A cross-species spatiotemporal proteomic analysis identifies UBE3A-dependent signaling pathways and targets.

Angelman syndrome (AS) is a severe neurodevelopmental disorder caused by the loss of neuronal E3 ligase UBE3A. Restoring UBE3A levels is a potential disease-modifying therapy for AS and has recently entered clinical trials. There is paucity of data regarding the molecular changes downstream of UBE3A hampering elucidation of disease therapeutics and biomarkers. Notably, UBE3A plays an important role in the nucleus but its targets have yet to be elucidated. Using proteomics, we assessed changes during postnatal cortical development in an AS mouse model. Pathway analysis revealed dysregulation of proteasomal and tRNA synthetase pathways at all postnatal brain developmental stages, while synaptic proteins were altered in adults. We confirmed pathway alterations in an adult AS rat model across multiple brain regions and highlighted region-specific differences. UBE3A reinstatement in AS model mice resulted in near complete and partial rescue of the proteome alterations in adolescence and adults, respectively, supporting the notion that restoration of UBE3A expression provides a promising therapeutic option. We show that the nuclear enriched transketolase (TKT), one of the most abundantly altered proteins, is a novel direct UBE3A substrate and is elevated in the neuronal nucleus of rat brains and human iPSC-derived neurons. Taken together, our study provides a comprehensive map of UBE3A-driven proteome remodeling in AS across development and species, and corroborates an early UBE3A reinstatement as a viable therapeutic option. To support future disease and biomarker research, we present an accessible large-scale multi-species proteomic resource for the AS community ( https://www.angelman-proteome-project.org/ ).

Elevated Fmr1 mRNA levels and reduced protein expression in a mouse model with an unmethylated Fragile X full mutation.

The human FMR1 gene contains a CGG repeat in its 5' untranslated region. The repeat length in the normal population is polymorphic (5-55 CGG repeats). Lengths beyond 200 CGGs (full mutation) result in the absence of the FMR1 gene product, FMRP, through abnormal methylation and gene silencing. This causes Fragile X syndrome, the most common inherited form of mental retardation. Elderly carriers of the premutation, defined as a repeat length between 55 and 200 CGGs, can develop a progressive neurodegenerative syndrome: Fragile X-associated tremor/ataxia syndrome (FXTAS). In FXTAS, FMR1 mRNA levels are elevated and it has been hypothesised that FXTAS is caused by a pathogenic RNA gain-of-function mechanism. We have developed a knock in mouse model carrying an expanded CGG repeat (98 repeats), which shows repeat instability and displays biochemical, phenotypic and neuropathological characteristics of FXTAS. Here, we report further repeat instability, up to 230 CGGs. An expansion bias was observed, with the largest expansion being 43 CGG units and the largest contraction 80 CGG repeats. In humans, this length would be considered a full mutation and would be expected to result in gene silencing. Mice carrying long repeats ( approximately 230 CGGs) display elevated mRNA levels and decreased FMRP levels, but absence of abnormal methylation, suggesting that modelling the Fragile X full mutation in mice requires additional repeats or other genetic manipulation.

The generation of a conditional Fmr1 knock out mouse model to study Fmrp function in vivo.

The FMR1 gene, mutated in Fragile X syndrome patients, has been modeled in mice with a neomycin cassette inserted in exon 5 of the mouse Fmr1 gene creating an Fmr1 knockout (Fmr1 KO) allele. This results in animals lacking Fmr1 protein (Fmrp) expression in all tissues. We have created a new, more versatile Fmr1 in vivo KO model (Fmr1 KO2) and generated conditional Fmr1 KO (CKO) mice by flanking the promoter and first exon of Fmr1 with lox P sites. This enables us to create a null allele in specific cell types and at specific time points by crossing Fmr1 CKO mice with tissue specific or inducible cre-recombinase expressing mice. The new Fmr1 KO2 line does not express any Fmrp and also lacks detectable Fmr1 transcripts. Crossing the Fmr1 CKO line with a Purkinje cell-specific cre-recombinase expresser produces mice that are null for Fmr1 in Purkinje neurons but wild type in all other cell types.

DNA adducts, mutant frequencies, and mutation spectra in various organs of lambda lacZ mice exposed to ethylating agents.

To investigate tissue-specific relations between DNA adducts and mutagenesis in vivo, lambda lacZ transgenic mice were treated i.p. with N-ethyl-N-nitrosourea (ENU), diethylnitrosamine (DEN), and ethyl methanesulphonate (EMS). In liver, bone marrow, and brain DNA from mice sacrificed at several time points after treatment O6-ethylguanine (O6-EtG) and N7-ethylguanine (N7-EtG) levels were determined as well as the mutant frequency (MF) in lacZ. In liver DNA of ENU- and DEN-treated mice, the bulk of O6-EtG was removed at 3 days after treatment, while the MF continued to increase thereafter. This suggests that O6-EtG is not the major premutagenic lesion in the liver. Indeed, sequence analysis of mutants showed only 24% GC-->AT transitions, consistent with the O6-EtG lesion, and 28% TA-->AT transversions, expected from O2-ethylthymine. In bone marrow after ENU treatment, a maximum mutation induction occurred at 3 days post-treatment, of which 43% were GC-->AT mutations and 22% were TA-->AT mutations. This suggests that in bone marrow O6-EtG may be a major premutagenic lesion at the 3-day time point. In liver and bone marrow, EMS treatment gave rise to a high level of N7-EtG and a low level of O6-EtG but no increase in MF. No adducts or mutation induction were observed in bone marrow of DEN-treated mice. No MF increase was observed in the brain of either ENU- or EMS-treated mice, although O6- and N7-adducts were present.

UVB-induced mutagenesis in hairless lambda lacZ-transgenic mice.

UVB-induced mutagenesis was studied in hairless 40.6 transgenic mice (MutaMouse), which contain the lambda gt10lacZ shuttle vector as a target for mutagenesis. Mice were exposed at the dorsal side to either single doses of 200, 500, 800, or 1000 J/m2 UVB or to two successive irradiations of either 200 and 800 J/m2 UVB, with intervals of 1, 3, or 5 days, or to 800 and 200 J/m2 UVB with a 5-day interval. At 23 days after the last exposure, lacZ mutant frequencies (MF) were determined in the epidermis. The lacZ MF increased linearly with increasing dose of UVB. The mutagenic effect of two successive irradiations appeared to be additive. The UV-induced mutation spectrum was dominated by G:C --> A:T transitions at dipyrimidine sites. DNA-sequence analysis of spontaneously mutated phages showed a diverse spectrum consisting of insertions, deletions and G:C --> A:T transitions at CpG sites. The results indicate that the hairless lambda lacZ-transgenic mouse is a suitable in vivo model for studying UVB-induced mutations.

Formation and persistence of O6-ethylguanine in genomic and transgene DNA in liver and brain of lambda(lacZ) transgenic mice treated with N-ethyl-N-nitrosourea.

LacZ transgenic mice are suitable for short-term mutagenicity studies in vivo. Mutagenicity in these mice is determined in the lacZ transgene. Since the lacZ gene is of bacterial origin the question has been raised whether DNA-adduct formation and repair in the transgene are comparable to those in total genomic DNA. Mice were treated with N-ethyl-N-nitrosourea (ENU) and killed at several time points following treatment. Some mice were pretreated with O6-benzylguanine to inactivate the repair protein O6-alkylguanine-DNA alkyltransferase (AGT). O6-ethylguanine (O6-EtG) was determined in lacZ in liver and brain by means of a monoclonal antibody-based immunoaffinity assay. In addition, O6-EtG and N7-ethylguanine (N7-EtG) were assayed in total genomic DNA of liver and brain with an immunoslotblot procedure. In liver, the initial O6-EtG level in total genomic DNA was 1.6 times that in lacZ. The extent of repair of O6-EtG during the first 1.5 h after treatment was 2.1 times that in lacZ. At later time points, O6-EtG repair was the same. N7-EtG repair in genomic DNA was evident. In contrast to the liver, little repair of O6-EtG in total genomic and lacZ DNA occurred in the brain while N7-EtG was repaired. No initial difference in O6-EtG levels were found in lacZ and genomic brain DNA. These findings indicate that in the liver, total genomic DNA is more accessible than lacZ to ENU and/or the AGT protein, during the first 1.5 h following treatment. Because the difference in O6-EtG levels in the transgene and genomic DNA in the liver is restricted to the first 1.5 h after treatment, while the fixation of mutations occurs at later time points, O6-EtG-induced mutagenesis most likely is also very similar in both types of DNA.

Comparison of the X-gal- and P-gal-based systems for screening of mutant lambda lacZ phages originating from the transgenic mouse strain 40.6.

The recent introduction of the phenyl-beta-D-galactopyranoside (P-gal)-based positive-selection system for screening of lambda lacZ phages originating from the lambda lacZ transgenic mouse (Muta Mouse) has made the determination of mutant frequencies (MF) a much simpler task. Previously, MF data from these mice have been collected by means of the 5-bromo-4-chloro-3-indolyl-beta-D-galactopyranoside (X-gal) colour-screening procedure. To determine whether data obtained with the two systems are comparable, the MF in lambda phages recovered from liver and brain of transgenic mice treated with N-ethyl-N-nitrosourea (ENU) and liver of benzo(a)pyrene (B(a)P)-treated mice was determined with both procedures. For the livers of mice treated with ENU, both methods yielded approximately the same MF values. No induction of mutants, relative to the control animals, was seen after 1.5 h, but a clear 4-fold increase was measured with both assays at the 14-day time point. No induction of mutants was found in the brain with either method. In the B(a)P-treated mice, both methods showed a substantial induction in MF after 21, 28 and 35 days. The values generated by the X-gal and P-gal methods were not significantly different, with the exception of the 35-day post-treatment point that appeared higher in the X-gal assay. When the mutants isolated by use of the X-gal method were tested in the P-gal assay, a number of these did not turn up as mutants, and the significance disappeared. In conclusion, the data obtained with the two screening procedures agree to such an extent as to permit a direct comparison between the earlier results generated with X-gal and P-gal values generated with the new positive-selection method. This is likely to apply also to other organs and mutagens than those studied here.

Inhibition of bacteriophage Mu transposition by Mu repressor and Fis.

In this paper we show that the Escherichia coli protein Fis has a regulatory function in Mu transposition in the presence of Mu repressor. Fis can lower the transposition frequency of a mini-Mu 3-80-fold, but only if the Mu repressor is expressed simultaneously. In this novel type of regulation of transposition by the concerted action of Fis and repressor, the IAS, the internal activating sequence, is also involved as deletion of this site lead to the loss of the Fis effect. As the IAS contains strong repressor binding sites these are probably the target for the repressor in the observed negative regulation by Fis and repressor. However, the role of Fis and repressor is not only to inactivate the IAS, since a 4 bp insertion in the IAS, which changes the spacing of the repressor-binding site, abolishes the enhancing function of the IAS but leaves the repressor-Fis effect intact. A likely target for Fis in this regulation is a strong Fis-binding site, which is located adjacent to the L2 transposase-binding site. However, when this Fis-binding sequence was substituted by a random sequence and Fis no longer showed specific binding to this site, the Fis effect was still observed. Although it is still possible that Fis can function by binding to this non-specific site in a particular complex, it seems more likely that Fis is directly or indirectly involved in determining the level of the repressor.