Ethambutol is toxic to retinal ganglion cells via an excitotoxic pathway.

PURPOSE: Ethambutol is an essential medication in the management of tuberculosis. However, it can cause an optic neuropathy of uncertain etiology. Ethambutol toxicity was therefore studied in rodent retinal cells, and agents that might block its toxicity were considered. METHODS: The toxicity of ethambutol and related agents was evaluated in rodent retinal dissociated cell preparations and whole eyes. Calcium fluxes and mitochondrial function were evaluated by fluorescent and staining techniques. For in vivo assays, adult rats were administered oral ethambutol over a 3-month period. Cell survival was assessed by stereology. RESULTS: Ethambutol is specifically toxic to retinal ganglion cells in vitro and in vivo. Endogenous glutamate is necessary for the full expression of ethambutol toxicity, and glutamate antagonists prevent ethambutol-mediated cell loss. Ethambutol causes a decrease in cytosolic calcium, an increase in mitochondrial calcium, and an increase in the mitochondrial membrane potential. CONCLUSIONS: The visual loss associated with ethambutol may be mediated through an excitotoxic pathway, inasmuch as ganglion cells are rendered sensitive to normally tolerated levels of extracellular glutamate. Ethambutol perturbs mitochondrial function. Its toxicity may depend on decreased ATPase activity and mitochondrial energy homeostasis. Glutamate antagonists may be useful in limiting the side effects seen with ethambutol.

NMDA sensitivity is neurite enhanced.

Glutamate toxicity in retinal ganglion cells has been well documented both in vitro and in vitro, and may play a role in both normal neuronal development and a variety of pathological states. Glutamate receptors are found on cell bodies and neuronal processes, both axons and dendrites. Other work has suggested that one or more of these locales may play a more pronounced role in glutamate-mediated toxicity. We now report that N-methyl-D-aspartate (NMDA) is more toxic to retinal ganglion cells with neurites. Cells without neurites were relatively unaffected by glutamate or NMDA. Cells with longer neurites or more neurite branch points were more likely to sustain NMDA-mediated neurotoxicity. These observations suggest that glutamate-mediated loss may be mediated through NMDA receptors found on neurites, rather than through a direct effect on the cell body.

Cation channel control of neurite morphology.

The development of neuronal polarity and morphology is essential for a functioning nervous system. The present study was undertaken to explore whether blockade of specific channels alter neuronal morphology. Retinal ganglion cells were cultured in the presence of antagonists to NMDA, AMPA/kainate, L-, N-, P-, and Q-type voltage-dependent calcium channels (VDCCs). Five parameters were measured under these conditions: the number of neurites at the cell body, total neurite length, the length of the longest neurite, the number of branch points per neurite, and the diameter of the cell soma. Antagonists to NMDA and L-type VDCCs reduce the number of neurites at the cell body; antagonists to P- and Q-type VDCCs increase the number of neurites. Antagonists to the N-type VDCCs increase total neurite outgrowth, while antagonists to the NMDA and P-type channels reduce total neurite length. Antagonists to the NMDA and L-type channels increase the length of a single neurite, while decreasing the number of branch points; antagonists to the P- and Q-type VDCCs do essentially the opposite-increase the number of neurites, while decreasing the length of each. Blockade of one or more cation channels in developing retinal ganglion cells significantly perturbs neurite morphology. This study may help elucidate part of the role that cation channel signaling plays in neuritic development.

Stimulated kidney tubular epithelial cells express membrane associated and secreted TNF alpha.

Tumor necrosis factor-alpha (TNF alpha) is a pleiotropic, pro-inflammatory peptide cytokine which promotes immune renal injury, and participates in T cell activation. It is produced by macrophages, T cells, and some non-hematopoietic cells, and is cytotoxic in picogram quantities. As renal tubular epithelial cells (TEC) bearing MHC class II (Ia) antigens and adhesion molecules (ICAM-1) can act as immune accessory cells, the ability of TEC to produce costimulatory cytokines could augment TEC accessory capacity in vivo. We report that transformed TEC express low levels of TNF alpha in response to LPS or IL-1 alpha as a secreted product and as a cytotoxic membrane associated molecule displayed on the cell surface. Surface labelling and immunoprecipitation studies of TEC detect a number of bands including a prominent 26 kD protein, which is the predicted size of TNF alpha precursor. TNF alpha mRNA transcripts were also detected by in situ hybridization in cortical tubules of C3H/FeJ mice injected with LPS, demonstrating the capacity of normal tubular epithelial cells to express TNF alpha in vivo. This report demonstrates for the first time the ability of kidney tubular cells to express TNF alpha protein and that membrane associated TNF alpha is not limited to hematopoietic cells. The function of small amounts of TNF displayed on the surface of tubular cells may be amplified by the abundance of these cells within the renal cortex, and may allow TEC to modulate immune responses within the kidney during inflammation.

An astrocytic binding site for neuronal Thy-1 and its effect on neurite outgrowth.

Thy-1, a member of the immunoglobulin superfamily, is one of the most abundant glycoproteins on mammalian neurons. Nevertheless, its role in the peripheral or central nervous system is poorly understood. Certain monoclonal antibodies to Thy-1 promote neurite outgrowth by rodent central nervous system neurons in vitro, suggesting that Thy-1 functions, in part, by modulating neurite outgrowth. We describe a binding site for Thy-1 on astrocytes. This Thy-1-binding protein has been characterized by immunofluroesence with specific anti-idiotype monoclonal antibodies and by three competitive binding assays using (i) anti-idiotype antibodies, (ii) purified Thy-1, and (iii) Thy-1-transfected cells. The Thy-1-binding protein may participate in axonal or dendritic development in the nervous system.