Robert Ietswaart, PhD.

Have you ever imagined a mathematical model of the human cell that can quantitatively explain and predict human cellular phenotypes from its molecular components? Dr. Robert Ietswaart has been motivated by this “big dream” for over a decade. It has propelled his research on genome-wide regulatory mechanisms that dictate global RNA expression profiles, which in turn translate to functional cell states. It is also one of the motivations for the recent establishment of Cellforma, a biotech startup that is developing novel regenerative cell therapies for the treatment of fatal lung diseases.

Dr. Ietswaart was introduced to the world of biology early on by his uncle who was a professor in biochemistry. Although the Netherlands did not have a degree program that combined biology with quantitative techniques back in the mid 2000s, Dr. Ietswaart acquired solid training in quantitative techniques by deliberately choosing a double major in physics and mathematics as an undergraduate at Leiden University. To further transfer his quantitative skill sets to biology and gain training in molecular biology, Dr. Ietswaart started his RNA journey by pursuing his PhD. degree co-mentored by Dr. Martin Howard and Dr. Caroline Dean at the John Innes Centre located in the United Kingdom. “I am still thankful that they opened up that opportunity, because it certainly wasn’t a deliberate choice from me at the time to work on RNA, but it turned out super exciting and a great foundation moving forward,” said Dr. Ietswaart.

During his PhD. studies, Dr. Ietswaart used mathematical modeling and molecular biology experiments to dissect the mechanism of transcriptional and non-coding RNA-mediated regulation of the single gene FLOWERING LOCUS C (FLC) in Arabidopsis. FLC is a special gene, as it is the major driver of flowering time. Hence Dr. Ietswaart’s favorite RNA: COOLAIR, a long non-coding RNA (lncRNA) antisense to the FLC gene. At that time, lncRNAs were relatively novel findings and there were debates about whether lncRNAs had any function or whether they were just “junk RNAs”. The Howard and the Dean labs were intensely focused on elucidating the function of COOLAIR. Together with colleagues in both labs, Dr. Ietswaart helped discover COOLAIR’s role in controlling Arabidopsis flowering time by modulating sense FLC transcription through COOLAIR transcription, and alternative splicing and 3’ processing, which in turn contribute to a transcriptionally repressed chromatin state at the FLC locus.  “It’s interesting to see these scientific cycles of discovery, hype, debate, systematic study, and conclusion on the functions of new RNA species,” commented Dr. Ietswaart. “Before lncRNAs, there were the small RNAs, and later came circular RNAs.”

When reflecting on his Ph.D. journey, Dr. Ietswaart never regretted starting with a background in math and physics before moving on to molecular biology to achieve quantitative biology. Although learning molecular biology skill sets via self-study and from experienced peers in the lab has its pros and cons, Dr. Ietswaart thinks he would not be able to do the RNA research he is doing now without his early math education serving as his foundation. As he progressed to doing RNA and cell biology research, he knows that the RNA biology field is where he feels ‘at home,’ in part because this is “where the exciting scientific progress and innovation for human health is being made.” However, he stresses that it hasn’t been easy. Although he strongly believed in solving biological problems by combining mathematical modeling and machine learning approaches with experimentation, such belief has certainly not always been mainstream. In fact, he encountered many very negative reviewers who were unfamiliar and/or skeptical of in silico modeling. At times, he had even considered calling both academia and science quits. But as time has passed, the fields of quantitative and computational biology have grown and gained more acceptance since they have demonstrated their power and applicability.

“Don’t let the naysayers bog you down. This is not saying one shouldn’t take into account constructive feedback. But if you have a vision and drive to pursue it, ignore those who respond negatively without good reason and keep going. Your hard work most likely will result in something worthy, and appreciation by others will follow eventually.” 

Dr. Ietswaart’s transition to a postdoctoral researcher in Dr. Stirling Churchman’s lab at Harvard Medical School was also a deliberate choice out of careful reasoning along the line of Quantitative Biology. He had realized that his passion lies with human health, as opposed to plants or other model organisms, as most phenotypes in human biology are complex and multigenic. However, at the time Dr. Ietswaart finished his Ph.D., a genome-wide understanding of gene regulation was absent due to a lack of tools to tackle the complexity of how 20,000 genes coherently function inside a cell. He reasoned that his quantitative skillset could make a much larger impact in big data sets, he decided to learn about machine learning and become an early advocate of machine learning’s capability to tackle the cellular complexity problem.

In the Churchman lab, Dr. Ietswaart achieved one significant step towards his long-dreamed quantitative model of human cells, by elucidating the flow of RNAs at a genome-wide scale at a subcellular resolution. Dr. Ietswaart prefers to study the whole expressed genome with a single genomic assay, where one can answer questions about all RNAs in the expressed genome at say 90% accuracy, versus a deep but low-throughput mechanistic assay, where questions about one single RNA can be addressed at near 100% accuracy. Therefore, in his recent publication Dr. Ietswaart and colleagues developed subcellular TimeLapse-seq to measure the rates at which RNAs are released from chromatin, exported from the nucleus, loaded onto polysomes, and degraded in the nucleus and/or cytoplasm. Dr. Ietswaart was partially inspired by earlier work from one of his favorite papers in RNA which used metabolic labeling to capture RNA dynamics, but he used chemical conversion of 4-thiouracil rather than a biotin enrichment step to avoid selection biases.

Seeing the improvement of human health as the ultimate goal, Dr. Ietswaart co-founded Cellforma to harness the predictive power of a quantitative understanding of RNA biology and cell biology. Like many startup stories, it began with the right people meeting at the right time. Dr. Ietswaart met his co-founders Dr. Carla Kim and Dr. Aaron Moye at the Harvard Medical School Genetics Retreat in 2019. As a faculty member, Dr. Kim came to judge Dr. Ietswaart’s poster on GeneWalk, a computational tool Dr. Ietswaart developed to identify context-specific relevant genes and their functions from a differentially expressed gene list. Dr. Moye was a postdoc in the Kim lab at the time and expressed interest in using GeneWalk for therapeutic target and function identification in early-stage non-small cell lung cancer. Their expertise meshed well with Dr. Ietswaart’s work on characterizing global gene regulation of the histone demethylase LSD1 in small cell lung cancer cells. “This encounter has really had a huge impact on my career trajectory,” said Dr. Ietswaart. “Our complementary expertise and a shared passion to build a startup for therapeutic discovery ultimately brought Cellforma to life.”

Cellforma is making great progress by combining scientific innovation with an emphasis on making a positive impact on human health, through developing regenerative cell therapies to cure fatal lung diseases. Indeed, Cellforma is ready to bring its lead asset, a potentially curative solution for Idiopathic Pulmonary Fibrosis, into first-in-human Phase I clinical trials, thanks to foundational academic work by Drs. Moye, Kim and colleagues. Another of Cellforma’s objectives is to develop graph-based machine learning models based on gene regulatory networks to discover novel cell therapies for other lung diseases. This would come close to Dr. Ietswaart’s “big dream” of making a mathematical “virtual cell”, and thus, the inception and progress of Cellforma has been like a dream come true to him.