Genetic reprogramming can help stem cells mature into desired cell types, but it is often kludgy, which is to say, clumsy and inefficient—or worse, inexact. It may produce cells that don’t mature quite as much as they should, or that fail to represent the exact right subtype. These shortcomings may be avoided if more elegant genetic programming methods become available, methods of the sort being developed by scientists in the Duke University laboratory led by Charles Gersbach, PhD, an associate professor of biomedical engineering.

According to a new study from the Gersbach laboratory, genetic reprogramming may be improved with a CRISPR-based method called CRISPR activation (CRISPRa). The Gersbach team used CRISPRa to identify factors that could improve the efficiency with which stem cells are turned into neurons. CRISPRa, the scientists pointed out, could have a more general application. That is, CRISPRa could be extended to other cell reprogramming applications and facilitate the production of cell types other than neurons. Ultimately, CRISPRa could help researchers generate cell sources that would be useful for disease modeling, drug screening, and regenerative medicine.

The CRISPRa work was published December 1 in Cell Reports, in an article titled, “Master Regulators and Cofactors of Human Neuronal Cell Fate Specification Identified by CRISPR Gene Activation Screens.” The article describes how CRISPRa screens were used to identify transcription factors that regulate human neuronal fate specification. These factors, the article’s authors emphasized, “influence conversion rate, subtype profile, and maturation.”

“Currently, the selection of fate-determining factors for cell reprogramming applications is typically a laborious and low-throughput process,” the article’s authors wrote. “[But we used] CRISPRa screens, which offer a high-throughput approach to profile thousands of gain-of-function perturbations in a pooled format.”

CRISPR technology is most often used for genome editing. In this application, the Cas9 protein is bound to a guide RNA that directs Cas9 to cut the DNA at a specific location, leading to changes in the DNA sequence. “DNA editing has been widely used to alter gene sequences,” Gersbach noted, “but that doesn’t help in situations where the gene is turned off.”

A deactivated Cas9 (dCas9) protein, though, will attach to the DNA without cutting it. In fact, it typically won’t do anything without another molecule attached or recruited to it. As Gersbach and colleagues have reported in previous studies, the dCas9 protein may be fused to molecular domains that allow it to activate a gene and remodel chromatin structure.

Back in 2016, Gersbach and graduate student Joshua Black reported an approach to use the CRISPR-based gene activators to turn on gene networks that would convert fibroblasts, an easily accessible cell type that makes up connective tissue, to neuronal cells. This study targeted gene networks that were known to be associated with neuronal specification, but they did not generate cells with all of the properties needed to make effective disease models. The right gene networks to generate those desired cells remained unknown. They were hidden among the thousands of possibilities encoded in the human genome. So, Gersbach and Black devised a strategy to test all of the networks in a single experiment.

“[We] developed a CRISPRa screening approach to profile the contribution of all putative human transcription factors (TFs) to neuronal cell fate specification of pluripotent stem cells (PSCs),” the authors of the current study reported. “We first performed a single-factor screen to identify master regulators of neuronal fate and identified many known and previously uncharacterized TFs. We subsequently performed paired gRNA screens and identified synergistic and antagonistic TF interactions that enhance or diminish neuronal differentiation, respectively.

“Importantly, through this method, we have uncovered TFs that increase conversion efficiency and modulate neuronal gene expression programs influencing subtype specificity and maturation of in vitro–derived neurons.”

The team engineered stem cells that fluoresced red once they became neuronal. The brighter the fluorescence, the stronger the push toward a neuronal fate. Then the team made a pooled library of thousands of guide RNAs targeted to all of the genes that encode transcription factors in the human genome. Transcription factors are the master regulators of gene networks, so to make the desired neurons, they have to get all of the right transcription factors turned on.

The scientists introduced the CRISPR gene activator and guide RNA library into the stem cells so that each cell received only a single guide RNA, and therefore turned on its particular corresponding transcription factor gene target. Then the scientists sorted the cells based on how red they became and sequenced the guide RNAs in the most and least red cells, which told them which genes, when turned on, made the cells more or less neuronal.

When the scientists profiled the gene expression from the stem cells engineered with the guide RNAs, the results suggested that the corresponding cells generated more specific and more mature types of neurons. The scientists also found genes that worked together when targeted simultaneously. Moreover, the experiment revealed factors that antagonized the neuronal commitment of the stem cells, and when they used CRISPR-based repressors of those genes, they could also enhance the neuronal specification.

To confirm that the engineered cells truly recapitulated the function of more mature neurons, the scientists tested their ability to transmit electrical signals. This task was completed by Shataakshi Dube, a graduate student in the Duke University laboratory of Scott Soderling, PhD, a professor of molecular biology and chair of the department of cell biology.

Using patch clamp electrophysiology, Dube determined that the neurons engineered to activate a particular pair of transcription factor genes emitted more action potentials more frequently. That is, these neurons were more functionally mature.

“I was curious but skeptical on how neuronal these stem cells could become,” Dube said, “but it was remarkable to see how much these programmed cells looked just like normal neurons.”

A CRISPR-based system was used to activate gene networks that had been shown to help human stem cells mature into neurons. CRISPR activation led to the generation of cells with neuronal shapes and markers (left) and enhanced function and electrophysiological properties, including more frequent action potentials (right). [Gersbach Lab, Duke University]
Often, programmed stem cells do not mature correctly when cultured in the lab, so researchers seeking adult neuron cells for an experiment might end up with embryonic neurons, which won’t be able to model late-onset psychiatric and neurodegenerative conditions.

“The cells might seem right at first glance,” Black explained, “but they are often missing some of the key properties you want in those cells.”

This shortcoming can be addressed, the scientists argued, with CRISPRa screening. In the current study, CRISPRa screening was used to identify a set of TFs that improved neuronal differentiation efficiency, maturation, and subtype specification. “Interestingly, the majority of these TFs did not possess neurogenic activity on their own, as assessed in our single-factor CAS-TF screen,” the scientists added. “This observation underscores the importance of synergistic TF interactions that govern cell differentiation and supports the use of unbiased methods to identify these TFs.”

The process from stem cell to mature neuronal cell took seven days, dramatically shortening the timeframe compared to other methods that take weeks or months. This faster timeline has the potential to significantly accelerate the development and testing of new therapies for neurological disorders.

“The key to this work is developing methods to use the power and scalability of CRISPR-based DNA targeting to program any function into any cell type,” Gersbach said. “By leveraging the gene networks already encoded in our genome, our control over cell biology is dramatically improved.”

Previous articleThe Future of Genomics
Next articleBrainExplain