Ataxia and peripheral nerve hypomyelination in ADAM22-deficient mice

Abstract

Background: ADAM22 is a member of the ADAM gene family, but the fact that it is expressed only in the nervous systems makes it unique. ADAM22's sequence similarity to other ADAMs suggests it to be an integrin binder and thus to have a role in cell-cell or cell-matrix interactions. To elucidate the physiological functions of ADAM22, we employed gene targeting to generate ADAM22 knockout mice.

Results: ADAM22-deficient mice were produced in a good accordance with the Mendelian ratio and appeared normal at birth. After one week, severe ataxia was observed, and all homozygotes died before weaning, probably due to convulsions. No major histological abnormalities were detected in the cerebral cortex or cerebellum of the homozygous mutants; however, marked hypomyelination of the peripheral nerves was observed.

Conclusion: The results of our study demonstrate that ADAM22 is closely involved in the correct functioning of the nervous system. Further analysis of ADAM22 will provide clues to understanding the mechanisms of human diseases such as epileptic seizures and peripheral neuropathy.

Figure 1

Targeted mutation of Adam22. (A) The genomic structure of the wild-type Adam22 allele (top), the targeting construct (middle) and the disrupted allele (bottom) are shown. ADAM22 expression was disrupted by the insertion of a termination codon and a PGKneo cassette into exon 8. An MC1/TK cassette at the end of the targeting vector allows for negative selection. The 3' probe represents the position of the external probe used for Southern blot analysis, and expected BamHI [B] fragments are indicated by arrows. (B) Southern blot analysis of mouse genomic DNA. The expected DNA fragments for the wild-type allele and disrupted allele are 7.5-kb and 2.5-kb, respectively. +/+, wild-type; +/-, heterozygote; -/-, homozygote. (C) Western blot analysis of ADAM22 expression in the mouse cerebellum. Absence of ADAM22 protein in the Adam22 (-/-) mutant cerebellum was shown using anti-ADAM22-cp (cytoplasmic domain) polyclonal antibody.

Figure 2

Uncoordinated movements in the Adam22 (-/-) mice at postnatal day 18 (P18). (A) Adam22 (-/-) mice were smaller than their wild-type (+/+) littermates. (B, C) Adam22 (-/-) mice at P18 were unable to support themselves on their hindlimbs.

Figure 3

Neuronal ADAM22 mRNA expression in the CNS. To determine the ADAM22 mRNA distribution, in situ hybridisation analysis using ³⁵S-labeled probe was performed. Coronal (A, B) and sagittal (C, D) sections of the mouse brain and spinal cord (E) were shown. Using the antisense probe (A, C), strong signals were obtained, especially in the hippocampus and the cerebellum, while no signals was detected by the sense probe (B, D). In the spinal cord, autoradiograms of ADAM22 mRNA was detected in the grey matter (E).

Figure 4

Normal neurodevelopment in the CNS of ADAM22-deficient mice. Sagittal sections of the cerebellum from wild-type mice (A, C, E, G) and homozygous mutants (B, D, F, H) at postnatal day 13 were stained for calbindin (C and D; green) or MBP (G and H; green), and were counterstained with DAPI (A, B, E, F; blue). Significant abnormalities were not observed in the homozygotes. Hippocampal neurons were stained by anti-Neu N antibody (I and J). Spinal myelin was analysed by MBP staining (K and L). These analyses showed no obvious differences between homozygotes (J, L) and wild-type littermates (I, K). Bar: (A, B, E, F) 100 μm, (K, L) 250 μm.

Figure 5

Hypomyelination of peripheral nerves in ADAM22-deficient mice. Epon embedded semithin cross-sections of the sciatic nerves (A, B), the trigeminal nerves (C, D) and the spinal cord [lateral funiculus] (E, F) of the indicated genotypes at postnatal day 10 were stained with toluidine blue. Note that the ADAM22-deficient mouse shows thin myelin or lack of myelin in the peripheral nerve fibres (B, D), but the spinal myelinated fibres are intact (F).

Figure 6

Electron microscopic analysis of sciatic nerves. Electron micrographs of the sciatic nerves from Adam22 +/- (A) and Adam22 -/- (B) mice at postnatal day 10 are shown. In the heterozygote (A), thick myelin was formed, while no myelin was formed in the ADAM22-deficient mouse (B). The axons looked normal in each genotype.

Figure 7

Marked delay in differentiation of the mutant Schwann cells. Transverse sections of the sciatic nerves of the indicated genotypes at postnatal day 13 were stained for MBP (C and D; green) or S100 (E and F; red), and were counterstained with DAPI (A, B; blue). In the mutant sciatic nerve, density of DAPI-signals were greatly increased (B) compared with those of the wild-type (A). Bar: (A, B) = 50 μm.

Figure 8

Adam22 gene structure and tissue specific transcripts. (A) RT-PCR analysis. Amplified DNA fragments were analysed by 1 % agarose gel electrophoresis. Lanes 1. 100 bp DNA ladder; 2. cerebellum; 3. spinal cord; 4. dorsal root ganglion; 5. sciatic nerve; 6. cultured Schwann cells; 7. distilled water (B) Exon organization of the mouse Adam22 gene is illustrated. Boxes indicate exons. The G01 transcript (orthologous to the human ADAM22 isoform 1 transcript) is composed of grey boxes. Boxes in black indicate non-coding region. (C) Summary of the isolated clones. The nucleotide sequence data have been deposited with the DDBJ/EMBL/GenBank Data Libraries under the accession numbers described in the table. (D) Number of clones isolated from each tissue is summarized. Cb; cerebellum, Sp; spinal cord, DRG; dorsal root ganglion, SN; sciatic nerve, cSC; cultured Schwann cells.