Disruption of Mbd5 in mice causes neuronal functional deficits and neurobehavioral abnormalities consistent with 2q23.1 microdeletion syndrome

Abstract

2q23.1 microdeletion syndrome is characterized by intellectual disability, motor delay, autistic-like behaviors, and a distinctive craniofacial phenotype. All patients carry a partial or total deletion of methyl-CpG-binding domain protein 5 (MBD5), suggesting that haploinsufficiency of this gene is responsible for the phenotype. To confirm this hypothesis and to examine the role of MBD5 in vivo, we have generated and characterized an Mbd5 gene-trap mouse model. Our study indicates that the Mbd5(+/) (GT) mouse model recapitulates most of the hallmark phenotypes observed in 2q23.1 deletion carriers including abnormal social behavior, cognitive impairment, and motor and craniofacial abnormalities. In addition, neuronal cultures uncovered a deficiency in neurite outgrowth. These findings support a causal role of MBD5 in 2q23.1 microdeletion syndrome and suggest a role for MBD5 in neuronal processes. The Mbd5(+/) (GT) mouse model will advance our understanding of the abnormal brain development underlying the emergence of 2q23.1 deletion-associated behavioral and cognitive symptoms.

Keywords: MBD5; autistic disorder; intellectual disability; mouse model.

Figure 1. Expression of mouse Mbd5 transcripts

A   Schematic of the Mbd5 transcripts depicting exons 11–15. B   Schematic of the predicted mouse MBD5 proteins. The black box (arrow) represents the MBD, the gray box (arrowheads) the PWWP domain, and the shaded areas represent alternative regions of the proteins. In brackets is the number of amino acids. C   RT–PCR from cDNA derived from mouse brain, heart, and liver at the indicated developmental time points. The numbered boxes represent the exons included in the amplicons. 12l and 12s depict the extended and short exon 12, respectively; and 11l depicts the extended exon 11. The data show that 11l is a terminal exon since we did not detect transcripts carrying exon 11l spliced with any downstream exon.

Figure 2. Insertion of a β-geo gene-trap cassette creates an Mbd5 hypomorph mouse

A   Schematic of the insertional mutation of Mbd5 in intron 2 with the β-geo gene-trap cassette pU-21B. Numbers correspond to the exons of Mbd5. Upward arrows indicate the start of the Orc4 and Mbd5 genes, respectively. Primers used in (B) are depicted by small arrows. B   Expression analysis of Mbd5 transcripts was measured by quantitative real-time PCR on E21 mouse brains (n = 6 Mbd5^+/+, n = 7 Mbd5^+/GT, and n = 3 Mbd5^GT/GT; Student's t-test; unpaired, two-tailed distribution, P-values are displayed above the bars in the figure).

Figure 3. Pattern of expression of mouse Mbd5

30-μm cryostat sections from brain (A and B) and pituitary (C) from Mbd5^+/+ and Mbd5^+/GT mice of 1 month of age were processed for X-gal staining (n = 2 Mbd5^+/+, n = 4 Mbd5^+/GT). A   Brain sections. Scale bar = 5 mm. B   Sections were also subjected to immunofluorescence for the detection of NeuN (red) and GFAP (green). Nuclei were stained with DAPI, and the images visualized using a confocal microscope. Transmittance image (X-gal) and fluorescent images from DAPI (blue), NueN (red), GFAP (green), and merged fluorescence (X-gal/DAPI/GFAP/NeuN) obtained by confocal acquisition. Magnified regions containing GFAP⁺ astrocytes and NeuN⁺ neurons are shown in the right panel to highlight the elevated β-gal activity in neuronal cells. Scale bar = 50 μm. C   X-gal staining of 30-μm pituitary sections showing expression in the intermediate (IL) and anterior (AL) lobes.

Figure 4. Reduced body weight and craniofacial abnormalities in Mbd5^+/GT mice

A   Body weight of male Mbd5^+/+ and Mbd5^+/GT mice measured biweekly from 4 to 8 weeks of age differ significantly (n = 9 Mbd5^+/+ and n = 13 Mbd5^+/GT; Student's t-test; unpaired, two-tailed distribution, P-values are displayed above the bars in the figure). B   Top view from skulls of male Mbd5^+/+ and Mbd5^+/GT mice at 1 month of age.

Figure 5. Mbd5^+/GT mice display abnormalities in motor performance

A   Mbd5^+/GT mice have decreased grip strength than Mbd5^+/+controls. B   Mbd5^+/GT mice were less capable than Mbd5^+/+ mice to hang from an elevated wire with their forepaws. C Mbd5^+/GT displayed impaired motor coordination compared with Mbd5^+/+ mice on the dowel test. Mbd5^+/GT mice reach the platform less times (arrivals) and fell of the dowel more frequently than Mbd5^+/+ mice. D   Performance on the accelerating rotarod is deficient in Mbd5^+/GT compared with Mbd5^+/+ over all trials (6-week-old male mice, n = 9 Mbd5^+/+ and n = 16 Mbd5^+/GT; Student's t-test; unpaired, two-tailed distribution, P-values are displayed above the bars in the figure).

Figure 6. Abnormal social behavior and deficient learning in Mbd5^+/GT mice

A–C   Evaluation of activities of male mice observed over 10 min in their own cage. Bars represent total time mice spent in the indicated activity. (A) Observation of unperturbed mice. (B) Evaluation of mice in the presence of a novel object (Eppendorf 1.5-ml tube). (C) Social interaction of male test mice in the presence of a stranger mouse. D   Mbd5^+/GT mice show impaired contextual and cued conditioning. Bars represent the amount of freezing behavior, which is used as an index for recollection of the context or tone. Age of male mice was between 12 and 13 weeks. Asterisks indicate a significant change in Mbd5^+/GT mice (n = 9 Mbd5^+/+ and n = 14 Mbd5^+/GT; Mann–Whitney U-test, P-values are displayed above the bars in the figure).

Figure 7. Mbd5^+/GT neurons have reduced neurite outgrowth and branching

A, B   Dissociated E16 cortical neuron cultures generated from Mbd5^+/GT and Mbd5^+/+ mice were plated, and cellular morphology was recorded every 3 h for 48 h using an IncuCyte live-cell imaging system located within the incubator. Phase images were analyzed for process length and branching by Incucyte's NeuroTrack software. Shown are average values for neurite length (A) and number of branching points (B) normalized to cell bodies of quadruplicate wells per embryo (n = 3 Mbd5^+/+ and n = 4 Mbd5^+/GT; Student's t-test; unpaired, two-tailed distribution, P-values are displayed above the line in the figure). Arrows indicate the removal of the plates from the incubator at 24 h to replace plating media for neurobasal media.

Figure 8. Mbd5 acts in vitro as a transcriptional activator

A   Schematic of the constructs for the transactivation studies. The black box represents the MBD, and the gray box the PWWP domain. B   Constructs encoding the Gal4 DNA-binding domain fused to Mbd5 and the Gal4-responsive luciferase reporter gene were co-transfected into HEK293 cells. A β-galactosidase expression plasmid was also co-transfected to allow normalization for transfection efficiency. Cells were collected, and luciferase activity was measured 24 h after transfection. Each bar represents the mean ± SD from three independent experiments. (Student's t-test; unpaired, two-tailed distribution compared with cells transfected with GAL4-DBD, P-values are displayed above the bars in the figure.) C   Localization of transfected Mbd5 in mouse neuroblastoma N2A cells. Mbd5 and the heterochromatic marker MeCP2 fused to GFP or dsRed at the N-terminus were transfected into N2A cells. DAPI was used for nucleic acid staining (blue). The images were taken with a confocal LSM 710 microscope from Zeiss. Scale bar = 10 μm.