Discovery vs. Hypothesis-Driven Cell Line Development

By Tyler Menichiello, contributing editor

While at the 16^th annual Bioprocessing Summit in August, I had the chance to speak with Susan Sharfstein, Ph.D., a professor of nanoscale science and engineering at the University at Albany. She gave a keynote presentation, “A Multiomics Perspective On Cell Line Development,” which highlighted some of the observed physiological differences between lower and higher productivity cell lines. Sharfstein explained how discovery-driven research can reveal more insights into cellular biology than hypothesis-driven research, insights that can ultimately be applied to cell-line engineering.

Using Omics To Better Understand Biology

Across the industry, researchers are united in the common goal of figuring out how to make biologics better, faster, and cheaper. However, biological limitations exist, and when it comes to engineering producer cell lines for productivity, we must ask ourselves, “How do we best understand the physiology of the cells that we’re using as our machines?”

Sharfstein believes the answer lies in omics, or taking a discovery-based research approach as opposed to hypothesis-driven research. The problem with hypothesis-driven research, she says, is that you only get what you look for. As the saying goes, when your only tool is a hammer, everything looks like a nail.

Seeking to understand biology through discovery-based research — i.e., using omics and bioinformatics to reveal what’s happening inside cells rather than trying to control them — can serve as a strong starting point and offer clues as to what factors should be controlled or investigated further in subsequent experiments.

“I would say that my research does not lead to answers, it leads to questions,” Sharfstein says with a chuckle. While that may be true, I don’t think it’s a bad thing. Questions are the heart of good science, and unveiling the right questions to answer is what discovery-based research is all about.  

What Can Omics Tell Us About Cell Productivity?

In Sharfstein’s summit presentation, as well as in our one-on-one conversation, she explained what discovery-based research looks like in practice. She and her team looked at two antibody-producing Chinese hamster ovary (CHO) cell lines, A0 (the parental cell line) and A1 (a DHFR/MTX-amplified cell line). Using omics, her team sought to find physiological difference between the high-productivity clones (A1) and the lower-productivity clones (A0).

While Sharfstein’s presentation highlighted multiple observed differences between the cell lines (e.g., in metabolism), proteomics and RNA-seq revealed a particularly interesting difference in ribosomal biosynthesis. They found that ribosomal proteins were upregulated at the proteomic level, but not at the transcriptional (RNA) level. In other words, ribosome biosynthesis increased despite mRNA levels staying the same. They also observed upregulation of the proteasome in the exponential phase, supported by both protein and RNA levels, as well as an overall upregulation of energetic pathways.

Taken together, these findings suggest that high-productivity cells upregulate their own cellular machinery.

 “What that says to us is that you probably don’t have to engineer cells, that the cells do this all on their own,” Sharfstein says. “It also suggests that engineering cells by changing one gene or another gene probably isn’t enough. You probably have to engineer many, many genes, and the cells are probably more likely to do that in a way that’s more coordinated than we could do.”

This sentiment made me think back on a conversation I had earlier this year with LyGenesis CEO, Michael Hufford, Ph.D., about working with nature as opposed to against it. As the industry works towards developing more complicated molecules (e.g., bispecific and trispecific antibodies), Sharfstein thinks taking a more holistic look at how cells make these molecules and regulate their expression will benefit bioprocess engineers. “Thinking about the interaction of the molecule with the cell and not just the bioprocessing is probably a good approach for things that are hard to make,” she says.

By taking a multiomics approach to cell line development, you gain a perspective on understanding factors that control productivity, she says. In this way, discovery-driven research can inform hypothesis-driven research and provide key insights into ways of optimizing cell line productivity.