Bergmann glia expression of polyglutamine-expanded ataxin-7 produces neurodegeneration by impairing glutamate transport.

Non-neuronal cells may be pivotal in neurodegenerative disease, but the mechanistic basis of this effect remains ill-defined. In the polyglutamine disease spinocerebellar ataxia type 7 (SCA7), Purkinje cells undergo non-cell-autonomous degeneration in transgenic mice. We considered the possibility that glial dysfunction leads to Purkinje cell degeneration, and generated mice that express ataxin-7 in Bergmann glia of the cerebellum with the Gfa2 promoter. Bergmann glia-specific expression of mutant ataxin-7 was sufficient to produce ataxia and neurodegeneration. Expression of the Bergmann glia-specific glutamate transporter GLAST was reduced in Gfa2-SCA7 mice and was associated with impaired glutamate transport in cultured Bergmann glia, cerebellar slices and cerebellar synaptosomes. Ultrastructural analysis of Purkinje cells revealed findings of dark cell degeneration consistent with excitotoxic injury. Our studies indicate that impairment of glutamate transport secondary to glial dysfunction contributes to SCA7 neurodegeneration, and suggest a similar role for glial dysfunction in other polyglutamine diseases and SCAs.

Expression of GABA(B) receptor in the avian auditory brainstem: ontogeny, afferent deprivation, and ultrastructure.

Nucleus magnocellularis (NM), nucleus angularis (NA), and nucleus laminaris (NL), second- and third-order auditory neurons in the avian brainstem, receive GABAergic input primarily from the superior olivary nucleus (SON). Previous studies have demonstrated that both GABA(A) and GABA(B) receptors (GABA(B)Rs) influence physiological properties of NM neurons. We characterized the distribution of GABA(B)R expression in these nuclei during development and after deafferentation of the excitatory auditory nerve (nVIII) inputs. We used a polyclonal antibody raised against rat GABA(B)Rs in the auditory brainstem during developmental periods that are thought to precede and include synaptogenesis of GABAergic inputs. As early as embryonic day (E)14, dense labeling is observed in NA, NM, NL, and SON. At earlier ages immunoreactivity is present in somas as diffuse staining with few puncta. By E21, when the structure and function of the auditory nuclei are known to be mature, GABA(B) immunoreactivity is characterized by dense punctate labeling in NM, NL, and a subset of NA neurons, but label is sparse in the SON. Removal of the cochlea and nVIII neurons in posthatch chicks resulted in only a small decrease in immunoreactivity after survival times of 14 or 28 days, suggesting that a major proportion of GABA(B)Rs may be expressed postsynaptically or on GABAergic terminals. We confirmed this interpretation with immunogold TEM, where expression at postsynaptic membrane sites is clearly observed. The characterization of GABA(B)R distribution enriches our understanding of the full complement of inhibitory influences on central auditory processing in this well-studied neuronal circuit.

Electron microscopy of degenerative changes in the chick basilar papilla after gentamicin exposure.

We present a sequential study of the substructural alterations in the chick basilar papilla at the earliest signs of hair cell degeneration. Three-day posthatch chicks received a single injection of gentamicin (300 mg/kg) and were killed at 6, 8, 12, 15, 18, 21, and 24 hours after the injection. The basilar papillae were studied by conventional transmission electron microscopy. Examination was limited to the basal region, where all hair cells are eliminated by this treatment. As early as 8 hours and clearly by 12 hours, altered fine structure was seen in hair cells. Changes included rounding and swelling of the hair cells, condensation of nuclear chromatin, dissolution of ribosomes, dilatation of the mitochondria, and accumulation of inclusion bodies and lysosomes. By 15-18 hours, lysosomes increased and became denser, afferent terminals appeared swollen, and the first cell extrusion was seen. Efferents were unaffected, and supporting cells, though having inclusion bodies now, retained normal intercellular junctions. By 21-24 hours, large regions of complete hair cell loss were composed of expanded supporting cell processes with normal-appearing intercellular junctions and portions of extruded hair cells, partially attached to the supporting cell surface. These observations demonstrate that auditory hair cells undergo a rapid and controlled process of hair cell extrusion that allows preservation of the reticular lamina and minimal contamination of surrounding structures by intracytoplasmic contents of the damaged hair cells.

Ultrastructural analysis of [3H]thymidine-labeled cells in the rat utricular macula.

Ototoxic drugs stimulate cell proliferation in adult rat vestibular sensory epithelia, as does the infusion of transforming growth factor alpha (TGFalpha) plus insulin. We sought to determine whether new hair cells can be regenerated by means of a mitotic pathway. Previously, studies have shown that the nuclei of some newly generated cells are located in the lumenal half of the sensory epithelium, suggesting that some may be newly generated sensory hair cells. The aim of this study was to examine the ultrastructural characteristics of newly proliferated cells after TGFalpha stimulation and/or aminoglycoside damage in the utricular sensory epithelium of the adult rat. The cell proliferation marker tritiated-thymidine was infused, with or without TGFalpha plus insulin, into the inner ears of normal or aminoglycoside-damaged rats for 3 or 7 days by means of osmotic pumps. Autoradiographic techniques and light microscopy were used to identify cells synthesizing DNA. Sections with labeled cells were re-embedded, processed for transmission electron microscopy, and the ultrastructural characteristics of the labeled cells were examined. The following five classes of tritiated-thymidine labeled cells were identified in the sensory epithelium: (1) labeled cells with synaptic specializations that appeared to be newly generated hair cells, (2) labeled supporting cells, (3) labeled leukocytes, (4) labeled cells that we have classified as "active cells" in that they are relatively nondescript but contain massive numbers of polyribosomes, and (5) labeled degenerating hair cells. These findings suggest that new hair cells can be generated in situ by means of a mitotic mechanism in the vestibular sensory epithelium of adult mammals.