(4cs) Development of Microbial Hosts for Low-Cost Manufacturing of Vaccines and Therapeutic Proteins

Research Interests

Microorganisms have evolved to occupy countless niches in nature. There should exist an organism, therefore, that is closely suited to any biotechnological application. The yeast Komagatatella phaffii (Pichia pastoris) is a promising alternative host for manufacturing recombinant proteins. In fermentation, yeasts grow to high cell densities on inexpensive media. K. phaffii, specifically, has a developed protein secretory pathway but few natively secreted proteins, reducing the number of required purification steps compared to bacterial lysates or mammalian cell supernatants. While these evolved features are sufficient for recombinant protein expression at lab-scale, further engineering is needed to meet productivity and quality requirements for manufacturing of clinical-quality therapeutic proteins. In my PhD research, I applied modern techniques from synthetic and systems biology to improve the manufacturing of therapeutic proteins in K. phaffii.

Manufacturing of low-cost vaccines

The COVID-19 pandemic revealed global inequity in access to biologic medicines like vaccines. As of July 2021, only 0.3% of COVID-19 vaccine doses were administered in low- and middle-income countries (LMICs) due to high costs, limited supply, and cold chain requirements. LMICs like India produce vaccines at large scale, but are primarily limited to viral vaccines or protein subunit vaccines, primarily from microbial fermentation, including in K. phaffii. Development of vaccines that can be manufactured in existing facilities in LMICs would improve global access.

I first worked to improve the manufacturing process for a trivalent subunit vaccine for rotavirus that is currently in clinical testing in Africa. Using -omics techniques like RNAseq and ribosome sequencing in recombinant K. phaffii, we hypothesized expression bottlenecks for the protein antigens, and made targeted changes to antigen sequences to improve the production of each antigen. When the COVID-19 pandemic began, we applied these same principles to develop a subunit vaccine for COVID-19. We developed a manufacturing process for the SARS-CoV-2 receptor binding domain (RBD) in K. phaffii and transferred this strain to the Serum Institute of India for manufacturing a COVID-19 vaccine candidate that is currently in clinical trials. We then engineered an RBD antigen with improved secreted titer from K. phaffii, and improved immunogenicity in mice. Yeast strains that produce the engineered RBD have also been transferred for GMP production.

Dalvie, N. C.*, Brady, J. R.*, ... & Love, J. C. (2021). Molecular engineering improves antigen quality and enables integrated manufacturing of a trivalent subunit vaccine candidate for rotavirus. Microbial cell factories, 20(1), 1-14.

Dalvie, N. C.*, Biedermann, A. M.*, ... & Love, J. C. (2021). Scalable, methanol-free manufacturing of the SARS-CoV-2 receptor binding domain in engineered Komagataella phaffii. bioRxiv.

Dalvie, N. C.*, Rodriguez-Aponte, S. A.*, ... & Love, J. C. (2021). Engineered SARS-CoV-2 receptor binding domain improves immunogenicity in mice and elicits protective immunity in hamsters. bioRxiv.

Dalvie, N. C.*, Tostanoski, L. H.*, Rodriguez-Aponte, S. A.*, ..., Love, J. C., & Barouch, D.H. (2021). SARS-CoV-2 receptor binding domain linked to HBsAg virus-like particles elicits protective immunity in non-human primates. In preparation.

Genome engineering for monoclonal antibody production

Complex therapeutic proteins like monoclonal antibodies also present challenges for global access due to intravenous administration, cold chain requirements, and high costs of manufacturing, quality assessments, and development. Chinese hamster ovary (CHO) cells, which are typically used to manufacture antibodies, require seed chains and expensive media to grow to high cell densities in fermentation, and generation of a master cell bank can take ~6 months. Alternative hosts like K. phaffii could lower costs and timelines, but engineering is required to achieve comparable product quality and titers. Antibodies have highly conserved framework sequences and post-translational modifications, so engineering efforts for manufacturing require modification of the host cell genome. At the start of my PhD, scalable genome editing tools did not exist for K. phaffii.

I developed an improved CRISPR-Cas9 genome editing tool for K. phaffii. We used small RNA sequencing to identify tRNA promoter cassettes for precise expression of single guide RNAs, enabling highly efficient, parallelizable genome editing. We first used this tool to engineer a strain with humanized glycosylation. Now, we are applying CRISPR-Cas9 to engineering the genome of K. phaffii to enhance secretion of antibodies and other human proteins.

Dalvie, N. C., Leal, J., ..., & Love, J. C. (2019). Host-informed expression of CRISPR guide RNA for genomic engineering in Komagataella phaffii. ACS synthetic biology, 9(1), 26-35.

Dalvie, N. C., Lorgeree, T., & Love, J. C. Pooled CRISPR screening facilities rapid remodeling of the secretory pathway in Komagataella phaffii. In preparation.

Future Directions

I am interested in the relationship between the sequence of a recombinant protein and the genotype of the host cell that manufactures it. Small changes to the protein sequence or host cell genome can have large impact on protein expression and manufacturability. In my PhD research, I engineered the protein sequence of vaccines to better match the manufacturing host, and engineered the manufacturing host to better match complex proteins like antibodies. Recombinant protein biology is critical in several fields of engineering beyond manufacturing therapeutic proteins, including the manufacturing of industrial and food enzymes, and expression of heterologous enzymes for metabolic engineering. In most settings, protein function is prioritized—it is not understood what protein domains could be engineered for manufacturability, or what host cell phenotypes and pathways are critical for expression of a particular protein.

I plan to explore this relationship with more systematic methods. Recent progress in large-scale DNA synthesis and genome transformation have enabled bottom-up construction of minimal microorganisms. Efforts so far have yielded minimal genomes of Mycoplasma mycoides and Escherichia coli, and effort is underway for a minimal Saccharomyces cerevisiae yeast cell. Expression of recombinant proteins of varied complexity in minimal cells would shed light on the relationship between protein motifs and cellular pathways and genes. Minimal cells, additionally, could eventually comprise efficient cellular hosts for manufacturing, with a large portion of cellular resources devoted to the recombinant function. K. phaffii, interestingly, has a small, compact native genome (~9 Mbp, ~5000 genes). Generation of a minimal K. phaffii genome that is devoted for the manufacturing of certain proteins is a tractable goal.

In the short term, I plan to pursue a post-doctoral position to gain experience with minimal cells, selection for complex protein function, or other tools for the bottom-up construction of cell factories. Capabilities like these would enable me to start an independent research program to work towards optimized manufacturing of complex proteins.

Teaching Interests

As a senior undergraduate at Northwestern University, I instructed a course in the School of Education and Social Policy on Asset-Based Community Development (ABCD)—communities are more productive when they rely more on internal strengths than on external forces. I led discussions about community development theory, and guided students in the ideation and planning of small projects to enhance the Northwestern community. I learned that developing community was an effective strategy to promote learning within the classroom as much as within any student organization or college campus.

I have applied ABCD theory to my teaching at MIT. I served as a teaching assistant for Introduction to Chemical Engineering, and observed that while students frequently struggled with common concepts, other students often had the needed expertise. Instructors that effectively leverage knowledge between students facilitate more efficient learning, and foster communities where students feel comfortable collaborating.

Later in graduate school, I taught a course closely related to my ongoing research efforts. As I worked in the lab to develop low-cost vaccines during the COVID-19 pandemic, I lectured on the development of immunotherapies, and beneficial concepts like manufacturability assessments and Quality-by-Design. The course was a graduate-level topical elective based in discussion, and while my own experiences provided sufficient case studies to cover the course material, students were more engaged in discussions when they shared their own industrial and academic experiences in the context of immunotherapy development.

In addition to classroom learning, I believe ABCD is an effective strategy for managing a research group. For my last three years of my PhD, I managed the daily activities and overall research direction of a team of 3-4 technical research staff, as well as younger graduate students. Surprisingly, managing the morale and internal communication of the team was more difficult than managing technical research challenges. I almost always found that solutions that came from team members were more effective than solutions from myself or my adviser. This year, I serve on the Diversity, Equity, and Inclusion committee at the MIT Koch Institute. My goal for the next year is to improve inclusion of technical research staff in community building efforts. In the future, I plan to continue practicing ABCD in the classroom and in my own research program.