The Best Functional Genomic Screening Method: CRISPR or RNAi?

With the advent of next-generation sequencing, genetic data from myriad organisms has increased exponentially but knowing the sequence data of a gene is only one piece of the puzzle. For the genetic data to be truly useful, gene function must be determined and the link between genotype and phenotype understood. Functional genomics aims to elucidate the relationship between genotype and phenotype, but, crucially, determine this on a genomic scale.

The classical method is to perform a forward genetic screen, where the function of a gene is modified and the resultant changes to the phenotype determined. Historically, forward screens relied upon random mutagenesis or viral transposons to modify gene function, but these methods were difficult to perform on a large-scale. The introduction of RNAi, and later CRISPR-Cas9 gene editing, gave researchers the ability to perform targeted, genome-wide disruption of gene function and functional genomic screening was within the technical capabilities of most laboratories. For the last decade, RNAi has been the go-to method for performing a functional genomic screen—but the recent debut of the CRISPR-Cas9 gene-editing technology seems to have usurped the dominance of RNAi. This article aims to compare RNAi with CRISPR-Cas9 and discuss its use as a functional genomic screening tool.

Gene silencing by RNAi

It had already been observed that the introduction of double-stranded RNA (dsRNA) into organisms elicits gene silencing, but it was not until the publication of work in C. elegans in 1998 by Fire and Mello¹ that the mechanism became clear and the term RNA interference, or RNAi, was coined. RNAi involves the regulation of gene expression by small RNA molecules, which play an important role in cellular development and immunity. Double-stranded RNA molecules are processed by the endogenous cellular proteins DICER and RISC to single-stranded RNA that target mRNA transcripts that share a complementary sequence for degradation, thereby resulting in reduction of protein levels at the post-transcriptional level. As well as exogenous dsRNA, gene silencing can also be triggered by the introduction of other RNA molecules, including small interfering RNA (siRNA), hairpin microRNAs (miRNA), and short hairpin RNA (shRNA).

RNAi and functional genomic screening

Following its discovery, RNAi was rapidly adopted to elicit targeted modulation of gene function in a myriad of cell types and organisms and became the primary method for gene modification, requiring only a simple transfection of short interfering RNA (siRNA) to elicit rapid loss-of-function. But genome-wide screening using siRNA is cumbersome, requiring targeting RNAs to be arrayed and assessed individually. This limitation can be overcome using delivery of shRNA libraries by lentiviral vector and the subsequent stable expression of the shRNA due to random integration into the genome. This approach, later applied to CRISPR-Cas9 screening, provides each cell with a “barcode” enabling easy identification of the sequence that elicited gene silencing by next-generation sequencing, allowing whole-genome shRNA libraries to be screened in a pooled format.

Unlike CRISPR-Cas9, which requires delivery of the Cas9 machinery alongside the sgRNA, a major advantage of RNAi is that the cellular machinery is conserved in the majority of mammalian cells. Transfection with siRNA remains the quickest and simplest method of eliciting phenotypic changes, and the time to phenotype onset is much less compared to shRNA and CRISPR-Cas9 gene editing, which require cloning of the appropriate plasmids. But the ability to utilize a pooled format makes the lentiviral approach highly attractive and cost effective for high-throughput screening of phenotypes such as proliferation or survival.

Off-target effects and the risk of false positives

Both RNAi and CRISPR-Cas9 are at risk of off-target effects that can lead to false positives as shRNA or sgRNA will target any sequence with a complementary seed region (sequence-dependent off target activity). RNAi also displays sequence-independent off-target activity with the introduction of RNA molecules causing upregulation of interferon-regulated genes and alteration in protein expression.² Off-target gene silencing results in altered phenotypes and is therefore disadvantageous in a functional genomic screen, making it difficult to identify and validate hits.

Sequence-dependent off-target effects of RNAi can be moderated by careful selection of unique shRNA sequences and the inclusion of multiple independent shRNA targeting the same gene to improve statistical significance and validate the phenotypic changes. This has allowed for robust datasets of RNAi-based positive selection screens, but negative selection screens remain a challenge due to the high background noise of off-target activity.

Sequence-dependent off-target activity was also observed in the early days of CRISPR-Cas9, as well as differences in sgRNA efficiency. However, development of design tools has, for the most part, mitigated these effects, and algorithms can now be used to aid selection of the most efficient and specific guides for a library. As with shRNA, multiple sgRNA targeting genes of interest are also used to build statistical confidence in observations. Off-target activity of RNAi has been somewhat reduced with the optimal siRNA design and chemical modifications, but a recent comparative study showed that CRISPR-Cas9 is less susceptible to off-target effects than RNAi.³

Knockdown or knockout?

The main distinction between RNAi and CRISPR-Cas9 is that RNAi reduces or knocksdown gene expression at the post-transcriptional level by targeting RNA, whereas CRISPR-Cas9 is a gene-editing tool, so targets DNA to permanently alter, or knockout, gene expression. Knockdown with RNAi produces a hypomorphic phenotype in contrast to the true null knockout possible with CRISPR-Cas9. Both options can be useful depending on the experimental design and question to be addressed. A complete knockout with CRISPR-Cas9 may in some cases be required, preventing any uncertain effects of residual low-level protein expression that remains after knockdown, which may mask certain phenotypes. It could prove useful in cell lines with polyploidy, such as cancer cell lines with multiple gene copies. However, in some cases knockdown could be advantageous. For example, knockout of essential genes is cell lethal, making effects on these difficult to study. Knockdown rather than knockout can also better mimic the inhibition of a target by a drug, allowing reduction of protein levels and intricacies of gene expression on phenotype to be studied.

But the development and expansion of the CRISPR-Cas9 toolkit means that the choice between knockout and knockdown is now not just between CRISPR-Cas9 and RNAi. The use of a catalytically inactive version of Cas9 (dCas9) tethered to a transcriptional repressor has provided the means to perform CRISPR-mediated knockdown. CRISPR interference (CRISPRi) combines the knockdown capacity of RNAi with the efficiency and specificity of CRISPR for functional genomic screening.⁴

shRNA versus sgRNA

The development of CRISPR-Cas9 screening technology in the preceding years has been rapid, with improvements in delivery vectors and sgRNA design, and an evolution to the gene-modification repertoire of CRISPR (i.e. CRISPRi, CRISPRa, base editing, epigenome editing) causing a shift away from RNAi-based screening. The efficiency of CRISPR gene knockout coupled with high specificity and low background stands in marked contrast to the often-confounding weaknesses of RNAi, namely incomplete knockdown and a propensity for off-target effects. There are some benefits of RNAi—by relying on endogenous RNA silencing machinery it remains the simplest method of eliciting gene modulation, and is ideal when screening requires knockdown rather than knockout. But as a result of the versatility and specificity of CRISPR-Cas9 screening technology, with a range of gene modifications now possible (CRISPR-KO, CRISPRi ,and CRISPRa), RNAi has now been surpassed as the dominant functional genomic screening tool.

If not now, then when?

• CRISPR-Cas9 has replaced RNAi as the go-to functional genomic screening tool in part due to the low switching costs for scientists wanting to make the move to CRISPR-Cas9 libraries.
• Many commercially available sgRNA libraries are available, and for those looking to synthesize new libraries, costs continue to fall.
• Protocols for viral preparation and transduction as well as post-screen analysis remain virtually identical, again eliminating barriers to entry for those already working in this manner.

References

1. Fire, A. et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811 (1998).

2. Bridge, A. J., Pebernard, S., Ducraux, A., Nicoulaz, A. L. & Iggo, R. Induction of an interferon response by RNAi vectors in mammalian cells. Nature Genetics 34, 263–264 (2003).

3. Smith, I. et al. Evaluation of RNAi and CRISPR technologies by large-scale gene expression profiling in the Connectivity Map. PLoS Biology 15, (2017).

4. Gilbert, L. A. et al. Genome-Scale CRISPR-Mediated Control of Gene Repression and Activation. Cell 159, 647–661 (2014).