CIRKO: A chemical-induced reversible gene knockout system for studying gene function in situ

  1. Hui Shi
  2. Qin Jin
  3. Fangbing Chen
  4. Zhen Ouyang
  5. Shixue Gou
  6. Xiaoyi Liu
  7. Lei Li
  8. Shuangshuang Mu
  9. Chengdan Lai
  10. Quanjun Zhang
  11. Yinghua Ye
  12. Kepin Wang
  13. Liangxue Lai

  • Curated by eLife

    eLife logo

    eLife assessment

    Conditional deletion and reactivation of a gene in situ remain challenging, and this study therefore addresses a gap in the genetic tool box. The authors introduce a reversible conditional gene inactivation and reactivation method using sequential expression of recombinases, with doxycycline treatment terminating gene transcription, while doxycycline and tamoxifen addition restore gene expression.

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Conditional loss and restoration of function are becoming important approaches for investigating gene function. Given that reversible conditional gene knockouts in cells required complicated manipulation, conditional inactivation and reactivation of a gene in primary somatic cells with limited proliferative capacity and in animal models remain difficult to achieve. Here, we first developed a reportable and reversible conditional intronic cassette (ReCOIN), wherein inactivation and reactivation of the gene are mediated via sequential expression of Cre and Flp recombinases, respectively. The expression pattern of the target gene can be monitored by direct visualization. To simply and tightly control temporal expression of the recombinases, on the basis of ReCOIN, we further presented a dual chemical-induced reversible gene knockout system (CIRKO) by insertion of reverse tetracycline transcriptional activator (rtTA) and tetracycline response element (TRE)-controlled Cre and FlpoERT2 recombinases cassettes into Rosa26 and Hipp11 loci of cells, respectively, in which transcription termination of the target gene can be induced at a specific stage in the presence of doxycycline, while gene restoration is achieved in the presence of doxycycline and tamoxifen simultaneously. This system provides a simple, rapid, and flexible gene switch for studying gene function in situ both in vitro and in vivo .

Impact statement

A novel chemical-induced reversible gene knockout system provides a simple, rapid, and flexible gene switch to facilitate the study of gene function in primary somatic cells in vitro , embryos in vitro or in vivo , and animals in vivo .

Article activity feed

  1. eLife

    eLife assessment

    Conditional deletion and reactivation of a gene in situ remain challenging, and this study therefore addresses a gap in the genetic tool box. The authors introduce a reversible conditional gene inactivation and reactivation method using sequential expression of recombinases, with doxycycline treatment terminating gene transcription, while doxycycline and tamoxifen addition restore gene expression.

  2. eLife

    Reviewer #1 (Public Review):

    The authors introduced a dual chemical-induced reversible gene knockout method (CIRKO) using a reportable and reversible conditional intronic cassette (ReCOIN). They could use Cre to delete genes, and another recombinase Flp to recover genes. This would provide a means to reversibly control gene expression rather than deletion. Another strength of this system is that GFP is built in to allow investigators to see if the gene is inactivated or activated, thus monitoring gene status by visible fluorescence. The authors have used this method mainly in pig gene manipulation. It would be great if this system could also be tested in mouse genes, as Cre-loxP system for gene deletion is mostly used in mouse. Thus demonstration of this method in mouse gene manipulation would broaden its future application. Overall, this work provides a flexible gene switch system for in vitro and in vivo gene function study.

  3. eLife

    Reviewer #2 (Public Review):

    Shi et al present novel genetic tools to carry out conditional reversible genetics. These allow the Cre-dependent inactivation of a given gene of interest, together with reporter expression, and the posterior Flp-recombinase-dependent deletion of the ReCOIN cassette and reactivation of the full gene expression. The process is based on the alternative splicing of an exon containing the ReCOIN cassette in the default mode, followed by sequential recombination/reversion of this cassette by Cre, leading to the expression of a reporter in the cells having a gene deleted. Subsequently, Flp recombinase can be used to delete the ReCOIN cassette, restoring the wild-type gene. The strategy is largely based on the XTR system (Robles-Oteiza et al, 2015) but with the difference that it also allows the targeting of genes composed by a single exon (without introns), as the construct is not targeted to an intron but instead integrated into one of the first exons of the gene of interest.

    The authors also design and generate a single genetic construct (CIRKO), that enables doxycycline-inducible expression of Cre and FlpOERT2, followed by tamoxifen activation of the latter. After doxycycline administration, inversion of the ReCOIN allele truncates the gene and fuses it with a reporter and a polyA cassette, and after tamoxifen induction of FlpOERT2, the ReCOIN cassette is deleted, thus restoring the wildtype gene.
    Although the authors provide convincing evidence of the sequential recombination process, several aspects of the data analysis need to be improved and controlled.

    1. Authors did not evaluate whether the integration of the ReCOIN construct in the exon of the gene of interest affects the gene's endogenous expression levels. This needs to be carefully assessed as it may generate loss-of-function or hypomorphic alleles, even in the absence of any other manipulation. Data presented in Fig.1-Sup 5 shows that the Cherry levels are much lower in the unrecombined allele containing the ReCOIN than after full recombination and expression of the native wildtype allele, suggesting that the simple integration of the ReCOIN cassette may decrease gene expression.
    2. One of the main problems of this and previous COIN or XTR systems is that the expression of the reporter after flipping the ReCOIN allele (to produce gene knockout) is often too weak because its expression is driven by the endogenous gene promoter and alternative splicing, which for most genes does not allow the clear separation between mutant (reporter+) and wildtype/reversed cells. Another caveat of this system is that wild-type and reversed (gene-reactivated) cells are indistinguishable.
    3. Although the authors use cell lines to demonstrate the expression of the targeted gene of interest before and after each of the sequential recombination events of the allele by immunofluorescence, there is no quantitative data reflecting the efficiency and reliability of each event in the entire cell population. Data is mainly obtained from a few single cell-derived clones, rather than the entire population of transfected cells.
    4. There is no data showing that their system works as predicted in vivo.
  4. eLife

    Reviewer #3 (Public Review):

    In the research described in this manuscript, Shi and colleagues were attempting to develop a versatile and flexible method for generating conditional and reversible gene knockouts. They wanted their method to be widely applicable and easily adapted to any target gene of interest. In addition, they wanted to demonstrate the use of their new method in several different experimental contexts, reinforcing their conclusions about its value. In pursuit of these goals, the authors modified a method (COIN) in which an artificial intron containing a Cre-dependent gene-trap cassette is inserted into an exon of the target gene. In the modified ReCOIN method, the gene trap cassette is flanked by target sites of Flp recombinase. Cre recombination inverts the gene trap cassette, resulting in the disruption of the targeted gene. Subsequent Flp recombination deletes the gene trap cassette, restoring the expression of the targeted gene. The authors also devised a strategy (CIRKO) to permit rapid, non-invasive control of the ReCOIN system. In general, the authors have achieved their goals. The experiments in the manuscript are well-designed and clearly described, and they highlight the strengths of the strategy. However, a few limitations of the strategy and the experimental analyses are also clear:

    1. The ReCoin module retains an antibiotic resistance cassette driven by the PGK promoter, which is a powerful ubiquitous promoter with bidirectional activity. In the original COIN module, the resistance cassette is deleted by Flp recombinase, but this is not possible in ReCOIN where Flp has been co-opted for gene regulation. In a variety of contexts, retained PGK-driven antibiotic cassettes have been shown to have unpredictable effects on the expression of surrounding genes. It would perhaps have been better if the ReCOIN module had been designed so that the resistance cassette was deleted by a third recombinase such as VCre or PhiC31. The possibility of ectopic gene expression or downregulation driven by the PGK promoter should be kept in mind when characterizing new ReCOIN alleles.

    2. Somewhat related to point 1, the authors performed an experiment in transiently transfected cells to demonstrate that insertion of the ReCOIN module does not affect the expression levels of an mCherry reporter. However, the metric they reported, % mCherry+ cells, speaks more to transfection efficiency than expression levels. Mean fluorescence intensity might have been more informative.

    3. In the section describing Cas9-ReCOIN, the authors mention the need to temporally control Cas9 expression, because persistent Cas9 expression can result in genomic instability. However, it is not clear that ReCOIN offers any advantage over the original COIN module in this context. In experiments where a Cas9 plasmid is transfected, Cre recombination allows the Cas9 to be switched off, but Flp recombination, turning Cas9 back on permanently, would seem to have no experimental value. Alternatively, in a cell line with Cas9 stably integrated into Rosa26 or a similar safe harbor locus, it would be desirable to have Cas9 temporarily turned on (Off-On-Off). Unfortunately, reCOIN seems to offer the ability to temporarily turn Cas9 off (On-Off-On).

    4. Although live pigs containing a ReCOIN allele of TP53 were generated, experiments showing recombination of ReCOIN alleles were all performed in cultured cells or pre-implantation embryos. As yet, the ReCOIN/CIRKO strategy has not been fully validated in postnatal animals.

    5. The CIRKO strategy allows rapid control of ReCOIN to turn gene expression off and on via dosing with doxycycline and tamoxifen. This non-invasive temporal control of gene expression has obvious value in both cultured cells and model organisms. However, as currently described CIRKO cannot be used for cell type-specific knockouts, because Cre and Flp expression is regulated by ubiquitous (though chemically inducible) promoters.

  5. Version published to 10.1101/2022.08.31.506064v1 on bioRxiv