Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

A polycistronic transgene design for combinatorial genetic perturbations from a single transcript in Drosophila

  • Alexander G. Teague,

    Roles Investigation, Methodology

    Current address: Adicet Bio, Redwood City, California, United States of America

    Affiliation Department of Biological Science, Florida State University, Tallahassee, Florida, United States of America

    ⨯
  • Maria Quintero,

    Roles Investigation, Writing – review & editing

    Affiliation Department of Biological Science, Florida State University, Tallahassee, Florida, United States of America

    ⨯
  • Fateme Karimi Dermani,

    Roles Investigation, Writing – review & editing

    Affiliation Department of Biological Science, Florida State University, Tallahassee, Florida, United States of America

    ⨯
  • Ross L. Cagan,

    Roles Conceptualization, Writing – review & editing

    Current address: Institute of Cancer Sciences, University of Glasgow, Glasgow, Scotland, United Kingdom

    Affiliation Department of Biological Science, Florida State University, Tallahassee, Florida, United States of America

    ⨯
  • Erdem Bangi

    Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing

    ebangi@bio.fsu.edu

    Affiliation Department of Biological Science, Florida State University, Tallahassee, Florida, United States of America

    ⨯

Abstract

Experimental models that capture the genetic complexity of human disease and allow mechanistic explorations of the underlying cell, tissue, and organ interactions are crucial to furthering our understanding of disease biology. Such models require combinatorial manipulations of multiple genes, often in more than one tissue at once. The ability to perform complex genetic manipulations in vivo is a key strength of Drosophila, where many tools for sophisticated and orthogonal genetic perturbations exist. However, combining the large number of transgenes required to establish more representative disease models and conducting mechanistic studies in these already complex genetic backgrounds is challenging. Here we present a design that pushes the limits of Drosophila genetics by allowing targeted combinatorial ectopic expression and knockdown of multiple genes from a single inducible transgene. The polycistronic transcript encoded by this transgene includes a synthetic short hairpin cluster cloned within an intron placed at the 5’ end of the transcript, followed by two protein-coding sequences separated by the T2A sequence that mediates ribosome skipping. This technology is particularly useful for modeling genetically complex diseases like cancer, which typically involve concurrent activation of multiple oncogenes and loss of multiple tumor suppressors. Furthermore, consolidating multiple genetic perturbations into a single transgene further streamlines the ability to perform combinatorial genetic manipulations and makes it readily adaptable to a broad palette of transgenic systems. This flexible design for combinatorial genetic perturbations will also be a valuable tool for functionally exploring multigenic gene signatures identified from omics studies of human disease and creating humanized Drosophila models to characterize disease-associated variants in human genes. It can also be adapted for studying biological processes underlying normal tissue homeostasis and development that require simultaneous manipulation of many genes.

Author summary

The rapid advancement and increased accessibility of genome profiling and big-data approaches have provided unprecedented access to human disease genome landscapes, opening the door to personalized approaches to disease diagnosis and therapy. These efforts continue to identify increasingly complex mutation profiles and gene expression signatures correlating with specific disease outcomes and responses to treatment, emphasizing the importance of experimental models that reflect the genetic complexity and diversity of human disease to investigate the functional relevance of these complex signatures and molecular mechanisms underlying their effects. Here we describe a flexible and streamlined strategy to introduce multiple genetic perturbations to cells in a cell-type and tissue-specific and temporally controlled manner from a single transgenic construct inserted into a predetermined genomic location in Drosophila. This new design represents a valuable addition to the arsenal of tools available in Drosophila to generate more representative human disease models, functionally interrogate individual human disease-associated gene variants and more complex profiles that include alterations in multiple genes.

Citation: Teague AG, Quintero M, Karimi Dermani F, Cagan RL, Bangi E (2023) A polycistronic transgene design for combinatorial genetic perturbations from a single transcript in Drosophila. PLoS Genet 19(6): e1010792. https://doi.org/10.1371/journal.pgen.1010792

Editor: Ken M. Cadigan, University of Michigan, UNITED STATES

Received: April 6, 2023; Accepted: May 18, 2023; Published: June 2, 2023

Copyright: © 2023 Teague et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting Information files.

Funding: This work was supported by grants R03 CA219321 (EB) and R21 GM141734 (EB) from the National Institutes of Health. The funders had no role in study design, data collection, and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Most common human diseases are polygenic; they arise due to the disruption of multiple genes and are also influenced by environmental factors [1]. For instance, genome landscapes of most solid tumors include concurrent somatic alterations in numerous oncogenes and tumor suppressors and many germline variants that can modify disease progression and response to therapy [2–4]. The ability to capture the genetic complexity and heterogeneity of human disease in experimental models is vital for the functional exploration of cancer omics datasets, improving our mechanistic understanding of disease biology and developing effective therapies. In addition to targeting multiple genes, in vivo disease modeling requires transgenes for tissue or cell-type specific and temporally regulated genetic manipulations and fluorescent labeling of targeted cells to study their interactions with their environment. Furthermore, parallel genetic perturbations within the local microenvironment or distant organs are often necessary to explore short-range and systemic interactions mediating disease phenotypes. While multiple such tools are available in genetically tractable model systems like Drosophila [5], bringing all these genetic tools together in experimental animals is challenging. It represents a significant barrier to in vivo disease modeling.

The most commonly used approach for genetic manipulations in Drosophila is the targeted transgene expression platform UAS/Gal4/Gal80ts[6,7]. In this method, genetic elements are cloned downstream of the Gal4 inducible UAS promoter, and transgene expression is induced using transgenic Gal4 lines with tissue-specific expression patterns. Temporal control of expression is achieved by Gal80ts, a temperature-sensitive allele of Gal80 that inhibits Gal4 activity [7]. Thousands of transgenic Drosophila Gal4 lines with tissue and cell-type-specific expression patterns and genome-wide collections for ectopic expression, knockdown, and editing of most genes in the Drosophila genome are publicly available [8–11]. We have previously used these transgenic resources to establish multigenic models of colorectal cancer by genetically manipulating Drosophila orthologs of genes recurrently mutated in human colon tumors [12]. These models were created by combining multiple transgenic insertions into a single genetic background using standard Drosophila genetic crosses. The presence of a large number of transgenic constructs—each inserted in a different genomic location—makes building and maintaining the stability of experimental animals a significant challenge and limits the additional genetic manipulations that can be performed for mechanistic studies.

To overcome these challenges, we modified the pUAST-attB vector, which allows targeted integration at predetermined locations within the Drosophila genome [13], to create a multigenic vector, pWALIUM 3xUAS attB (S1A and S1B Fig), that contained three UAS cassettes for Gal4-mediated transgene expression [14], including one from the pWALIUM vector designed for short hairpin RNA (shRNA) expression [15,16]. Each cassette included a unique multiple cloning site (MCS) flanked by the Gal4 inducible UAS enhancer sequence, followed by the basal promoter of the hsp70 gene upstream and the SV40 terminator sequence downstream. To simultaneously reduce the activity of multiple tumor suppressor genes, we generated synthetic multi-hairpin shRNA clusters inspired by the microRNA (miRNA) clusters found in most genomes [17,18]. This multigenic vector design allowed targeted expression of up to two oncogenes, each from its own UAS cassette, and simultaneous knockdown of as many as eight genes from a single synthetic multi-hairpin cluster. In other words, a 10-hit cancer model could be generated by a multigenic construct as a single transgenic insertion in the Drosophila genome. We have used this approach successfully to generate and drug screen personalized Drosophila models built based on tumor genome landscapes of individual patients in the context of a clinical study [14,19].

The multigenic vector design significantly improves our ability to capture the genetic complexity of human disease. Here, we further streamline and expand upon this platform by consolidating the genetic manipulations requiring all three UAS cassettes of pWALIUM 3xUAS attB into a single inducible UAS transgene. This new design combines intron-mediated short hairpin expression [20] and T2A-mediated polycistronic protein expression [21,22] strategies to target multiple genes from a single polycistronic transcript. It frees the other two UAS cassettes of the multigenic vector for additional genetic manipulations to build more sophisticated models from a single vector. We also show that the number of short hairpins targeting each gene of interest can be increased in a single cluster to ensure robust knockdown without limiting the number of genetic manipulations in cancer models; this is particularly useful as validated shRNA sequences with strong in vivo efficacy are not available for all genes. Notably, this design can also be used with the standard P-element and phiC31-based vectors for Gal4 regulated transgene expression (pUAST or pUASTattB) and adapted to other targeted expression systems available in Drosophila [23–25].

Results

Evaluating knockdown efficacy and positional effects within multi[sh] clusters

We have previously demonstrated that eight genes can be effectively knocked down from a single synthetic short-hairpin (sh) cluster [14]. To determine whether knockdown efficacy is altered in the context of a multi[sh] cluster compared to single shRNA expression, we built a series of 4[sh] test clusters by stitching together short hairpin sequences available as transgenic UAS-shRNA fly lines from the TRiP collection [26]. To this end, we selected three hairpin sequences targeting the Drosophila white (w), singed (sn), and the exogenous GFP gene, all of which provided strong knockdown of their target genes at the RNA, protein, or phenotypic level upon ubiquitous expression. We also included a short hairpin targeting the Drosophila p53 gene, which did not provide significant knockdown as a single hairpin, as a control (Fig 1 and S1A Table). To determine whether there are positional effects within a multi-hairpin cluster that might influence knockdown efficacy, we stitched together these four short hairpins to create four different 4[sh] clusters, UAS-tester1-4, where each short hairpin occupied a different position in the cluster (Fig 1A and S1B and S1C Table). We cloned these clusters into our multigenic pWALIUM 3xUAS attB vector [14] (S1A and S1B Fig) and generated transgenic flies. Driving ubiquitous expression of the 4[sh] clusters using tubulin-gal4 (tub-gal4) resulted in a strong knockdown of all targeted genes (Fig 1B–1D and S2A and S2B Table). We did not see a significant difference in knockdown efficiency among the four clusters (Fig 1B–1D), indicating that there are no positional effects, at least in the context of a 4[sh] cluster. Furthermore, cluster-mediated knockdown efficacy was comparable to single hairpin controls, demonstrating that hairpin expression from a cluster does not reduce its efficacy. Notably, p53 shRNA expression from all four 4[sh] clusters resulted in strong p53 knockdown, while the same sequence as a single hairpin did not (Fig 1C). While not generalizable to other hairpins (e.g. Fig 1D), our findings demonstrate that, at least in some cases, short hairpins can be more effective in the context of a synthetic cluster.

thumbnail
Download:
Fig 1. Evaluating knockdown efficacy and positional effects in the context of synthetic multi-hairpin clusters.

A. 4[sh] tester cluster design. The position of each short hairpin is varied so that all four short hairpins occupy all four possible positions within the cluster. B. Evaluation of sn and w knockdown efficacy using bristle and eye phenotypes observed in response to the ubiquitous expression of tester clusters and single hairpin controls. All phenotypes were fully penetrant (n = 50 animals). tub-gal4-only animals were used as baseline controls to demonstrate the wild-type eye and bristle phenotypes. C,D. Evaluation of p53 (C) and GFP (D) knockdown efficacy in response to cluster and single hairpin expression by western blotting. Error bars represent the standard error of the mean (SEM). *:p ≤0.05, **:p≤ 0.01, ***: p≤ 0.001, ****: p≤ 0.0001 (ordinary one-way ANOVA with Tukey’s multiple comparisons). Differences between the four tester constructs are not statistically significant.

https://doi.org/10.1371/journal.pgen.1010792.g001

To determine whether knockdown efficacy changes in longer synthetic clusters, we generated 8[sh], 12[sh], and 16[sh] clusters by adding additional short hairpins to tester-1 (Fig 2A and S1A and S1D Table), cloned them into pWALIUM 3xUAS attB (S1A and S1B Fig) and established transgenic flies. As simultaneous knockdown of many genes can result in lethality and hinder the evaluation of knockdown efficacy, we prioritized short hairpins targeting exogenous genes frequently used as markers in Drosophila, where transgenic UAS constructs to induce their expression for evaluating knockdown efficacy are available. To fill the remaining positions, we selected hairpins targeting Drosophila genes whose ubiquitous expression did not result in any phenotype or lethality. Despite these efforts, ubiquitous expression of the longer, 8[sh], 12[sh], and 16[sh] clusters during development resulted in organismal lethality. As a result, we evaluated knockdown efficacy by transiently inducing cluster expression for three days using tub-gal80ts, a temperature-sensitive allele of the Gal4 inhibitor Gal80.

thumbnail
Download:
Fig 2. Evaluating knockdown efficacy in the context of longer clusters.

A. Longer short hairpin cluster designs. 8[sh], 12[sh], 16[sh] clusters were generated by adding additional hairpins to the 4[sh] tester-1 cluster. Exogenous genes targeted by the clusters are in orange. B-G. Knockdown efficacy of short hairpins at indicated positions within longer clusters by qPCR (B,C,E-G) and western blot (D) analysis. Clusters that do not include the hairpin being evaluated are used as negative controls (4[sh] for D and E, 8[sh] for F, 12[sh] for G). Error bars represent the standard error of the mean (SEM). *:p ≤0.05, **:p≤ 0.01, ***: p≤ 0.001, ****: p≤ 0.0001 (ordinary one-way ANOVA with Tukey’s multiple comparisons).

https://doi.org/10.1371/journal.pgen.1010792.g002

We found that increasing cluster length did not affect the efficacy of hairpins present in the original tester-1 cluster, indicating that longer clusters are equally effective in gene knockdown. For instance, both GFP and sn hairpins, at positions 1 and 3, respectively, provided strong and comparable knockdown in 4[sh], 8[sh], 12[sh] and 16[sh] clusters (Fig 2B and 2C). Hairpins that were only present in longer constructs were also effective: For instance, short hairpins targeting β-galactosidase (β-gal) and sevenless (sev) genes, present in positions 5 and 7 respectively, resulted in comparable knockdown in response to 8[sh], 12[sh] and 16[sh] cluster expression (Figs 2D and 2E and S2C). We had similar results with hairpins that are present in 12[sh] and 16[sh] (Fig 2F) and in 16[sh] only (Fig 2G). In each case, the expression of a cluster that did not include the tested hairpin was used as a negative control (Fig 2D–2G). Overall, these results demonstrate that cluster length does not negatively impact knockdown efficacy, and at least 16 short hairpins can be expressed from a single synthetic cluster for combinatorial gene knockdowns.

Given the lethality induced by ectopic expression of the longer clusters, we next tested the possibility that the expression of long multi[sh] clusters with many inverted repeats may have some non-specific toxic effects. To this end, we generated two new constructs by inverting the orientation of DNA sequences encoding 8[sh] and 16[sh] clusters to change the transcribed strand and produce a transcript that has the same number and sequence of short hairpins as the original constructs but the opposite orientation of the passenger and guide sequences (S3A Fig and S1 Table). We cloned the inverted clusters into pWALIUM 3xUAS attB and generated transgenic flies. Ubiquitous expression of these inverted clusters had no effect on viability (S3B Fig), suggesting that the lethality observed with the original constructs is likely a synergistic response to the simultaneous knockdown of multiple genes rather than general toxicity of long multi[sh] cluster expression. Furthermore, an 8[sh] cluster targeting the p53 gene, which is not required for development [27], did not result in any lethality despite providing strong p53 knock-down at the protein and RNA level (S3C–S3F Fig), further supporting the conclusion that long hairpin clusters are not inherently toxic to the cell.

These findings demonstrate that while feasible, simultaneous and ubiquitous knock-down of a large number of genes could result in synthetic lethality, potentially limiting the use of longer clusters in biological studies. However, long clusters have additional, more broadly relevant applications, including as an in vivo hairpin discovery tool to identify effective hairpins for a large number of genes for subsequent studies, calibrating the level of knockdown of a gene by varying the number of hairpins, or targeting fewer genes with multiple short hairpins each, as an in vivo, transgenic equivalent of siRNA pools commonly used in cell-based assays. We next explored some of these applications focusing on Drosophila orthologs of genes commonly mutated in colorectal cancer.

Using the multi[sh] design to screen hairpins targeting colon cancer tumor suppressors

Sequencing and genome-wide association studies often identify genes whose orthologs are not well-characterized and for which functionally validated short hairpins are not available. Multi[sh] clusters offer a streamlined way to screen short hairpins targeting several genes for in vivo efficacy without the need for generating separate transgenic animals for each. To test our approach using disease-relevant genes, we built a series of multi[sh] constructs targeting Drosophila orthologs of four tumor suppressors recurrently mutated in colon cancer: TP53, APC, SMAD4 and SMAD2 (Drosophila orthologs p53, apc, Med and smox, respectively). First, we generated four different 4[sh] clusters where each gene is targeted with a different hairpin sequence, including one, pams-4, constructed using hairpins that were predicted to be moderately effective as a control (Fig 3A and S2A Table). Hairpins were selected using the publicly available shRNA prediction tool DSIR [28,29] as previously described [15]. We cloned each cluster into pWALIUM 3xUAS attB (S1A and S1B Fig) and generated transgenic flies. We then tested the knockdown efficiency of each short hairpin in vivo by ubiquitous transient expression of the clusters. Using 4 transgenic fly lines, this approach allowed in vivo evaluation of 16 different short hairpins to identify multiple short hairpins that provided a significant knockdown for each gene (Fig 3B). As predicted, the pams-4 control was the least effective cluster.

thumbnail
Download:
Fig 3. Multi[sh] cluster design as a streamlined screening strategy to identify effective short hairpins for multiple tumor suppressors.

A. Four different 4[sh] clusters targeting Drosophila orthologs of four recurrently mutated colorectal cancer tumor suppressors TP53 (p53), APC (apc), SMAD4 (Med) and SMAD2 (smox). Each cluster expresses a different short hairpin targeting the same set of four genes. B. Evaluation of knockdown efficacy of p53, apc, Med and smox in response to the ubiquitous expression of the 4[sh] pams clusters by qPCR. C. An 8[sh] cluster generated by stitching pams-2 and pams-3 4[sh] clusters. D. Evaluation of knockdown efficacy of p53, apc, Med and smox in response to the ubiquitous expression of the 8[sh] pams cluster compared to pams-2 and pams-3 by qPCR. E. A 16[sh] cluster which includes 4 different hairpin sequences targeting each gene. F. Evaluation of knockdown efficacy of p53, apc, Med and smox in response to the ubiquitous expression of the 16[sh] pams cluster by qPCR. B,D,F. Error bars represent the standard error of the mean (SEM). *:p ≤0.05, **:p≤ 0.01, ***: p≤ 0.001, ****: p≤ 0.0001 (ordinary one-way ANOVA with Tukey’s multiple comparisons).

https://doi.org/10.1371/journal.pgen.1010792.g003

Next, we tested whether we could calibrate the level of knockdown by varying the number of short hairpins targeting a gene. To this end, we stitched pams-2 and pams-3 clusters (Fig 3A) together by PCR to generate the pams[8sh2-3] construct that includes two different short hairpins for each targeted gene (Fig 3C). We cloned the new cluster into pWALIUM 3xUAS attB (S1A and S1B Fig) and generated transgenic flies. We then compared the level of knockdown induced by the pams[8sh2-3] and the two 4[sh] clusters pams-2 and pams-3. Bringing together two hairpins targeting each gene did not increase the level of knockdown (Fig 3D), suggesting that there are no additive effects between multiple hairpins targeting the same gene, at least in the case of these four genes we analyzed. Regardless, these results demonstrate that the multi[sh] clusters are useful to evaluate the efficacy of multiple hairpins in vivo and identify optimal hairpin sequences to build new clusters for disease modeling or functional studies.

Even if including multiple hairpins targeting a single gene does not result in a progressively stronger knockdown, it may increase the likelihood that at least one hairpin with strong in vivo efficacy will be present in the final cluster, eliminating the need to create new clusters using hairpins identified from the initial screen. To further explore this possibility, we created a new 16[sh] cluster targeting the same 4 genes with four different short hairpin sequences for each gene, using a combination of short hairpins from previous pams constructs and new hairpins designed using DSIR (Fig 3A and 3E and S2B Table). We cloned the pams16[sh] cluster into pWALIUM 3xUAS attB (S1A and S1B Fig) and generated transgenic flies. The pams16[sh] construct provided a strong knockdown of all four genes (Fig 3F), demonstrating that longer clusters can be beneficial for targeting a small number of genes, as multiple hairpins targeting each gene of interest can be included in the design. Overall, these findings demonstrate the utility of this approach in evaluating the efficacy of a large number of hairpins in vivo to optimize knockdown efficacy and as a genetically encoded pooled shRNA strategy to increase the likelihood of strong knockdown.

A polycistronic design for combinatorial gene expression and knockdown

Our multigenic vector design allows multi-gene expression and knockdown from a plasmid with three separate UAS cassettes (S1A Fig). We next sought to consolidate these into a polycistronic transgene that expresses a multi[sh] cluster from within an intron placed upstream of coding sequences of multiple proteins separated by a T2A sequence from a single UAS cassette. Intron-mediated expression of single short hairpins has been documented before [20], and T2A-mediated expression of multiple proteins is a relatively common strategy in Drosophila [21,22]. To combine these two elements with our multi[sh] cluster strategy, we first designed a new UAS-cassette with two distinct multiple cloning sites, one within the Drosophila ftz intron to allow intron-mediated shRNA expression [20], and a second one to introduce protein-coding sequences downstream of the intron (S1C Fig). We then generated a new multigenic vector, pWALIUM-intron 3xUAS attB, that includes this polycistronic UAS-cassette along with two standard ones to allow expression of additional transgenes (S1D Fig).

To determine whether intron-mediated multi[sh] cluster expression altered knockdown efficacy, we cloned the tester-1 cluster (Fig 1A) into the cluster cloning site within the ftz intron and the mCherry coding sequence into the CDS cloning site downstream of the intron (Fig 4A) in pWALIUM-intron 3xUAS attB (S1C and S1D Fig) and generated transgenic flies. Ubiquitous expression of this transgene resulted in strong gene knockdown comparable to non-intron mediated expression of the same cluster from the original pWALIUM 3xUAS attB vector (Fig 4B), demonstrating that the ftz intron can be used as a structural design element to facilitate co-expression of a hairpin cluster and a protein as a single transcript.

thumbnail
Download:
Fig 4. A polycistronic design for simultaneous gene knockdown and expression.

A. The construct designed for intron-mediated expression of the 4[sh] tester 1 cluster. B. Evaluation of sn and GFP knockdown in response to intron-mediated cluster expression. C. The bicistronic construct designed for T2A-mediated expression of Drosophila orthologs of two oncogenes commonly altered in human colon tumors, an oncogenic KRAS[G12V] (dRASG12V) and IRS2 (chico). The control construct where each coding sequence is cloned downstream of its own UAS enhancer/promoter in the multigenic vector. D. Evaluation of dRASG12V and Chico protein expression from the bicistronic construct by western blot analysis. E. The polycistronic construct designed for knockdowns of the 4 colorectal cancer tumor suppressors by intron-mediated expression of a 4[sh] cluster and T2A-mediated expression of two proteins from a single transcript. F,G. Evaluation of knockdown efficacy of p53, apc, Med and smox by qPCR (F) and dRASG12V and Chico protein expression (G) upon ubiquitous expression of the 6-hit polycistronic construct (E). B,F. Error bars represent the standard error of the mean (SEM). **:p≤ 0.01, ***: p≤ 0.001, ****: p≤ 0.0001 (ordinary one-way ANOVA with Tukey’s multiple comparisons).

https://doi.org/10.1371/journal.pgen.1010792.g004

To test the ectopic expression of multiple proteins, we selected two oncogenes commonly altered in human colon tumors, dRASG12V, which represents a commonly observed oncogenic mutation in KRAS, and chico, the Drosophila ortholog of IRS2, a gene that frequently shows copy number gain in human colon tumors [30,31]. dRASG12V and chico coding sequences were tagged by FLAG and HA epitopes, respectively, to facilitate detection by western blot analysis. We created a single bicistronic transgene where the two coding sequences, separated by the T2A sequence (chico—HA—T2A - FLAG—dRASG12V), were tandemly cloned into MCS1 of pWALIUM 3xUAS attB (Figs 4C and S1A). As a control, we separately cloned each tagged coding sequence downstream of its own UAS promoter into MCS1 and MCS2 of pWALIUM 3xUAS attB (UAS-chico–HA; UAS-FLAG—dRASG12V, Figs 4C and S1A) and generated transgenic flies. Transient ubiquitous expression of the bicistronic and control transgenes resulted in strong expression of both proteins (Fig 4D).

Next, we tested the final polycistronic transgene design by combining the short hairpin cluster and T2A-mediated bicistronic protein expression into a single UAS-inducible transgene that represents a 6-hit colorectal cancer model expressing 2 oncogenes and knockdown of 4 tumor suppressors. To this end, we cloned the 4[sh] cluster pams-1 and the [chico—HA—T2A - FLAG—dRASG12V] fragment into the polycistronic UAS cassette of our new pWALIUM-intron 3xUAS attB vector (Fig 4E and S1C and S1D Fig) and generated transgenic flies. Transient ubiquitous expression of this polycistronic transgene resulted in significant knockdown of p53, apc, Med and smox (Fig 4F) and ectopic expression of both Chico and dRASG12V proteins (Fig 4G). The use of a single transgene to manipulate 6 different genes from a single polycistronic transcript frees the remaining two cassettes of the multigenic vector for additional transgenes. These cassettes can be used to clone additional polycistronic transgenes to further increase the complexity of disease models or incorporate some of the components required for inducible genetic tools to further reduce the number of independent transgenic insertions required in experimental animals. Lastly, the polycistronic transgene design is compatible with the standard single-cassette vectors typically used for transgenesis in Drosophila and can be adapted for other experimental systems to perform combinatorial genetic manipulations without the need for a multigenic vector.

Discussion

In recent years, in vivo combinatorial genetic perturbations have become increasingly essential to generate experimental models that capture the genetic complexity and heterogeneity of human diseases and for detailed mechanistic explorations of local and systemic interactions underlying disease states. Here, we report a new design that combines elements from different genetic manipulation tools for targeted, inducible expression and knockdown of multiple genes. This highly flexible and adaptable tool consolidates multiple genetic perturbations into a single, inducible polycistronic transgene to further increase the genetic sophistication of disease models. The ability to use a single UAS cassette for multiple genetic manipulations provides an opportunity to introduce additional transgenes, reporters, and lineage-tracing constructs into the multigenic vector. It also makes it broadly adaptable to other targeted expression, site-specific recombination, and gene editing platforms for mechanistic studies.

Multi-omics approaches to studying disease states often identify complex gene signatures as reporters of disease progression, recurrence, drug response, and resistance [32,33]. Exploring the functional relevance of multigenic signatures and their potential contributions to disease in experimental models that reflect the genetic complexity and heterogeneity of human disease can be challenging. Using this design, combinatorial genetic perturbations reflecting such multigenic signatures can be more easily generated to investigate their roles in disease development or drug response.

The rapid advancement of sequencing technologies and the concomitant reduction of cost led to the identification of a large number of variants in human genes associated with human disease [34]. However, functional exploration of these variants to investigate causal relationships in disease development and to explore underlying molecular mechanisms has been difficult. This is a particular challenge in cancer: tumor sequencing studies routinely identify large numbers of novel variants in potentially cancer-relevant genes whose functional relevance can be challenging to investigate [35–37] as they require representative disease models in genetically tractable experimental systems. “Humanization” of fly genes has been a successful approach for functional studies of disease-associated human genes and variants, where a disease-associated human gene or variant is expressed while simultaneously removing its Drosophila ortholog [38–40]. The polycistronic design described here offers another useful strategy for the humanization of Drosophila genes by inducible expression of a single transcript that expresses the human variant while simultaneously knocking down its Drosophila ortholog. It can be particularly beneficial in cases where multiple Drosophila orthologs with redundant functions must be simultaneously removed to study the function of a human gene. It could also be used to express multiple human genes and reduce the activity of their Drosophila orthologs to “humanize” disease-relevant regulatory nodes and investigate their functions.

Recent advances in CRISPR-based technologies have made somatic, inducible gene editing possible in multiple model systems, and multiplexed DNA editing technologies that allow simultaneous editing of multiple genes are promising for modeling polygenic diseases [41]. However, significant challenges associated with editing multiple genes remain. These include reduced efficiency of editing due to competition between individual guide RNAs [42–44], undesired chromosomal rearrangements as a result of cuts at multiple genomic locations [45–47], and potential activation of DNA damage response pathways [48,49] that can reduce the fitness of targeted cells and modify disease phenotypes. Importantly, as with CRISPR-based multi-gene editing, multiplexed shRNA approaches for gene knockdown can result in enhanced off-target effects [46,47]; therefore, regardless of the approach, carefully controlled experiments to validate these tools and rule out potential off-target effects are essential. Notably, our polycistronic design allows us to design rescue constructs that co-express one or more genes targeted by the hairpin clusters for validation purposes. As the research community attempts to address and overcome technical and biological challenges associated with multiplexed gene-editing technologies, shRNA-mediated knockdown and targeted gene expression strategies in this design offer a well-established complementary approach to combinatorial gene editing, which can be adapted for use in other experimental systems.

Materials and methods

Cloning and Transgenesis

The 4[sh] clusters tester1-4 were synthesized by four rounds of PCRs using the primers listed in S1C Table. Guide and spacer sequences used in the construction are shown in S1A Table. Digitally assembled clusters and individual hairpins are shown in S1B Table. PCR products representing fully assembled [4sh] clusters were digested with XbaI (5’) and EcoRI (3’) introduced during the final round of PCR and cloned into the pWALIUM-Multiple Cloning Site (MCS) of pWALIUM 3xUAS attB, which is designed for shRNA expression (S1A and S1B Fig). To generate the longer test clusters 8[sh], 12[sh], and 16[sh], three additional 4[sh] clusters representing hairpins 5–8, 9–12, and 13–16 were digitally assembled (S1D Table) and generated by gene synthesis (Genewiz). The sequence-confirmed fragments were PCR-amplified and stitched to the 4[sh] cluster tester-1 by sequential PCR reactions. The resulting 8[sh], 12[sh], and 16[sh] clusters were digested with XbaI (5’) and NotI (3’) and cloned into the pWAL-MCS of pWALIUM 3xUAS attB (S1A and S1B Fig). Inverted constructs were generated by PCR-amplifying the 8[sh] and 16[sh] clusters from the multigenic vectors used for transgenesis using a new set of primers designed to append NotI and EcoRI restriction enzyme sites at the 5’ and the 3’ ends of the clusters respectively so that the clusters could be re-cloned into the MCS of the multigenic vector in the inverted orientation. The final products were sequence confirmed (Genewiz), and transgenic flies were generated by PhiC31-mediated targeted integration into the attp2 landing site on the third Drosophila chromosome [13] (BestGene).

Synthesis of 4[sh] pams1-4 clusters (S2A Table) and the 4[sh] clusters targeting p53, apc, Med, and smox that make up the pams [16sh] cluster (S2B Table) were outsourced to Genewiz. Sequence confirmed 4[sh] pams clusters were cloned into the multigenic vector using XbaI (5’) and EcoRI (3’). 4[sh] p53, apc, Med, and smox clusters were stitched together by sequential PCRs to create the final [16sh] pams cluster, which was then cloned into the pWAL-MCS of pWALIUM 3xUAS attB using XbaI (5’) and NotI (3’) (S1A and S1B Fig). pams-2 and pams-3 4[sh] clusters were similarly stitched together to create the pams[8sh2-3] cluster, which was cloned into the pWAL-MCS of pWALIUM 3xUAS attB using XbaI (5’) and EcoRI (3’) (S1A and S1B Fig). All constructs were sequence verified before transgenesis. Transgenic flies were generated by PhiC31-mediated targeted integration [13]. Vectors carrying the 4[sh] pams1-4 and the pams[8sh2-3] clusters were inserted into the attp40 landing site on the 2nd chromosome and the pams [16sh] vector into the attp2 landing site on the third chromosome (Bestgene).

Validation constructs for T2A mediated protein expression were generated as follows: First, a fragment that contained the T2A sequence flanked by the 3xHA tag sequence at the 5’ end and the 3xFLAG tag sequence at the 3’ end was digitally assembled using sequences from Addgene plasmids #24355, #47948 and #32426 for 3xHA, 3xFLAG, and T2A respectively and generated by gene synthesis (Genewiz). The sequence-confirmed final product [AgeI - 3xHA—T2A - 3xFLAG], and the dRASG12V coding sequence from a vector available in-house were PCR-amplified separately and stitched together in another PCR reaction. The stitched product, [AgeI - 3xHA—T2A - 3xFLAG—dRASG12V - PacI], was cloned into the MCS1 of pWALIUM 3xUAS attB using AgeI (5’) and PacI (3’) (S1A Fig) and sequence verified. Next, the chico coding sequence was amplified from a vector available in-house using primers that appended a FseI site to the 5’ end, removed the stop codon, and appended an AgeI site to the 3’ end. The resulting product was cloned into the multigenic vector upstream of [AgeI - 3xHA—T2A - 3xFLAG—dRASG12V - PacI] using FseI (5’) and AgeI (3’) to create the final [UAS-chico- 3xHA—T2A - 3xFLAG—dRASG12V - SV40] construct. The sequence-verified final construct was also used to PCR amplify individual tagged coding sequences [FseI—chico- 3xHA—PacI] and [BsiWI 3xFLAG—dRASG12V - PmeI] for cloning into the MCS1 and MCS2 of pWALIUM 3xUAS attB respectively (S1A Fig) to create [UAS—FseI—chico- 3xHA—PacI—SV40; UAS—BsiWI 3xFLAG—dRASG12V - PmeI—SV40]. Transgenic flies carrying each construct were generated by PhiC31-mediated targeted integration into the attp2 landing site on chromosome three [13] (BestGene).

pWALIUM-intron 3xUAS attB vector construction

For intron-mediated cluster and protein co-expression, we designed a new UAS cassette, [10XUAS—hsp70 basal promoter—ftz intron(cluster MCS: Acc65I, KpnI, XhoI)–CDS MCS: NotI, XbaI], which included 1) a longer ftz intron sequence previously used for intron-mediated short hairpin expression with an MCS within the intron to clone short hairpin sequences [20], and 2) another MCS for cloning a protein-coding sequence (CDS) immediately downstream of the intron (S1C Fig). The new cassette was generated by gene synthesis (Genewiz) and used to create a new multigenic vector, pWALIUM-intron 3xUAS attB, along with two additional UAS cassettes with their own unique MCS for cloning additional transgenes (S1C and S1D Fig).

To test intron-mediated cluster expression, mCherry coding sequence was PCR-amplified from a transgenic UAS-mCherry line available in-house using primers designed to append NotI and XbaI restriction sites to the 5’ and 3’ends of the fragment, respectively, and cloned into the CDS cloning site downstream of the ftz intron (S1C Fig). The tester-1 4[sh] cluster was PCR-amplified using a new set of primers designed to append KpnI and XhoI to the 5’ and 3’ ends, respectively, and cloned into the cluster cloning site within the intron (S1C Fig). The final product vector was confirmed by sequencing. Transgenic flies were generated by targeted insertion into the attp2 landing site on the 3rd chromosome (Bestgene).

To generate the polycistronic construct that combines colon cancer tumor suppressor knockdown and oncogene expression, the pams-1 [4sh] cluster was PCR-amplified from pWALIUM 3xUAS attB with a new set of primers to append Kpn I and XhoI restriction sites at the 5’ and 3’ ends of the cluster respectively and cloned into the cluster cloning MCS within the ftz intron of pWALIUM-intron 3xUAS attB using Kpn I(5’) and XhoI (3’) (S1C Fig). The [chico- 3xHA—T2A - 3xFLAG—dRASG12V] fragment was PCR amplified from pWALIUM 3xUAS attB using primers designed to append NotI and XbaI sites to the 5’ and 3’ ends of the fragment and cloned into the CDS cloning site of pWALIUM-intron 3xUAS attB downstream of the ftz intron using those enzymes (S1C Fig). The final sequence verified vector was used to generate transgenic flies by PhiC31-mediated targeted integration into the attp2 landing site on chromosome three [13] (BestGene).

Drosophila crosses and phenotypic analyses

Experimental animals used in this study were generated by crossing male flies carrying the UAS-transgenes described in the previous section to virgin females with the genotype w1118; tub-gal4 tub-gal80ts /TM6b, Hu, Tb. tub-gal4 drives UAS-transgene expression ubiquitously and tub-gal80ts[7] was used temporally regulate Gal4 activity. All crosses were set up on Bloomington’s semi-defined medium. Crosses to generate adult experimental animals for phenotypic and survival analyses were set up at 29°C to induce transgene expression throughout development. Experimental animals for qPCR and western blot analyses were generated from crosses set up at 18°C to keep transgenes silent during early development and shifted to 29°C to transiently induce transgene expression for 3 days during larval development.

qPCR and western blot analysis

RNA and protein were extracted from whole experimental and control larvae at the late 3rd instar stage using previously established protocols [14,50] (6 larvae/biological replicate; 3 biological replicates/genotype). For evaluating the knockdown efficacy of exogenous proteins, animals ubiquitously expressing GFP (Fig 1D), β-gal (Fig 2D), and CD2 (Fig 3F) were used as controls. For all others, w1118; tub-gal4 tub-gal80ts /+ animals were used as controls. Western blot analyses were performed as described previously [14,50] using the following primary and secondary antibodies: mouse anti-OctA(FLAG) antibody (1:1000, Santa Cruz Biotechnology, sc-166355), rat anti-HA High-Affinity Antibody (1:1000, ROAHAHA, MilliporeSigma), rabbit anti-GFP N-terminal antibody (1:4000, G1544, MilliporeSigma), mouse anti-p53 antibody (1:1000, Developmental Studies Hybridoma Bank, #dmp53 H3), and mouse anti-Syntaxin antibody (1:1000, Developmental Studies Hybridoma Bank, #8C3) as a loading control, goat anti-mouse and goat-anti-rat IgG HRP-linked secondary antibodies (1:5000, Cell Signaling Technology #7076 and #7077, respectively). qPCR analyses were performed as described previously [14] using the housekeeping gene rpl32 as an internal control. qPCR data were analyzed using the double-ΔCT method [51].

Supporting information

S1 Fig. Multigenic vectors used in the study.

A. Map of our previously published pWALIUM 3xUAS attB vector, which includes three UAS cassettes, each with its unique multiple cloning site (MCS). B. Detailed map of the pWALIUM-derived cassette of pWALIUM 3xUAS attB, which was used to clone the test clusters in this study. C. The new polycistronic UAS cassette designed for intron-mediated expression of a short-hairpin cluster and a protein-coding sequence as a single transcript generated in this study. D. Map of the new pWALIUM-intron 3xUAS attB multigenic vector which includes the new cassette (C) along with two other standard cassettes (UAS-MCS1 and UAS-MCS2). Complete sequences of both plasmids are provided in S1 Data and S2 Data.

https://doi.org/10.1371/journal.pgen.1010792.s001

(TIFF)

S2 Fig. Western blots used for quantification in Figs 1 and 2.

Western blot images used for generating the data presented in Fig 1C (A), 1D (B), and 2D (C). Syntaxin (Syt) was used as loading control.

https://doi.org/10.1371/journal.pgen.1010792.s002

(TIFF)

S3 Fig. Evaluating long cluster toxicity using control clusters.

A. The inverted cluster design is illustrated using the 8[sh]-INV construct as an example. The inverted constructs are designed to generate the same number and sequence of hairpins when expressed but do not target any genes. B. Ubiquitous expression of the inverted 8[sh] and 16[sh] clusters does not result in organismal lethality. C-F. An 8-hairpin cluster targeting p53 with strong efficacy at the RNA (C), and protein level (D,E) does not result in organismal lethality upon ubiquitous expression during development (F).

https://doi.org/10.1371/journal.pgen.1010792.s003

(TIFF)

S1 Table. Tester construct designs and sequences.

A. Guide sequences and their sources for each gene targeted in test 4[sh], 8[sh], 12[sh], 16[sh] clusters (top) and 39 nucleotide-long sequences selected as spacers to separate individual hairpins (bottom). Spacer sequences are derived from well-characterized Drosophila microRNA genes as indicated. B. 4[sh] test cluster design. Sequences that make up each short hairpin (5’ flank, passenger, loop, guide and 3’ flank), concatenated short-hairpin sequences, and the full clusters, including the spacer sequences between individual short hairpins are provided. C. Primers for 4 sequential PCR reactions used to synthesize the 4[sh] test clusters. D. Additional 4[sh] clusters designed to append to tester-1 to generate 8[sh], 12[sh] and 16[sh] test clusters.

https://doi.org/10.1371/journal.pgen.1010792.s004

(XLSX)

S2 Table. Proof of concept pams construct designs and sequences.

A. Hairpin and cluster sequences of the four 4[sh] pams constructs pams1-4. B. Hairpin and cluster sequences of the 16[sh] pams constructs. Spacer sequences that separate individual hairpins can be found in S1A Table.

https://doi.org/10.1371/journal.pgen.1010792.s005

(XLSX)

S1 Data. Complete sequence and map of the pWALIUM 3xUAS attB vector.

https://doi.org/10.1371/journal.pgen.1010792.s006

(DNA)

S2 Data. Complete sequence and map of the pWALIUM-intron 3xUAS attB vector.

https://doi.org/10.1371/journal.pgen.1010792.s007

(DNA)

S3 Data. Numerical data underlying graphs and summary statistics presented in the main figures.

https://doi.org/10.1371/journal.pgen.1010792.s008

(XLSX)

Acknowledgments

We thank Dr. Brian Washburn, Dr. Diego Zorio, and Cheryl Pye at the Florida State University Molecular Cloning Facility and Jason Cassara for technical support. This study used transgenic RNAi lines (Office of the Director R24 OD030002: "TRiP resources for modeling human disease") obtained from the Bloomington Drosophila Stock Center (NIH P40OD018537).

References

  1. 1. Kumuthini J, Zick B, Balasopoulou A, Chalikiopoulou C, Dandara C, El-Kamah G, et al. The clinical utility of polygenic risk scores in genomic medicine practices: a systematic review. Hum Genet. 2022;141: 1697–1704. pmid:35488921
  2. 2. Chatrath A, Przanowska R, Kiran S, Su Z, Saha S, Wilson B, et al. The pan-cancer landscape of prognostic germline variants in 10,582 patients. Genome Med. 2020;12: 15. pmid:32066500
  3. 3. Hoadley KA, Yau C, Hinoue T, Wolf DM, Lazar AJ, Drill E, et al. Cell-of-Origin Patterns Dominate the Molecular Classification of 10,000 Tumors from 33 Types of Cancer. Cell. 2018;173: 291–304.e6. pmid:29625048
  4. 4. Sanchez-Vega F, Mina M, Armenia J, Chatila WK, Luna A, La KC, et al. Oncogenic Signaling Pathways in The Cancer Genome Atlas. Cell. 2018;173: 321–337.e10. pmid:29625050
  5. 5. Ugur B, Chen K, Bellen HJ. Drosophila tools and assays for the study of human diseases. Dis Model Mech. 2016;9: 235–244. pmid:26935102
  6. 6. Brand AH, Perrimon N. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development. 1993;118: 401–415. pmid:8223268
  7. 7. McGuire SE, Mao Z, Davis RL. Spatiotemporal Gene Expression Targeting with the TARGET and Gene-Switch Systems in Drosophila. Science Signaling. 2004. pp. l6–pl6. pmid:14970377
  8. 8. Zirin J, Hu Y, Liu L, Yang-Zhou D, Colbeth R, Yan D, et al. Large-Scale Transgenic Resource Collections for Loss- and Gain-of-Function Studies. Genetics. 2020;214: 755–767.
  9. 9. Jory A, Estella C, Giorgianni MW, Slattery M, Laverty TR, Rubin GM, et al. A survey of 6,300 genomic fragments for cis-regulatory activity in the imaginal discs of Drosophila melanogaster. Cell Rep. 2012;2: 1014–1024. pmid:23063361
  10. 10. Ogienko AA, Andreyeva EN, Omelina ES, Oshchepkova AL, Pindyurin AV. Molecular and cytological analysis of widely-used Gal4 driver lines for Drosophila neurobiology. BMC Genet. 2020;21: 96. pmid:33092520
  11. 11. Lee P-T, Zirin J, Kanca O, Lin W-W, Schulze KL, Li-Kroeger D, et al. A gene-specific library for. Elife. 2018;7. pmid:29565247
  12. 12. Bangi E, Murgia C, Teague AGS, Sansom OJ, Cagan RL. Functional exploration of colorectal cancer genomes using Drosophila. Nat Commun. 2016;7: 13615. pmid:27897178
  13. 13. Bischof J, Maeda RK, Hediger M, Karch F, Basler K. An optimized transgenesis system for Drosophila using germ-line-specific phiC31 integrases. Proc Natl Acad Sci U S A. 2007;104: 3312–3317. pmid:17360644
  14. 14. Bangi E, Ang C, Smibert P, Uzilov AV, Teague AG, Antipin Y, et al. A personalized platform identifies trametinib plus zoledronate for a patient with KRAS-mutant metastatic colorectal cancer. Sci Adv. 2019;5: eaav6528. pmid:31131321
  15. 15. Ni J-Q, Zhou R, Czech B, Liu L-P, Holderbaum L, Yang-Zhou D, et al. A genome-scale shRNA resource for transgenic RNAi in Drosophila. Nat Methods. 2011;8: 405–407. pmid:21460824
  16. 16. Ni J-Q, Markstein M, Binari R, Pfeiffer B, Liu L-P, Villalta C, et al. Vector and parameters for targeted transgenic RNA interference in Drosophila melanogaster. Nat Methods. 2008;5: 49–51. pmid:18084299
  17. 17. Altuvia Y. Clustering and conservation patterns of human microRNAs. Nucleic Acids Research. 2005. pp. 2697–2706. pmid:15891114
  18. 18. Kabekkodu SP, Shukla V, Varghese VK, Souza JD, Chakrabarty S, Satyamoorthy K. Clustered miRNAs and their role in biological functions and diseases. Biological Reviews. 2018. pp. 1955–1986. pmid:29797774
  19. 19. Bangi E, Smibert P, Uzilov AV, Teague AG, Gopinath S, Antipin Y, et al. A Drosophila platform identifies a novel, personalized therapy for a patient with adenoid cystic carcinoma. iScience. 2021;24: 102212. pmid:33733072
  20. 20. Haley B, Foys B, Levine M. Vectors and parameters that enhance the efficacy of RNAi-mediated gene disruption in transgenic Drosophila. Proc Natl Acad Sci U S A. 2010;107: 11435–11440. pmid:20534445
  21. 21. Daniels RW, Rossano AJ, Macleod GT, Ganetzky B. Expression of multiple transgenes from a single construct using viral 2A peptides in Drosophila. PLoS One. 2014;9: e100637. pmid:24945148
  22. 22. Diao F, White BH. A novel approach for directing transgene expression in Drosophila: T2A-Gal4 in-frame fusion. Genetics. 2012;190: 1139–1144. pmid:22209908
  23. 23. Lai S-L, Lee T. Genetic mosaic with dual binary transcriptional systems in Drosophila. Nat Neurosci. 2006;9: 703–709. pmid:16582903
  24. 24. Bello B, Resendez-Perez D, Gehring WJ. Spatial and temporal targeting of gene expression in Drosophila by means of a tetracycline-dependent transactivator system. Development. 1998;125: 2193–2202. pmid:9584119
  25. 25. Potter CJ, Tasic B, Russler EV, Liang L, Luo L. The Q system: a repressible binary system for transgene expression, lineage tracing, and mosaic analysis. Cell. 2010;141: 536–548. pmid:20434990
  26. 26. Hu Y, Comjean A, Rodiger J, Liu Y, Gao Y, Chung V, et al. FlyRNAi.org-the database of the Drosophila RNAi screening center and transgenic RNAi project: 2021 update. Nucleic Acids Res. 2021;49: D908–D915. pmid:33104800
  27. 27. Lee JH, Lee E, Park J, Kim E, Kim J, Chung J. In vivo p53 function is indispensable for DNA damage-induced apoptotic signaling in Drosophila. FEBS Lett. 2003;550: 5–10. pmid:12935877
  28. 28. Filhol O, Ciais D, Lajaunie C, Charbonnier P, Foveau N, Vert J-P, et al. DSIR: assessing the design of highly potent siRNA by testing a set of cancer-relevant target genes. PLoS One. 2012;7: e48057. pmid:23118925
  29. 29. Vert J-P, Foveau N, Lajaunie C, Vandenbrouck Y. An accurate and interpretable model for siRNA efficacy prediction. BMC Bioinformatics. 2006;7: 520. pmid:17137497
  30. 30. Day E, Poulogiannis G, McCaughan F, Mulholland S, Arends MJ, Ibrahim AEK, et al. IRS2 is a candidate driver oncogene on 13q34 in colorectal cancer. Int J Exp Pathol. 2013;94: 203–211. pmid:23594372
  31. 31. Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature. 2012;487: 330–337. pmid:22810696
  32. 32. Pilarczyk M, Fazel-Najafabadi M, Kouril M, Shamsaei B, Vasiliauskas J, Niu W, et al. Connecting omics signatures and revealing biological mechanisms with iLINCS. Nat Commun. 2022;13: 4678. pmid:35945222
  33. 33. Kreitmaier P, Katsoula G, Zeggini E. Insights from multi-omics integration in complex disease primary tissues. Trends Genet. 2023;39: 46–58. pmid:36137835
  34. 34. McCarthy MI, MacArthur DG. Human disease genomics: from variants to biology. Genome Biology. 2017. pmid:28137298
  35. 35. Federici G, Soddu S. Variants of uncertain significance in the era of high-throughput genome sequencing: a lesson from breast and ovary cancers. J Exp Clin Cancer Res. 2020;39: 46. pmid:32127026
  36. 36. Medendorp NM, Hillen MA, Murugesu L, Aalfs CM, Stiggelbout AM, Smets EMA. Uncertainty related to multigene panel testing for cancer: a qualitative study on counsellors’ and counselees’ views. Journal of Community Genetics. 2019. pp. 303–312. pmid:30430454
  37. 37. Morash M, Mitchell H, Beltran H, Elemento O, Pathak J. The Role of Next-Generation Sequencing in Precision Medicine: A Review of Outcomes in Oncology. Journal of Personalized Medicine. 2018. p. 30. pmid:30227640
  38. 38. Bellen HJ, Wangler MF, Yamamoto S. The fruit fly at the interface of diagnosis and pathogenic mechanisms of rare and common human diseases. Hum Mol Genet. 2019;28: R207–R214. pmid:31227826
  39. 39. Marcogliese PC, Deal SL, Andrews J, Harnish JM, Bhavana VH, Graves HK, et al. Drosophila functional screening of de novo variants in autism uncovers damaging variants and facilitates discovery of rare neurodevelopmental diseases. Cell Rep. 2022;38: 110517. pmid:35294868
  40. 40. Savitsky M, Solis GP, Kryuchkov M, Katanaev VL. Humanization of Gαo to Model Paediatric Encephalopathies. Biomedicines. 2020;8. pmid:33036271
  41. 41. McCarty NS, Graham AE, Studená L, Ledesma-Amaro R. Multiplexed CRISPR technologies for gene editing and transcriptional regulation. Nat Commun. 2020;11: 1281. pmid:32152313
  42. 42. Zhang S, Voigt CA. Engineered dCas9 with reduced toxicity in bacteria: implications for genetic circuit design. Nucleic Acids Res. 2018;46: 11115–11125. pmid:30289463
  43. 43. Qian Y, Huang H-H, Jiménez JI, Del Vecchio D. Resource Competition Shapes the Response of Genetic Circuits. ACS Synth Biol. 2017;6: 1263–1272. pmid:28350160
  44. 44. Chen P-Y, Qian Y, Del Vecchio D. A Model for Resource Competition in CRISPR-Mediated Gene Repression. 2018 IEEE Conference on Decision and Control (CDC). 2018. pp. 4333–4338.
  45. 45. Shou J, Li J, Liu Y, Wu Q. Precise and Predictable CRISPR Chromosomal Rearrangements Reveal Principles of Cas9-Mediated Nucleotide Insertion. Mol Cell. 2018;71: 498–509.e4. pmid:30033371
  46. 46. Tycko J, Wainberg M, Marinov GK, Ursu O, Hess GT, Ego BK, et al. Mitigation of off-target toxicity in CRISPR-Cas9 screens for essential non-coding elements. Nat Commun. 2019;10: 4063. pmid:31492858
  47. 47. Morgens DW, Wainberg M, Boyle EA, Ursu O, Araya CL, Tsui CK, et al. Genome-scale measurement of off-target activity using Cas9 toxicity in high-throughput screens. Nat Commun. 2017;8: 15178. pmid:28474669
  48. 48. Liu M, Zhang W, Xin C, Yin J, Shang Y, Ai C, et al. Global detection of DNA repair outcomes induced by CRISPR-Cas9. Nucleic Acids Res. 2021;49: 8732–8742. pmid:34365511
  49. 49. Bowden AR, Morales-Juarez DA, Sczaniecka-Clift M, Agudo MM, Lukashchuk N, Thomas JC, et al. Parallel CRISPR-Cas9 screens clarify impacts of p53 on screen performance. Elife. 2020;9. pmid:32441252
  50. 50. Datta I, Linkous B, Odum T, Drew C, Taylor A, Vassel T, et al. A targeted genetic modifier screen in Drosophila uncovers vulnerabilities in a genetically complex model of colon cancer. bioRxiv. 2022. p. 2022.09.23.508936.
  51. 51. Sopko R, Foos M, Vinayagam A, Zhai B, Binari R, Hu Y, et al. Combining genetic perturbations and proteomics to examine kinase-phosphatase networks in Drosophila embryos. Dev Cell. 2014;31: 114–127. pmid:25284370