The power of expression data in rare disease research

In 2022, Rodney Bowling Jr was stuck. Three years earlier, he had taken on an ambitious project in the hope that he might save the life of Rose McPherson, then a three year old child suffering from a rare neurodevelopmental condition. Success was unlikely for a number of reasons. The disease’s molecular origins may be unclear; and, even if they’re made clear, there’s no guarantee that they can be safely drugged. Against the odds, however, Bowling Jr found success. As the Chief Scientific Officer at the non-profit To Cure A Rose organisation, he was able to lead a team in identifying the cause of her ailment and zeroing in on a promising therapeutic approach. 

And so it was all the more frustrating in 2022 when the project began to stall, not because of biological challenges, but because critical data was either not available or needlessly locked behind commercial walls. The team needed a robust in vitro cell model for Rose’s condition to start therapeutic screening. Yet, the expression of their target protein in relevant commercially available cells was unknown, and vendors were unwilling to help. This left Bowling Jr’s team with the burden of purchasing and generating expression data for each available cell type—a task that would quickly become untenable with their limited budget and urgent timeline. 

Then they found bit.bio. With transcriptomic data freely available for each of its human iPSC-derived ioCells, bit.bio was able to help Bowling Jr’s team quickly home in on an appropriate and scalable cell type for their studies. Now three years later, the team is preparing for clinical trials and working to repeat their success with a wide range of other rare disease types. Central to this effort is their partnership with bit.bio, whose high-quality iPSC-derived cells and freely available expression data help to both remove critical barriers and enable effective, rapid therapeutic development. 

Rare disease drug development

Developing therapeutics for rare disease patients is a uniquely challenging task in large part due to financial limitations. 

Typical drug development pipelines are fuelled by a complex cost-benefit calculation which considers the total number of people who might benefit from—and pay for—therapeutic intervention. Therefore, development efforts often focus on moulding therapeutics to treat the largest number of patients. Considerable time and resources are then gathered and spent in preclinical stages, ensuring that therapeutic candidates are likely to be safe and effective across large, heterogeneous patient populations. 

Yet, by its nature, a rare disease affects fewer than 1 in 2000 people¹, making it difficult to attract the funding needed for both basic research and therapeutic development. Accordingly, researchers like Bowling Jr are often working with very limited resources, most of which are provided by the families of affected individuals. Fortunately, they have one advantage in their favour: “[Families] don’t need to spend time making sure this drug works for everyone,” explained Bowling Jr in a recent interview. “They just need it to work for their child.” 

Rather than spending time and resources proving a therapeutic is safe and effective across a large patient population, the team simply needs to prove safety and efficacy for a single patient. However, the success of this approach hinges on the team’s ability to model key aspects of each patient’s disease in vitro, and to do so at the scale needed for iterative drug screening.

“When you are studying a rare disease,” explained Bowling Jr, “you really have to break it down and explore it from the gene level back up to the cell line and then extrapolate that to the patient.” 

An unexpected challenge

Rose’s condition is driven by a missense mutation in the HNRNPH2 (H2) gene, which disrupts the normal function of H2, a nuclear RNA-binding protein that plays a critical role in pre-mRNA splicing. In healthy cells, H2 associates with specific RNA transcripts, facilitating their recruitment to the spliceosome — the macromolecular complex responsible for removing introns and generating mature mRNA. However, the mutation alters H2’s ability to deliver its bound RNAs to the spliceosome, preventing proper intron removal. As a result, transcripts that depend on H2 remain incompletely spliced, leading to widespread mRNA processing defects and likely contributing to the neuronal dysfunction observed in Rose’s condition. Previous studies of similar mutations reported that the gene’s homologue (HNRNPH1, H1) can compensate for the loss of H2, but only when H2 levels are significantly reduced². Because Rose’s cells express a mutant form of H2, H1 compensation doesn’t occur, and disease ensues. 

This discovery quickly led the team to a potential therapeutic approach: With small molecules, ASOs, or gene therapies, they could potentially prevent mutant H2 expression in her cells, thereby enabling natural compensation by H1. To test this approach, the team needed an in vitro model. 

Patient-derived cells—a gold standard in rare disease research—couldn’t be effectively used, in part because of Rose’s neurodevelopmental limitations and the difficulty of collecting a sufficient number of cells. Alternatively, the team could use neurons derived from induced pluripotent stem cells (iPSCs). In theory, patient-derived iPSCs can proliferate indefinitely and be differentiated into any desired cell type. In reality, however, methods for differentiating iPSCs are challenging in their complexity, time-consuming, and result in batch-to-batch inconsistencies, consequently impacting the scalability of the cell model^3,4. 

“We needed cells that gave us more control and scale, something that is absolutely uniform and consistent. We hoped we could get that by ordering iPSC-derived cell lines from one prominent vendor,” explained Bowling Jr. “Unfortunately, the cost was quite high, and they wouldn’t collaborate with us to identify the right neuronal cell line [one that expresses H2]. In the end, we didn’t get enough cells from them, customer service was very poor, and the cells we did get didn’t express our protein.” 

In other words, the team was left adrift. Sifting through vendors and the cells they offered to find a source of neurons that is scalable, robust, and expresses H2 was likely to cost significant time and capital. Then they found bit.bio. 

 “We needed cells that gave us more control and scale, something that is absolutely uniform and consistent.”

Rodney Bowling Jr., PhDFounder and Chief Scientific Officer,
To Cure A Rose Foundation

 

 

 

The rare effect of transparency

bit.bio is known for its ability to precisely program human iPSCs, producing high-quality differentiated cell types for research and drug discovery. In contrast to the high variability and heterogeneous output of commonly used iPSC differentiation approaches, bit.bio’s technology produces consistent batches of defined differentiated cells on a large scale, and the technology can be applied to any cell type, including multiple neuronal cell types. 

While the scale and consistency of bit.bio’s ioCells are critical, these features would mean little to the To Cure A Rose team if they weren’t paired with transparency. 

They knew Rose’s condition was affecting her central nervous system and that H2 is highly expressed in the brain. But it is not ubiquitous among cell types. The team needed help identifying neuronal types that express H2. Fortunately, bit.bio makes this type of question easy to answer before you make a purchase. 

bit.bio routinely generates a comprehensive characterisation data package for its cell types that includes representative morphology images, protein expression for key markers, and bulk RNA sequencing data. From the bulk RNA sequencing data, researchers who request information on their specific genes of interest will receive an easy to read gene expression heatmap along with the associated transcripts per million (TPM) values. This data provides immediate insight, allowing scientists like Bowling Jr to confirm the expression of their target gene(s) and ultimately decide on the right cell types to work with.

“Right away, we could tell a difference with bit.bio,” Bowling Jr reflected. “They helped us identify the neuronal populations expressing H2 (ioGlutamatergic and ioGABAergic Neurons); they worked with us to ensure we grew these cells properly; and they helped customise our sample size to ensure we had enough to carry out screening. It was just incredible customer service.”

Using bit.bio’s ioGABAergic Neurons, the team was able to bring 57 antisense oligonucleotides (designed in silico) into the wet lab for screening. Four of these candidates were capable of reducing H2 expression by at least 90%, three of which also prompted a compensatory increase in H1 activity. 

Iterative screening of the 57 ASOs in GABAergic neurons revealed 7 promising ASOs, 4 of which were capable of reducing H2 expression by at least 90% and 3 of which also prompted increased H1 activity⁵. Subsequent animal studies led the team to focus on 2 of the ASO candidates. Both show strong activity and have thus far shown no signs of toxicity, despite being administered in concentrations 10-fold higher than would be given to human patients. The team hopes to begin clinical trials in the coming year. 

Accessing expression data: a small step with mighty impact

Researchers like those at To Cure A Rose are often working with limited funding and are racing against the clock as they try to find a cure. Under these circumstances, being able to quickly and efficiently find the tools you need can be immeasurably valuable. As Bowling Jr puts it, “these families are operating with a ‘Time Is Life’ urgency.” 

Easy access to expression data for bit.bio’s portfolio of iPSC-derived ioCells is a small step towards making researchers’ lives easier, but one that can have a large impact.

Read more about To Cure a Rose’s race to treatment for rare disease patients here

References

1. The Lancet Global Health 2024. The landscape for rare diseases in 2024. The Lancet Global Health, 12. doi:10.1016/S2214-109X(24)00056-1.
2. Korff, A. et al. 2023. A murine model of hnRNPH2-related neurodevelopmental disorder reveals a mechanism for genetic compensation by Hnrnph1. Journal of Clinical Investigation, 133(14), e160309. doi:10.1172/JCI160309.
3. Volpato, V. & Webber, C. 2020. Addressing variability in iPSC-derived models of human disease: guidelines to promote reproducibility. Disease Models & Mechanisms, 13(1), dmm042317. doi:10.1242/dmm.042317.
4. Volpato, V. et al. 2018. Reproducibility of molecular phenotypes after long-term differentiation to human iPSC-derived neurons: a multi-site omics study. Stem Cell Reports, 11(4), pp.897–911. doi:10.1016/j.stemcr.2018.08.013.
5. Zhu, H., Bowling, R., Cabrera, R., Banks, C. & Finnell, R. 2024. P147: Precision medicine approaches to treatment for HNRNPH2 mutations. Genetics in Medicine Open, 2(Suppl 1), 101044. doi:10.1016/j.gimo.2024.101044.