Genetic interaction screen for severe neurodevelopmental disorders reveals a functional link between Ube3a and Mef2 in Drosophila melanogaster

Abstract

Neurodevelopmental disorders (NDDs) are clinically and genetically extremely heterogeneous with shared phenotypes often associated with genes from the same networks. Mutations in TCF4, MEF2C, UBE3A, ZEB2 or ATRX cause phenotypically overlapping, syndromic forms of NDDs with severe intellectual disability, epilepsy and microcephaly. To characterize potential functional links between these genes/proteins, we screened for genetic interactions in Drosophila melanogaster. We induced ubiquitous or tissue specific knockdown or overexpression of each single orthologous gene (Da, Mef2, Ube3a, Zfh1, XNP) and in pairwise combinations. Subsequently, we assessed parameters such as lethality, wing and eye morphology, neuromuscular junction morphology, bang sensitivity and climbing behaviour in comparison between single and pairwise dosage manipulations. We found most stringent evidence for genetic interaction between Ube3a and Mef2 as simultaneous dosage manipulation in different tissues including glia, wing and eye resulted in multiple phenotype modifications. We subsequently found evidence for physical interaction between UBE3A and MEF2C also in human cells. Systematic pairwise assessment of the Drosophila orthologues of five genes implicated in clinically overlapping, severe NDDs and subsequent confirmation in a human cell line revealed interactions between UBE3A/Ube3a and MEF2C/Mef2, thus contributing to the characterization of the underlying molecular commonalities.

Figure 1

Genes of interest and study outline. (a) Five syndromic NDDs with considerable phenotypic overlap were selected. AS, Angelman syndrome; MOWS, Mowat-Wilson syndrome; MRD20, mental retardation, autosomal dominant 20 (MEF2C haploinsufficiency syndrome); MRHXF1, mental retardation-hypotonic facies syndrome/ATRX, alpha-thalassemia-mental retardation syndrome; PTHS, Pitt-Hopkins syndrome (b) Schematic drawing of the study outline and the work flow. (c) Overview of results from genetic interaction studies in Drosophila. Black lines indicate gene-gene connections investigated in this study by one (thin line), two (middle line) or four (thick line) different combinations. Magenta lines indicate genetic interactions between two genes, respectively. Genetic interaction was defined here as the observation of multiple phenotypic modifications upon pairwise dosage manipulation in several tissues. Solid magenta line indicates strongest evidence for genetic interaction with several consistent phenotypic modifications observed in different tissues. Dashed line in magenta indicates evidence for a possible genetic interaction with some phenotypic modifications observed. Single or inconsistent phenotypic modifications are not indicated.

Figure 2

Genetic interaction of Ube3a and Mef2 or XNP in the Drosophila wing (ms1096-GAL4). (a) Knockdown (KD) of Ube3a does not cause a wing phenotype, KD of Mef2 causes abnormally curled wings in male flies with additional cross vein defects, such as missing anterior cross veins and/or ectopic cross veins (marked with an arrow). Simultaneous KD of Ube3a and Mef2 results in a milder phenotype with significantly more flies with both cross veins present and fewer flies with ectopic cross veins, quantified in (b). (c) Overexpression (OE) of Ube3a causes abnormally curled wings in females, while KD of Mef2 does not cause a phenotype in female flies (male phenotype see above). Simultaneous OE of Ube3a and KD of Mef2 results in male lethality and in a more severe disorganization of wing architecture in about 75% of females, as quantified in (d). (e) OE of XNP causes abnormally curled wings in female flies with additional cross vein defects, such as missing anterior cross veins and/or ectopic cross veins (marked with an arrow). Simultaneous KD of Ube3a (normal) and OE of XNP results in a more severe phenotype with more flies with ectopic cross veins and fewer flies with both cross veins intact, as quantified in (f). Statistical analysis was performed using Fisher’s Exact test, **p ≤ 0.001; ***p ≤ 0.0001). Flies are counted towards the more severe phenotype if at least one wing was affected. These results are from an independent experiment than in Supplementary Table S2, thus numbers are different.

Figure 3

Genetic interaction of Mef2 and Ube3a in the Drosophila eye (GMR-GAL4). (a) Knockdown (KD) of Ube3a does not cause a phenotype, KD of Mef2 causes rough eyes with more severely affected flies also displaying a bubble-like appearance. Simultaneous KD of Ube3a and Mef2 results in a milder phenotype with significantly fewer eyes with bubble-like appearance, quantified in (b) (***: p ≤ 0.001, Fisher’s Exact test). (c) Overexpression (OE) of Ube3a causes rough eyes, OE of Mef2 results in a mildly reduced number of bristles but grossly intact ommatidial structure. Simultaneous OE of both results in a severe phenotype with reduced eye size and dissolved ommatidia structure in all eyes as quantified in (d). (e) Simultaneous OE of Ube3a (rough eye) and KD of Mef2 (rough eyes and occasionally bubble-like appearance) results in a different and more severe phenotype with disorganized ommatidia structure and progressive necrosis as quantified for male and female flies in (f). Pictures and quantifications are from male flies if not indicated otherwise.

Figure 4

Physical interaction of human MEF2C and UBE3A in cells. (a) Co-localization studies of overexpressed human tagged MEF2C (HA, red) and UBE3A (Myc, green) in HeLa cells show diffuse nuclear localization of both proteins. (b) Co-immunoprecipitation experiments show physical interaction of overexpressed human MEF2C and UBE3A in HEK293 cells. Immunoprecipitation of Myc-tagged UBE3A also precipitates MEF2C-HA and vice versa. Please note, that the figure panels are cropped from two different blots with different exposure times. For full blots see Supplementary Fig. S6.