Five Results Unmasked by Endogenous Gene Targeting (That Overexpression Missed!)

Chris Thorne Gene Editing Community Specialist Horizon Discovery

Historically, the roles of genes and proteins in cell function have often been explored using overexpression experiments (also known as transgenesis or ectopic expression). While these approaches have resulted in huge leaps in our understanding of many biological processes, conclusions can in some instances be misleading due to the protein of interest being expressed at concentrations much higher than those found in an endogenous context.

This overabundance can result in a number of potential artifacts. For example, miscompartmentalization and trafficking can lead to proteins interacting with binding partners that they would never normally see. This in turn can cause aberrant downstream effects that are simply a consequence of the overexpression.

Alternatively, the abundance can produce signaling pathway activation at levels beyond those seen in physiological systems, or lead to down-regulation of other signaling pathways in an effort to maintain homeostasis and avoid apoptosis or senescence, which again can confound results.

Endogenous gene targeting on the other hand allows researchers to investigate specific genetic modifications in an endogenous context. By “knocking in” these modifications under the control of their existing regulatory elements, expression occurs at physiological levels, and the true functional contribution of this mutation can be studied. This becomes all the more important when considering the role of genes in disease.

The pitfalls of using overexpression models are most stark when directly contrasted with the endogenously modified equivalents. This is illustrated well by the five publications listed below:

• KRAS, one of the most extensively characterized oncogenes, is a prime example of where overexpression can lead to incorrect conclusions. This was highlighted in two key papers, which demonstrate that unlike overexpression of oncogenic variants of KRAS (which singularly transformed normal cells), endogenous knock-in of the same variants, either KRAS G12D (Arena, et al.¹) or KRAS G12V (Konishi, et al.²), does not result in transformative phenotypes. This is consistent with accepted wisdom that cancer is a multistep disease resulting from numerous genetic mutations.
• Similarly, overexpressed oncogenic PIK3CA mutants such as H1047R have been found to induce a large growth induction phenotype in normal mammary epithelial cells. However, Di Nicolantonio, et al.³ found that knock-in of the same oncogenic mutations at the endogenous loci results in a much milder, nontransformative phenotype.
• Endogenous tagging of both clathrin light chain A (CLTA) and dynamin 2 (DNM2) by Doyon, et al.⁴ found that the efficiency of clathrin-mediated endocytosis in mammals shows far greater similarity to yeast models than had originally been reported using overexpression systems (with genome-edited cells exhibiting enhanced endocytic function, dynamics, and efficiency relative to the transgenic counterparts).
• Most recently Smurnyy, et al.⁵ used endogenously targeted cells to explore the contribution of ERCC3 mutations to triptolide (an antiproliferative agent) resistance. The recessive nature of these mutations meant that overexpression had no effect on drug sensitivity. Endogenous, homozygous knock-ins however resulted in a 10 to 20 fold increase in IC50 relative to wild-type cells.

What these studies emphasize are the risks of working with nonphysiological overexpression models when investigating the role of genes and mutations in cellular processes. In the case of oncogenes it is apparent that overexpression can dramatically overestimate the importance of single genes in tumorigenesis.

It is worth noting the scarcity of examples where overexpression models have been directly compared with their endogenous knock-in counterparts. Perhaps this is because historically genome editing has been a relatively complex and therefore niche pursuit.  It begs the question, therefore: What else has been missed?

Hopefully the answer to this question is not so far away. The recent and rapid adoption of CRISPR/Cas9 as a genome editing approach, as well as the refinement of rAAV for precision knock-ins, mean it is reasonable to assume that the number of examples will grow. As more labs get to grips with genome editing it will be interesting to see which stones are overturned by the proliferation of these biologically relevant models. 

Chris Thorne ([email protected]) is the Gene Editing Community Specialist at Horizon Discovery.

References:
1. Arena, S et al., 2007. Knock-in of oncogenic Kras does not transform mouse somatic cells but triggers a transcriptional response that classifies human cancers. Cancer Res. 67(18), p8468-76
2. Konishi, H et al., 2007. Knock-in of mutant K-ras in nontumorigenic human epithelial cells as a new model for studying K-ras mediated transformation. Cancer Res. 67(18), p8460-7
3. Di Nicolantonio, F et al., 2008. Replacement of normal with mutant alleles in the genome of normal human cells unveils mutation-specific drug responses. PNAS. 105(52), p20864-9
4. Doyon, J.B et al., 2011. Rapid and efficient clathrin-mediated endocytosis revealed in genome-edited mammalian cells. Nat Cell Biol. 13(3), p331-7
5. Smurnyy, Y et al., 2014. DNA sequencing and CRISPR-Cas9 gene editing for target validation in mammalian cells. Nat Chem Biol. 10(8), p623-5