Abstract
Histone H4 is a highly conserved eukaryotic protein and an essential structural component of the nucleosome, around which DNA is wrapped to form chromatin. Recently, de novo missense variants in H4 were identified in a large cohort of patients with a novel neurodevelopmental disorder. Canonical histone H4 is encoded by 14 paralogous genes which are expressed at different levels during neurodevelopment; eight genes are affected across the patient cohort. Rare missense H4 variants have also been recorded in neurodevelopmentally healthy individuals. A number of these encode identical missense changes to patient variants, however they occur in H4 genes not associated with disease. This thesis hypothesised a dosage-dependent model of variant H4 pathogenesis. Highly expressed variant H4 is likely to have pathogenic effects during neurodevelopment, whereas a lower expressed variant may be tolerated. In this model, there must be a threshold of variant H4 protein required for pathogenesis in neurodevelopment.
To interrogate the model, this thesis investigated expression levels of canonical histone H4 genes and the influence of variant H4 dosage on normal cell function. In HEK293FT cells, knockdown of the H4 transcription factor HINFP resulted in a 55% decrease in H4C5 mRNA and a 40% decrease in total H4 protein but had minimal effect on the 13 additional H4 paralogues. Both in vitro and in silico analyses of histone gene expression revealed high expression of H4C5, the most frequently affected gene in H4 patients. The 14 canonical H4 genes were found to be expressed at different levels in the brain across neurodevelopmental stages and discrepancies between datasets suggested that some paralogues had tissue-specific expression. Whether an H4 variant is pathogenic or can be tolerated in the developing brain may be partly attributed to this complex regulation of the H4 gene family.
A library of plasmids was generated which each encoded four copies of variant and/or wild type H4 at varying dosages. These were expressed in H4 knockdown cells to recapitulate the effects of different H4 levels and effects on the cell cycle were investigated using an EdU pulse-chase followed by flow cytometry. Pro32Leu and Arg40Cys H4 caused a decrease in the rate of progression through S phase which relied on dosage: at minimum three copies of Pro32Leu or two copies of Arg40Cys were required to impede cell cycle progression. A third H4 variant, Lys91Gln, caused an identical decrease in S phase progression irrespective of dosage. Additionally, Lys91Gln H4, which is linked to severe neurodevelopmental defects in patients, caused the formation of a distinctive population of cells with irregular cell division. This novel link to cell cycle progression in H4 pathogenesis provides insight into the molecular mechanisms underpinning the neurodevelopmental impairments that affect patients. Further investigation into the threshold between tolerated and pathogenic H4 variants in brain development is required, but this preliminary evidence implicates dosage as a key contributor.