NEW YORK – Researchers at the National Institutes of Health and the University of California, San Francisco have developed a CRISPR interference (CRISPRi)-based platform that can be used to perform genetic screens in human neurons derived from induced pluripotent stem cells (iPSCs).
In a study published today in Neuron, the researchers described three genetic screening experiments that they said demonstrated robust and durable knockdown of endogenous genes in iPSC-derived neurons. First, they performed a survival-based screen which found neuron-specific essential genes and genes that improved neuronal survival upon knockdown. A second screen with a single-cell transcriptomic readout uncovered several examples of genes whose knockdown conferred cell-type-specific consequences. Third, a longitudinal imaging screen detected consequences of gene knockdown on neuronal morphology.
"Our results highlight the power of unbiased genetic screens in iPSC-derived differentiated cell types and provide a platform for systematic interrogation of normal and disease states of neurons," the authors wrote.
Because most previous CRISPR-based screens have been implemented in cancer cell lines or stem cells rather than healthy differentiated human cells, this has limited the potential insights into cell-type-specific roles of human genes. In developing their new platform, the researchers integrated CRISPRi technology with their previously developed Neuron platform, which yields large quantities of highly homogeneous neurons.
They decided to use CRISPRi rather than CRISPR-Cas9 because Cas9-associated DNA damage is highly toxic to iPSCs and untransformed cells. Further, CRISPRi perturbs gene function by partial knockdown rather than knockout, enabling the investigation of the biology of essential genes.
In an experiment to identify cell-type-specific genetic modifiers of survival in a pooled genetic screen in iPSCs and iPSC-derived neurons, the team analyzed the depletion or enrichment of single-guide RNAs targeting specific genes at different time points and identified hit genes for which knockdown was toxic or beneficial to either iPSCs or neurons at different time points. The researchers found that knockdown phenotypes of hit genes were strongly correlated between neurons at different time points but distinctly less correlated between neurons and iPSCs.
Next, they compared genes that were essential in iPSCs and/or neurons in their screens with so-called gold-standard essential genes that were previously identified through genetic screens in cancer cell lines. Although this analysis identified a shared core set of essential genes, it also found additional iPSC-specific and neuron-specific essential genes.
"As a group, neuron-essential genes were expressed at significantly higher levels than non-essential genes in iPSC-derived neurons," the authors wrote. "The vast majority of neuron-essential genes were detectable at the transcript level, further supporting the specificity of our screen results."
Interestingly, they found several genes that specifically enhanced neuronal survival when knocked down, including MAP3K12, MAPK8, CDKN1C, and EIF2AK3.
To characterize how gene knockdown altered transcriptomes of iPSCs and neurons, the researchers then performed a differential expression analysis. They found that the knockdown of some genes induced the expression of cell-death related genes, but that there was no generic signature of dying cells that dominated the differentially expressed genes. Rather, knockdown of different genes resulted in gene-specific transcriptomic signatures.
"By clustering gene knockdown groups based on the signature of differential gene expression, we found transcriptomic signatures associated with knockdown of functionally related genes," they wrote. "For some genes, knockdown resulted in upregulation of functionally related genes."
They also found that knockdown of most genes induced distinct transcriptomic responses in iPSCs and iPSC-derived neurons. This suggests that either gene knockdown caused different stress states in the two cell types or that gene regulatory networks are wired differently in iPSCs and iPSC-derived neurons.
To further dissect these cell-type-specific phenotypes, the researchers ranked genes by the similarity of their knockdown phenotypes in iPSCs and neurons with respect to survival and transcriptomic response. They found that for some genes, both survival and transcriptomic phenotypes were similar in iPSCs and neurons.
However, they also found examples of genes that were essential in both neurons and iPSCs but caused substantially different transcriptomic phenotypes when knocked down. For example, knockdown of the essential E1 ubiquitin activating enzyme UBA1 caused neuron-specific induction of a large number of genes, including those encoding heat shock proteins, the researchers reported. This suggested to them that compromised UBA1 function triggered a broad proteotoxic stress response in neurons, but not iPSCs. This is consistent with the role of UBA1 in several neurodegenerative diseases. This would suggest that even ubiquitously expressed housekeeping genes can play distinct roles in different cell types.
Finally, they discovered that some genes differed with respect to both survival and transcriptomic phenotypes in neurons and iPSCs.
"These proof-of-concept screens have generated a wealth of phenotypic data, which will provide a rich resource for further analysis and the generation of mechanistic hypotheses," the authors concluded. "The combination of CRISPRi functional genomics and iPSC-derived neuron technology leverages the strengths of both approaches. … We anticipate that functional genomics approaches, such as ours, may hold the key to improving protocols that lead to ever more faithful models of mature human neurons."