Associate Professor of Molecular and Cellular Biology

Images from: An equilibrium-dependent retroviral mRNA switch regulates translational recoding

It’s hard to know how something works if you don’t know what it looks like. Biomolecules are no exception: decipher their structure and their function will come into better view. But solving the architecture and shape of biomolecules, the work of the field of structural biology, can pose difficult challenges. That’s particularly true when the research involves nuclear magnetic resonance (NMR), an analytical method that reveals molecular structures according to how their constituent atoms absorb electromagentic radiation.  Victoria D’Souza knows this better than most. A new MCB Assistant Professor, D’Souza comes to Harvard from the University of Maryland at Baltimore County (UMBC), where, in 2004, she used NMR to solve the largest RNA structure to date: the encapsidation signal of the murine leukemia virus. At 101 nucleotides, this molecule, required for retroviral replication, is so big that it took D’Souza and her advisor Michael F. Summers seven years to understand it. "I’d go to conferences and [other scientists] would tell me, ‘This is insane, don’t even try, you’re wasting your time,’" she recalls with a modest smile. "But we worked really hard on it and came up with a method we could use to solve the whole thing.  And then people started calling our lab asking for advice on their own projects."


D'Souza now brings her expertise in structural biology to Harvard, where she hopes her fundamental investigations will lead to better drugs against retroviral pathogens, especially those linked to cancer. Her work will be sustained in part by a three-year, $300,000 scholar award given by the prestigious Damon Runyon Cancer Research Foundation (http://drcrf.org/apScholar.html), which supports outstanding scientists as independent investigators during the initial years of their first faculty position.

Most anti-retroviral drugs used today target just one protein in the microbe's life cycle. Reverse transcriptase, which converts viral RNA to DNA, is a well-known example. But viruses often outwit compounds focused on single protein targets, with drug resistance as the regrettable outcome. The better approach, D'Souza suggests, is to design drugs that target multiple sites at once, such as protein-protein interactions, or RNA-protein complexes crucial to the virus' survival. Structural biology offers a path to this goal, she says. By mapping the three-dimensional TOPOLOGY of viral nucleic acids, D'Souza aims to learn more about how their molecular components interact. In turn, that information could inform efforts to design drugs that block multiple targets. "Viruses would find it much harder to develop resistance mutations these drugs," she explains.

From India to Maryland

Born to a farming family in Goa, India, D’Souza thought of becoming a doctor, but decided against it after working with children sickened by thalassemia, an often-fatal type of anemia. "I was more emotional about these children than I should have been," she recalls wistfully. "It was hard for me to let go; I realized at a young age that I needed to change my focus."

Research offered a less emotionally demanding alternative to clinical practice, but proved otherwise rewarding. D’Souza studied life science and biochemistry at Sophia College in Bombay (now Mumbai), and stayed in that city for an MS in biochemistry, which she completed in 1997, from the Institute of Science. As a graduate student, she studied ways to improve the nutritional content of soy-based baby food, and cultivated a growing interest in cancer biology.

Her dream of becoming a cancer researcher was realized upon coming to the United States for a PhD at UMBC. There, she joined forces with Michael F. Summers, a Howard Hughes Investigator, who was using NMR to study genome recognition, i.e., how new viral particles in a cell distinguish their own RNA from that of their host. In Summers’ lab, D’Souza studied the murine leukemia retrovirus (MLV), an organism widely applied in gene therapy and cancer research.

Her studies eventually led her to the MLV’s encapsidation signal, which coordinates the virus’ genome packaging activities. HIV and MLV are similar in that their genomes contain only RNA. To reproduce, they hijack the host cell’s genome, and infiltrate it with DNA produced from their own RNA by reverse transcriptase. Upon infection, host cell’s genome unwittingly produces more viral RNA and viral proteins, which agglomerate into new viral particles, thus perpetuating the pathogen’s life cycle. But without the encapsidation signal, retroviruses can’t package their new genomes correctly, and thus lose their ability to replicate. For that reason, the signal offers an optimal target for anti-retroviral drugs.

D’Souza became captivated by the goal of resolving the signal’s structure. Doing so posed a daunting challenge: the largest RNA structures solved to date were just 40-50 nucleotides, about half the encapsidation signal’s size. "And doubling the size doesn’t double the effort," D’Souza stresses. "Larger molecules are exponentially harder to solve." 

Five years into her research, with the structure still unknown, she faced a difficult choice: D’Souza could, after getting her PhD, leave Summers’ lab for a different post-doc elsewhere, as would most aspiring academics, or stay in the lab for two more years as a Howard Hughes Fellow.

Looking out for her career, Summers himself suggested that she move on, but D’Souza chose to stay. "I didn’t want to leave when I was so close [to completing the structure]," she says. "I knew I as on the verge of solving something really big." But to do so, D’Souza would have to embark on new approaches, one of which involved making multiple retroviral RNA samples that were specifically labeled with different nucleotides. For instance, in one, just the adenosines was labeled, while the other three nucleotides were left unlabeled. In a final step, the nucleotide-specific data sets were linked together to arrive at the original structure. Those efforts paid off within a year. and  D’Souza published the encapsidation structure to great fanfare.

D’Souza then moved quickly towards her next quest. Like other retroviruses, MLV packages two identical copies of its own genome, linked together in mirror image. The encapsidation signal, therefore, exists as a 202-nucleotide complex, composed of identical structures bound as if each were a reflection of the other.  D’Souza set out to solve the entire complex, which is now completed and awaiting publication. "[My findings] lend support for the hypothesis of  a structural RNA switch mechanism for genome encapsidation, in which protein binding sites are sequestered by base pairing in the monomeric RNA (single genome) and then become exposed upon dimerization (when two copies come together) to promote packaging of the diploid genome," D'Souza says.

A novel A-Minor K-Turn motif that

causes a 54 degree

bend in the encapsidation signal

A New Challenge

D’Souza now plans to tackle the complex to which reverse trancriptase binds while viral DNA is produced. The complex is made up of four separate molecules: a primer binding site, a primer called tRNA, the enzyme reverse transcriptase, and a protein called nucleocaspid that aligns the molecules in a required conformation. D’Souza emphasizes that solving the entire complex could vastly increase knowledge of its function, which varies dynamically according to the behavior of its molecular components. "Getting to this quaternary complex will be difficult," she concedes. "But once we have the general structure, we can collaborate with chemists to find drugs that block these things from interacting. To produce cures, or make to drugs that improve quality of life, people need to be willing to take risks. If something has even the remotest possibility of working, then it’s worth giving it a try."

D’Souza’s arrival in many ways illustrates MCB’s rapidly growing commitment to structural biology. To support her research, the Department purchased an NMR machine that cost in excess of $1 million.  She says she can now collaborate readily with other MCB experts in the field, including Rachelle Gaudet, who focuses on signaling and transport in biological membranes, and David Jeruzalmi, who studies assemblies involved in RNA replication. The group meets for a common lab meetings once a week to discuss their work and to share findings. "It’s great to be here," D’Souza says. "We’re all structural biologists, so we speak the same language, but we’re all doing different types of research. I feel the Department is really committed to what we’re trying to do."