The N-terminal BRCT domain determines MCPH1 function in brain development and fertility

MCPH1 is a causal gene for the neurodevelopmental disorder, human primary microcephaly (MCPH1, OMIM251200). Most pathogenic mutations are located in the N-terminal region of the gene, which encodes a BRCT domain, suggesting an important function of this domain in brain size determination. To investigate the specific function of the N-terminal BRCT domain in vivo, we generated a mouse model lacking the N’-BRCT domain of MCPH1 (referred as Mcph1-ΔBR1). These mutant mice are viable, but exhibit reduced brain size, with a thinner cortex due to a reduction of neuroprogenitor populations and premature neurogenic differentiation. Mcph1-ΔBR1 mice (both male and female) are infertile; however, almost all female mutants develop ovary tumours. Mcph1-ΔBR1 MEF cells exhibit a defect in DNA damage response and DNA repair, and show the premature chromosome condensation (PCC) phenotype, a hallmark of MCPH1 patient cells and also Mcph1 knockout cells. In comparison with Mcph1 complete knockout mice, Mcph1-ΔBR1 mice faithfully reproduce all phenotypes, indicating an essential role of the N-terminal BRCT domain for the physiological function of MCPH1 in the control of brain size and gonad development as well as in multiple cellular processes.


Introduction
Human primary microcephaly (MCPH, OMIM251200) is a neurodevelopmental disorder characterised by a marked reduction of brain size, especially reduced size of the cerebral cortex, yet with normal brain architecture and nonprogressive intellectual disability. MCPH has a 1:30,000 to 1:250,000 incidence per live-birth, depending on the population [1][2][3] . So far, 25 gene loci have been identified to be potentially involved in the occurrence of MCPH [4][5][6][7] .
The MCPH1/BRIT1 gene is the first gene reported to be responsible for MCPH type 1 (MCPH1, OMIM251200) and also for another syndrome called premature chromosome condensation syndrome (PCC, OMIM 606858), because MCPH1 patient cells exhibit prophaselike and premature chromosome condensation 1,[8][9][10][11] . The human MCPH1 gene, located in human chromosome 8p23.1, contains 14 exons and encodes 835aa (amino acid) of the full length MCPH1 protein. MCPH1 contains several functional domains; the mid region binds to condensin II, TopBP1, Chk1 and βTrCP2 12,13 and three BRCA1 C-terminus (BRCT) domains-an evolutionarily conserved phospho-peptide interacting domain-which are frequently found in proteins involved in DNA damage response (DDR) and cell cycle control 1,14,15 . The Nterminal BRCT domain (13aa to 89aa) is implicated in DNA repair via interaction with chromatin remodelling complex SWI/SNF 16 and allows MCPH1 to localise to centrosomes in response to DNA damage 17 . The two Cterminal tandem BRCT domains encoded by exons12-14, ranging from 672aa to 730aa and from 751aa to 833aa respectively 1,15,18,19 , are necessary for oligomer formation and ionising radiation (IR)-induced foci (IRIF) by interacting with γH2AX during the DDR and also in E2F1mediated transcriptional regulation of Chk1 and BRCA1 in G2/M checkpoint 14,18,20,21 .
MCPH1 has been shown to play an important role in brain development, gonad formation and tumorigenesis in transgenic mouse models. Mice carrying a deletion of exon 2 (designated hereafter as Mcph1-ko) 22 showed no obvious microcephaly phenotype, but exhibited growth retardation and infertility. Mutant male mice had atrophic testes, characterised by thinner testicular tubules and absence of spermatocytes, due to impaired DNA repair that leads to failure of chromosomal synapsis, meiosis arrest and apoptosis. Moreover, Mcph1-ko female ovaries were small and often lacked ovarian follicles. About 17% of Mcph1-ko mice developed malignant tumours, originated from lymphomas and ovary tumours over a period of 2.5 years 22 . We generated Mcph1 null mice (designated hereafter Mcph1-Δ) 23 by deleting exons 4 and 5. These mice were viable and showed a microcephaly phenotype associated with a thinner neocortex and premature differentiation of neuroprogenitors, due to MCPH1-TrCP2-Cdc25 mediated premature mitosis entry and the switch of division mode from symmetric to asymmetric 12,23 . Like human MCPH1 cells, which are defective in the DNA damage response 20,24,25 , Mcph1-Δ neuroprogenitors are sensitive to irradiation-induced DNA damage, which enhanced apoptosis in Mcph1-Δ cortex 26 ; Mcph1-Δ cells showed a classic PCC phenotype. Moreover, Mcph1-Δ mice were infertile and strikingly exhibited a very high incidence (94.4%) of ovary tumours (granulosa cell tumours and Sertoli-Leydig cell tumour) in female mice over a period of 18 months 27 . Another mouse model called Mcph1 tm1a , generated by deleting exon 4 of the Mcph1 gene, exhibited less brain weight and a moderate defect of hearing loss 28 . It is worth mentioning that a mouse model lacking the C-terminal BRCT domain (Mcph1 gt/gt mice) showed a reduced survival rate, without affecting the brain size 29 , suggesting that the C-terminal BRCT domain is not important for brain development. Nevertheless, Mcph1 gt/gt cells showed a PCC phenotype-a characteristic of MCPH1 29 -indicating that this C'-BRCT is responsible for PCC, which can separate MCPH1 from its function in brain development. Combined with findings in human MCPH1 cells, these mouse model studies indicate that MCPH1 is a multi-faceted protein and different regions may control the specific function of MCPH1, ranging from brain development, cell cycle progression, fertility and tumorigenesis.
A total of 12 mutations were identified in MCPH1 patients, with all located in exons 1 to 6. Seven of these mutations caused code frameshift or loss of start transcription site leading to loss of MCPH1 proteins; whereas 5 missense mutations leading to a single amino acid substitution at the N-terminal BRCT domain of MCPH1, were found in human MCPH1 patients 1,10,13,27,[30][31][32][33] . The mutation spectrum in the N-terminal domain of the MCPH1 protein, strongly suggests an important function of N-terminal BRCT domain in preventing the microcephaly phenotype.
This study investigated the function of the N-terminal BRCT domain by specifically deleting it in the mouse genome (Mcph1-ΔBR1 mice). These mutant mice developed primary microcephaly, accompanied with a thinner neocortex, a reduction of the neuroprogenitor pool and premature neuronal differentiation during brain development. Mcph1-ΔBR1 mice showed a block in reproductive organ development, characterised by testis atrophy and lacking ovaries. Intriguingly, almost all mutant females developed ovary tumours. At the cellular level, Mcph1-ΔBR1 MEF (mouse embryonic fibroblast) cells exhibited the PCC phenotype and defects in cell proliferation and the DDR. All these phenotypes identify the N-terminal BRCT domain as a major regulator of brain size, gonad development and tumour repression.

Mice
Mice were maintained in the mouse facility of Fritz Lipmann-Institute (FLI, Jena, Germany) and fed ad libitum with standard laboratory chow and water in ventilated cages under a 12 hr light/dark cycle. All animal experiments and breeding were conducted according to the German animal welfare legislation and approved by the Thüringer Landesverwaltungsamt. Genotype analysis for Mcph1-Δ mice were determined by PCR on DNA extracted from tail tissue, as previously described 23 .

Gene targeting in ES cells and generation of Mcph1-ΔBR1 mice
For construction of a mouse model carrying a deletion of the N-terminal BRCT domain of MCPH1, the targeted vector was constructed with 5′ homologous arm (5HR) containing exon1 and 3HR containing the entire sequence from exon 4 to exon 6. The targeted mice were obtained through electroporation of this vector into embryonic stem cells, screening for corrected targeted ES clones and injection of targeted ES clones into blastocysts as previously described 23 . The knocked-in allele was obtained by crossing with Stra8-Cre mice to delete the Neo cassette. Homozygous Ki/Ki mice were generated by breeding between + /Ki mice. All animal experiments were conducted according to the German animal welfare legislation.

Southern blotting
Southern blot was performed as previously described 23 . Briefly, 10 μg of genome DNA was digested with 30 units of the corresponding enzymes at 37°C overnight. The samples were electrophoresed in 0.8% agarose gel. After denaturation and depurination, the gel was transferred to Amersham Hybond membrane XL (#RPN2020S, GE healthcare life science), which then was hybridised with indicated probes after labelling with α-32 P dCTP (Hartmann Analytic) using Prime-a-Gene® Labelling System (#U1100, Promega), as per manufacturer's protocol and the Amersham NICK Columns (108-0855-02, GE healthcare life science). The blot was visualised by exposure to FUJIFLIM or by Phospho-imaging machine (FUJI).

Probes for Southern blotting
Probes were designed by using software and NCBI Blast online, amplified by PCR using both forward and reverse primers, cloned to pGEM®-T Easy Vector (Promega) and sequenced. The probe vectors were digested by EcoRI and purified for Southern blotting.
5P4 was an external probe which is located at 5′-prime upstream of the targeted region (intron 1). The sequence of 5P4 was as follows: TTGGTGCTGGCCTCTAAATCTGACGTTCAGGAG  CTTCTGGAAACCCAGGTCTAGGCCCCGCTGTCTT  CAGTGTCTGGGCCACACTGTTGAGAGTGCAGGTT  TCTGGGGGAACTTCAGACTGTGGGGAGCCAGTG  TTCCCCCTCTGATGCAGCGGACCCTTTAGACCCT  TTTTATCTGACTTGTTTGCCAGGACTTCTTGGAG  ACTCGACTGCTGCTGATTCCTGCCTGCTGTCCAA  ATGCTGTCAGGGTGTTAAGAAGAAATCTCTGAAA  GCTAACCTCCCGCTGAGCCCACGTGTGCGCCTCA  CAAAAGTCAATATTTTCAGCTTTCCCTCGGCCCC  AGTGTGTTTAAAGGGCTTTCTAGTGTACATCCTT  TGATTGTGGGTGAGTGTTCATAAGGAAAATGGC  TTCCATCTTTTGTCTGCAGAGAGTGAAAAAGAAG  CAGAGTCGCTCTAGTTTCGTCTTCACCGGATTCT  TTCTTTAGATGAAGTTA 3P4 was an external probe which is located at 3′-prime downstream of the targeted region (intron 7). The sequence of 3P4 was as follows: ACACCCAGGCTTCTGACATAAGTAGTCTAAGAA TTTATGCTCTTAGTGATGTGCCACACTGAATCGT  GGCTTATTGAAGAGTGATTAATGCATTCAGTTCC  TAAGAGTAAATCTCAGTATCACTCCAACTGCTGT  CTTGGAAGCACTTCAGCAGAGGCCCAGTATCAGC  TGACCTTTGCTTTTGCCCCTCCCTTCCCAGAGCA  CAGCTTCAAGGACAGGATATCTCTGTTGCCACCA  CTCTGCTCTGAAAGCCGAGCTCGAGTTGTTTAAG  TTAGAATCCCAGCACTGCCTCCAGATGGTGTCTT  TTCTCTAAAAAAGGAAGCCTTCAGCTGGGCAGCA  CTGTACCAGTAGAACTGGGTCTGAGTAATGACCA  CAGGTTCAGGAAAGTGAGCAGAAAGTGCTGACT  GCCCTCTAGTACTTCAGTATAATGAGAGGACAGT  AGCATGCTGAGTCCCAGGAGGGGCTAGCACTGT  GACCTGACAGGATCAGGAGAGTCACAGGTGACA  CTGACACTGAGCTGGTTTTGAAGGATGAGTTGTG  TTTACCCACAGCAAAATGTGTCAGAAAGCTGGGG  TCAGAGTGGGGAGCCTGCAGAGGTGATGGACAG  CGTATCATTACTTCTGAGAGAAGTGCAGACTTCA  TGAGAGATCGGAATATTGAGAGCCCCACATGCTT  CATTTGAAATATAAATAAAACATGAGGACTTGAT GGAAGAACAGGGGACAGATATGAGGAGCCGTGT TCTGTTT Preparation of peptides from E13.5 mouse cortex for mass spectrometry analysis E13.5 cortex isolation was performed in PBS and the tissue lysed in lysis buffer (4% SDS (w/v), 100 mM HEPES, 50 mM DTT in milliQ water), then sonicated in a Bioruptor Plus (Diagenode) for 5 cycles of 60 sec ON and 30 sec OFF, followed by heating at 95°C for 10 min. For reduction and alkylation of cysteines, lysates were incubated with 10 mM DTT at 45°C for 15 min and incubated with 15 mM iodacetamide at room temperature for 30 min. Each sample was precipitated with ice-cold acetone (Biosolve #010306) overnight at −20°C. Following centrifugation at 20,800 × g for 30 min at 4°C, the pellets were washed twice with ice-cold 80% acetone and allowed to air-dry, before being dissolved in digestion buffer at 1 mg protein/ml using 1 M GuHCl in 100 mM HEPES, pH8. The suspension of the protein was digested for 4 hr at 37°C using 1:100 (w/w) LysC (Wako Chemicals GmbH #125-05061). Samples were then diluted to 0.5 M GuHCl by milliQ water and digested with 1:100 (w/w) trypsin (Promega #V5111) for 16 hr at 37°C. Digested peptide solutions were acidified with 10% TFA, then desalted in Waters Oasis® HLB μElution Plate (30 μm, Waters 186001828BA) under a slow vacuum. Before this process, the plate was pre-conditioned with 3 × 100 μl solvent B (80% acetonitrile from Biosolve #06914143) and equilibrated with 3 × 100 μl solvent A (0.05% formic acid in milliQ water). Samples were loaded, washed 3 times with 100 μl solvent A, and eluted into PCR tubes with 50 μl solvent B. Eluates were dried with a speed vacuum centrifuge and dissolved in 5% acetonitrile, 0.1% formic acid to a peptide concentration of 2 mg/ml.

Parallel Reaction Monitoring (PRM) for MCPH1 peptides
To prepare standard peptides for PRM, proteotypic peptides for mouse MCPH1 were synthesised and isotopically labelled by incorporation of heavy Arginine (U-13C6;U-15N4) or Lysine (U-13C6; U-15N2) at the Cterminus (JPT Peptide Technologies GmbH, Berlin, Germany). Lyophilised peptides were reconstituted in 20% (v/ v) acetonitrile, 0.1% (v/v) formic acid before being pooled together in equal amounts. An aliquot of the pooled peptides (~160 fmol per peptide on column) was analysed by shotgun liquid chromatography tandem mass spectrometry using both data-dependent and dataindependent acquisition. The data were used for PRM assay development using SpectroDive v.9 (Biognosys AG) at default settings. PRM assays were successfully developed for 14 peptides derived from MCPH1 (see below).

Protein Access
GeneID Peptide sequence Charge state After assay development, the pool of heavy synthetic peptides was spiked into cortex derived peptides. iRT peptides (Biognosys AG) were also spiked into each sample prior to analysis by PRM-MS for retention time calibration. Peptides (1 μg) were separated using a nanoAcquity UPLC M-Class system (Waters) with a trapping (nanoAcquity Symmetry C18, 5 µm, 180 µm × 20 mm) and analytical column (nanoAcquity BEH C18, 1.7 µm, 75 µm × 250 mm). The outlet of the analytical column was coupled directly to a Q-exactive HF-X (Thermo Fisher) using the Proxeon nanospray source.
Peptides were eluted via the analytical column with a constant flow of 0.3 μl/min. During this step, the percentage of acetonitrile (in 0.1% in formic acid) was increased in a nonlinear fashion from 0% to 40% over 40 min. PRM acquisition was scheduled for the duration of the entire gradient using the "DIA" mode with the following settings: resolution 120,000 FWHM, AGC target 3 × 10 6 , maximum injection time (IT) 250 ms, isolation window 0.4 m/z. For each acquisition cycle, a "full MS" scan was acquired with the following settings: resolution 120,000 FWHM, AGC target 3 × 10 6 , maximum IT 10 ms, scan range 350 to 1650 m/z. Peak group identification and quantification were performed using SpectroDive v9. The summed height of all identified transitions was used to estimate the quantity of each peptide. Peptide quantities were normalised by dividing their integrated intensity with summed intensity of all samples of the Base Peak Chromatogram extracted for each sample from "full MS" scans, using Xcalibur v4.1. Only peptides reliably quantified across all replicates (identification q < 0.05 and at least 5 transitions identified) were retained for genotype comparison and protein quantification.

Organ isolation and preparation for paraffin sections and cryosection
Adult mice were sacrificed by CO 2 asphyxiation. Testes, ovaries and brains were isolated and fixed with 4% PFA at 4°C overnight. The newborn brains (P1) were isolated by decapitation and the pregnant mice at E17.5 were sacrificed by cervical dislocation and embryonic brains isolated and fixed overnight with 4% PFA at 4°C. The tissues were embedded in paraffin and sectioned. For preparation of cryosections, the fixed brains were transferred to a solution of 30% sucrose, embedded in a plastic mould using a cryomatrix (Neg −50, Richard-Allan Scientific) and sectioned using cryostate (Leica, Wetzlar, Germany). The cryosections were stored at −80°C and processed for immunostaining.

Histological and immunochemical analyses of gonads
Testes and ovaries from control and mutant mice were harvested, fixed in Bouin's solution, processed and paraffin embedded. Specimens were sectioned at 5 µm and H&E staining completed by the histology core facility of the FLI. Morphology and quantitative histological analyses were performed as described 34 . Briefly, only circular seminiferous tubules were included in the analysis and diameter of control and mutant tubular cross-sections measured using Nikon's NIS-Elements imaging software. At least 30 tubular cross-sections per animal were analysed. To determine the number of damaged seminiferous tubules, vacuolised tubular cross-sections were counted and calculated as number per total seminiferous tubular cross-sections of a testis.

Primary MEFs isolation and culture
Pregnant mice at E13.5 were sacrificed by cervical dislocation and the embryos removed and collected in a 6well dish containing ice cold PBS. The foetal liver was collected for DNA isolation and PCR-genotyping to determine the genotype of the embryos. Embryos were transferred into a cryovial containing ice-cold PBS and chopped into small pieces. The chopped tissue was digested with trypsin-EDTA-solution (#25300, life technologies) for 4 min and the supernatant transferred to tubes containing MEF medium. This trypsinization procedure was repeated three more times and the cell supernatant collected in the same tube. Said cell supernatant was spun down and the cell pellets cultured in MEF medium (Dulbecco's modified Eagle's medium supplemented with 10% foetal bovine serum, 2 mM glutamine, 1 mM sodium pyruvate, β-mercaptoethanol, 50 units/ml penicillin, and 50 mg/ml streptomycin) at 37°C in a 5% CO 2 incubator.

Proliferation assay of primary MEF cells
Primary MEF cells of indicated genotypes were seeded with the same number (1.5 × 10 5 ) in a 6-well dish. Cells were passaged every three days and cell numbers were counted. The same number of cells (1.5 × 10 5 ) were seeded at each passage.

Analysis of γH2AX by immunofluorescence
MEFs were seeded on gelatin-coated coverslips and γirradiated with 2 Gy in Gammacell GC40 24 h after plating. At the indicated timepoints, cells were washed with PBS and fixed in PBS containing 4% PFA (Sigma-Aldrich). The blocking was performed using PBS containing 1% BSA (Sigma Aldrich), 0.4% Triton X-100 and 5% normal donkey serum for 2 hr at room temperature. The primary antibody anti-γH2AX (1:200, #05-636, Millipore) was applied overnight at 4°C. Secondary antibody containing 0.2 µg/ml DAPI (Sigma Aldrich) incubated at room temperature for 1 hr followed by three times of washing before mounting with ProLong Gold Antifade Mountant (P10144, Thermo Fisher). Images of γH2AX foci were examined using Apotome microscope (Carl Zeiss). Foci were counted inside DAPI mask. 300 cells of each samples were randomly chosen for analysis.

Western blotting
Immunoblotting was carried out as previously described 12

Statistical analysis
The sampling distribution was tested using chi-squared test. Depending on the normality and variance equality, two-tailed Student's t-test was used. The specific statistical analysis is described in the respective figure legend.

Generation of Mcph1-ΔBR1 mice
The N-terminal BRCT domain of MCPH1 (designated as BR1) is mainly encoded by exon 2 and exon 3 ( Figure S1A). To delete this domain in a mouse model, the targeting vector was designed with 5′ homologous arm (5HA) containing exon 1 and 3HA containing the entire sequence from exon 4 to exon 6. A neomycin cassette was flanked by two loxP sites, which could be deleted after Cre-mediated recombination (Fig. 1A). In order to avoid the frameshift from exon 1 to exon 4 after deletion of exons 2-3, one nucleobase adenine (A, insA) was inserted at the end of exon 1, prompting an exchange of the sequence GAA (Aspartic acid) with GAT (Glutamic acid) (Fig. 1A). The targeting vector was transfected into embryonic stem cells (ES, E14.1) by electroporation and the targeted events confirmed by Southern blotting. The 5′ external probe 5P4 detected a fragment of 7.1Kb in wildtype (WT) allele, 9.3Kb of targeted (Tg) allele and 8.1Kb in case of knock-in (Ki) allele after SacI digestion of genomic DNA (Fig. 1b, S1B). The 3′ external probe 3P4 detected a fragment of 7.1Kb of WT allele, 9.3Kb of Tg allele and 8.1Kb of Ki allele after EcoNI/AhdI double digestion (Fig. 1b, S1B). Internal probes were used to control random integration of the targeting vectors ( Figure S1B). Upon blastocyst injection of the targeted ES clone 2-8E, the Mcph1 Tg allele successfully went through the mouse germline, which was confirmed by Southern blotting (Fig.  1B) and by PCR genotyping ( Figure S1C) To verify whether the mutation is correctly introduced into the germline, we performed cDNA sequencing using primer pair Ex1-Fwd located in exon 1 and Ex8-R located in exon 8 and detected designated insA at the end of exon 1 (Fig. 1C). Western blot analysis revealed a truncated MCPH1 protein from Mcph1 Ki/Ki brain extracts, due to a loss of 70 aa encoded by exons 2 and 3 (Fig. 1D). To further determine the nature of the truncated MCPH1 protein, we employed targeted mass spectrometry based on parallel reaction monitoring (PRM-MS) using synthetic, isotopically labelled standard peptides covering different regions of MCPH1. PRM-MS was performed on E13.5 (embryonic day 13.5) neocortex of Mcph1 Ki/Ki mutant mice and confirmed the deletion of the N'-BRCT domain (lacking Pep1) and the presence of the rest of the MCPH1 protein at least from 149aa peptide (determined by Pep2 and Pep3) (Fig. 1E). Of note, the expression level of Mcph1 Ki/Ki mutant protein is lower, approximately 70% of wildtype (full length) MCPH1, whereas, the Mcph1-Δ cortex contained no MCPH1 protein (Fig. 1E) Figure S1A). Mcph1-ΔBR1 mice had smaller body weight in both female and male compared to wildtype (WT) mice, readily at P5 (postnatal day 5) after birth ( Figure S1B, C), indicating that N'-BRCT deletion results in growth retardation, as in MCPH1 complete knockout mice 22,23 . Given the lack of phenotypes among heterozygous + /Ki or + /Δ and WT mice (data not shown, see below), we used heterozygous as well as WT mice as controls in the following experiments.
Similar to Mcph1-Δ mice 23,26 , Mcph1-ΔBR1 mice displayed smaller brain size and their brain weight was significantly smaller compared to control mice at P1 ( Fig.  2A, B). Hematoxylin and Eosin (H&E) staining revealed another similarity in that Mcph1-ΔBR1 mice had a thinner neocortex compared to controls (Fig. 2C, D). The brain weight of Mcph1-ΔBR1 mice at E17.5 was already smaller (Fig. 3A). These findings demonstrate a primary microcephaly in Mcph1-ΔBR1 mice. We examined neuroprogenitors and newborn neurons at E17.5, to investigate the effect of the deletion of the N-terminal BRCT domain in brain development. By staining the cerebral cortex with Sox2 (a marker for neuroprogenitors in layer I), Tbr2 (a marker for intermediate neuroprogenitors in layer II) and Ctip2 (a marker for neurons in layer VI), we found that the percentage of Sox2+ cells was significantly lower in the Mcph1-ΔBR1 cortex (Fig. 3B); whereas, the percentage of Tbr1+ and Ctip2+ cells was significantly higher in the Mcph1-ΔBR1 neocortex compared to controls (Fig. 3C, D), reminiscent of the Mcph1-Δ brain 23 . Thus, the N'-BRCT is as important as the entire MCPH1 protein in maintaining the neuroprogenitor pool and preventing primary microcephaly.

PCC and DDR defects of Mcph1-ΔBR1 MEF cells
PCC is a prominent cellular characteristic of both human microcephaly patients and the Mcph1-Δ mouse model 9,11,23 . To study the cellular effect of N'-BRCT deletion, we isolated primary MEF cells from E13.5 embryos of WT, Mcph1-ΔBR1 and Mcph1-Δ genotypes. The cells were stained with mitotic marker phosphohistone H3 (pS10-H3) and counterstained with DAPI, which allows visualisation of chromosome condensation. We scored PCC cells that display chromosome condensation by DAPI staining, but negative for pS10-H3 signals. The incidence of PCC cells was significantly higher in Mcph1-ΔBR1 MEFs, similar to Mcph1-Δ, compared to controls (Fig. 4A, B). Next we examined the proliferation capacity of Mcph1-ΔBR1 primary MEF cells by a 3T3 protocol. Initially, 1.5 × 10 5 cells of the desired genotypes were seeded, cells counted every three days and reseeded at the same number. As shown in Fig. 4C, Mcph1-ΔBR1 MEFs proliferated at a similar rate to Mcph1-Δ MEFs, but much slower than control cells, with a loss of proliferation capacity two passages earlier than controls. Thus, MCPH1 mutated cells have impaired proliferation and prematurely enter cellular senescence.
MCPH1 is involved in the DDR 24,25,35 and Mcph1-Δ neuroprogenitors are hypersensitive to ionising radiation (IR) 26 . We next studied the DDR of Mcph1-ΔBR1 by treating MEF cells with 2 Gy IR and stained with antibodies against γH2AX, a DNA damage marker, at 0 hr, 0.5 hr or 3 hr after irradiation. The Mcph1-ΔBR1 cells and Mcph1-Δ cells displayed a significantly lower number of γH2AX foci at 0 hr and 0.5 hr post-IR, but maintained a higher γH2AX level at 3 hr post-IR, when compared to controls (Fig. 4D, E). Western blotting confirmed the kinetics of γH2AX accumulation after IR ( Figure S3A). These findings together indicate an impaired DDR signalling and compromised DNA repair in both Mcph1-ΔBR1 cells and Mcph1-Δ cells.

Mcph1-ΔBR1 mice are infertile because of gonad developmental defects
As with Mcph1-Δ mice 23 , we could not obtain any offspring from intercrosses or backcrosses of Mcph1-ΔBR1 mice, suggesting that the N'-BRCT domain is required for mouse fertility. We then investigated whether testicular germ cells, as well as somatic cells of mutant animals are affected by the loss of the N-terminal BRCT. Histological analysis of 9.5-12 month-old males revealed Mcph1-ΔBR1 mutant male mice to have smaller testes, similar to Mcph1-Δ mice, in stark contrast to control mice (Fig. 5A, B). No mature spermatozoa could be found in the lumen of testicular tubules of Mcph1-Δ and Mcph1-ΔBR1 mice (Fig. 5C, S4A, B). H&E staining illustrated significantly smaller diameter of seminiferous tubules (average diameter of controls: 131,7 µm versus Mcph1-ΔBR1: 78,7 µm) and a strong vacuolisation of the germinal epithelium in Mcph1-ΔBR1 mice (Fig. 5C, D, S4A, B). In order to study whether germ cells are still present in seminiferous tubules of adult Mcph1-ΔBR1 mice, immunohistochemical analyses were performed on testis sections. DEAD box helicase 4 (Ddx4) staining revealed a strong reduction of germ cell-containing mutant tubules (~43% Ddx4+ tubules and 57% Sertoli cell-only tubules) (Fig. 5E). Interestingly, Lin28+ spermatogonial stem cells (SSCs) were still present in Mcph1-ΔBR1 testes (approx. 38% of mutant tubules with Lin28+ SSCs versus approx. 93% control tubules with Lin28+ SSCs); although the number of Lin28+ cells in seminiferous tubules of mutant mice was significantly lower and the ratio of Lin28+ SSCs to SRY-box 9 positive (Sox9+) Sertoli cells is 2.8-fold lower in Mcph1-ΔBR1 mice compared to controls (Fig.  5E). A strongly reduced number of Ddx4-positive seminiferous tubules per testicular cross-section accompanies an increase in Sertoli cell-only tubules in mutant mice (Sox9 staining). Quantification of Sox9 and Lin28 markers shows a significant reduction of the number of Ddx4+ and Lin28+ SSCs and progenitors in testes of Mcph1-ΔBR1 mice compared to controls.
Of note, the presence of Sertoli cells that express the androgen receptor (Ar) was normal in mutant mice ( Figure S4C) and these cells still stained positive for connexin 43 (Cx43), a gap junction protein critical to formation of the blood-testis-barrier ( Figure S4D). However, the blood-testis barrier in Mcph1-ΔBR1 testes could not be established compared to controls, due to germ cell degeneration ( Figure S4D). Moreover, we observed a normal pattern of testosterone-producing Leydig cells positive for 3β-hydroxysteroid dehydrogenase (3β-Hsd), an enzyme involved in the steroid hormone synthesis ( Figure S4E). The seemingly higher number of 3β-Hsd+ cells might be a compensatory response to the degeneration or disruption of testicular seminiferous epithelium. These results indicate that both Sertoli and Leydig cells are functional and unaffected by the deletion. All in all, these results demonstrate an impaired meiotic  progression and a strongly degenerated testicular seminiferous epithelium in mice lacking the N-terminal BRCT, while testicular niche cells are not affected.
When examining female infertility, we failed to observe the ovary macroscopically in 2-month-old females of both Mcph1-ΔBR1 and Mcph1-Δ mice (Fig. 6A, B). So as to study the reason for infertility, histological analyses were performed on H&E stained ovary sections of mutant and control mice. Ovaries of Mcph1-ΔBR1 and Mcph1-Δ young female mice were significantly smaller and lacking obvious follicles and oocytes (Fig. 6A, B). Strikingly, we noticed the occurrence of ovarian tumours in aged Mcph1-ΔBR1 mice (Fig. 6C-E) with a high penetrance (Fig. 6F), as observed in Mcph1-Δ females 27 . Of note, no testicular tumours were observed. These findings indicate that the N-terminal BRCT domain is required for gonad development and prevents ovary tumorigenesis in old female mice.  35 , our cellular analyses demonstrate that the Nterminal BRCT domain is indeed required for the efficient initiation of the DDR and DNA repair. In conclusion, the N-terminal BRCT domain has a functional importance equivalent to full length MCPH1 in vivo.

Discussions
In humans, four isoforms of MCPH1 are expressed in cells, including MCPH1-FL, MCPH1Δe1-3, Δe9-14, and Δe8 36 . MCPH1Δe1-3 lacking N-terminal BRCT domain may not be physiologically expressed, while MCPH1Δe9-14 that miss two C-terminal BRCT domains are highly expressed at the foetal stage of the brain, heart, and thymus, similar to MCPH1-FL 36suggesting the isoform lacking the C-terminal BRCT domains remains functional. In this regard, it is notable that mice lacking the C-terminal BRCT domain (Mcph1 gt/gt mice) did not exhibit microcephaly, although mutant cells showed PCC 29 . Altogether, these findings suggest that the entire protein is required for repressing PCC and the N-terminal BRCT is essential for brain and gonad development.
It is interesting to note that deletion of the N-terminal BRCT domain faithfully recapitulates the microcephaly and other phenotypes in mice lacking the entire MCPH1 protein. We showed previously that a complete depletion of MCPH1 caused premature mitotic entry with hypophosphorylation of Cdk1 via Chk1-dependent and βTrCP2-mediated Cdc25 regulation 12,23 . These molecular pathways are responsible for the neuroprogenitor division mode and maintenance of neuroprogenitor pools, thereby preventing microcephaly 12,23 . Yet, both Chk1 and βTrCP2 bind the middle region of MCPH1, but not N-terminal BRCT domain 12,17 . Of note there is a recently published paper reporting that two novel compound heterozygous missense variants (c.982 G > A and c.1273 T > A) in the exon 8 of MCPH1 gene caused primary microcephaly in a Saudi family; indicating an important role of middle Testis size is strongly reduced in Mcph1-ΔBR1 mice compared to controls (Ctr). B Ratio of paired testis weight to body weight (TW/BW) is significantly lower in Mcph1-ΔBR1 mice compared to controls. The number of mice analysed is indicated within the bar. C Representative pictures of H&E stained testis sections demonstrate a strong degeneration of the testicular germinal epithelium in Mcph1-ΔBR1 mice with lack of post-meiotic cells. Note that mutant testis tubules contain either few or lack of spermatocytes (sc) and vacuolised lumens (lu). D Quantitative histological analyses of H&E stained testis sections reveal significantly smaller diameter of seminiferous tubules and a strong vacuolisation of the germinal epithelium in Mcph1-ΔBR1 mice. Each circle represents one Ctr mouse and each diamond one Mcph1-ΔBR1 mouse. Per mouse, 30 circular tubular cross-sections were analysed to determine the diameter, and a minimum of 70 tubular cross-sections per mouse have been evaluated for presence or absence of vacuoles (va). The number of mice analysed is indicated within the bar and the number of analysed (circular) tubular cross-sections is indicated under the genotype in parentheses. E Representative pictures of immunostaining of testis sections by the indicated antibodies. DEAD box helicase 4 (Ddx4) germ cell marker staining detects residual germ cells (red arrows) in mutant testes. Lin-28 homologue A (Lin28) is expressed by spermatogonial stem cells and progenitors (green arrows). SRY-box 9 (Sox9) marker protein is expressed normally in Sertoli cells (blue arrows) of mutants and controls. Intense brown colour indicates positive marker staining, blue colour depicts hematoxylin counterstaining. Quantification of marker proteins is shown on the right panel of respective immunostaining. The number of mice analysed is indicated within the bar. The number of tubuli analysed is indicated under the genotype in parentheses. All experiments have been performed on testicular cross-sections of adult mice (9.5-12 months old). Bars represent Means ± SEM of a minimum of 3 determinations. Statistical analysis was performed by Student's t-test. **p < 0.01, ***p < 0.001. regions of MCPH1 for brain development in addition to the N-terminal BRCT domain 37 . Because Mcph1-ΔBR1 mutation maintained the intact middle region required for the interaction between MCPH1 and βTrCP2, these data suggest that the N-terminal BRCT domain has a function to regulate neurogenesis likely beyond the MCPH1-βTrCP2-Cdc25 axis. But we cannot disregard the possibility that a physical cross-talk between the N'-BRCT and the middle region is necessary in a conformationdependent manner.
Both our Mcph1-Δ and Mcph1-ΔBR1 models and another Mcph1-ko model 22 , all exhibited infertility with atrophy in reproductive organs, i.e. testes and ovaries 12,22 . Currently, no case report exists on human primary microcephaly associated with infertility, or testicular or ovarian atrophy. This raises an interesting question as to whether positive selection of MCPH1 during primate and human lineage evolution links brain size expansion with the germline fitness. Interestingly, Mcph1-Δ and Mcph1-ΔBR1 develop a very high incidence of tumours, yet only in ovaries, in great contrast to 17.1% Mcph1-ko mice, which developed malignant tumours originated from lymphomas and granulosa ovary tumours 22 . Different tumour spectrum and penetrance between these mice are unclear, but are likely attributable to genetic background and mutation introduced in these mice. Nonetheless, a high tumour incidence of these Mcph1-Δ and Mcph1-ΔBR1 mice supports tumour repressor function of MCPH1, corresponding to previous studies showing that MCPH1 is down-regulated or mutated in human cancer patients, including those suffering ovarian cancer 24,38,39 .
The identical phenotypes -small brain size and neuroprogenitor defects, PCC, DDR defects, the developmental block of gonads, as well as tumorigenesisamong N'-BRCT-deleted (Mcph1-ΔBR1) and MCPH1 complete knockout (Mcph1-ko, Mcph1-Δ) mice, strongly suggest an essential function of the N-terminal BRCT domain of the MCPH1 protein in brain size determination and other biological processes. MCPH1 has no enzymatic activity and functions most likely as a scaffold protein participating in various cellular activities. Therefore, the interaction partners of the N-terminal BRCT domain seem to coordinate the neuroprogenitor fate and pools in brain size determination, gonad development and tumorigenesis.