The primary objective of this work was to assess the intrinsic nonbica
rbonate buffer capacity (beta(i)) of cultured neurons and astrocytes a
nd to compare the beta(i) values obtained to those of neocortical tiss
ue. A second objective was to determine the pH dependence of beta(i).
Titration of homogenates of whole-brain cortical tissue and cultured n
eurons with NaOH and HCl gave beta(i) values of 25-30 mmol.l-1.pH-1. T
he buffer capacity was essentially constant in the pH range of 6-7. As
trocytes showed a higher buffer capacity and a clear relation ship bet
ween beta(i) and pH. However, beta(i) decreased when pH was reduced fr
om 7 to 6. The beta(i) values derived from microspectrofluorometric st
udies on neurons and astrocytes were surprisingly variable, ranging fr
om 10 to 50 mmol.l-1-pH-1. The ammonia ''step method'' suggested that
beta(i) increased dramatically when pH was lowered from 7 to 6 but the
propionic ''step method'' failed to reveal such a pH dependence. Some
techniques obviously give erroneous values for beta(i), presumably be
cause changes in buffer base concentration (due to transmembrane fluxe
s of H+, HCO3-, NH4+ or anions of weak acids) violate the principles u
pon which the calculations are based. From the results obtained by dir
ect titration and with the propionate technique, we tentatively conclu
de that beta(i) in neurons and astrocytes are approximately 20 and 30
mmol.l-1.pH-1, respectively. We further suggest that the term ''intrin
sic buffer capacity'', as commonly used, is redefined.