REGULATION OF THE GLIAL NA-DEPENDENT GLUTAMATE TRANSPORTERS BY CYCLIC-AMP ANALOGS AND NEURONS()

Sodium-dependent transport into astrocytes is critical for maintaining the extracellular concentrations of glutamate below toxic levels in t he central nervous system. In this study, the expression of the glial glutamate transporters GLT-1 and GLAST was studied in primary cultures derived from cortical tissue. In primary astrocytes, GLAST protein le vels were approximately one half of those observed in cortical tissue, but GLT-1 protein was present at very low levels compared with cortic al tissue. Maintenance of these astrocytes in medium supplemented with dibutyryl-cAMP (dbcAMP) caused a dramatic change in cell morphology, increased GLT-1 and GLAST mRNA levels approximate to 5-fold, increased GLAST protein approximate to 2-fold, and increased GLT-1 protein grea ter than or equal to 8-20-fold. These increases in protein expression were accompanied by 2-fold increases in the V-max and K-m values for N a+-dependent L-[H-3]glutamate transport activity. Although GLT-1 is se nsitive to inhibition by dihydrokainate in heterologous expression sys tems, no dihydrokainate sensitivity was observed in astrocyte cultures that expressed GLT-1. Biotinylation with a membrane-impermeant reagen t, separation of the biotinylated/cell surface proteins, and subsequen t Western blotting demonstrated that both GLT-1 and GLAST were present at the cell surface. Coculturing of astrocytes with neurons also indu ced expression of GLT-1, which colocalized with the glial specific mar ker, glial fibrillary acidic protein. Neurons induced a small increase in GLAST protein. Several studies were performed to examine the mecha nism by which neurons regulate expression of the glial transporters. T hree different protein kinase A (PKA) antagonists did not block the ef fect of neurons on glial expression of GLT-1 protein, but the addition of dbcAMP to mixed cultures of neurons and astrocytes did not cause G LT-1 protein to increase further. This suggests that neurons do not re gulate GLT-1 by activation of PKA but that neurons and dbcAMP regulate GLT-1 protein through convergent pathways. As was observed with GLT-1 ,the increases in GLAST protein observed in cocultures were not blocke d by PKA antagonists, but unlike GLT-1, the addition of dbcAMP to mixe d cultures of neurons and astrocytes caused GLAST protein to increase approximate to 2-fold. Neurons separated from astrocytes with a semipe rmeable membrane increased GLT-1 protein, indicating that the effect o f neurons was mediated by a diffusible molecule. Treatment of cocultur es with high concentrations of either N-methyl-D-aspartate or glutamat e killed the neurons, caused GLT-1 protein to decrease, and caused GLA ST protein to increase. These studies suggest that GLT-1 and GLAST pro tein are regulated independently in astrocyte cultures and that a diff usible molecule secreted by neurons induces expression of GLT-1 in ast rocytes.