An Adapted GeneSwitch Toolkit for Comparable Cellular and Animal Models: A Proof of Concept in Modeling Charcot-Marie-Tooth Neuropathy

Abstract

Investigating the impact of disease-causing mutations, their affected pathways, and/or potential therapeutic strategies using disease modeling often requires the generation of different in vivo and in cellulo models. To date, several approaches have been established to induce transgene expression in a controlled manner in different model systems. Several rounds of subcloning are, however, required, depending on the model organism used, thus bringing labor-intensive experiments into the technical approach and analysis comparison. The GeneSwitch™ technology is an adapted version of the classical UAS-GAL4 inducible system, allowing the spatial and temporal modulation of transgene expression. It consists of three components: a plasmid encoding for the chimeric regulatory pSwitch protein, Mifepristone as an inducer, and an inducible plasmid. While the pSwitch-containing first plasmid can be used both in vivo and in cellulo, the inducible second plasmid can only be used in cellulo. This requires a specific subcloning strategy of the inducible plasmid tailored to the model organism used. To avoid this step and unify gene expression in the transgenic models generated, we replaced the backbone vector with standard pUAS-attB plasmid for both plasmids containing either the chimeric GeneSwitch™ cDNA sequence or the transgene cDNA sequence. We optimized this adapted system to regulate transgene expression in several mammalian cell lines. Moreover, we took advantage of this new system to generate unified cellular and fruit fly models for YARS1-induced Charco-Marie-Tooth neuropathy (CMT). These new models displayed the expected CMT-like phenotypes. In the N2a neuroblastoma cells expressing YARS1 transgenes, we observed the typical "teardrop" distribution of the synthetase that was perturbed when expressing the YARS1^CMT mutation. In flies, the ubiquitous expression of YARS1^CMT induced dose-dependent developmental lethality and pan-neuronal expression caused locomotor deficit, while expression of the wild-type allele was harmless. Our proof-of-concept disease modeling studies support the efficacy of the adapted transgenesis system as a powerful tool allowing the design of studies with optimal data comparability.

Keywords: Charcot–Marie–Tooth neuropathy; Drosophila melanogaster; GeneSwitch™; aminoacyl tRNA-synthetase; disease modelling.

Figure 1

Schematic illustration of the adapted pSwitch system. (A) Simplified scheme of the pSwitch regulatory cassette containing the yeast specific GAL4 DNA binding domain, the truncated human progesterone receptor ligand binding domain (hPR), and an activation domain from the human NF-KB p65 protein (p65). This cassette encodes for the (B) pSwitch regulatory fusion protein in a monomeric inactive state. (C) By adding Mifepristone (red), a human progesterone receptor antagonist, the pSwitch fusion protein undergoes a conformational change leading to homodimerization. (D) The homodimer binds to the yeast UAS sequence to activate transgene expression. We adapted the GeneSwitch™ system by generating (E) the inducible pSwitch-multi vector containing five UAS (5 × UAS), a T7 promoter, a TK minimal promoter, the pSwitch regulatory cassette (GAL4-hPR-p65), a simian virus 40 polyadenylation site (SV40), an attB sequence for site-directed insertion, an SV40 promoter, an EM7 promoter, an Hygromycin resistance gene, and the mini-white gene. (F) The inducible pUAS-attB-multi plasmid contains the same grey features as the pSwitch regulatory vector, the cDNA sequence of the transgene flanked by three tags (HA, V5 and Flag) and a Zeocin resistance gene. (G) Schematic representation of the experimental layout for which expression of the transgene in the pUAS-attB-multi vector can now be controlled in cells and flies using Mifepristone or GAL4 activation, respectively, to ensure optimal data comparability without the need for subcloning.

Figure 2

Induction of YARS1 expression using the pSwitch-multi system in mammalian and invertebrate cell lines. Western blot analysis of protein extracts from (A) N2a, (B) CHO-K1, (C) HeLa, (D) HEK293T and (E) S2 stable and non-stable cell lines that have been transiently transfected with 1 µg of pUAST-attB-multi-YARS1. Four hours after transfection, a recommended concentration of Mifepristone (10 nM) was added to the medium in the conditions containing the two plasmids, for 3 days. (F) Western blot analysis of protein extract from S2 cells 24 h after transient co-transfection with pUAST-attB-multi-YARS1 and/or Actin-GAL4 driver (1:1 transfection ratio). Expression of exogenous YARS1 was detected using mouse monoclonal HA-tag antibody. Endogenous YARS1 expression could be detected in mammalian cells by using mouse monoclonal YARS1 antibody. Equal loading was validated by using mouse monoclonal α-tubulin antibody (n = 3). Note the exogenous YARS1 signal detection in the condition containing both vectors but without mifepristone induction in HeLa, CHO-K1 and HEK293T cells.

Figure 3

Concentration of inducible plasmid to be transfected in stable cell lines for pSwitch-multi regulatory vector. N2a cells stably expressing the pSwitch-multi regulatory vector have been transiently transfected with increased concentrations of inducible plasmid (ranging from 0 to 1 µg) containing (A) YARS1 or (B) AARS1 cDNA sequence. Four hours after transfection, a recommended Mifepristone concentration (10 nM) was added to the medium to induce transgene expression for 24 h. The transgene signal was first determined by using the HA-tag antibody. Exogenous and endogenous YARS1 and AARS1 expression was detected with mouse monoclonal YARS1 or AARS1 antibodies, respectively. Equal loading was validated by using mouse monoclonal α-tubulin antibody. Each graph represents the relative quantification of each band’s intensity compared to the expression level at 0.5 µg inducible plasmid without the presence of Mifepristone (n = 3). Bar charts are represented with s.e.m. for endogenous (black), exogenous in absence (grey) or presence of Mifepristone (red) YARS1 and AARS1, respectively. Statistical significance (** p < 0.01, **** p < 0.0001, ns—not significant) was determined using one-way ANOVA analysis.

Figure 4

Characterization of unified cellular and fly models for YARS^CMT. (A) Immunofluorescence analysis of differentiated N2a cells transfected with wild-type and mutant YARS1 inducible plasmids. YARS1—yellow, nuclei—cyan, actin cytoskeleton—magenta (Scale bar = 20 µM). A zoom-in on the YARS1 staining is represented by a white box. The teardrop effect is indicated by a white arrowhead. (B) Quantification of cells displaying the teardrop effect. Statistical significance ** p < 0.01, * p < 0.1, ns—not significant was determined after one-way ANOVA analysis (n = 4). (C) Strong ubiquitous expression of YARS1^WT has no effect with Actin5c-GAL4weak (light blue) and Actin5c-GAL4strong (dark blue) on the expected 1:1 adult flies eclosion ratio. YARS1^E196K ubiquitous expression has detrimental effects in a dosage-dependent manner in both new and previously published models. Attp40 flies were used as a negative control. The number of adult flies ecloding is indicated above each graph bar. Dashed line marks the expected 1:1 genotypes’ eclosion ratio. Statistical significance (*** p < 0.0001) was determined after One-Way ANOVA analysis which compares the odd ratios (Actin5c-GAL4/transgene ON vs. Balancer/transgene OFF) of flies in three independent crosses. (D) Pan-neuronal expression of mutant YARS1 with nSyb-GAL4 induces locomotor performance deficits as determined in a negative geotaxis climbing assay. NSyb-GAL4 flies were used as a negative control (grey). The Y-axis indicates the time needed for the fastest fly to climb a vertical wall to a height of 82 mm. ** p < 0.01 was determined after one-way ANOVA analysis (n = 3). Both lethality and negative geotaxis assays were performed on flies carrying the transgenic construct on the second chromosome.