A de novo missense variant in EZH1 associated with developmental delay exhibits functional deficits in Drosophila melanogaster

Abstract

EZH1 ( Enhancer of Zeste, homolog 1) , a Polycomb Repressive Complex-2 (PRC2) component, is involved in a myriad of cellular processes through modifying histone 3 lysine27 (H3K27) residues. EZH1 represses transcription of downstream target genes through H3K27 trimethylation (H3K27me3). Genetic mutations in histone modifiers have been associated with developmental disorders, while EZH1 has not yet been linked to any human disease. However, the paralog EZH2 is associated with Weaver syndrome. Here we report a previously undiagnosed individual with a novel neurodevelopmental phenotype identified to have a de novo variant in EZH1 , p.Ala678Gly, through exome sequencing. The individual presented in infancy with neurodevelopmental delay and hypotonia and was later noted to have proximal muscle weakness. The variant, p.A678G, is in the SET domain, known for its methyltransferase activity, and was the best candidate variant found in the exome. Human EZH1 / 2 are homologous to fly Enhancer of zeste E(z) , an essential gene in flies, and the residue (A678 in humans, A691 in Drosophila ) is conserved. To further study this variant, we obtained Drosophila null alleles and generated transgenic flies expressing wild-type (E(z) ^WT ) and the variant (E(z) ^A691G ) . The E(z) ^A691G variant led to hyper H3K27me3 while the E(z) ^WT did not, suggesting this is as a gain-of-function allele. When expressed under the tubulin promotor in vivo the variant rescued null-lethality similar to wild-type but the E(z) ^A691G flies exhibit bang sensitivity and shortened lifespan. In conclusion, here we present a novel EZH1 de novo variant associated with a neurodevelopmental disorder. Furthermore, we found that this variant has a functional impact in Drosophila . Biochemically this allele leads to increased H3K27me3 suggesting gain-of-function, but when expressed in adult flies the E(z) ^A691G has some characteristics of partial loss-of-function which may suggest it is a more complex allele in vivo .

Figure 1:. Patient information:

Photos and clinical features: Patient information is covered as per medRxiv rules please contact authors for the further information.

Figure 2:

A: Domain structure schematic of human EZH1(Hsa_EZH1) and Drosophila melanogaster E(z)(Dmel_E(z)) B-B’: The conservation of EZH1-pA678 throughout the species and also in both human paralogs EZH1 and EZH2 (The structures are derived from UniProt.) [Hsa_EZH1: Homo Sapiens EZH1; Dis R: Disordered Region; SANT: SANT-Myb- Protein binding Domain, CXC: cysteine rich DNA binding Domain; Dme_E(z) : Drosophila melanogaster Enhancer of Zaste (E(z)] C-C’: Western blot assay for H3K27me3 of UAS-E(z)^wt and UAS-E(z)^A691G overexpression. Tri-methylation of H3K27 was evaluated in the GAL4 driven E(z)^WT and E(z)^A691G progenies. C’ is the graphical representation of the normalised H3K27 values to H3 values. UAS-LacZ was used as a positive control. p-values were determined by unpaired t-tests.

FIGURE 3:. Functional assessment of the transgenic flies:

A-A’. Genetic crosses for the rescue of heteroallelic lethality: Under constitutively active tubulin promotor (ptub), the WT and the Variant constructs$\underset{{}^{{}^{{}^{{}^{{}^{{}^{,}}}}}}}{\text{-}}$ were both rescued the E(z)⁷³¹/E(z)⁶³ heteroallelic lethality (Table A). A’ represents the graphical representation of the observed/expected ratio of the F1 progenies explained in Table A. B: Western blot assay for H3K27-trimethylation: Under constitutively active tubulin promotor, the H3K27me3 levels in the variant were compared to WT (p=0.0002). In the sensitized background, when one copy of E(z) is removed using a null allele (E(z)⁷³¹), the variant continues to show hyper tri-methylation as compared to WT (p<0.0001). Even in the fly null background, when compared with WT, the variant maintains the hyper-trimethylation (p<0.0001). p-value were determined by unpaired t-tests.

Figure 4:. Behavior study with the transgenic flies.

A-A’: Bang sensitivity performed at 5 days after eclosion (DAE): ptub::E(z)^A691G transgenic flies in the fly null background observed to be bang sensitive than ptub::E(z)^WT(FIGURE 4A, p=0.0041). When similar experiment is carried out in the wild-type background ptub::E(z)^A691G do not behave different than the wild-type ptub::E(z)^WT(FIGURE 4A’, p=ns). B-B’: Bang sensitivity performed at 16 DAE: ptub::E(z)^A691G transgenic flies in the fly null background continues to show bang sensitivity at 16 DAE when compared to ptub::E(z)^WT(FIGURE 4B, p=0.0004). In the wild-type background the ptub::E(z)^A691G starts showing the bang sensitivity when compared to ptub::E(z)^WT(FIGURE 4B’, p=0.0002). C-C’: Life span assay : ptub::E(z)^A691G transgenic flies in the fly null background live shorter than the ptub::E(z)^WT(FIGURE 4C, Median Survival: E(z)^WT = 47, E(z)^A691G = 35, p<0.0001). However, in the fly wild-type background, ptub::E(z)^A691G transgenic flies shows a similar trend in life span assay when compared to ptub::E(z)^WT(FIGURE 4C’, Median Survival: E(z)^WT = 58, E(z)^A691G = 51, p<0.0001). p-value were determined by unpaired t-tests.