The ability to control gene expression and analyze protein coding gene function inside cells is a critical step in our understanding of normal physiology and disease pathology. CRISPR/Cas9 technology has revolutionized gene editing approaches and makes generation of gene knock-outs or knock-ins straightforward (Doudna and Charpentier, 2014). Combination of CRISPR with single-stranded DNA (ssDNA) donors has proven to be the most precise and efficient method of on-target integration (Quadros et al., 2017; Li et al., 2017), although knock-in approaches still suffer from low efficiencies, especially when aiming to achieve homozygosity (Gurumurthy et al., 2019). Fluorescent protein (FP) knock-in at endogenous loci leading to protein-fusion allows functional enrichment of complete knock-in events and study of native protein function using fluorescent microscopy (Roberts et al., 2017; Leonetti et al., 2016). However, they only complement loss-of-function/gain-of-function methods, and their usefulness is limited by brightness of the chosen FP. Conditional loss-of-function methods mainly rely on CRISPR/Cas9 combination with FLP-frt or Cre-LoxP systems (Quadros et al., 2017). However, these are labor intensive and mediated by recombination, which is by nature irreversible thus complicating attempts to rescue gene expression. In addition, to understand the full spectrum of gene functions, it is advantageous to reversibly tune expression levels, visualized by fluorescence microscopy, in a spatiotemporal context and not only depend on binary on/off approaches. Such levels of control can be partly offered by the auxin-inducible degron (AID) system (Nishimura et al., 2009) introduced as a tag into the endogenous locus via CRISPR/Cas9 (Röth et al., 2019). Here, an exogenously expressed plant auxin receptor F-box protein triggers ubiquitylation and degradation of target proteins fused to an AID tag on addition of the plant hormone auxin. However, this technology has its own limitations and caveats; for example, poor spatial control due to diffusion of auxin, significant basal degradation without auxin, and/or incomplete auxin-induced protein depletion.

Ideally, gene expression control and visualization of its protein product should be generated by a single gene modification (e.g. CRISPR knockin) at the endogenous locus and allow temporal, conditional, and reversible gene inactivation of all protein-coding transcripts, on demand, without the drawbacks of AID system. Such an approach is particularly useful for studying highly similar protein coding genes where their complementary and specific functions remain poorly understood, such as for members of Rab11 family of small GTPases: Rab11a, Rab11b, and Rab25 (Rab11c). Rab11s share high amino acid identity (Rab11a:Rab11b 89%, Rab11a:Rab25 66%, Rab11b:Rab25 61%), are known to play key roles in membrane transport, localize to recycling endosomes, and have been identified as important players in the cellular basis of an ever-increasing number of human disorders, including cell migration/invasion and cancer (Wilson et al., 2018; Kelly et al., 2012). Several Rab11 family isoforms have been reported (Ota et al., 2004), and specific antibodies are hard to obtain without cross-reactivity. Moreover, much what is known about their regulation and dynamics in live cells has been achieved with the aid of transiently transfected cDNA (complementary DNA)-mediated overexpression which may lead to non-physiological artifacts (Ratz et al., 2015; Gibson et al., 2013).

To overcome the limitations of previously established gene expression control technologies, we generated tractable single CRISPR knock-in based strategies for tunable and reversible gene expression control capable of knock-out/rescue and endogenous gene re-activation. These modifications are ideal for microscopy when combined with FP knock-in, are free of transfection artifacts, preserve post-transcriptional and post-translational control, and avoid comprehensive and labor-intensive genotyping post-editing of clones. Specifically, we harnessed CRISPR/Cas9 technology for targeted knock-in of the third-generation tetracycline-inducible system (Tet-On 3G) promoter TRE3GS followed by an FP coding region. This promoter lacks binding sites for endogenous mammalian transcription factors, is effectively silent in the absence of induction, and any transcription driven by endogenous active promoter generates non-sense frameshift. Simultaneous or sequential silencing of additional alleles was achieved by additional ssDNA donor coding antibiotic resistance to attain complete and selectable functional knock-out. Spatiotemporal control was achieved by combination with a photoactivatable (PA)-Tet-OFF/ON system (Yamada et al., 2018), even in cells within 3D microenvironments. We further established a tunable dual-input system combining TetON and AID, termed DExogron (DExCon [Doxycycline-mediated endogenous gene Expression Control] combined with auxin-mediated targeted protein degradation), that enhanced switch off kinetics of the TetON system, allowing rapid complete protein depletion or adjustment of expression to physiological levels on demand.

Our optimized pipeline to deliver this suite of CRISPR tools was applied to modulate endogenous gene expression of Rab11 family members, demonstrating that re-activation of endogenous Rab25 expression can promote the invasive migration of cancer cells, and revealing new insight into the localization and functions of Rab11a, Rab11b, and Rab25.

Deconvolving the relative functions of genes within highly related gene families has proven difficult, for example, the Rab11 family has a complex relationship with cancer, with several potential oncogenic and tumor suppressive functions identified (Kelly et al., 2012; Jin et al., 2021; Cheng et al., 2004). However, gene editing offers potential solutions. In order to identify the best knockin strategies to visualize the Rab11 gene family, we first analyzed the expression patterns of the Rab11s across tissues and in cancer using UCSC Xena, an online exploration tool for multi-omic data (Goldman et al., 2020). Rab11a, Rab11b, and Rab25 were significantly enriched in ovarian cancer, both for primary and recurrent tumors (Figure 1—figure supplement 1A). Bioinformatic analyses suggested the existence of additional potentially protein-coding human Rab11a/b splice variants (Figure 1—figure supplement 1B). Analysis of expression quantitative trait loci (eQTLs) revealed abundant expression of two Rab11b-coding and five Rab11a-coding transcripts across multiple healthy tissues while Rab25 is more restricted, but expression is increased across multiple types of cancer (Figure 1—figure supplement 1C, https://doi.org/10.48420/16988617). Similarly, several Rab11a/Rab11b/Rab25 splicing isoforms are enriched in ovarian cancer compared to normal tissue (Figure 1—figure supplement 1D). All significantly expressed transcripts share an intact N-terminus; however, for some splice variants, amino acids important for GTP/GDP binding (which allow oscillation between active/inactive states) or for the attachment to the outer leaflet of endosomal membranes are missing, which could therefore profoundly alter Rab11s function (Figure 1—figure supplement 1E). Thus, to capture the full physiological picture of the combination of Rab11s protein-coding transcripts, we fluorescently tagged the N-terminus of Rab11s using CRISPR knockin (Figure 1A). We used long ssDNA donor for precise and efficient on-target integration (Quadros et al., 2017; Li et al., 2017; Miura et al., 2018) with flanking homologous arms 150–300 bases prepared via RNA intermediates (Figure 1B). Initial testing of different approaches led to our optimized protocol (Figure 1C) combining magnetofection of long ssDNA and high fidelity (Hifi) Cas9 delivered as ribonucleoprotein (RNP) complex with guideRNA (crRNA:tracrRNA). This combination reduces off-target effects to non-detectable levels and leads to Hifi hetero/homozygous knockin (Vakulskas et al., 2018). To avoid comprehensive and labor-intensive genotyping post-editing of clones (Miura et al., 2018), we took the advantage of fluorescent readouts to monitor and enrich correct homology directed repair (HDR) by fluorescence-activated cell sorting (FACS) (Li et al., 2017; Leonetti et al., 2016). Pilot testing led to a high rate of successful knock-in human embryonic kidney (HEK)293T cells with expected localization of Rab11a (10–20%) or Rab11b (2–10%), respectively (Figure 1—figure supplement 1F). For further work, we selected a commonly used cell line model for ovarian cancer, A2780, and knocked-in mNeonGreen or mCherry after the start codon of Rab11a or Rab11b, respectively (Figure 1D). This integration, although less efficient (0.2–1%) compared to HEK293T, was specific as all sorted cells exhibited correct Rab11a/b localization, and our protocol was also successful in a second ovarian cancer cell line that has defective homologous recombination (BRCA-1 mutant COV362 Beaufort et al., 2014; Figure 1E). Western blotting with antibodies that recognize Rab11a, or both Rab11a and Rab11b, confirmed knock-in (Figure 1F). Consistent with the literature (Li et al., 2017; Canaj et al., 2019), a proportion of cells showed truncation of the linker flanking the FP (Figure 1G) but not in the integrated homologous arms, indicating that sorting cells for fluorescence (or another functional protein) is important for establishing in frame knock-in at the N-terminus preserving protein function. This potential caveat was demonstrated by our effort to knock-in an additional non-fluorescent component, antiGFPnanobody, together with mCherry. Although knock-in to the Rab11a locus showed the expected localization of mCherry-Rab11a in all sorted cells (Figure 1—figure supplement 2A), integration of the antiGFPnanobody sequence was not complete in ~90% of mCherry-positive cells (Figure 1—figure supplement 2B-C).

Figure 1 with 2 supplements see all

Download asset Open asset

DExCon supresses endogenous gene expression and re-activates in a doxycycline-dependent manner

Being able to visualize endogenous Rab11a/b, we analyzed the sequences of different promoters for their ability manipulate gene expression at the endogenous locus. The most promising candidate was the advanced non-leaky TRE3GS promoter of the third-generation tetracycline-inducible system (Tet-On 3G) as it offered conditional and reversible control regulated by the small molecule doxycycline (dox). Analysis of 365 bp TRE3GS promoter revealed four potential translation initiation ATG codons, all of which could be designed out of frame to the endogenous start codon, leading to multiple premature stop codons in transcripts driven by basal expression of the endogenous active promoter (Figure 2A). Tet-On 3 G transactivator was stably delivered to A2780 cells by lentivirus in tandem with mTagBFP (separated by the self-cleaving peptide T2A) and cells selected for mTagBFP fluorescence. The TRE3GS promoter sequence was introduced into the donor sequence upstream of mNeonGreen or mCherry and delivered by long ssDNA-mediated knock-in to the Rab11a or Rab11b locus, respectively, in A2780 cells. Sorted cells showed expression of corresponding FP-Rab11 with a perinuclear vesicle-enriched localization matching the endogenous protein, but only in the presence of dox, and decreased levels of non-modified endogenous Rab11a/b indicating that DExCon blocks expression at the endogenous locus (Figure 2B and C; Figure 2—figure supplement 1A).

Figure 2 with 2 supplements see all

Download asset Open asset

To determine whether the DExCon system preserves the expression of splice variants of the N-terminally modified genes from the endogenous locus, we analyzed Rab11b splice variants in A2780 cells by quantitative PCR (qPCR) and semi-quantitative reverse transcription PCR (Figure 2—figure supplement 1B). We identified two Rab11b splice variants that could produce protein coding products of similar molecular weight to the single band observed in A2780 cells by western blotting (201 [24.5 kDa] and 202 [20 kDa]). The main Rab11b protein-coding splice variant 201 is ~4× more abundant than the protein-coding splice variant 202, consistent with eQTLs analysis of ovarian cancer (Figure 1—figure supplement 1D), and DExCon modification did not significantly influence their relative abundance (Figure 2—figure supplement 1B). Specific primers that detect splice variants 201, 202, and a third (short and likely untranslated) 204 variant indicated that the latter is expressed, but again the DExCon module did not change the ratio of Rab11b splice variants 201/202/204 upon dox induction. This suggests that the DExCon module can recapitulate the spectrum of splice variant expression from the Rab11b locus and provides an advantage over generation of a knock-out cell line and subsequent tet-inducible protein-coding rescue from a viral vector containing cDNA.

To exemplify the versatility of our CRISPR knock-in pipeline, we generated double mNeonGreen/mCherry knock-in at Rab11a/Rab11b loci, respectively, with and without the DExCon module (Figure 2E and F). In all combinations, endogenous Rab11a/Rab11b localization driven by dox-induced or endogenous promoter was significantly correlated, and both Rab11a and Rab11b concentrated in vesicles at the perinuclear recycling compartment (Figure 2E and F). By contrast, transient transfection of plasmids coding FP-Rab11a/Rab11b gave very bright overlapping signals, but large vesicles were observed particularly toward the cell periphery (Figure 2G). These findings were confirmed when red/green-FPs were switched (Figure 2—figure supplement 2A, B) and suggest that DExCon can induce expression from the endogenous gene locus that preserves physiological distribution of the protein within cells with signal intensities compatible with low light/exposure live microscopy (Figure 2—figure supplement 2C,D; Videos 1–2). This was further demonstrated by imaging Rab11a/b vesicles transporting the recognized cargo transferrin (TFN) (Figure 2—figure supplement 2D; Video 3) and by imaging cells in 3D environments. In 3D cell-derived matrix (CDM), higher exposures were required for visualization of conventional double knock-ins, and this led to a lag in capture between channels and significant bleaching of mCherry over time (Figure 2—figure supplement 2F; Video 4). However, DExCon knock-ins showed higher intensity and could therefore be visualized at lower exposure, avoiding caveats associated with lag (mis-localization) and photobleaching, revealing that while Rab11a vesicle intensity is consistent within polarized cells, Rab11b intensity is increased at the tips of protrusions relative to the region of the cell in front of the nucleus (Figure 2—figure supplement 2E).

Video 1

Download asset

Video 2

Download asset

Video 3

Download asset

Video 4

Download asset

We next explored the ability of DExCon to deliver full-length knock-in of GFP nanobody-mCherry sequence to Rab11 gene loci, as its flanking by an upstream TRE3GS sequence could improve the likelihood of complete integration and positive selection. Upon dox induction, all sorted cells exhibited the expected localization of Rab11a or Rab11b (Figure 2H) and full-length size on western blot (Figure 2I) indicating that encoding functionality at each of the 5′ and 3′ ends of ssDNA donor allows enrichment of accurately modified cells even for non-fluorescent coding genes.

Re-activation of silenced genes by DExCon and LUXon (light responsive DExCon)

A2780 ovarian cancer cells show abundant Rab11a/b expression while the more restricted family member Rab25 is not expressed (Figure 3A). To study endogenous Rab25 in this cell line, we hypothesized that our DExCon strategy could re-activate Rab25 expression from the genome (Figure 3B). TRE3GS-mNeonGreen was delivered by long ssDNA-mediated knock-in to the Rab25 locus in A2780 cells expressing Tet-On 3G transactivator. Dox treatment led to mNeonGreen fluorescence with expected perinuclear localization of mNeonGreen-Rab25 in all sorted cells (in 2D and 3D-CDM; Figure 3C) and specificity of integration was confirmed by western blot (Figure 3—figure supplement 1A) and sequencing (Figure 3—figure supplement 1B).

Figure 3 with 1 supplement see all

Download asset Open asset

We next explored tunability of Rab25 expression. Exposing cells to different dox concentrations appeared to have a dose-dependent effect when analyzed by western blotting (Figure 3—figure supplement 1C); however, fluorescence microscopy revealed that the difference was primarily due to the number of responding cells (Figure 3D). Nevertheless, we noticed a positive correlation between expression of Tet-On 3G transactivator, reported as mTagBFP fluorescence, and mNeonGreen-Rab25 (Figure 3—figure supplement 1D; R=0.43, p=0.0003). Taking advantage of this, we sorted for mNeonGreen-Rab25 expression and analyzed Rab25 and mTagBFP levels. Dox-induced Rab25 expression levels in ‘low’ sorted cells were comparable to endogenous Rab25 expression levels in OVCAR-3 ovarian cancer cell lines over 48 hr (Figure 3E) and Rab25 expression levels correlated with Tet-On 3G levels as expected (Figure 3F). The ability to re-induce previously sorted expression levels was preserved over several months under standard cell culture conditions and reached maximal fluorescence after 72 hr (Figure 3E). These results indicate that the DExCon approach was able to re-activate endogenous silenced Rab25 on demand and levels could be tuned to match endogenous, physiologically relevant expression levels.

While DExCon was able to temporally control Rab25 expression, this method lacks spatial control. We therefore explored the newly developed PA-Tet-OFF/ON system (Yamada et al., 2018), which has not previously been applied at an endogenous locus, to establish a light inducible ‘LUXon’ variant of DExCon. This system integrates the cryptochrome 2 cryptochrome-interacting basic helix-loop-helix 1 light-inducible binding switch with Tet-binding domains of Tet repressor (TetR) (residues 1–206) as the split DNA-binding domain, and the transcription activation domain of p65 (p65 AD) (Figure 3G). We tested PA-Tet-OFF or PA-Tet-ON and versions with improved blue-light sensitivity (ON2; OFF2), stably delivered by lentivirus. Cells stably expressing PA-Tet-ON2, PA-Tet-OFF1, or two were then edited by mNeonGreen DExCon knock-in to the Rab25 locus, illuminated by blue light for 24 hr and sorted, all showing the expected Rab25 endosomal localization (Figure 3H; Figure 3—figure supplement 1E-G). Cyclical exposure to light/dark led to repeatable re-silencing of Rab25, and the level of re-activation could be modulated by timing of blue light which is one of the major advantages of a light-inducible gene expression system (Figure 3—figure supplement 1E-G). As these systems responded to the weak blue-light exposure used (~6.8 W/m²), we found the additional layer of dox control to prevent leakiness of the system very practical. Dox addition with blue light was necessary to re-activate Rab25 in PA-Tet-ON2 system (Figure 3—figure supplement 1E), while dox addition to PA-Tet-OFF1/2 completely blocked expression (Figure 3—figure supplement 1F,G). Interestingly, in addition to the expected Rab25 localization, we observed some Rab25 positive vesicles in close proximity of the nucleus. These vesicles were significantly less mobile compared to others suggesting that the nucleus could under some circumstances sterically block movement of vesicles (Figure 3H; Videos 5 and 6).

Video 5

Download asset

Video 6

Download asset

To demonstrate local induction of gene expression from the endogenous locus, we utilized a modified 3D cell-zone exclusion assay (Gemperle et al., 2019). In this assay, a monolayer of the cells grown on a thick layer of collagen supplemented with non-labeled fibronectin is wounded and then overlaid with another layer of collagen supplemented with far-red labeled fibronectin (Figure 3I). The advantage of this approach is that system can be partly illuminated and Rab25 expressing cells can be monitored migrating toward the far-red fibronectin labeled cell-free area. This approach allowed us to demonstrate that Rab25 expression can be induced locally in 3D matrices and is suggestive of a link between Rab25 re-expression levels and increased invasiveness (Figure 3J; Video 7). To quantify this, we next illuminated cell spheroids overlaid by 3D matrix, modulating the exposure to blue light (Figure 3K–N). Longer exposure to blue light led to brighter mNeonGreen signal in both PA-Tet-OFF/ON systems with expected Rab25 localization in polarized/elongated cells interacting with matrix fibrils (Figure 3K; Figure 3—figure supplement 1H). When all cells were pre-illuminated 2 days before invasion to induce Rab25 expression, we saw a significant dose-dependent effect of Rab25 induction via light exposure for over 6 hr on invasion that was not observed in the presence of dox in the TET-OFF system (Figure 3L–N). Adjusting the illumination protocol (Figure 3—figure supplement 1I), activating Rab25 expression with constant blue light for 24 hr before spheroid formation/invasion, led to a light-dose dependent effect where the most invasive cells were those illuminated the longest approach (Figure 3—figure supplement 1J,K). This demonstrates that tight control of gene expression levels and their impact on complex cell behavior can be easily achieved/studied by the LUXon system.

Video 7

Download asset

Triple knock-in determines Rab11 family colocalization

After generating mNeonGreen-Rab11a/mCherry-Rab11b double knock-in cells, we challenged ourselves to generate triple knock-in A2780 cells (Rab11a/Rab11b/Rab25; Figure 4A). To do so, mNeonGreen-Rab11a/mCherry-Rab11b double knock-in cells were further genetically manipulated to introduce mTagBFP2 as part of the DExCon module controlling Rab25 expression. Dox treatment successfully specifically re-activated Rab25, visible as mTagBFP2, on the background of double knock-in mNeonGreen-Rab11a/mCherry-Rab11b cells (Figure 4—figure supplement 1A,B). Interestingly, activation of Rab25 expression increased the localization of Rab11a and Rab11b at ERC spots, suggesting that Rab11a/b/25 networks are interconnected and can be modulated by Rab25 expression (Figure 4—figure supplement 1C), consistent with redistribution of Rab11a in Madin-Darby canine kidney (MDCK) cells upon the introduction of Rab25 expression (Casanova et al., 1999). Because DExCon-mTagBFP2-Rab25 fluorescence intensity was less intense than that of DExCon-mNeonGreen-Rab25 and Rab11a/b knock-ins, triple knock-in cells were re-sorted for bright mTagBFP2 signal and dox treatment kept over 94 hr prior fluorescence microscopy highlighting the tunability of expression with DExCon (Figure 4—figure supplement 1A,B).

Figure 4 with 1 supplement see all

Download asset Open asset

Rab11 family members have been reported to colocalize with TFN receptor and control its recycling (Ullrich et al., 1996; Schlierf et al., 2000). To compare the colocalization of Rab11 family members and their potential contribution to the TFN receptor recycling, we allowed cells to internalize Alexa-647 labeled TFN and analyzed colocalization using AiryScan confocal and lattice light sheet microscopy (Figure 4B–F; Figure 4—figure supplement 1B-E; Videos 8–12). While there was clear colocalization between Rab11a/b/25 and TFN, some vesicles were highly enriched for Rab25 compared to Rab11a/b but lacked TFN. Rab25 enriched vesicles negative for TFN were found in close proximity to the nucleus in ∼90% of triple knock-in cells (Figure 4B), both in 2D and 3D environments (Figure 4C; Videos 9–12). To quantify the colocalization of vesicles, we tracked ~46,000 vesicles per Rab11 family member. This analysis revealed that there is no significant difference in TFN colocalization with Rab11 family members, but Rab11a colocalizes with Rab11b to a significantly higher extent than with Rab25, or Rab11b with Rab25 (Figure 4D). Rab11a/b/25 largely colocalize with TFN, one another, or all at the same time (16.1% of all Rab11 vesicles/26.5% of TFN vesicles; Figure 4E) and that differences between Rab11 family members are mainly due to the unique Rab25 recruitment to a subset of vesicles (Figure 4F; Figure 4—figure supplement 1D,E).

Video 8

Download asset

Video 9

Download asset

Video 10

Download asset

Video 11

Download asset

Video 12

Download asset

These results demonstrate that DExCon is a tool well suited for microscopy that allows multiple gene modifications simultaneously, including re-activation of previously silenced genes. Furthermore, complete integration can be verified using fluorescence-based sorting in polyclonal populations, and the approach provides a relatively simple platform for conditional rescues/knock-outs that are, in contrast to FLP/Cre-driven recombination, reversible.

Protein expression kinetics, stability, and dynamics revealed using DExCon

Intrigued by slow expression kinetics of re-activated mNeonGreen-Rab25 DExCon in A2780 cells, we generated Rab11a and Rab11b DExCon cells to compare kinetics. Western blot suggested that we can detect mNeonGreen-Rab11a from 4 hr and mCherry-Rab11b from 6 hr after dox induction, while Rab25 DExCon was not detected before 8 hr (Figure 5A). The rate of dox induced expression was then quantified on individual DExCon Rab11 family gene modifications. mNeonGreen-Rab11a DExCon reached ~75% of its maximal expression peak at 24 hr, significantly faster than other family members: at 24 hr mCherry-Rab11b DExCon reached ~50% peak expression, whereas mNeonGreen-Rab25 DExCon only reached ~40% (Figure 5B). Similar results were also obtained from nanobody-mCherry-DExCon modifications (Figure 5—figure supplement 1A,B), suggesting that observed differences can be explained by differences in the regulation of Rab11 gene expression and not by different FPs.

Figure 5 with 3 supplements see all

Download asset Open asset

Since Rab25 is silenced in A2780, we performed bioinformatic analysis of its methylation status. We found that there is an inverse relationship between methylation of the Rab25 promoter and its expression, where decreased methylation is correlated with increased gene expression, at both mRNA and protein levels across all cancer types (Figure 5C; Figure 5—figure supplement 1C). ATAC-seq dataset analysis further suggested that the chromatin structure of the silenced Rab25 locus, both in promoter and enhancer sequences, is less accessible for transcriptional machinery to drive transcription (Figure 5—figure supplement 1C). To prove that slower Rab25 DExCon expression kinetics in A2780 cells can be simply explained by the decrease of DNA accessibility for transcription, we knocked in TRE3GS-mNeonGreen to the Rab11a or Rab25 loci of OVCAR-3 cells, an ovarian cancer line where both Rab11a and Rab25 are endogenously expressed. Dox induction again led to the correct localization of Rab11a or Rab25, as expected (Figure 5—figure supplement 1D), but this time there was no delay in the normalized expression kinetics of Rab25, as was observed for A2780 (Figure 5D). Interestingly, in both A2780 and OVCAR-3 cells Rab11a DExCon always led to greater fluorescence intensity compared to Rab25 DExCon, suggesting additional mechanisms of protein level control (Figure 5—figure supplement 1E).

We further explored the possibility of simultaneous visualization of different alleles of the same gene (Figure 5E; Figure 5—figure supplement 1F, Figure 5—figure supplement 2A-C). We were able to successfully modify a second allele of mCherry-Rab11a, mCherry-Rab11b, or DExCon-mCherry-Rab11b modified cells by knocking in DExCon-mNeonGreen modules (Figure 5F; Figure 5—figure supplement Figure 5—figure supplements 1F and 2A-C). Dox-mediated induction led to consistent almost identical localization in all cells with alleles modified with mCherry and DExCon-mNeonGreen for both Rab11a and Rab11b (Figure 5—figure supplement 2A-C). Consistent with previous results (Figure 2E; Figure 2—figure supplement 2E), DExCon-mNeonGreen-Rab11a was slightly more diffuse than DExCon-mCherry-Rab11b in the same cells (Figure 5—figure supplement 2D). This demonstrates that DExCon-induced Rab11 expression levels do not appreciably alter localization of the GTPases and generate signals compatible with routine imaging.

Unexpectedly, expression kinetics of double DExCon-mCherry/DExCon-mNeonGreen Rab11b cells were noticeably different between mCherry and mNeonGreen modified Rab11b (Figure 5H; Video 13). While the mNeonGreen kinetics of Rab11a expression were still slightly faster than mNeonGreen of Rab11b, the mCherry Rab11b fluorescence displayed significant delay similar in both Rab11a/b and Rab11b double DExCons cells. When DExCon cells were pre-induced with dox for 24 hr, followed by dox removal, a decrease in fluorescence was observed over time (Figure 5I). mCherry-Rab11b showed remarkably stable fluorescence, whereas mNeonGreen-Rab11b was lost more rapidly. Stable fluorescence of mCherry was also observed for nanobody-mCherry DExCon Rab11s (Figure 5—figure supplement 1G). By contrast, mNeonGreen fluorescence continuously declined overtime, where Rab11a and Rab11b decreased in intensity more rapidly than Rab25 following dox removal (Figure 5J). This suggests that the FP is an important consideration when using FP-knock-in to understand protein dynamics but opens the possibility of modifying individual (e.g. mutant vs non-mutant) alleles within cells to study their stability and function.

Video 13

Download asset

The localization of Rab GTPases is fundamental to their function, but the relative stability of Rab11 family members at their endogenous location is not known. We therefore combined DExCon with knock-sideways, a method of protein relocalization within live cells (Robinson et al., 2010), to give insight into the dynamics of Rab11a, Rab11b, and Rab25 localization. Relocalization to mitochondria was achieved by rapamycin (rap) induced dimerization of FKBP-mGFP with mitochondrially targeted mito-FRB-iRFP670 in GFPnanobody-mCherry DExCon modified A2780 cells (Figure 5K) and the dynamics of Rab11 family members analyzed by time-lapse imaging (Figure 5L; Figure 5—figure supplement 3A,B). Direct comparison of Rab11a/b/25 dynamics revealed an ultra-fast relocalization of Rab11a (t_1/2=209 s), significantly slower Rab11b relocalization (t_1/2=362 s), and slower still dynamics of Rab25 (t_1/2=476 s) (Figure 5M; Figure 5—figure supplement 3C).

Taken together, these results show our DExCon approach can be used to reveal subtle but significant differences in protein expression kinetics and protein dynamics, and that despite their similarity at the protein level, such differences are observed within the Rab11 family.

DExogron: a tunable approach to modify protein levels within cells

Because removal of dox did not rapidly decrease DExCon-driven protein levels, we added a second layer of control utilizing the improved AID system reported with low-basal degradation (Li et al., 2019). We combined our DExCon module with an auxin-inducible destabilizing domain (short degron miniIAA7) that allows proteasomal degradation in an auxin-dependent manner to generate DExogron gene modifications (Figure 6A).

Figure 6 with 4 supplements see all

Download asset Open asset

Because we were able to independently target both alleles of Rab11a/b, we first combined our knock-in strategy with knock-in of a cassette composed of puromycin (PuroR) or blasticidine (BlaR) resistance markers and PolyA signal to block endogenous expression (Figure 6B; Figure 6—figure supplement 1A). In A2780 cells, inclusion of an additional promoter sequence (SFFV) was required because endogenous promoter-driven expression was insufficient to yield antibiotic resistance, in contrast to HEK293T cells. Sequential targeting of the second Rab11a or Rab11b allele in DExCon cells using PuroR or BlaR, significantly improved knock-in outcome both in DExogron, DExCon Rab11b, or double DExCon Rab11a/b cells (Figure 6—figure supplement 1B-D) and thus successfully generated near-complete knock-in/knock-out cells with only one allele of Ra11a or b under dox control (rescue). Notably, simultaneous targeting of two Rab11a alleles by DExCon and PuroR dramatically increased the proportion of DExCon knock-in cells surviving PuroR treatment, from 0.2 to 25% (Figure 6—figure supplement 1E). We successfully generated combinations of DExCon/DExogron-Rab11a/b, with additional PuroR/BlaR knock-ins (Figure 6C; Figure 6—figure supplement 1F). Dox treatment led to rescue of protein expression as expected and simultaneous addition of Indole-3-acetic acid (IAA as a source of auxin) efficiently depleted dox-induced expression only when the degron tag was present (DExogron minilAA7-mCherry-Rab11b) (Figure 6C; Figure 6—figure supplement 1D, F; Videos 14 and 15). Flow cytometry analysis confirmed that IAA/dox co-treatment (24–48 hr) decreases DExogron-mCherry-Rab11b fluorescence to the level of un-treated Rab11b miniIAA7-mCherry knock-in cells with degradation half time kinetics of 1.7±0.1 hr (Figure 6D and E; Figure 6—figure supplement 1G; Video 14).

Video 14

Download asset

Video 15

Download asset

Introducing a degron tag using one donor without additional DExCon modification led to a decrease in mCherry fluorescence (Figure 6D) and revealed two additional drawbacks. A significant proportion of cells did not respond to IAA likely due to the truncation of the minilAA7 sequence during HDR (Figure 6—figure supplement 1H). This could be overcome by including negative sort after treatment by IAA in our protocol (see diagram Figure 6—figure supplement 1I). However, treatment with IAA resulted in 90–70% mCherry removal, which was close to the control level but not complete, limiting the effectiveness as complete removal tool (Figure 6D).

Rab11 family GTPases are implicated in cell migration, invasion, and cancer progression (Kelly et al., 2012; Jin et al., 2021; Caswell et al., 2009; Caswell et al., 2007; Howe et al., 2020). To test the functionality of DExCons/DExogrons, we performed a 2D scratch wound migration assay and found that Rab11a or Rab11b DExCon or DExogron modified cells closes the wound significantly later than control wt cells in the absence of dox-induced expression, and this phenotype is further amplified by Rab11a/b knock-out combinations (Figure 6F; Figure 6—figure supplement 2A; Video 16). Knock-out phenotypes were reversed when Rab11a/b expression was re-activated with dox; however, induction of degradation with IAA did not oppose this rescue, suggesting that IAA induced Rab11b depletion is incomplete in the presence of dox and is not sufficient to mimic the knock-out phenotype. The importance of Rab11a or Rab11b for cell migration was further verified with A2780 stably expressing shRNA anti-Rab11a or b (Figure 6F; Figure 6—figure supplement 2A,B). The effect of proliferation on wound healing closure was negligible, as any change in Rab11a/b expression did not alter cell growth (Figure 6—figure supplement 2C). Notably, we observed a toxic effect of IAA (100 μg/ml) from 10 hr of exposure when A2780 cells seeded sparsely.

Video 16

Download asset

To investigate how manipulation of Rab11a/b expression influences cell migration/invasion into 3D collagen/FN matrix, we performed spheroid invasion assays. Reducing Rab11a or b expression using shRNA, DExCon, DExogron, or their combination decreased the ability of cells/spheroids to invade FN-rich collagen, and this was significantly rescued by dox treatment in Rab11a DExCon single knock-in cells (Figure 6G; Figure 6—figure supplement 2D,E). Dox-induced re-expression of Rab11b had a dose-dependent dominant-negative effect on invasion under all combinations, sufficiently blocking even Rab11a rescue. The negative effect of Rab11b rescue was reversed by simultaneous dox/IAA co-treatment and led to cell invasion comparable to wt control cells or even significantly higher (Figure 6H–I) despite the cytotoxic effect of IAA observed on individually migrating cells (Figure 6—figure supplement 2D,E vs Figure 6H). This confirms Rab11b functionality and suggests that cells are sensitive to Rab11b re-expression levels and that balanced dosage is needed for coordination of membrane trafficking important for 3D cell migration.

We next tested whether induction of degradation with IAA can improve the AID pitfall of incomplete Rab11b depletion by withdrawing dox following re-activation of Rab11b expression. Indeed, mCherry fluorescence intensity of DExogron Rab11b cells or DExogron-Rab11b/DExCon-Rab11a cells was slightly, but significantly, decreased below the fluorescence of dox/IAA co-treated cells as early as 1–4 hr after switching to dox-free media supplemented only with IAA (Figure 6J; Figure 6—figure supplement 3A) and returned to non-detectable levels in 24–48 hr (Figure 6K). After 48 hr, expression of DExCon mNeonGreen-Rab11a was reduced, but still detectable (Figure 6K). These data show that the DExogron is an important modification that can lead to complete removal of protein following withdrawal of dox-induced gene expression.

Because IAA appeared to influence cell proliferation, we dropped its concentration from 100 µg/ml to 0.5 µg/ml and found that degradation was significantly slowed but still occurred to a similar extent up to 2.5 µg/ml (Figure 6K and L; Figure 6—figure supplement 3B). Dox/IAA co-treatment was sufficient to tune DExogron mCherry-Rab11b to endogenous levels (IAA 0.5 µg/ml; Figure 6L). DExogron-Rab11a/DExCon-Rab11b similarly resulted in tuned steady-state fluorescence intensity of mNeonGreen Rab11a with degradation halftime kinetics around 2.5±0.1 hr (Figure 6—figure supplement 4A-E; Video 17).

Video 17

Download asset

Because IAA induced depletion of DExogron Rab11b was incomplete when expression was driven in the presence dox, we pre-induced DExogron-Rab11b for 24 hr (Figure 6—figure supplement 2F) and then transferred cells to dox-free media with IAA (10 g/ml) for 24 hr before scratch. As expected, this condition led to complete mCherry removal (not detectable by fluorescent microscopy) and was sufficient to reverse the ‘rescued’ motility of DExogron-mCherry-Rab11b or DExogron-mCherry-Rab11b/DExCon-mNeonGreen-Rab11a cells (Figure 6—figure supplement 2F).

These data therefore indicate that DExogron can eliminate the shortcomings of the AID system, allowing fine-tuning of expression levels up to complete depletion following CRISPR-based modification. Furthermore, our results suggest that although Rab11b and Rab11a may have a complementary function in migration and invasion of A2780 ovarian cancer cells, they differ substantially in their post-transcriptional regulation.

Tuning Rab11 expression levels with DExCon and DExogron modulates TFN receptor recycling

Recycling of TFN-R is typically biphasic with an initial rapid direct route from early endosomes to the plasma membrane (under control of Rab4) and a slower non-essential route that requires additional trafficking through perinuclear recycling endosomes via Rab11 (Daro et al., 1996; Sheff et al., 2002; Sheff et al., 1999; Figure 7A). Our previous experiments demonstrated that the DExCon/DExogron system induces expression of Rab11 family members to promote cell motility (Figure 3N; Figure 6F–H), the expressed proteins show endogenous localization within cells (2E; Figure 5—figure supplement 2A-D) and they traffic together with TFN (Figure 2—figure supplement 2D; Video 3), suggesting that they are fully functional. We further tested how DExCon/DExogron controlled Rab11 family expression level influenced recycling of internalized TFN to test the function of these re-expressed proteins in a different context. DExCon or DExogron modified Rab11a or Rab11b cells which lack expression of the targeted GTPase(s) had little impact on TFN-R recycling at 30 min (Figure 7B), suggesting that fast recycling is likely predominant in A2780 cells as in other cell types (Sheff et al., 1999). Re-expression of Rab25 did not impact upon recycling, indicating that this Rab11 family member may have a limited influence on trafficking this cargo (Figure 7B). However, dox-induced high-level expression of either Rab11a or Rab11b under DExCon/DExogron control decreased the level of recycling at 30 min (Figure 7B), consistent with the literature (Ren et al., 1998), perhaps by favoring the slow-recycling route. DExCon/DExogron Rab11a/b cells selected for expression levels comparable to endogenous Rab11a/b knock-in did not have any significant effect on TFN-R recycling rate compared to the non-modified control, indicating that it is the expression level that slows recycling. Similarly, IAA/Dox co-treatment of DExogron-Rab11b cells led to near-physiological levels of Rab11b which had no significant effect on TFN-R recycling (Figure 7B). Surprisingly, TFN-R recycling slowed by high Rab11a DExCon levels was restored to the level of double knock-out cells by lowering Rab11b levels with DExogron (Rab11a DExCon/Rab11b DExogron, IAA/dox co-treatment). These data suggest that Rab11a/Rab11b can recycle TFN-R via the slow recycling pathway, and that increasing levels of Rab11a/b could favor the slow-recycling route to delay recycling of TFN.

Figure 7

Download asset Open asset

To test the effect of an acute decrease of Rab11a/Rab11b levels toward the expression driven by endogenous locus, we pre-adapted A2780 cells to high Rab11a/Rab11b levels by treating them with dox over >7 days before decreasing Rab11 levels by switching to IAA; dox+IAA or dox-positive media for 24–48 hr prior analysis. TFN-R recycling was assessed after 10 min to highlight competition between fast and slow recycling routes (Figure 7C). Cells induced to have high-level expression of DExCon-mNeonGreen-Rab11a or/and DExogron-mCherry-Rab11b again showed lower levels of TFN-R recycling compared to control (mainly evident for Rab11a), but tempering expression of Rab11a and Rab11b (or Rab11b alone in DExogron-mCherry-Rab11b cells) by withdrawing dox/adding IAA demonstrated that lower Rab11 levels correlated with increased recycling rates (7C). These data further support the notion that competition between fast and slow recycling likely controlled by Rab4 and Rab11a/b, respectively, underpins the differences in recycling rates induced by increasing Rab11 expression level.

Taken together, this suggests that the DExCon/DExogron system affords control over the timing and level of gene expression of Rab11 family members, and points toward a more complex relationship between fast Rab11-independent and slow Rab11-dependent recycling.

Here, we describe DExCon, LUXon, and DExogron tools (see Figure 8) to reversibly modulate gene expression from the endogenous locus. These systems allow effective knock-out of gene expression, but allow subsequent re-activation of transcription (or activation of transcriptionally silent genes). LUXon permits the spatial and temporal control of gene expression, and DExogron adds a further level of control such that expression levels can be tuned appropriately, or complete degradation induced.

Figure 8

Download asset Open asset

Significant effort has been made to achieve conditional control of gene expression at the endogenous locus in order interrogate gene function in a native biological and pathological context. Binary recombination approaches, such as the Cre/loxP system, allow inducible inactivation (Quadros et al., 2017) or activation through conditional excision of a stop element in the target gene (Ventura et al., 2007). However, the ability to control gene expression is lost following irreversible deletion, and this system does not allow evaluation of the phenotypic effects across a spectrum of different expression levels. Our DExCon/Dexogron/LUXon approaches are attractive alternatives as they offer, in addition to reversible inactivation of protein-coding genes, conditional and (for LUXon) spatiotemporal expression control of re-activated genes which are visualized by an introduced FP.

Other approaches to modulate endogenous gene expression have used components of lac repression and tet activation systems introduced to the endogenous promoter sequence or downstream, including ‘Reversible Manipulation of Transcription at Endogenous loci-control’ (Lee et al., 2017). Our approach differs because it bypasses the function of the endogenous promoter and does not require insertion of effector binding sites that can perturb target gene expression. Our one-step CRISPR knock-in based method is best suited for any genes whose protein-coding transcripts share the initial ATG codon and can be N-terminally tagged, or introduce the DExCon module upstream of alternatively edited genes. For genes where the fluorophore would have to be introduced internally or C-terminally, two-step CRISPR knock-in approaches are necessary (one for TRE3GS promoter, second for FP). Although our system does not silence transcription driven by an endogenous active promoter, it generates non-sense frameshifts in transcripts originating from the endogenous promoter and can activate silenced transcripts. This offers advantages over methods that alter promoter methylation (e.g. dCas9 fused to transcriptional activators or repressors Pflueger et al., 2019, which can create modifications over several tens of kilobases and impact off-target genes Peng et al., 2009; Groner et al., 2010) and overcomes the limit of classical knock-in where low expression confounds detection or protein products. Furthermore, in contrast to transient transfection of exogenous vectors, DExCon preserves splicing control and allows analysis of the full complement of physiologically localized isoforms within cells (Figure 2E–G; Figure 2—figure supplement 2A,B; Figure 5—figure supplement 2A-D).

The AID system is an excellent method of protein down-regulation (Nishimura et al., 2009), which has since been further developed (Röth et al., 2019; Li et al., 2019; Yesbolatova et al., 2020) allowing rapid degradation of target proteins. Using CRISPR, we successfully applied the improved AID system to the endogenous locus of Rab11b (Li et al., 2019); however, the incorporation of degron miniIAA7 tag in front of mCherry at the Rab11b locus resulted in a profound decrease in mCherry fluorescence without induction by IAA, and induction did not lead to complete protein removal (Figure 6D). Our combination of degron and DExCon (DExogron), however, can generate robust expression and allows additional fine temporal control of the DExCon system, providing a system that can reveal acute phenotypes that otherwise might be masked by compensation or a dominant terminal phenotype.

For broad application of strategies such as DExCon, feasibility and efficiency are critical. One of the main obstacles to gene editing via HDR is its low efficiency, which complicates targeting all alleles, generation, and selection of homozygous knock-in modifications (Gurumurthy et al., 2019). Our optimized magnetofection mediated delivery of Hifi Cas9 as a RNP complex with single-stranded donor DNA ensured precise and on-target integrations to endogenous Rab11 family gene loci reflected by expected localization in cells (Figure 1C). Although truncated integration events are often reported (Canaj et al., 2019), functional flanking components (tet-responsive promoter and FP) allow selection of polyclonal populations with full integration of the ssDNA template (including for example the nanobody anti-GFP domain), negating the need for comprehensive and labor-intensive genotyping (Figure 2H–I). We demonstrated the robustness of our approach by generating single, double, and triple knock-in DExCon at the endogenous loci of Rab11 family genes in A2780 cells without the need for clonal selection. Our strategy further offers sufficient expression modulation from monoallelic modifications; therefore, we could independently target alleles of the same gene using combinations of different donors (DExCon-mNeonGreen, DExCon-mCherry, antibiotic resistance [Figure 5F; Figure 6—figure supplement 1B-F]). As proof of principle, we imaged protein products of both Rab11b alleles to demonstrate new possibilities, for example, studying the dynamics of mutant and wildtype alleles of oncogenes and tumor suppressors. When simultaneously delivering DExCon together with antibiotic resistance to a second allele, we observed a dramatic increase in selection efficiency from 0.2 to 25% (FP-positive dox-responsive cells surviving antibiotic treatment) allowing for easy enrichment of homozygous edits (Figure 6—figure supplement 1E). Sequential delivery of antibiotic resistance to DExCon-Rab11a and -Rab11b (Figure 6—figure supplement 1B-F) further improved editing outcome, effectively generating a knockout of one allele to ensure that all protein generated by cells from the second allele is tagged appropriately (mNeonGreen/mCherry in this case), full knock-out with dox-controlled expression. The conventional Tet-OFF/ON system is widely used in various research models, such as transgenic animals (Zhu et al., 2002) and can be easily combined with our DExCon/DExogron/LUXon strategy for a huge range of applications across the field of biology. Notably, DExogron harbors promising application also for essential genes with knock-out lethal phenotypes. Here, the continual presence of dox or the transition from Tet-on to Tet-off system would be required, and IAA treatment could induce immediate knock-out effect.

To demonstrate the potency of our strategies on closely related genes, we deployed them to study members of the Rab11 family (Rab11a, Rab11b, and Rab25) in an ovarian cancer cell line that lacks endogenous Rab25 expression (Figure 3A). We demonstrated that spatiotemporal re-activation of silenced endogenous Rab25 expression can promote invasive migration (Figure 3J–O), and that Rab11b and Rab11a may have complementary functions in migration and invasion (Figure 6F–I). We identified and reversibly rescued two Rab11b protein-coding splice variants (201 [24.5 kDa] and 202 [20 kDa]) in a physiologically proportional manner (Figure 2—figure supplement 1B). Mutants of Rab11 that lack the C-terminus, similar to Rab11b variant 202, have previously been shown localize diffusely and are unable to bind and recycle TFN-R (Schlierf et al., 2000). However, we observed specific endosomal localization of CRISPR/Cas9-modified Rab11b (Figure 2E; Figure 5—figure supplement 2B) likely corresponding with the 201 variant and suggesting that the 202 variant is not a significant protein product. Interestingly, cells migratory/invasive responses appear to be sensitive to expression levels, particularly for Rab11b. Cells handle recycling cargoes by numerous routes simultaneously, and these differences could be explained by favoring trafficking of a cargo along Rab11b routes by increasing expression, for example, Rab11/Arf6 and Rab4-dependent trafficking of α5β1 and αvβ3 integrins appear to act as competing pathways in migrating cells to modulate migration (Caswell et al., 2008; Morgan et al., 2013). The tunable Rab11 family expression control further revealed a pleiotropic effect of Rab11a/b, and Rab25, on their recognized cargo TFN (Figure 7A–C) and further suggests that cells may be sensitive to Rab11 re-expression levels that can change the coordination of membrane trafficking (Daro et al., 1996; Sheff et al., 2002; Sheff et al., 1999). Rab11a/b seem to have little influence on the rate of TFN recycling when knocked-out by DexCon/DExogron, suggesting that Rab4 rapid recycling (or other routes) predominate in the cells used in this study (Figure 7B and C). The ability of Rab11a/b to divert TFN via a longer recycling route to delay return to the surface likely stems from TFN receptor carrying capacity, as higher DexCon/DExogron Rab11 levels have physiological endosomal membrane localization (Figure 2E; Figure 5—figure supplement 2A-D). Interestingly, lowering/tuning Rab11b expression restored TFN-R recycling rate slowed by high Rab11a levels (Figure 7B and C), and Rab25 expression, even at high level, has no discernible effect on TFN recycling (Figure 7A), as found previously in MDCK cells Casanova et al., 1999. Rab11a/b/25 colocalize with TFN to a similar extent (Figure 2—figure supplement 2D, E), and it is therefore interesting to speculate that this could be underpinned by their different relocalization/recycling kinetics (Figure 5M) or by differences in effector recruitment pathways/cargo selectivity.

DExCon/DExogron tools allowed us to quantify surprising differences between Rab11 family members. We found that our system reflects the accessibility for transcriptional machinery to drive transcription at the endogenous locus, demonstrating that Rab11a has significantly faster expression kinetics and compared to Rab11b and re-activated Rab25 expression is slower still (Figure 5A–C). However, in cells with endogenous Rab25 expression, expression kinetics are demonstrable faster (Figure 5D). We also demonstrated that the ability to relocalize differed significantly between Rab11 family members with Rab25 as the least dynamic (Figure 5K–M; Figure 5—figure supplement 3A-C). The faster relocalization rate of Rab11a over Rab11b could be explained by its more diffuse localization (Figure 5—figure supplement 2C,D), which could suggest it is less stably associated with endosomal membranes. Furthermore, we observed a population of Rab25 specific vesicles in the nuclear periphery which appeared to be less mobile (Figure 3I), and these vesicles were always negative for TFN receptor (Figure 4B–F). Such vesicles with limited mobility were recently described to be located on the nuclear envelope, where they exhibit propensity to undergo fusion and act as a novel endosomal route from the cell surface to the nucleoplasm that facilitates the accumulation of extracellular and cell surface proteins (e.g. epidermal growth factor receptor (EGFR)) in the nucleus (Shah et al., 2019; Chaumet et al., 2015). Rab25-mediated EGFR recycling (Caswell et al., 2007; Caswell et al., 2008) and elevated levels of nuclear EGFR correlate with poor prognosis in ovarian cancer (Xia et al., 2009), and our data which suggest that Rab25 may be connected with nuclear envelope-associated endosomes warrants further investigation.

Overall, the work presented here opens new avenues not only for the precise expression control of endogenous genes, but also for the manipulation and imaging of gene function. All the benefits of gene editing can be obtained in one CRISPR based knock-in. DExCon, DExogron, and LUXon are powerful tools that allow new insight into the function of individual genes and closely related gene families.

Detailed list of antibodies, chemicals, sequences, plasmids, and resources used are described in supplementary excel file S1 or accessible from https://doi.org/10.48420/14999550.

Code and analysis tools developed are freely available via Figshare.

1. 1. Harrison T

   (2022) figshare

   Tools for temporal knocksideways analysis.

   https://doi.org/10.6084/m9.figshare.17085632
2. 1. Gemperle J

   (2022) Figshare

   GraphPrism files with raw/normalized data supporting DExCon study.

   https://doi.org/10.48420/16878904
3. 1. Gemperle J

   (2022) figshare

   Macro for spheroid invasion assay including protocol.

   https://doi.org/10.48420/16878829
4. 1. Harrison T

   (2022) figshare

   Colocalisation Quantifier Python Script.

   https://doi.org/10.6084/m9.figshare.16810546

1. 1. Beaufort CM
   2. Helmijr JCA
   3. Piskorz AM
   4. Hoogstraat M
   5. Ruigrok-Ritstier K
   6. Besselink N
   7. Murtaza M
   8. van IJcken WFJ
   9. Heine AAJ
   10. Smid M
   11. Koudijs MJ
   12. Brenton JD
   13. Berns EMJJ
   14. Helleman J

   (2014) Ovarian cancer cell line panel (OCCP): clinical importance of in vitro morphological subtypes

   PLOS ONE 9:e103988.

   https://doi.org/10.1371/journal.pone.0103988

   • PubMed
   • Google Scholar
2. 1. Bennett H
   2. Aguilar-Martinez E
   3. Adamson AD

   (2021) CRISPR-mediated knock-in in the mouse embryo using long single stranded DNA donors synthesised by biotinylated PCR

   Methods (San Diego, Calif.) 191:3–14.

   https://doi.org/10.1016/j.ymeth.2020.10.012

   • PubMed
   • Google Scholar
3. 1. Brinkman EK
   2. Kousholt AN
   3. Harmsen T
   4. Leemans C
   5. Chen T
   6. Jonkers J
   7. van Steensel B

   (2018) Easy quantification of template-directed CRISPR/Cas9 editing

   Nucleic Acids Research 46:e58.

   https://doi.org/10.1093/nar/gky164

   • PubMed
   • Google Scholar
4. 1. Canaj H
   2. Hussmann JA
   3. Li H
   4. Beckman KA
   5. Goodrich L
   6. Cho NH
   7. Li YJ
   8. Santos DA
   9. McGeever A
   10. Stewart EM
   11. Pessino V
   12. Mandegar MA
   13. Huang C
   14. Gan L
   15. Panning B
   16. Huang B
   17. Weissman JS
   18. Leonetti MD

   (2019) Deep Profiling Reveals Substantial Heterogeneity of Integration Outcomes in CRISPR Knock-in Experiments

   Genomics 1:841098.

   https://doi.org/10.1101/841098

   • Google Scholar
5. 1. Casanova JE
   2. Wang X
   3. Kumar R
   4. Bhartur SG
   5. Navarre J
   6. Woodrum JE
   7. Altschuler Y
   8. Ray GS
   9. Goldenring JR

   (1999) Association of Rab25 and Rab11a with the apical recycling system of polarized Madin-Darby canine kidney cells

   Molecular Biology of the Cell 10:47–61.

   https://doi.org/10.1091/mbc.10.1.47

   • PubMed
   • Google Scholar
6. 1. Caswell PT
   2. Spence HJ
   3. Parsons M
   4. White DP
   5. Clark K
   6. Cheng KW
   7. Mills GB
   8. Humphries MJ
   9. Messent AJ
   10. Anderson KI
   11. McCaffrey MW
   12. Ozanne BW
   13. Norman JC

   (2007) Rab25 associates with alpha5beta1 integrin to promote invasive migration in 3D microenvironments

   Developmental Cell 13:496–510.

   https://doi.org/10.1016/j.devcel.2007.08.012

   • PubMed
   • Google Scholar
7. 1. Caswell PT
   2. Chan M
   3. Lindsay AJ
   4. McCaffrey MW
   5. Boettiger D
   6. Norman JC

   (2008) Rab-coupling protein coordinates recycling of alpha5beta1 integrin and EGFR1 to promote cell migration in 3D microenvironments

   The Journal of Cell Biology 183:143–155.

   https://doi.org/10.1083/jcb.200804140

   • PubMed
   • Google Scholar
8. 1. Caswell PT
   2. Vadrevu S
   3. Norman JC

   (2009) Integrins: masters and slaves of endocytic transport

   Nature Reviews. Molecular Cell Biology 10:843–853.

   https://doi.org/10.1038/nrm2799

   • PubMed
   • Google Scholar
9. 1. Chaumet A
   2. Wright GD
   3. Seet SH
   4. Tham KM
   5. Gounko NV
   6. Bard F

   (2015) Nuclear envelope-associated endosomes deliver surface proteins to the nucleus

   Nature Communications 6:1–9.

   https://doi.org/10.1038/ncomms9218

   • PubMed
   • Google Scholar
10. 1. Cheeseman LP
    2. Harry EF
    3. McAinsh AD
    4. Prior IA
    5. Royle SJ

    (2013) Specific removal of TACC3-ch-TOG-clathrin at metaphase deregulates kinetochore fiber tension

    Journal of Cell Science 126:2102–2113.

    https://doi.org/10.1242/jcs.124834

    • PubMed
    • Google Scholar
11. 1. Cheng KW
    2. Lahad JP
    3. Kuo W-L
    4. Lapuk A
    5. Yamada K
    6. Auersperg N
    7. Liu J
    8. Smith-McCune K
    9. Lu KH
    10. Fishman D
    11. Gray JW
    12. Mills GB

    (2004) The RAB25 small GTPase determines aggressiveness of ovarian and breast cancers

    Nature Medicine 10:1251–1256.

    https://doi.org/10.1038/nm1125

    • PubMed
    • Google Scholar
12. 1. Cukierman E
    2. Pankov R
    3. Stevens DR
    4. Yamada KM

    (2001) Taking cell-matrix adhesions to the third dimension

    Science (New York, N.Y.) 294:1708–1712.

    https://doi.org/10.1126/science.1064829

    • PubMed
    • Google Scholar
13. 1. Daro E
    2. van der Sluijs P
    3. Galli T
    4. Mellman I

    (1996) Rab4 and cellubrevin define different early endosome populations on the pathway of transferrin receptor recycling

    PNAS 93:9559–9564.

    https://doi.org/10.1073/pnas.93.18.9559

    • PubMed
    • Google Scholar
14. 1. Doudna JA
    2. Charpentier E

    (2014) Genome editing. The new frontier of genome engineering with CRISPR-Cas9

    Science 346:1258096.

    https://doi.org/10.1126/science.1258096

    • PubMed
    • Google Scholar
15. 1. Gemperle J
    2. Dibus M
    3. Koudelková L
    4. Rosel D
    5. Brábek J

    (2019) The interaction of p130Cas with PKN3 promotes malignant growth

    Molecular Oncology 13:264–289.

    https://doi.org/10.1002/1878-0261.12401

    • PubMed
    • Google Scholar
16. 1. Gibson TJ
    2. Seiler M
    3. Veitia RA

    (2013) The transience of transient overexpression

    Nature Methods 10:715–721.

    https://doi.org/10.1038/nmeth.2534

    • PubMed
    • Google Scholar
17. 1. Goldman MJ
    2. Craft B
    3. Hastie M
    4. Repečka K
    5. McDade F
    6. Kamath A
    7. Banerjee A
    8. Luo Y
    9. Rogers D
    10. Brooks AN
    11. Zhu J
    12. Haussler D

    (2020) Visualizing and interpreting cancer genomics data via the Xena platform

    Nature Biotechnology 38:675–678.

    https://doi.org/10.1038/s41587-020-0546-8

    • PubMed
    • Google Scholar
18. 1. Groner AC
    2. Meylan S
    3. Ciuffi A
    4. Zangger N
    5. Ambrosini G
    6. Dénervaud N
    7. Bucher P
    8. Trono D

    (2010) KRAB-zinc finger proteins and KAP1 can mediate long-range transcriptional repression through heterochromatin spreading

    PLOS Genetics 6:e1000869.

    https://doi.org/10.1371/journal.pgen.1000869

    • PubMed
    • Google Scholar
19. 1. Gross LA
    2. Baird GS
    3. Hoffman RC
    4. Baldridge KK
    5. Tsien RY

    (2000) The structure of the chromophore within DsRed, a red fluorescent protein from coral

    PNAS 97:11990–11995.

    https://doi.org/10.1073/pnas.97.22.11990

    • PubMed
    • Google Scholar
20. 1. Gurumurthy CB
    2. O’Brien AR
    3. Quadros RM
    4. Adams J Jr
    5. Alcaide P
    6. Ayabe S
    7. Ballard J
    8. Batra SK
    9. Beauchamp M-C
    10. Becker KA
    11. Bernas G
    12. Brough D
    13. Carrillo-Salinas F
    14. Chan W
    15. Chen H
    16. Dawson R
    17. DeMambro V
    18. D’Hont J
    19. Dibb KM
    20. Eudy JD
    21. Gan L
    22. Gao J
    23. Gonzales A
    24. Guntur AR
    25. Guo H
    26. Harms DW
    27. Harrington A
    28. Hentges KE
    29. Humphreys N
    30. Imai S
    31. Ishii H
    32. Iwama M
    33. Jonasch E
    34. Karolak M
    35. Keavney B
    36. Khin N-C
    37. Konno M
    38. Kotani Y
    39. Kunihiro Y
    40. Lakshmanan I
    41. Larochelle C
    42. Lawrence CB
    43. Li L
    44. Lindner V
    45. Liu X-D
    46. Lopez-Castejon G
    47. Loudon A
    48. Lowe J
    49. Jerome-Majewska LA
    50. Matsusaka T
    51. Miura H
    52. Miyasaka Y
    53. Morpurgo B
    54. Motyl K
    55. Nabeshima Y-I
    56. Nakade K
    57. Nakashiba T
    58. Nakashima K
    59. Obata Y
    60. Ogiwara S
    61. Ouellet M
    62. Oxburgh L
    63. Piltz S
    64. Pinz I
    65. Ponnusamy MP
    66. Ray D
    67. Redder RJ
    68. Rosen CJ
    69. Ross N
    70. Ruhe MT
    71. Ryzhova L
    72. Salvador AM
    73. Alam SS
    74. Sedlacek R
    75. Sharma K
    76. Smith C
    77. Staes K
    78. Starrs L
    79. Sugiyama F
    80. Takahashi S
    81. Tanaka T
    82. Trafford AW
    83. Uno Y
    84. Vanhoutte L
    85. Vanrockeghem F
    86. Willis BJ
    87. Wright CS
    88. Yamauchi Y
    89. Yi X
    90. Yoshimi K
    91. Zhang X
    92. Zhang Y
    93. Ohtsuka M
    94. Das S
    95. Garry DJ
    96. Hochepied T
    97. Thomas P
    98. Parker-Thornburg J
    99. Adamson AD
    100. Yoshiki A
    101. Schmouth J-F
    102. Golovko A
    103. Thompson WR
    104. Lloyd KCK
    105. Wood JA
    106. Cowan M
    107. Mashimo T
    108. Mizuno S
    109. Zhu H
    110. Kasparek P
    111. Liaw L
    112. Miano JM
    113. Burgio G

    (2019) Reproducibility of CRISPR-Cas9 methods for generation of conditional mouse alleles: A multi-center evaluation

    Genome Biology 20:171.

    https://doi.org/10.1186/s13059-019-1776-2

    • PubMed
    • Google Scholar
21. 1. Howe EN
    2. Burnette MD
    3. Justice ME
    4. Schnepp PM
    5. Hedrick V
    6. Clancy JW
    7. Guldner IH
    8. Lamere AT
    9. Li J
    10. Aryal UK
    11. D’Souza-Schorey C
    12. Zartman JJ
    13. Zhang S

    (2020) Rab11b-mediated integrin recycling promotes brain metastatic adaptation and outgrowth

    Nature Communications 11:1–15.

    https://doi.org/10.1038/s41467-020-16832-2

    • PubMed
    • Google Scholar
22. 1. Jacobi AM
    2. Rettig GR
    3. Turk R
    4. Collingwood MA
    5. Zeiner SA
    6. Quadros RM
    7. Harms DW
    8. Bonthuis PJ
    9. Gregg C
    10. Ohtsuka M
    11. Gurumurthy CB
    12. Behlke MA

    (2017) Simplified CRISPR tools for efficient genome editing and streamlined protocols for their delivery into mammalian cells and mouse zygotes

    Methods (San Diego, Calif.) 121–122:16–28.

    https://doi.org/10.1016/j.ymeth.2017.03.021

    • PubMed
    • Google Scholar
23. 1. Jin H
    2. Tang Y
    3. Yang L
    4. Peng X
    5. Li B
    6. Fan Q
    7. Wei S
    8. Yang S
    9. Li X
    10. Wu B
    11. Huang M
    12. Tang S
    13. Liu J
    14. Li H

    (2021) Rab GTPases: Central Coordinators of Membrane Trafficking in Cancer

    Frontiers in Cell and Developmental Biology 9:648384.

    https://doi.org/10.3389/fcell.2021.648384

    • PubMed
    • Google Scholar
24. 1. Kelly EE
    2. Horgan CP
    3. McCaffrey MW

    (2012) Rab11 proteins in health and disease

    Biochemical Society Transactions 40:1360–1367.

    https://doi.org/10.1042/BST20120157

    • PubMed
    • Google Scholar
25. 1. Lee KH
    2. Oghamian S
    3. Park JA
    4. Kang L
    5. Laird PW

    (2017) The REMOTE-control system: a system for reversible and tunable control of endogenous gene expression in mice

    Nucleic Acids Research 45:12256–12269.

    https://doi.org/10.1093/nar/gkx829

    • PubMed
    • Google Scholar
26. 1. Leonetti MD
    2. Sekine S
    3. Kamiyama D
    4. Weissman JS
    5. Huang B

    (2016) A scalable strategy for high-throughput GFP tagging of endogenous human proteins

    PNAS 113:E3501–E3508.

    https://doi.org/10.1073/pnas.1606731113

    • PubMed
    • Google Scholar
27. 1. Li H
    2. Beckman KA
    3. Pessino V
    4. Huang B
    5. Weissman JS
    6. Leonetti MD

    (2017) Design and Specificity of Long SsDNA Donors for CRISPR-Based Knock-In

    Molecular Biology 1:e8905.

    https://doi.org/10.1101/178905

    • Google Scholar
28. 1. Li S
    2. Prasanna X
    3. Salo VT
    4. Vattulainen I
    5. Ikonen E

    (2019) An efficient auxin-inducible degron system with low basal degradation in human cells

    Nature Methods 16:866–869.

    https://doi.org/10.1038/s41592-019-0512-x

    • PubMed
    • Google Scholar
29. 1. Miura H
    2. Quadros RM
    3. Gurumurthy CB
    4. Ohtsuka M

    (2018) Easi-CRISPR for creating knock-in and conditional knockout mouse models using long ssDNA donors

    Nature Protocols 13:195–215.

    https://doi.org/10.1038/nprot.2017.153

    • PubMed
    • Google Scholar
30. 1. Morgan MR
    2. Hamidi H
    3. Bass MD
    4. Warwood S
    5. Ballestrem C
    6. Humphries MJ

    (2013) Syndecan-4 phosphorylation is a control point for integrin recycling

    Developmental Cell 24:472–485.

    https://doi.org/10.1016/j.devcel.2013.01.027

    • PubMed
    • Google Scholar
31. 1. Nishimura K
    2. Fukagawa T
    3. Takisawa H
    4. Kakimoto T
    5. Kanemaki M

    (2009) An auxin-based degron system for the rapid depletion of proteins in nonplant cells

    Nature Methods 6:917–922.

    https://doi.org/10.1038/nmeth.1401

    • PubMed
    • Google Scholar
32. 1. Ota T
    2. Suzuki Y
    3. Nishikawa T
    4. Otsuki T
    5. Sugiyama T
    6. Irie R
    7. Wakamatsu A
    8. Hayashi K
    9. Sato H
    10. Nagai K
    11. Kimura K
    12. Makita H
    13. Sekine M
    14. Obayashi M
    15. Nishi T
    16. Shibahara T
    17. Tanaka T
    18. Ishii S
    19. Yamamoto J
    20. Saito K
    21. Kawai Y
    22. Isono Y
    23. Nakamura Y
    24. Nagahari K
    25. Murakami K
    26. Yasuda T
    27. Iwayanagi T
    28. Wagatsuma M
    29. Shiratori A
    30. Sudo H
    31. Hosoiri T
    32. Kaku Y
    33. Kodaira H
    34. Kondo H
    35. Sugawara M
    36. Takahashi M
    37. Kanda K
    38. Yokoi T
    39. Furuya T
    40. Kikkawa E
    41. Omura Y
    42. Abe K
    43. Kamihara K
    44. Katsuta N
    45. Sato K
    46. Tanikawa M
    47. Yamazaki M
    48. Ninomiya K
    49. Ishibashi T
    50. Yamashita H
    51. Murakawa K
    52. Fujimori K
    53. Tanai H
    54. Kimata M
    55. Watanabe M
    56. Hiraoka S
    57. Chiba Y
    58. Ishida S
    59. Ono Y
    60. Takiguchi S
    61. Watanabe S
    62. Yosida M
    63. Hotuta T
    64. Kusano J
    65. Kanehori K
    66. Takahashi-Fujii A
    67. Hara H
    68. Tanase T
    69. Nomura Y
    70. Togiya S
    71. Komai F
    72. Hara R
    73. Takeuchi K
    74. Arita M
    75. Imose N
    76. Musashino K
    77. Yuuki H
    78. Oshima A
    79. Sasaki N
    80. Aotsuka S
    81. Yoshikawa Y
    82. Matsunawa H
    83. Ichihara T
    84. Shiohata N
    85. Sano S
    86. Moriya S
    87. Momiyama H
    88. Satoh N
    89. Takami S
    90. Terashima Y
    91. Suzuki O
    92. Nakagawa S
    93. Senoh A
    94. Mizoguchi H
    95. Goto Y
    96. Shimizu F
    97. Wakebe H
    98. Hishigaki H
    99. Watanabe T
    100. Sugiyama A
    101. Takemoto M
    102. Kawakami B
    103. Yamazaki M
    104. Watanabe K
    105. Kumagai A
    106. Itakura S
    107. Fukuzumi Y
    108. Fujimori Y
    109. Komiyama M
    110. Tashiro H
    111. Tanigami A
    112. Fujiwara T
    113. Ono T
    114. Yamada K
    115. Fujii Y
    116. Ozaki K
    117. Hirao M
    118. Ohmori Y
    119. Kawabata A
    120. Hikiji T
    121. Kobatake N
    122. Inagaki H
    123. Ikema Y
    124. Okamoto S
    125. Okitani R
    126. Kawakami T
    127. Noguchi S
    128. Itoh T
    129. Shigeta K
    130. Senba T
    131. Matsumura K
    132. Nakajima Y
    133. Mizuno T
    134. Morinaga M
    135. Sasaki M
    136. Togashi T
    137. Oyama M
    138. Hata H
    139. Watanabe M
    140. Komatsu T
    141. Mizushima-Sugano J
    142. Satoh T
    143. Shirai Y
    144. Takahashi Y
    145. Nakagawa K
    146. Okumura K
    147. Nagase T
    148. Nomura N
    149. Kikuchi H
    150. Masuho Y
    151. Yamashita R
    152. Nakai K
    153. Yada T
    154. Nakamura Y
    155. Ohara O
    156. Isogai T
    157. Sugano S

    (2004) Complete sequencing and characterization of 21,243 full-length human cDNAs

    Nature Genetics 36:40–45.

    https://doi.org/10.1038/ng1285

    • PubMed
    • Google Scholar
33. 1. Padilla-Rodriguez M
    2. Parker SS
    3. Adams DG
    4. Westerling T
    5. Puleo JI
    6. Watson AW
    7. Hill SM
    8. Noon M
    9. Gaudin R
    10. Aaron J
    11. Tong D
    12. Roe DJ
    13. Knudsen B
    14. Mouneimne G

    (2018) The actin cytoskeletal architecture of estrogen receptor positive breast cancer cells suppresses invasion

    Nature Communications 9:2980.

    https://doi.org/10.1038/s41467-018-05367-2

    • PubMed
    • Google Scholar
34. 1. Peng H
    2. Ivanov AV
    3. Oh HJ
    4. Lau YFC
    5. Rauscher FJ

    (2009) Epigenetic gene silencing by the SRY protein is mediated by a KRAB-O protein that recruits the KAP1 co-repressor machinery

    The Journal of Biological Chemistry 284:35670–35680.

    https://doi.org/10.1074/jbc.M109.032086

    • PubMed
    • Google Scholar
35. 1. Pflueger C
    2. Swain T
    3. Lister R

    (2019) Harnessing targeted DNA methylation and demethylation using dCas9

    Essays in Biochemistry 63:813–825.

    https://doi.org/10.1042/EBC20190029

    • PubMed
    • Google Scholar
36. 1. Quadros RM
    2. Miura H
    3. Harms DW
    4. Akatsuka H
    5. Sato T
    6. Aida T
    7. Redder R
    8. Richardson GP
    9. Inagaki Y
    10. Sakai D
    11. Buckley SM
    12. Seshacharyulu P
    13. Batra SK
    14. Behlke MA
    15. Zeiner SA
    16. Jacobi AM
    17. Izu Y
    18. Thoreson WB
    19. Urness LD
    20. Mansour SL
    21. Ohtsuka M
    22. Gurumurthy CB

    (2017) Easi-CRISPR: a robust method for one-step generation of mice carrying conditional and insertion alleles using long ssDNA donors and CRISPR ribonucleoproteins

    Genome Biology 18:92.

    https://doi.org/10.1186/s13059-017-1220-4

    • PubMed
    • Google Scholar
37. 1. Ratz M
    2. Testa I
    3. Hell SW
    4. Jakobs S

    (2015) CRISPR/Cas9-mediated endogenous protein tagging for RESOLFT super-resolution microscopy of living human cells

    Scientific Reports 5:1–6.

    https://doi.org/10.1038/srep09592

    • PubMed
    • Google Scholar
38. 1. Ren M
    2. Xu G
    3. Zeng J
    4. De Lemos-Chiarandini C
    5. Adesnik M
    6. Sabatini DD

    (1998) Hydrolysis of GTP on rab11 is required for the direct delivery of transferrin from the pericentriolar recycling compartment to the cell surface but not from sorting endosomes

    PNAS 95:6187–6192.

    https://doi.org/10.1073/pnas.95.11.6187

    • PubMed
    • Google Scholar
39. 1. Roberts B
    2. Haupt A
    3. Tucker A
    4. Grancharova T
    5. Arakaki J
    6. Fuqua MA
    7. Nelson A
    8. Hookway C
    9. Ludmann SA
    10. Mueller IA
    11. Yang R
    12. Horwitz R
    13. Rafelski SM
    14. Gunawardane RN

    (2017) Systematic gene tagging using CRISPR/Cas9 in human stem cells to illuminate cell organization

    Molecular Biology of the Cell 28:2854–2874.

    https://doi.org/10.1091/mbc.E17-03-0209

    • PubMed
    • Google Scholar
40. 1. Robinson MS
    2. Sahlender DA
    3. Foster SD

    (2010) Rapid inactivation of proteins by rapamycin-induced rerouting to mitochondria

    Developmental Cell 18:324–331.

    https://doi.org/10.1016/j.devcel.2009.12.015

    • PubMed
    • Google Scholar
41. 1. Röth S
    2. Fulcher LJ
    3. Sapkota GP

    (2019) Advances in targeted degradation of endogenous proteins

    Cellular and Molecular Life Sciences 76:2761–2777.

    https://doi.org/10.1007/s00018-019-03112-6

    • PubMed
    • Google Scholar
42. 1. Schlierf B
    2. Fey GH
    3. Hauber J
    4. Hocke GM
    5. Rosorius O

    (2000) Rab11b is essential for recycling of transferrin to the plasma membrane

    Experimental Cell Research 259:257–265.

    https://doi.org/10.1006/excr.2000.4947

    • PubMed
    • Google Scholar
43. 1. Shah P
    2. Chaumet A
    3. Royle SJ
    4. Bard FA

    (2019) The NAE Pathway: Autobahn to the Nucleus for Cell Surface Receptors

    Cells 8:915.

    https://doi.org/10.3390/cells8080915

    • PubMed
    • Google Scholar
44. 1. Sheff DR
    2. Daro EA
    3. Hull M
    4. Mellman I

    (1999) The receptor recycling pathway contains two distinct populations of early endosomes with different sorting functions

    The Journal of Cell Biology 145:123–139.

    https://doi.org/10.1083/jcb.145.1.123

    • PubMed
    • Google Scholar
45. 1. Sheff D
    2. Pelletier L
    3. O’Connell CB
    4. Warren G
    5. Mellman I

    (2002) Transferrin receptor recycling in the absence of perinuclear recycling endosomes

    The Journal of Cell Biology 156:797–804.

    https://doi.org/10.1083/jcb.20111048

    • PubMed
    • Google Scholar
46. 1. Song J
    2. Yang D
    3. Xu J
    4. Zhu T
    5. Chen YE
    6. Zhang J

    (2016) RS-1 enhances CRISPR/Cas9- and TALEN-mediated knock-in efficiency

    Nature Communications 7:10548.

    https://doi.org/10.1038/ncomms10548

    • PubMed
    • Google Scholar
47. 1. Ullrich O
    2. Reinsch S
    3. Urbé S
    4. Zerial M
    5. Parton RG

    (1996) Rab11 regulates recycling through the pericentriolar recycling endosome

    The Journal of Cell Biology 135:913–924.

    https://doi.org/10.1083/jcb.135.4.913

    • PubMed
    • Google Scholar
48. 1. Vakulskas CA
    2. Dever DP
    3. Rettig GR
    4. Turk R
    5. Jacobi AM
    6. Collingwood MA
    7. Bode NM
    8. McNeill MS
    9. Yan S
    10. Camarena J
    11. Lee CM
    12. Park SH
    13. Wiebking V
    14. Bak RO
    15. Gomez-Ospina N
    16. Pavel-Dinu M
    17. Sun W
    18. Bao G
    19. Porteus MH
    20. Behlke MA

    (2018) A high-fidelity Cas9 mutant delivered as A ribonucleoprotein complex enables efficient gene editing in human hematopoietic stem and progenitor cells

    Nature Medicine 24:1216–1224.

    https://doi.org/10.1038/s41591-018-0137-0

    • PubMed
    • Google Scholar
49. 1. Ventura A
    2. Kirsch DG
    3. McLaughlin ME
    4. Tuveson DA
    5. Grimm J
    6. Lintault L
    7. Newman J
    8. Reczek EE
    9. Weissleder R
    10. Jacks T

    (2007) Restoration of p53 function leads to tumour regression in vivo

    Nature 445:661–665.

    https://doi.org/10.1038/nature05541

    • PubMed
    • Google Scholar
50. 1. Wilson BJ
    2. Allen JL
    3. Caswell PT

    (2018) Vesicle trafficking pathways that direct cell migration in 3D matrices and in vivo

    Traffic (Copenhagen, Denmark) 19:899–909.

    https://doi.org/10.1111/tra.12605

    • PubMed
    • Google Scholar
51. 1. Xia W
    2. Wei Y
    3. Du Y
    4. Liu J
    5. Chang B
    6. Yu Y-L
    7. Huo L-F
    8. Miller S
    9. Hung M-C

    (2009) Nuclear expression of epidermal growth factor receptor is a novel prognostic value in patients with ovarian cancer

    Molecular Carcinogenesis 48:610–617.

    https://doi.org/10.1002/mc.20504

    • PubMed
    • Google Scholar
52. 1. Yamada M
    2. Suzuki Y
    3. Nagasaki SC
    4. Okuno H
    5. Imayoshi I

    (2018) Light Control of the Tet Gene Expression System in Mammalian Cells

    Cell Reports 25:487–500.

    https://doi.org/10.1016/j.celrep.2018.09.026

    • PubMed
    • Google Scholar
53. 1. Yesbolatova A
    2. Saito Y
    3. Kitamoto N
    4. Makino-Itou H
    5. Ajima R
    6. Nakano R
    7. Nakaoka H
    8. Fukui K
    9. Gamo K
    10. Tominari Y
    11. Takeuchi H
    12. Saga Y
    13. Hayashi KI
    14. Kanemaki MT

    (2020) The auxin-inducible degron 2 technology provides sharp degradation control in yeast, mammalian cells, and mice

    Nature Communications 11:5701.

    https://doi.org/10.1038/s41467-020-19532-z

    • PubMed
    • Google Scholar
54. 1. Zhu Z
    2. Zheng T
    3. Lee CG
    4. Homer RJ
    5. Elias JA

    (2002) Tetracycline-controlled transcriptional regulation systems: advances and application in transgenic animal modeling

    Seminars in Cell & Developmental Biology 13:121–128.

    https://doi.org/10.1016/s1084-9521(02)00018-6

    • PubMed
    • Google Scholar

Author details

1. Jakub Gemperle

   Wellcome Trust Centre for Cell-Matrix Research, School of Biological Sciences, Faculty of Biology Medicine and Health, Manchester Academic Health Science Centre, The University of Manchester, Manchester, United Kingdom

   Contribution

   Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Software, Supervision, Writing - original draft, Writing - review and editing

   For correspondence

   jakub.gemperle@gmail.com

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8360-7075

2. Thomas S Harrison

   Wellcome Trust Centre for Cell-Matrix Research, School of Biological Sciences, Faculty of Biology Medicine and Health, Manchester Academic Health Science Centre, The University of Manchester, Manchester, United Kingdom

   Contribution

   Data curation, Formal analysis, Investigation, Methodology, Software, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2938-1831

3. Chloe Flett

   Wellcome Trust Centre for Cell-Matrix Research, School of Biological Sciences, Faculty of Biology Medicine and Health, Manchester Academic Health Science Centre, The University of Manchester, Manchester, United Kingdom

   Contribution

   Formal analysis, Investigation, Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0950-3128

4. Antony D Adamson

   Wellcome Trust Centre for Cell-Matrix Research, School of Biological Sciences, Faculty of Biology Medicine and Health, Manchester Academic Health Science Centre, The University of Manchester, Manchester, United Kingdom

   Contribution

   Conceptualization, Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

5. Patrick T Caswell

   Wellcome Trust Centre for Cell-Matrix Research, School of Biological Sciences, Faculty of Biology Medicine and Health, Manchester Academic Health Science Centre, The University of Manchester, Manchester, United Kingdom

   Contribution

   Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing - original draft, Writing - review and editing

   For correspondence

   patrick.caswell@manchester.ac.uk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2633-2324

• 1,957

  views

• 240

  downloads

• 1

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.