gene having its own volume setting and adjusting how much a certain mutation is expressed (Figure 1)¹.

But what exactly controls ASE? Does recombination, the rearrangement of genetic material across chromosomes, contribute to ASE? These were the questions that Dr. Michelle Harwood sought to address during her PhD, aiming to improve our understanding of how ASE influences the risk and severity of diseases.

Education and Early Career

 While pursuing her undergraduate degree in Biology at Queen’s University, Dr. Harwood took on a summer research position at the Lougheed lab—a pivotal moment in her journey. There, she developed an appreciation for conservation genomics as she collected environmental DNA samples to survey turtles, preserved different types of snakes to map their diversity across Ontario, and led a pilot study to track ticks near Queen’s University Biological Station. Captivated by the work, she extended her involvement, choosing to conduct her fourth-year thesis in the same lab. For her thesis, she analyzed the population structure of polar bears in the Canadian Arctic, investigating how climate change was impacting them. Additionally, she contributed to developing a new non-invasive method for quantifying DNA from polar bear fecal samples. Her exposure to population genetics sparked a curiosity about its connection to human health and disease, leading her to Dr. Philip Awadalla’s lab for her PhD, where she transitioned from studying wildlife genomics to human genetics.

Unraveling the Link Between Recombination and ASE

Dr. Harwood’s work on ASE began when researchers in Dr. Awadalla’s lab were uncovering clues about how ASE differs between individuals. The lab had access to CARTaGENE, a large population health cohort in Québec, offering a unique opportunity to explore ASE in depth^3,4. In this cohort, Dr. Harwood focused on 884 individuals of multiple ancestries from three major Québec cities, quantifying ASE using advanced sequencing techniques and publicly available reference data for tissue expression (Figure 2)⁵. While these tools allowed her to measure ASE across different genetic variations and tissues, something was still missing. A key piece of the puzzle, recombination, had already been studied in the lab, but no one had yet connected it to ASE. This fueled a question in Dr. Harwood’s mind: “How can we actually intersect ASE and recombination, [given the] interesting insights already demonstrated with recombination and deleterious mutations, [while] ASE was beginning to be explored from the same data?” Determined to find out, she set out to uncover the mechanisms linking these two processes, ultimately revealing new insights into how human genetic variation is shaped.

A major finding in her work was how recombination modulates ASE, offering protection against harmful mutations. Interestingly, this discovery reinforces the findings of a previous paper in the Awadalla lab where Hussin et al.⁶ found that deleterious mutations tend to accumulate in low recombination regions. This is consistent in these regions because the lack of recombination makes it harder for natural selection to eliminate deleterious mutations (Figure 3)⁵. These mutations then stay linked to other genes and ultimately do not get weeded out by natural selection. Digging further, Dr. Harwood then sought the burning question: Why do some people develop disease from a genetic mutation while others remain completely healthy? Her findings suggest that the answer lies in ASE. In regions of high recombination, regulatory variants can act as a protective mechanism by reducing the expression of harmful mutations⁵. This means that if someone inherits a disease-causing mutation along with a protective regulatory variant in these regions, they may stay healthy⁵. However, in low recombination regions, there is less genetic diversity, making protective regulatory variants less common. As a result, harmful mutations in these regions are more likely to be highly expressed, increasing disease risk⁵.  For Dr. Harwood, seeing this theory play out in the data was a pleasant surprise as it strengthens the relationship between recombination and ASE regulation of deleterious mutations.

When asked about areas of future exploration in ASE, Dr. Harwood mentioned using ASE in developing biomarkers and integrating them with precision medicine. By analyzing ASE across individuals, researchers may develop more tailored and effective treatments. Dr. Harwood emphasized that this approach is relatively new as traditional methods mainly focus on methylation profiles, single-nucleotide polymorphisms, or total gene expression. However, by integrating ASE and its connection to recombination into these models, more novel and precise insights into disease mechanisms and treatment responses can be provided.

Transforming Precision Medicine with ASE: What’s Next?

Another potential area of research is the use of ASE in oncology, where identifying allele-specific biomarkers could improve cancer detection and prognosis. Dr. Harwood highlighted that, “not only does [the presence of a] cancer-associated mutation matter, but whether that person has over or under-expression of that mutation may also matter.” Therefore, future therapies could be tailored to not only target genetic mutations, but also to account for how these mutations are expressed in different genomic contexts, leading to more refined and personalized treatment strategies. Furthermore, determining changes in ASE between different cancer mutations can pinpoint differences in treatment response while potentially influencing disease progression.

 While these insights are promising, further research is needed to effectively apply ASE data within clinical settings. Dr. Harwood reflected, “the results that we showed, although are significant statistically, [require] a lot more work to prove in a specific disease context.” This indicates that while the importance of ASE was proven in multiple cohorts, there is still a need for experimental and clinical validation before pharmaceutical companies can consider ASE-based biomarkers in drug development. These key considerations, integral for pharmaceutical companies, are now central to Dr. Harwood’s professional career.

Venturing Into Industry

Dr. Harwood now applies her expertise at Roche, where she works as a Bioinformatics Scientist. Dr. Harwood operates in the Translational Research Bioinformatics group, where she specifically develops bioinformatic assays and workflows that are directly involved in improving human health. One aspect of her work that she enjoys is its direct clinical impact and emphasis on teamwork. Many professionals who have moved from academia to industry share this perspective, as they gain a broader understanding of their work and how different teams collaborate toward a common goal. In an industry setting, team members also have equally vested interests, fostering an environment where everyone is eager to share their insights and consider diverse perspectives. Through these aspects of working in the industry, Dr. Harwood can clearly envision how her projects will be used to improve patients’ lives. An example of this is the development of assays for identifying small numbers of cancer cells remaining after treatment. These assays are designed to detect circulating tumour DNA (ctDNA), a cancer biomarker, in the blood of cancer patients. CtDNA is a booming field in cancer diagnostics with the potential to detect cancer relapse earlier while reducing the need for frequent scans, which can be less sensitive and harmful to the patient. The use of ctDNA as a cancer biomarker is also widely explored in academia, with Wong et al.⁸ discovering that ctDNA can be used

for early cancer detection in Li-Fraumeni syndrome. Hence, seeing this concept cement itself at Roche highlights the crucial role of research and development, as well as commercialization, in bringing diagnostic tools to the bedside. Having experienced both ends of academia and the commercial sector, Dr. Harwood offered unique insights into navigating both worlds.

Lessons from the Frontlines of Genomics

Dr. Harwood recounted several challenges during her PhD and how she overcame them. At the beginning of the project, Dr. Harwood recounted the initial exploratory phase as she struggled to find a direct research question, often going in circles and asking a lot of questions. Rather than getting overwhelmed by the study’s complexities, Dr. Harwood emphasized how important it was to “let the data lead you.” Ultimately, she trusted that the data accurately reflected patient information, allowing her to stay grounded in the evidence. This allowed her to approach the analysis with an open and objective mindset, realizing that the research question became clearer as patterns naturally began to emerge. Along with the ambiguity in the initial exploratory phase, Dr. Harwood struggled to navigate the complications of large patient datasets. Given the substantial amount of time that programming code may take to run, careful thought needs to go into parsing through datasets in an efficient and accurate manner. Although Dr. Harwood faced many hardships during her PhD, she gained valuable lessons along the way.

Reflecting on defining moments in academia, Dr. Harwood recalled gaining a deeper appreciation for her work. While the first two years of her PhD posed to be a huge learning curve, it was not until her third year that she became reinvigorated in her research again. After receiving opportunities to attend talks, presentations, and conferences, she realized that people were not only listening to but also understanding her work. This moment of clarity made her recognize how worthwhile research can be, realizing that “[she] may struggle initially to form the ideas, but eventually proving that they are valid ideas.” These sentiments are not exclusive to her PhD chapter as she also gained knowledge from working in the industry, leaving her with lots of wisdom to share.

Towards the end of the interview, Dr. Harwood detailed essential skills and advice that were instrumental to her success in both academia and industry. In hindsight, Dr. Harwood realized that she benefitted greatly from working at the Ontario Institute for Cancer Research during her PhD. While she did not directly work on oncology projects at the time, she stated that “even just being surrounded by those kinds of projects and conversations on a daily basis in my PhD were very helpful when I’m now working on related projects.” From this experience, Dr. Harwood emphasized the ability to apply what is learned from one context into another as this is a relevant skill in all different career stages. As a result, her exposure to different fields outside of her PhD helped her later in her career at Roche, encouraging students in academia to participate in learning experiences outside of their domain of expertise. Lastly, Dr. Harwood stressed the importance of having a strong appetite to learn and adapt quickly. Echoing advice she received from someone during her PhD, Dr. Harwood expressed how alongside PhD students being expected to become expert learners in their field, they also become excellent problem solvers. Dr. Harwood underscored the essence of problem-solving, emphasizing that the most important skill graduate students develop is the ability to say: “I don’t know the answer, but I know exactly where to go to find the answer.”

References

1. St. Pierre, C. L. et al. Genetic, epigenetic, and environmental mechanisms govern allele-specific gene expression. Genome Res. 32, 1042–1057 (2022).
2. Kukurba, K. R. et al. Allelic Expression of Deleterious Protein-Coding Variants across Human Tissues. PLOS Genet. 10, e1004304 (2014).
3. Dummer, T. J. B. et al. The Canadian Partnership for Tomorrow Project: a pan-Canadian platform for research on chronic disease prevention. CMAJ Can. Med. Assoc. J. 190, E710–E717 (2018).
4. Awadalla, P. et al. Cohort profile of the CARTaGENE study: Quebec’s population-based biobank for public health and personalized genomics. Int. J. Epidemiol. 42, 1285–1299 (2013).
5. Harwood, M. P. et al. Recombination affects allele-specific expression of deleterious variants in human populations. Sci. Adv. 8, eabl3819 (2022).
6. Hussin, J. G. et al. Recombination affects accumulation of damaging and disease-associated mutations in human populations. Nat. Genet. 47, 400–404 (2015).
7. Peñalba, J. V. & Wolf, J. B. W. From molecules to populations: appreciating and estimating recombination rate variation. Nat. Rev. Genet. 21, 476–492 (2020).
8. Wong, D. et al. Cell-free DNA from germline TP53 mutation carriers reflect cancer-like fragmentation patterns. Nat. Commun. 15, 7386 (2024).