Highly efficient cell-type-specific gene inactivation reveals a key function for the Drosophila FUS homolog cabeza in neurons

Abstract

To expand the rich genetic toolkit of Drosophila melanogaster, we evaluated whether introducing FRT or LoxP sites in endogenous genes could allow for cell-type-specific gene inactivation in both dividing and postmitotic cells by GAL4-driven expression of FLP or Cre recombinase. For proof of principle, conditional alleles were generated for cabeza (caz), the Drosophila homolog of human FUS, a gene implicated in the neurodegenerative disorders amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). Upon selective expression in neurons or muscle, both FLP and Cre mediated caz inactivation in all neurons or muscle cells, respectively. Neuron-selective caz inactivation resulted in failure of pharate adult flies to eclose from the pupal case, and adult escapers displayed motor performance defects and reduced life span. Due to Cre-toxicity, FLP/FRT is the preferred system for cell-type-specific gene inactivation, and this strategy outperforms RNAi-mediated knock-down. Furthermore, the GAL80 target system allowed for temporal control over gene inactivation, as induction of FLP expression from the adult stage onwards still inactivated caz in >99% of neurons. Remarkably, selective caz inactivation in adult neurons did not affect motor performance and life span, indicating that neuronal caz is required during development, but not for maintenance of adult neuronal function.

Figure 1. Generation and characterization of caz knock-out and conditional knock-out alleles.

(a), The caz² allele was generated by imprecise excision of the P-transposable element P^Mae-UAS.6.11 DP00882 in the caz gene promoter, which resulted in excision of exons 1 through 3. Caz^KO was generated by homologous recombination, replacing caz exons 2 through 7 by a white marker gene flanked by AttP sites. ΦC31-mediated RMCE was subsequently used to reintroduce caz, flanked by either LoxP (caz^lox) or FRT sites (caz^FRT). (b), Quantitative real-time PCR revealed that caz² and caz^KO mutants are caz transcript null. The ‘revertant' allele is generated by precise excision of P^Mae-UAS.6.11 DP00882, and serves as a control for caz². N = 5 (c), No caz protein could be detected in caz² or caz^KO mutants by Western blot. The full-length blot is shown in Supplementary Figure S5. (d–l), Immunostaining for caz and the neuronal nuclear marker elav revealed caz expression in all neuronal nuclei, as well as in non-neuronal cells in third instar larval ventral cord of control animals. In caz² and caz^KO mutants, no caz protein was detected. Scale bar: 20 μm.

Figure 2. Caz² and caz^KO mutants display developmental delay and pupal lethality, which can be rescued by neuron-selective caz reintroduction.

(a), At the third instar larval stage, caz² and caz^KO mutant animals are present at the expected Mendelian ratios. Results of three independent experiments were pooled, with 300–400 L3 larvae per experiment. (b), caz mutants display an increased time to pupation, indicative of developmental delay. N = 80–110. (c), caz² and caz^KO mutant animals died during the pupal stage, with very few adult escapers. Pupal lethality could be rescued by selective expression of wild type caz in neurons (elav-GAL4). N = 92–393.

Figure 3. Neuron-selective caz inactivation induces pupal lethality, motor deficits and reduced life span.

(a–l), elav-GAL4 was used to drive panneuronal expression of FLP (a–f) or Cre (g–l), either in a wild type or conditional caz KO background. Immunostaining for caz and elav on third instar larval ventral nerve cord revealed caz inactivation in all neurons, but not in non-neuronal cells. Scale bar: 20 μm. (m), Quantification confirmed caz inactivation in 100% of neurons. N = 7–13. (n, o), Adult offspring frequencies of control genotypes versus neuron-selective caz KO mediated by FLP/FRT (n) or Cre/LoxP (o) revealed developmental lethality with a variable fraction of adult escapers. N = 514–665 (n) and 414–540 (o). (p), Neuronal caz KO adult escaper flies display significant motor performance deficits, as revealed by a negative geotaxis climbing assay. N = 100. (q), Neuronal caz KO adult escaper flies display reduced life span. N = 74–90.

Figure 4. Neuronal caz knock-down by transgenic RNAi does not induce developmental lethality and does not or only slightly impair motor performance.

(a), Panneuronal expression of caz-RNAi with nsyb-GAL4 does not induce significant developmental lethality. Data normalized to control (w1118) are shown. N = 83–144. (b), Panneuronal caz knock-down by nsyb-GAL4 induced no or only slight motor performance defects in a negative geotaxis climbing assay. (c), caz transcript levels were quantified by qPCR in heads of adult flies ubiquitously expressing caz-RNAi from the adult stage onwards. Three distinct transgenic caz-RNAi lines were tested, either as a single transgene or in combinations of two transgenes. Significant caz knock-down of ≈ 50 to 80% was observed. caz-RNAi-1: VDRC100291; caz-RNAi-2: HMS00790; caz-RNAi-3: HMS00156.

Figure 5. Muscle-selective caz inactivation.

(a–o), Mef2-GAL4 was used for muscle-selective expression of FLP or Cre, either in wild type or conditional caz KO background. GFP co-expression was used to visualize muscle cells and DAPI to label nuclei. Immunostaining for caz and GFP on third instar larval body wall musculature revealed caz inactivation in all muscle cells. Arrowheads in panels h and i indicate caz expression in a non-muscle cell. Scale bar: 50 μm. (p), Quantification confirmed caz inactivation in 100% of muscle cells. Muscle 7 (indicated in panel a) in segments A1 and A2 was used for quantification. N = 7–10. (q, r), Defects in muscle morphology induced by Cre expression are documented by reduced width of muscle 7 (q), whereas muscle length was not significantly altered (r). N = 8–9.

Figure 6. Temporal control over neuron-selective caz inactivation.

Tubulin-GAL80^ts>elav-GAL4 was used for induction of FLP expression in neurons of adult caz^FRT flies. (a-f), Immunostaining for caz and elav on adult ventral cord revealed highly efficient caz inactivation in adult neurons (scale bar: 10 μm). (g), Quantification documented caz inactivation in >99% of adult neurons. N = 9–11. (h), Loss of neuronal caz function from the adult stage onwards does not affect motor performance at 4 weeks of age, as revealed by a negative geotaxis climbing assay. N = 90–100. (i), Loss of caz in adult neurons does not reduce life span in a biologically relevant manner. N = 100.