Rationally-engineered reproductive barriers using CRISPR & CRISPRa: an evaluation of the synthetic species concept in Drosophila melanogaster

Abstract

The ability to erect rationally-engineered reproductive barriers in animal or plant species promises to enable a number of biotechnological applications such as the creation of genetic firewalls, the containment of gene drives or novel population replacement and suppression strategies for genetic control. However, to date no experimental data exist that explores this concept in a multicellular organism. Here we examine the requirements for building artificial reproductive barriers in the metazoan model Drosophila melanogaster by combining CRISPR-based genome editing and transcriptional transactivation (CRISPRa) of the same loci. We directed 13 single guide RNAs (sgRNAs) to the promoters of 7 evolutionary conserved genes and used 11 drivers to conduct a misactivation screen. We identify dominant-lethal activators of the eve locus and find that they disrupt development by strongly activating eve outside its native spatio-temporal context. We employ the same set of sgRNAs to isolate, by genome editing, protective INDELs that render these loci resistant to transactivation without interfering with target gene function. When these sets of genetic components are combined we find that complete synthetic lethality, a prerequisite for most applications, is achievable using this approach. However, our results suggest a steep trade-off between the level and scope of dCas9 expression, the degree of genetic isolation achievable and the resulting impact on fly fitness. The genetic engineering strategy we present here allows the creation of single or multiple reproductive barriers and could be applied to other multicellular organisms such as disease vectors or transgenic organisms of economic importance.

Gene activation by dCas9-CBP and the SAM system differ in target preference

“…However, transgenic expression of dCas9-CBP resulted in sterility and non-specific wing phenotypes with tub-Gal4 or vg-Gal4, respectively. Similarly, expression of dCas9-CBP in the eye was recently shown to cause a phenotype in the absence of gRNA 25 . Thus, finding appropriate conditions for CBP HAT domain expression in vivo awaits further experiments.…”

Transgene Biocontainment Strategies for Molecular Farming

“…Attempts have also been made at constructing a synthetic species of D. melanogaster, where an artificial reproductive barrier is engineered, however the goal of complete genetic isolation wasn't achieved. The main difficulty proved to be getting strong activation of a lethal gene without the fitness costs associated with broad expression of the transactivating CRISPR machinery (Waters et al, 2018).…”

Recent advances in threshold-dependent gene drives for mosquitoes

“…Lethality is prevented in the homozygous transgenic line by recoding both copies of the target promoter region such that it is no longer recognised by the CRISPRa complex. This system has promise as a drive system, in tested organisms (yeast and fruitflies) [ 36 , 40 ] there are a large number of promoter regions that readily respond to CRISPRa [ 41 ], resulting in a wide array of potential targets. However, evidence so far indicates that fine-tuning activation against basal toxicity, in order to make a viable transgenic strain for gene drive, is less straightforward.…”

“…CRISPRa in D. melanogaster showed only partial efficacy — as they could not be readily assembled into a homozygous viable line in order for a drive to be tested. Those lines where the CRISPRa transgenes were homozygous viable did not show complete lethality when crossed to wildtype, slowing the potential spread of this drive system [ 40 ].…”

“…Looking forward, development of alternative strains may improve the performance of the system. In comparison, UD gene drives are markedly less developed, with no field trials or established systems in Aedes , but many promising laboratory proof-of principles in Drosophila [ 22 , 23 , 30 , 40 ]. Invasion thresholds for UD systems are higher than those predicted for Wolbachia at equal fitness, making the system more local, controllable and reversible, at the cost of needing somewhat higher initial release numbers ( Figure 2 ).…”