Introduction
Rubinstein-Taybi syndrome (RTS or RSTS; OMIM #180,849, #613,684), first described in 1963, is a rare but archetypal developmental disorder with multiple congenital anomalies, intellectual disability and a prevalence of 1:100,000 to 1:125,000 at birth [
22,
23,
48]. It is characterized by microcephaly and intellectual disability, postnatal growth impairment, specific facial abnormalities and broad, angulated thumbs and big halluces [
9]. Genetic analyses uncovered heterozygous mutations in the highly homologous
CREB binding protein (CBP) and
p300 genes (also called
EP300 or
E1A binding protein p300) genes to be causative for the development of RSTS [
48,
53]. Mutations of
CBP can be found in 40–60% of patients with RSTS and mutations of
p300 are observed in approximately 10% of cases [
52]. Though showing an autosomal dominant character, RSTS is mostly caused by de novo mutations. The causative mutations in the
CBP gene include point mutations, small deletions and duplications, which may lead to premature translational stops as well as large deletions, including
CBP and flanking genes [
14,
36,
38,
48]. CBP, as well as its homolog p300, is a ubiquitously expressed transcriptional coactivator known to play an important role in embryonic development, growth control and cell homeostasis [
20]. It has an intrinsic lysine acetyltransferase (KAT) activity and stabilizes protein interactions with the transcription complex, thus mediating chromatin remodeling and transcription factor recognition [
31,
44]. It was shown to integrate signals from a multitude of signaling pathways, interacting with more than 400 transcription factors and other regulatory proteins, and to be present at promoters of more than 16,000 human genes [
5,
50]. Mouse models for RSTS with a conventional global heterozygous loss of CBP have been established and delivered indications for the causative role of the heterozygous loss of CBP or its KAT activity [
4,
33,
45,
59]. However, despite the heterozygous loss being the appropriate resemblance of the human situation, conventional global heterozygous knockout mouse models did not help to explain the intellectual disability in RSTS patients [
4,
62].
The knock-in and conditional knock-out models generated so far contributed largely to the existing knowledge about the mechanisms behind RSTS [
15,
27,
62]. The models include for example a knockout in postmitotic neurons as well as a central nervous system (CNS)-specific knockout and reveal an effect on long-term memory formation in CBP-deficient postmitotic neurons and effects of CBP on neuron morphology.
Thus, although a cognitive deficit was observed in all mouse models generated so far, the mechanisms underlying the intellectual disability could not be determined. Therefore, the effects of a complete loss of CBP during embryonic development in a CNS-specific conditional homozygous CBP knockout mouse model driven by the human glial fibrillary acidic protein (hGFAP) promoter (hGFAP-cre::CBPFl/Fl) were studied both in vivo and in vitro. A focus was thereby placed on the analysis of development and integrity of the forebrain structures neocortex, hippocampus, and olfactory bulb (OB). Additionally, postnatal neurogenesis and developmental processes such as neural precursor cell (NPC) proliferation and viability, neural differentiation and precursor cell migration were investigated.
Using this homozygous knockout approach, we were able to demonstrate that CBP function is crucial for proper brain development, cell differentiation and NPC migration as well as establishment of adult neurogenesis.
Materials and methods
Gene mutation type & location in published cases of RSTS
To investigate the role of the KAT and other domains of the CBP gene for the development of RSTS, 193 published cases of RSTS listed in the Human Gene Mutation Database (Qiagen Bioinformatics) were grouped by type and localization of the pathogenic mutation as follows: 97 cases of RSTS caused by point mutations; 34 cases of RSTS caused by duplications, insertions & indels; 62 cases of RSTS caused by small deletions; 59 cases of RSTS caused by large deletions (not used); 54 cases of RSTS caused by missense mutations regardless the location within the CBP gene; 36 cases of RSTS with missense mutations within KAT domain; 9 cases of RSTS with missense mutations in exons 1–17 (before KAT domain), 43 cases of RSTS with nonsense mutations regardless the location; 16 cases of RSTS with nonsense mutations in the KAT domain, 21 cases of RSTS with nonsense mutations in exons 1–17 (before KAT domain). We focused on point mutations for a more precise allocation concerning different domains and exons. To compare the mutation distribution between different areas of the gene, the numbers of point mutations per 100 bp (base pairs) rates were calculated for exons, certain domains, or regions of interest to take the different region sizes into account. Cases, in which the pathogenic mutation did not lie within the CBP exons, were not included.
Transgenic animals
The generation of both
hGFAP-cre [
8,
75] and
CBPFl/Fl [
73] transgenic mouse lines has been described previously. All animal procedures were performed in accordance with applicable animal protection laws. All experiments were approved by the state of Bavaria under license number 55.2–1-54-2532-10-14 and the state of Hamburg (Reference 113/16). Animal handling was done in accordance with local governmental and institutional animal care regulations.
Animal treatments
For measuring the proliferation rate in vivo, 25 μg bromodeoxyuridin (BrdU) or 5-ethynyl-2′-deoxyuridine (EdU) per gram bodyweight were injected intraperitoneally 2 h before sacrificing the animal. If indicated, a BrdU/EdU double pulse fate-mapping method was utilized. The double pulse method consisted of two injection steps. First, BrdU was injected and after a chosen interval, EdU was applied and the animal was sacrificed. For electron microscopy an in vivo perfusion with 4% paraformaldehyd was used as described elsewhere [
19].
Behavior testing
Different behavior tests were conducted to characterize hGFAP-cre::CBPFl/Fl mice. All tests were conducted between 9 am and 9 pm. The animals were accustomed to the test room for at least 24 h prior to testing. Gender-matched litter mates were used as a control group. All tests were performed with animals at P30. The animals’ genotypes were unknown to the tester during testing. Females and males were tested separately but indiscriminately included in the analysis.
An anxiety/curiosity light/dark test was utilized as published with a standard box size (60 × 40 × 40 cm, length x width x height). Latency until the first transition from the dark to bright compartment, the number of transitions and the total time spent in each compartment were measured.
An open field test was performed to analyze anxiety. Mice were subjected to an empty cage and their behavior was videotaped for 2 min. Afterwards, the video was analyzed for total distance travelled, time spent in center and vertical activity. The mouse path was tracked manually.
To investigate, whether structural and histological findings in
hGFAP-cre::CBPFl/Fl transgenic mice reflected disorders of functional systems, a modified buried food and an olfactory habituation/dishabituation test [
35,
72] were used. In the buried food test (BFT), the time until the mice had dug up a buried piece of food, was measured. For habituation/dishabituation testing plastic cartridges carrying a piece of cotton impregnated with 20 μl of either almond or banana extract were presented to the mice repeatedly. For the first 6 trials, almond extract was used for examining habituation, and in a 7th trial, banana extract was used as a novel scent to trigger dishabituation.
Genotyping
For genotyping, tail or ear biopsies were used. DNA was extracted by tissue lysis with Laird’s buffer (200 mM NaCl, 100 mM Tris-HCl pH 8.3, 5 mM EDTA, 0.2% SDS, 200 μg/ml protein kinase K in ddH2O) and Isopropanol precipitation. DNA was dissolved in TE buffer (20 mM Tris-HCL pH 8.3, 1 mM EDTA in ddH2O) and stored at 4 °C. Genotype-specific regions of the genome were amplified via PCR utilizing primers described in the original publications (Cre: TCCGGGCTGCCACGACCAA, GGCGCGGCAACACCATTTT, CBP: CCTCTGAAGGAGAAACAAGCA, ACCATCATTCATCAGTGGACT) and a TAQ-Polymerase (Promega) based standard reaction mixture.
Mouse MRI data
All measurements were conducted in T2 weighted images on freshly sacrificed animals. At least 3 animals per genotype were used. Pictures were analyzed by manual quantification using MRIcro software (Chris Rorden, Version 1.40).
Human MRI
The procedure and study design were approved by the ethics committee of the faculty of medicine of the Ludwig-Maximilians-University in Munich. In this retrospective analysis, images were analyzed in a pseudonymized manner. The children’s age at the time of MRI ranged from 1 month to 5 years. All measurements were performed on T2–weighted and T1-weighted sequences in axial, coronal and sagittal orientation. To perform matched-pair statistical analysis analogue measurements were conducted on cranial MR imaging data from patients with known RSTS and age-matched control subjects retrospectively identified in the institutional database.
Histology and immunohistochemistry
After dissection of mice, the brain was prepared for staining procedures with standard procedures. Tissue that was meant to be analyzed through confocal microscopy for 3D cell reconstruction was fixated overnight in 4% paraformaldehyde in PBS at 4 °C and then processed to 100 μm slices using a VT1000S microtome (Leica Biosystems). Tissue destined for light or fluorescence microscopy was fixated in 4% formaldehyde solution, embedded in paraffin and 3 μm sections were cut. General morphology was analyzed by hematoxylin/eosin (H&E) staining, following a standard protocol.
For immunohistochemical procedures, standard procedures including dewaxing and rehydration, antigen retrieval, endogenous peroxidase inactivation, antigen blocking (I-Block casein-based blocking reagent (ThermoFisher Scientific)) and detection were performed. Primary antibodies used in this study were as follows: BrdU (Roche #11170376001, 1:500), BrdU clone Mobu-1 (Invitrogen #B35128, 1:100), Calbindin (Chemicon #AB1778, 1:100), Caspase 3 (Cell Signaling Tech #9664, 1:00), CBP (Biozol #LS-B3360, 1:50), Cre (Covance #PRB-106P, 1:3000), HuB (Sigma #H1538, 1:200), Ki67 (Abcam #ab16667, 1:200), MBP (Abcam #ab40390, 1:100), NeuN (Abcam #ab104224, 1:300), Pax6 (DSHB #Pax6, 1:25), Prox1 (Abcam #ab199359, 1:500), Sox2 (Abcam #ab79351, 1:200), Tbr2 (Millipore #AB2283,1:300), Wfs1 (Proteintech #11558–1-AP, 1:50). Detection was achieved by using the DAKO EnVision™Plus System, HRP following the manufacturer’s instruction or for immunofluorescence staining, with species-specific fluorophore linked secondary antibodies (Alexa 546 Invitrogen # A-11003 and DAPI Roth #28718–90-3).). EdU positive cells were stained by using the Click-IT® assay (ThermoFisher #C10637). The 100 μm slices for 3D cell reconstruction were stained directly with NeuroTrace 530/615 (1:100; ThermoFisher #N21482), and Hoechst (1:1000; Invitrogen #H3570) as a nuclear counter staining. VECTASHIELD HardSet Antifade mounting medium (VECTOR Laboratories) was used for mounting.
Stereological measurements
To increase validity and reduce variability a stereological approach was chosen for analyzing the different forebrain structures. For each structure of interest three section planes were analyzed in every animal of the hGFAP-cre::CBPFl/Fl- or control group. In each section plane, several measurements were executed and averaged for parameters. Section planes were chosen by recognizable landmarks in light microscopy instead of predetermined intervals as brain size was not a stable constant. All measurements were conducted on pseudorandomized images with the help of the open source image processing program ImageJ.
3D cell volume reconstruction of layer V giant pyramidal cells
For 3D cell reconstruction z-stack series in layer V of the neocortex were acquired from 100 μm slices stained with NeuroTrace & Hoechst using a Zeiss LSM780 confocal microscope and ZEN microscope software (Zeiss). The z-stacks acquired through confocal microscopy were further analyzed through custom-written Matlab analysis using the microscopy image analysis software Imaris (Bitplane) to result in 3D reconstructions of neual cell somata and cell cores.
Golgi-Cox staining and quantification
Golgi-Cox staining of adult mouse brains was performed using the FD Rapid GolgiStain Kit (FD NeuroTechnologies, INC.) according to the instructions from the manufacturer. Briefly, freshly dissected brains were impregnated with a premade solution of mercuric chloride, potassium dichromate and potassium chromate for 1 week, cut in 200 μm slices with a vibratome (Leica) and the impregnation was visualized with a solution provided by the manufacturer. For quantification of dendrite length, numbers of branches per dendrite and spine density, 10 representative pictures per mouse were taken and three mice per genotype were used. For analysis of spine types, 10 neurons per mouse were analyzed.
Electron microscopy
Electron microscopy was performed on adult mice perfused with 4% PFA. The brain was dissected immediately after perfusion and stored in 2% PFA + 2% Glutaraldehyde. A piece of cerebral cortex was cut from a frontal slice of the brain and prepared for EM. Samples were washed in 0.1 M cacodylate buffer (Sigma-Aldrich), incubated for 2 h in 1% osmium tetroxide (Science Services, Munich, Germany), dehydrated in an ascending series of ethanol, and embedded in Epon 812 (Serva). Ultrathin sections were counterstained with uranyl acetate (Polyscience, Eppelheim, Germany) and lead citrate (Riedel-de Haën, Seelze, Germany), and analyzed with a LEO 912 AB OMEGA electron microscope (Leo Elektronenmikroskopie, Oberkochen, Germany).
SVZ and OB explant culture
Cell explants of SVZ were prepared as previously described in [
68]. Briefly, brains of 2 to 5 day old mice were dissected, freshly cut with a vibratome (Leica) and the SVZ dissected. The SVZ was cut and put in 50 μl Matrigel (Corning). The explants were cultured in 500 μl culture medium (Neurobasal (Gibson), B27, 0.5 mM L-Glutamine, 50 U/ml Pen/Strep) in a 24 well plate. OB explants were kept in the same medium in 30 μl Matrigel. Cultures were grown for 48 h in a humidified incubator at 37 °C and 5% CO
2. In case of medium exchange cultures, the medium was changed after 24 h. After 48 h, pictures of the cultures were taken and migration distance or single cell number was determined. Per experiment, at least 6 explants per condition were analyzed and three independent experiments were conducted. If indicated, 200 ng recombinant IGF1 (PeproTech) were added to the culture medium.
RNA sequencing
After isolation of total RNA the RNA integrity was analyzed with the RNA 6000 Nano Chip on an Agilent 2100 Bioanalyzer (Agilent Technologies). From total RNA, mRNA was extracted using the NEBNext Poly(A) mRNA Magnetic Isolation module (New England Biolabs) and RNA-Seq libraries were generated using the NEXTFLEX Rapid Directional qRNA-Seq Kit (Bioo Scientific) as per the manufacturer’s recommendations. Concentrations of all samples were measured with a Qubit 2.0 Fluorometer (Thermo Fisher Scientific) and fragment lengths distribution of the final libraries was analyzed with the DNA High Sensitivity Chip on an Agilent 2100 Bioanalyzer (Agilent Technologies). All samples were normalized to 2 nM and pooled equimolar. The library pool was sequenced on the NextSeq500 (Illumina) with 1 × 75 bp, with 16.1 to 18.6 mio reads per sample.
For each sample the sufficient quality of the raw reads was confirmed by
FastQC v0.11.8 [
54]. Afterwards, the reads were aligned to the mouse reference genome GRCm38 with
STAR v2.6.1c [
16] and simultaneously counted per gene by employing the
--quantmode GeneCounts option. Counts are based on the Ensembl annotation release 95. Differential expressed genes were estimated with
DESeq2 v1.22.2 [
37].
Statistical analysis
Statistical analysis was conducted with Prism (Versions 5.0, 6.0 and 7.0) software (GraphPad). If not stated otherwise, all data presented are mean ± s.e.m., with n = 3 for each group and each data point represents an individual animal or an independent experiment. P values < 0.05 were considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). By default, the unpaired t test (two-tailed) was applied to compare the means of two groups if not stated otherwise in the figure legend. The Χ2 (Chi-squared) test was used for comparing point mutation frequencies of specific domains and regions of the CBP gene. Further, a two-way repeated measurements ANOVA test and Bonferroni’s multiple comparisons post hoc test as well as nonlinear regression using an exponential decay model were applied for evaluation of the habituation test.
Acknowledgments
We are indebted to Michael Schmidt, Silvia Occhionero, Pia Schindler, Margarethe Gregersen, Kristin Hartmann, Gundula Pilnitz-Stolze, Anne Reichstein and Jacqueline Kolanski for excellent technical support. We also thank the HPI Next generation sequencing platform for performing the RNA Sequencing. This work was supported by grants from the Deutsche Krebshilfe, the Roggenbuck Stiftung, and the Wilhelm Sander Stiftung. U.S. was supported by the Fördergemeinschaft Kinderkrebszentrum Hamburg. M.S. was supported by the Hans Brökel Stiftung für Wissenschaft und Kultur.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.