Emergent interactions due to resource competition in CRISPR-mediated genetic activation circuits

Abstract

CRISPR-mediated gene regulation has gained considerable attention due to its scalability, allowing to create increasingly large genetic circuits. Unintended interactions due to competition for the dCas9 resource among different small guide RNAs have been characterized extensively for CRISPRmediated repression (CRISPRi). Such an analysis is to a large extent missing for CRISPR-mediated activation (CRISPRa). In this paper, we model CRISPRa considering two required shared resources (dCas9 and an activator protein), and identify the interaction graphs that emerge through resource competition. The presence of two shared resources among multiple scaffold RNAs (scRNA) is responsible for two main phenomena. First, we mathematically prove the existence of a “self-sequestration” effect, wherein an scRNA represses its own target gene instead of activating it, thereby negating the CRISPRa function. Second, we demonstrate that unwanted repression of non-target genes is substantially stronger when compared to a scenario with a single resource. These results indicate that new control approaches to concurrently regulate multiple resources will be useful for mitigating the undesirable effects of resource competition in CRISPRa.