Sodium-dependent high-affinity uptake of taurine in cultured cerebellar granule cells and astrocytes

Taurine uptake in cultured cerebellar granule cells and astrocytes consisted of a saturable high-affinity component and nonsaturable diffusion. The transport constant (Km) was significantly lower and the maximal velocity (V) higher in granule cells than in astrocytes. The uptakes were strictly sodium dependent and also moderately decreased in potassium-free medium. The specificity profile of taurine uptake was similar in both cell types, hypotaurine, beta-alanine, and guanidinoethanesulphonic acid being the most potent inhibitors, followed by GABA and homotaurine. Glutamate inhibited taurine uptake more in astrocytes than in granule cells. In principle, the uptake systems were similar in granule cells and astrocytes, exhibiting features characteristic of uptake of a neurotransmitter or -modulator.     

Uptake,release and metabolism of alaume in neurons and astrocytes in primary cultures

The uptake, release, and metabolism of alanine were studied in primary cultures of cerebral cortical neurons or astrocytes and cerebellar granule neurons. All three cell types exhibited a saturable, sodiumdependent uptake of alanine with K_M values (ULM) of 256 ± 30, 463 ± 39, and 292 ± 39, respectively, and V values (nmol/min/mg) of 15.9 ± 0.7,7.9 ± 0.01, and 17.4 ± 0.8, respectively. The corresponding val' ues < nmol/min/mg for the specific activity of alanine aminotransferase were 4.7 ± 0.4, 17.1 ± 2.5, and 4.5 ± 0.9 (all values represent the mean ± SEM). Release of alanine from the cells was rectilinear with time over a 10 hr period in case of astrocytes (40 nmol/hr/ mg) and cerebellar granule neurons (21 nmol/hr/mg). In cortical neurons the release rate declined from an initial value of 19 nmol/hr/mg during the first 3 hr to a value of less than 3 nmol/hr/mg during the subsequent 7 hr of incubation. Metabolism of [¹⁴C]alanine to ¹⁴CO₂ was found to have a lag period of 15 min and subsequently the rate of CO₂ production was constant over a 45 min period with a value of 0.5 nmol/min/mg in granule neurons and about 0.3 nmol/ min/mg in the other two cell types. Altogether the results show that alanine is preferentially produced in and released from astrocytes and accumulated into both GABAergic cortical neurons and gintamatergic cerebellar granule neurons. © 1993 Wiley-Liss, Inc.     

Stimulus-coupled release of amino acids from cerebellar granule cells in culture

Cerebellar cultures greatly enriched in excitatory granule neurones were depolarized by exposure to either elevated K⁺ or veratrine. Stimulation of the release of not only Glu, but also of certain amino acids, including Gly, Ala and Ser, was observed. The effect was specific, as depolarization did not induced the release of all the estimated amino acids or of lactate dehydrogenase. In comparison with the characteristics of the evoked release of Glu, those of the responsive neutral amino acids were similar in terms of Ca²⁺-dependence, but differences were also noted. Thus, upon stimulation, the relative rise was smaller than for Glu and the degree of depolarization causing maximal release was lower. The questions of whether stimulus-coupled release of the non-transmitter amino acids from granule cells may play a neuromodulatory role in the cerebellum is discussed.     

Taurine-deficient cultured cerebellar astrocytes and granule neurons obtained by treatment with guanidinoethane sulfonate.

Mouse cerebellar granule neurons and astrocytes grown in the presence of 2 mM guanidinoethane sulfonate (GES) exhibited a progressive and rapid decrease in taurine concentration. A reduction of 20% was observed as early as 1 hr after exposure to GES and the loss of cell taurine continued until the taurine pool was reduced by about 90%. This remaining taurine persisted without further decrease even after 3 weeks of exposure to GES. Taurine reduction caused by GES was similar in both types of cells. The effect of GES was dose-dependent, with significant decreases in taurine levels already detected at 100 microM. It was selective for taurine, since none of the other free amino acids were affected. Taurine depletion induced by GES was totally reversible. Intracellular taurine was not mobilized by GES. Taurine uptake in both astrocytes and granule neurons, examined at the taurine concentration present in the culture medium, was practically abolished by 2 mM GES. This approach represents an in vitro model of taurine depletion that may be useful to investigate the cell abnormalities responsible for the failure of differentiation and migration of granule cells and astrocytes observed in taurine-deficient cats.     

Potassium-stimulated release of taurine from cultured cerebellar granule neurons is associated with cell swelling

Effects of increased concentrations of potassium and of hyposmolar conditions on release of taurine were investigated in cerebellar granule neurons cultured from mice. It was found that increases in the external potassium concentration as well as decreases in osmolarity dose-dependently increased release of exogenously supplied [3H]-taurine and endogenous taurine from the neurons. The release of endogenous taurine elicited by a reduction of the osmolarity of the incubation media to 70% or 50% was much more pronounced than that of other amino acids, particularly glutamine, the release of which was not affected at all. The potassium-stimulated release of [3H]-taurine was strictly chloride dependent and it was inhibited by an increase of the osmolarity of the media as well as by 4,4'-diisothyocyanostilbene-2,2'-disulfonate (DIDS) (100 microM). Moreover, a similar increase in the potassium concentration led to an increase in intracellular volume (swelling), a process which was also chloride dependent. It is concluded that potassium-stimulated taurine release from cerebellar granule neurons is associated with cell volume changes and that taurine is likely to play a role as an osmotically active substance in these neurons.     

Autoradiographic localization and depolarization-induced release of acidic amino acids in differentiating cerebellar granule cell cultures

Granule cells from 8-day-old rat cerebella were grown in basal Eagle's medium with 10% fetal calf serum, for 2,5,8 or 12 days in vitro (DIV), in conditions giving a purity > 90%. The results obtained can be summarized as follows: (1) Light microscopic autoradiography showed that cultured granule cells and their processes can accumulate the glutamate analog [³H]d-aspartate once they have reached an advanced degree of morphological differentiation (8 and 12 DIV), but, even then, only a limited number of cells washeavily labeled. In contrast, astrocytes were heavily labeled at all stages. (2) Calcium-dependent, high [K⁺]-induced release, or tetrodotoxin-sensitive, veratridine-induced release of [³H]d-aspartate from granule cell-enriched cultures was detectable only in cultures of 8 or 12 DIV. (3) When subject to 3 consecutive depolarizations, cultured granule cells maintained their ability to release [³H]d-aspartate and endogenous glutamate almost unchanged. (4)_Newly synthesized [³H]glutamate was autoradiographically localized in both neurons and astrocytes (the latter, however, were not preferentially labeled as with [³Hd-aspartate), but was specifically release from neuronal structures (perikarya and processes) by depolarizing stimuli.     

Thiophene is toxic to cerebellar granule cells in culture after bioactivation by rat liver enzymes

Several compounds that are not neurotoxic by themselves can cause toxic effects in vivo after enzymatic bioactivation. Thiophene is an industrial solvent known to produce degeneration primarily of the granule cells in the cerebellum when administered to animals in vivo. The mechanism for thiophene toxicity is not known, although it has been suggested that thiophene metabolism may lead to formation of oxidative intermediates that could function as the ultimate toxicants. In the present work we have used rat cerebellar granule cells (CGCs) in culture combined with rat liver postmitochondrial (S9) fraction as a source of biotransformation enzymes to test the toxicity of thiophene in vitro. The results demonstrate that thiophene is toxic to rat cerebellar granule cells in culture only after biotransformation. Furthermore, the toxic effects were reduced by cytochrome P450 (CYP) inhibitors and by scavengers of reactive molecules (alpha-tocopherol, reduced glutathione, and phenyl-N-tert-butylnitrone). These findings support the hypothesis that thiophene requires metabolism to produce the ultimate toxicant, and that the cytochrome P450 enzyme system is involved in the metabolism.     

Maturation of cerebellar granule cells is delayed in cultures derived from ethanol-treated chick embryos: Survival and proliferation studies

Previous studies from this laboratory have shown that ethanol administration to chick embryos during embryonic days 1–3, a critical period of neuroembryogenesis, differentially affects primordial CNS structures. In this study, chick embryos were treated in ovo with ethanol (10 mg/50 μl/day) at E1 to E3. At 14 days of embryonic age cerebellar (E14CE) granule cell cultures were prepared from both control and ethanol-treated embryos. Growth patterns were evaluated morphologically and the neuronal nature of these cultures was evaluated immunocytochemically. E14CE granule cell cultures exhibited neurofilament immunoreactivity demonstrating the neuronal-nature of these cultures. In addition E14CE granule cultures contained numerous glutamateric neurons as assessed by positive glutamate immunoreactivity and also some GABAergic neurons as assessed by positive GABA immunoreactivity. Cultures derived from both control and ethanol-treated embryos were labeled with ³H-thymidine and assessed for effects on survival and proliferation in culture. Cultures derived from ethanol-treated embryos showed a higher rate of proliferation and survival during the first 3 days in culture as compared to those derived from controls. However, after 3 days in culture, survival was lower in the cultures from ethanol-treated embryos as compared to those derived from controls. We interpret these findings to mean that (a) ethanol arrested cerebellar granule cell development at an immature state; (b) immature neurons have a higher survival capacity than differentiated neurons; and (c) ethanol accelerates normal neuronal cell death as previously reported.     

Glutamine transport in cerebellar granule cells in culture

In the present study, uptake of glutamine by rat cerebellar granule cells, a predominantly glutamatergic nerve cell population, has been investigated. Glutamine is taken up by granule cells via at least three transport systems, A, ASC and L. The L-type low affinity system (K_m=2.6 mM) is the major transport system in the absence of Na⁺. The systems A and ASC represent the Na⁺-dependent transport routes, both with almost identical high affinity for glutamine (K_m=0.26 mM). Similar transport systems for glutamine are also found in cerebral cortical neurons, a predominantly GABAergic nerve cell population, and cerebral cortical astrocytes. The glutamine transport properties in granule cells, however, show a series of differences from that of cortical neurons and astrocytes: (1) uptake of glutamine by granule cells is primarily mediated by system A (54%), while contributions by system A in cortical neurons and astrocytes are less than 30%; (2) granule cells exhibit strikingly higher transport efficiency for glutamine (V_max/K_m=20 min⁻¹ for system A as compared to the V_max/K_m ratio of 5 min⁻¹ in cortical neurons and astrocytes), and (3) the initial uptake rates and the steady-state accumulation levels of glutamine are two- to threefold higher in granule cells than that of cortical neurons and astrocytes. These results taken together suggest that in accordance with the important need to replenish the neurotransmitter pool of glutamate, glutamatergic neurons exhibit highly efficient transport systems to accumulate glutamine, one of the major precursors of glutamate.     

Ethanol-induced increase of sphingosine recycling for ganglioside biosynthesis: a study performed on cerebellar granule cells in culture

The effect of ethanol on ganglioside metabolism was assessed in cultured rat cerebellar granule cells. Cells were incubated in the presence of tritiated serine or galactose, and the synthesis of radioactive gangliosides was followed. The rate of de novo biosynthesis of gangliosides labeled in the oligosaccharide moiety (deriving from tritiated galactose) was not affected by the presence of ethanol. On the contrary, the biosynthesis of gangliosides labeled in the ceramide long chain base moiety (deriving from tritiated serine), dramatically decreased in the presence of alcohol. These results suggest that the gap between the extent of the biosynthesis of lipid and polar portions observed in the presence of ethanol, is filled by an increased recycling of sphingosine produced from ganglioside degradation. This hypothesis was confirmed by pulse-chase experiments with GM1 ganglioside, tritiated in the sphingosine moiety, and following radiolabeled gangliosides deriving from its metabolic processing. In fact, the radioactivity carried by gangliosides whose labeling could derive exclusively (GD1b + GT1b) or partially (GD1a) from the recycling of catabolic radiolabeled sphingosine, dramatically increased in ethanol-treated cells during the chase period. Taken together, these results suggest that ethanol increases ceramide sphingosine recycling for ganglioside biosynthesis.     

Modulation of calcium entry and glutamate release in cultured cerebellar granule cells by palytoxin

A channel open on the membrane can be formed by palytoxin (PTX). Ten nanomolar PTX caused an irreversible increase in the cytosolic calcium concentration ([Ca(2+)](c)), which was abolished in the absence of external calcium. The increase was eliminated by saxitoxin (STX) and nifedipine (NIF). Calcium rise is secondary to the membrane depolarization. PTX effect on calcium was dependent on extracellular Na(+). Li(+) decreased the PTX-evoked rise in [Ca(2+)](c); replacement of Na(+) by N-methyl-D-glucamine (NMDG) abolished PTX-induced calcium increase. [Ca(2+)](c) increase by PTX was strongly reduced after inhibition of the reverse operation of the Na(+)/Ca(2+) exchanger, in the presence of antagonists of excitatory amino acid (EAA) receptors, and by inhibition of neurotransmitter release. PTX did not modify calcium extrusion by the plasma membrane Ca(2+)-ATPase (PMCA), because blockade of the calcium pump increased rather than decreased the PTX-induced calcium influx. Extracellular levels of glutamate and aspartate were measured by HPLC and exocytotic neurotransmitter release by determination of synaptic vesicle exocytosis using total internal reflection fluorescence microscopy (TIRFM). PTX caused a concentration-dependent increase in EAA release to the culture medium. Ten nanomolar PTX decreased cell viability by 30% within 5 min. PTX-induced calcium influx involves three pathways: Na(+)-dependent activation of voltage-dependent sodium channels (VDSC) and voltage-dependent calcium channels (VDCC), reverse operation of the Na(+)/Ca(2+) exchanger, and indirect activation of EAA receptors through glutamate release. The neuronal injury produced by the toxin could be partially mediated by the PTX-induced overactivation of EAA receptors, VDSC, VDCC and the glutamate efflux into the extracellular space.     

Pharmacological characterisation of voltage-sensitive calcium channels and neurotransmitter release from mouse cerebellar granule cells in culture

Using subtype-specific Ca-channel blockers, we have characterised the voltage-sensitive Ca²⁺ currents as well as neurotransmitter release from cultured mouse cerebellar granule cells. The whole cell version of the patch clamp technique was adapted to monitor the isolated Ca-channel currents. The currents were activated at potentials more positive than −40 mV and were composed of at least four pharmacological distinct components being sensitive to nifedipine (35%), ω-conotoxin GVIA (10%), and ω-agatoxin IVA (42%) corresponding to L-, N-, and P-channel-mediated currents. The insensitive fraction (13%) possibly represented R channels. High potassium-evoked release of ³H-D-aspartate was used as a model of synaptic release. These studies were performed at relatively mild stimulation conditions (30 mM K⁺, 0.4 mM Ca²⁺), and 85% of the evoked release was Ca²⁺ dependent as well as tetrodotoxin and Cd²⁺ sensitive. Nifedipine and ω-agatoxin IVA dose dependently (IC₅₀ values of 10 nM and 0.7 nM, respectively) blocked most of the release, whereas ω-conotoxin MVIIA (IC₅₀ = 5 nM) caused partial blockage. The results indicate that several subtypes of voltage-sensitive Ca channels are present in mouse cerebellar granule cells. Furthermore, the data suggest that L, N, and P channels act in concert in the neurotransmitter release process. J. Neurosci. Res. 48:43–52, 1997. © 1997 Wiley-Liss, Inc.     

GABA influences the ultrastructure composition of cerebellar granule cells during development in culture

The influence of GABA on cerebellar granule cells in culture was followed morphometrically during in vitro development by growing the cells in the absence or presence of 50 μM GABA. The presence of GABA in the culture media increased the number of neurite-extending cells by 50% after 7 days in culture. At the ultrastructural level GABA treatment led to an increased density of neurotubules, rough endoplasmic reticulum, Golgi apparatus, coated vesicles and other vesicles, whereas structures such as mitochondria and smooth endoplasmic reticulum were not affected by GABA. The density of free ribosomes showed a more pronounced tendency to decrease as a function of the culture period (1–7 days) when the cells were grown in the presence of GABA as compared to control cultures. The results strongly indicate that GABA in addition to being an important neurotransmitter serves as a trophic factor in the development of at least certain types of neurons.     

Effect of glutamine and GABA on [U-(13)C]glutamate metabolism in cerebellar astrocytes and granule neurons.

To probe the effect of glutamine and GABA on metabolism of [U-(13)C]glutamate, cerebellar astrocytes were incubated with [U-(13)C]glutamate (0.5 mM) in the presence and absence of glutamine (2.5 mM) or GABA (0.2 mM). It could be shown that consumption of [U-(13)C]glutamate was decreased in the presence of glutamine and release of labeled aspartate and [1,2,3-(13)C]glutamate decreased as well, whereas the concentrations of these metabolites increased inside the cells. Glutamine decreased energy production from [U-(13)C]glutamate presumably by substituting for glutamate as an energy substrate. No additional effect was seen in the presence of both glutamine and GABA. When cerebellar granule neurons were incubated with [U-(13)C]glutamate (0.25 mM) and GABA (0.05 mM), less [U-(13)C]glutamate was used for energy production than in controls. Because the barbiturate thiopental did not elicit such response (Qu et al., 2000, Neurochem Int 37:207-215) it appears that GABA also has a metabolic function in the glutamatergic cerebellar granule neurons in contrast to the astrocytes.     

Transport of 5-formyltetrahydrofolate into primary cultured cerebellar granule cells

Transport of 5-formyltetrahydrofolate (5-FTHF) into primary cultured cerebellar granule cells (CGC) was studied. Uptake of 5-FTHF into CGC was saturable with K(m)=2.86 microM and V(max)=40.8 pmol/mg protein/45 min in pH 7.4 medium. Uptake of 5-FTHF in the astrocytes has a similar style in the time curve. Uptake of 5-FTHF is characterized by countertransport because adding unlabeled 5-FTHF in the medium resulted in the efflux of labeled 5-FTHF. Uptake of 5-FTHF was inhibited by the structural analogs 5-methyltetrahydrofolate, methotrexate and folic acid (K(i)=6.64, 7.69, and 19.38 microM, respectively). Uptake was significantly decreased by high concentrations of sodium azide and sodium arsenate but not by sodium cyanide. Uptake was also inhibited by p-chloromercuriphenylsulfonate and by the anions probenecid and 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid. Acute exposure of the cells to ethanol (100 mM) did not affect the uptake. It is concluded that CGC have a carrier-mediated system for the uptake of 5-FTHF and other folates.     

Mechanisms of MPP incorporation into cerebellar granule cells

Exposure of cerebellar granule cells to 1-methyl-4-phenylpiridinium (MPP(+)) results in cell death. We have studied the implication of various membrane transporter systems on MPP(+) neurotoxicity, including the dopamine transporter system (DAT) and cationic amino acid transporters (CAT). We have showed a partial protection against MPP(+) toxicity when the dopamine transporter is inhibited by 1-[2-[bis(4-fluorophenyl)methoxy]ethyl]4-(3-phenylpropyl)piperazinedihydrochloride (GBR-12909). However, almost full protection is only achieved by the simultaneous addition of GBR-12909 and cationic amino acids. These results suggest two ways system of MPP(+) entrance into cerebellar granule cells: the DAT with high activity and the CAT with low activity. We also demonstrated that 5,7-dichlorokynurenic acid (MK-801) failed to protect against MPP(+) exposure, evidencing that N-methyl-D-aspartate (NMDA) receptor is not involved in the MPP(+)-induced cell death.     

Hyposmolarity-induced taurine release in cerebellar granule cells is associated with diffusion and not with high-affinity transport.

The effects of hyposmotic conditions on taurine uptake and release were studied in mice cultured cerebellar granule cells. The effect of DIDS (4,4'-diisothiocyanatostilbene-2,2'-disulfonate) and of the divalent cations Mg++ and Mn++ on the hyposmolarity-induced changes in these parameters was investigated. Mg++ (20 mM) and Mn++ (5 mM) inhibited by 25% and 41%, respectively, the release of taurine observed in 30% hyposmolar media. DIDS (100 microM) inhibited this release by 46%. Taurine efflux evoked by 50% hyposmolar solutions was reduced about 40% by Mg++ and 55% by Mn++. Taurine uptake into the granule cells could be resolved into a high-affinity carrier-mediated component plus a nonsaturable diffusion component. The kinetic constants (Km and Vmax) for the high-affinity uptake were unaffected by a 50% decrease in the osmolarity. The diffusion constant for the nonsaturable taurine uptake was increased from 1.5 x 10(-4) in isosmotic media to 4.6 x 10(-4) ml x min-1 x mg-1 in hyposmotic (50%) media. This increase in the diffusional component of taurine uptake elicited by the hyposmotic condition was inhibited approximately 25% by either 100 microM DIDS or 5 mM Mn++. These results strongly suggest that the increase in taurine release induced by swelling under hyposmotic conditions is mediated by a diffusional process and not by a reversal of the high-affinity taurine carrier.     

Effects of the marine phycotoxin palytoxin on neuronal pH in primary cultures of cerebellar granule cells

Palytoxin (PTX) is a potent marine phycotoxin that binds to the Na,K-ATPase, converting this pump into an open channel. We have recently shown (Vale et al., 2006) that PTX causes an irreversible increase in the cytosolic calcium concentration ([Ca(2+)](c)) in primary cultures of cerebellar granule cells (CGC). In this work, we investigated the effect of PTX on the intracellular pH (pH(i)) in the same cellular model. PTX-induced changes in pH(i) were studied in CGC by using the fluorescent probe 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein acetoxymethyl ester (BCECF-AM). PTX caused an irreversible intracellular acidification of CGC. This acidification was due to an influx of extracellular calcium, inasmuch as it was completely abolished by the use of Ca(2+)-free medium. Different mechanisms that could be involved in the PTX-induced pH(i) decrease such as displacement of H(+) by Ca(2+) from a common intracellular binding site, PTX-induced alteration of pH(i) regulation mechanisms, and a possible acidification caused by an increase of mitochondrial Ca(2+) uptake by PTX were excluded. PTX-induced intracellular acidification was completely prevented by several inhibitors of the plasma membrane calcium ATPase (PMCA), including orthovanadate, lanthanum, high extracellular pH, and caloxin 2A1. Our results indicate that the PMCA is involved in the PTX-induced intracellular acidification in primary cultures of CGC. The PTX-evoked increase in [Ca(2+)](c) will activate the calcium extrusion mechanisms through the PMCA, which, in turn, will decrease pH(i) by countertransport of H(+) ions. The effect of PTX on neuronal pH could be a potential factor to contribute to the high cytotoxicity of this toxin in cultured cerebellar neurons.     

Activation of caspase-3 and apoptosis in cerebellar granule cells

Caspase-3 activity increased dramatically in cytosolic extracts of rat cerebellar granule cells exposed to apoptotic conditions (basal medium Eagle (BME) containing 5 mM K⁺ without serum) when assayed with Ac-DEVD-amc,but not with Ac-YVAD-afc, a preferred substrate for caspase-1. This provided a basis to examine relationships between enzyme activity and cell viability for purposes of selecting an optimal time for comparing neuroprotective agents or strategies. Exposure of neurons to an apoptotic medium containing 5 mM K⁺ in absence of serum led to a rapid 5- to 10-fold increase in caspase-3 within 2-4 hr but without significant cell loss, or morphological alterations. Exposure to apoptotic medium followed by replacement with maintenance medium containing 25 mM K⁺and serum led to a rapid fall in caspase-3 and prevention of cell death. This strategy was not effective after 13 hr exposure despite a large fall in enzyme activity. These temporal changes infer systems for rapid enzyme turnover and/or activation of cytoplasmic components linked to later DNA degradation. The effects of cycloheximide point to requirements for protein synthesis, and those of Glu exclude a caspase-3 dependent pathway for necrotic cell damage. Brief treatment with 10 μM LIGA₂₀, an anti-necrotic agent, also attenuated cell loss and caspase-3 activity, indicating a broad spectrum of neuroprotection. Rapid and long-lasting effects, together with its biophysical properties, suggest that this semisynthetic ganglioside acted upstream at or near a membrane site. As such, gangliosides provide useful agents to further probe pathways relevant to neuronal death in culture. J. Neurosci. Res. 52:334–341, 1998. © 1998 Wiley-Liss, Inc.     

Distribution of nuclear pores and chromatin organization in neurons and glial cells of the rat cerebellar cortex

Nuclear pores were assessed on freeze-fracture replicas from different neuronal and glial cell types of the rat cerebellar cortex. Nuclear diameter and perimeter were measured on semithin sections, and nuclear surface area and volume were calculated from these data. The proportion of inner nuclear membrane in apposition to condensed chromatin was measured on thin sections. The values of nuclear pore numerical density (number/micron2) were as follows (mean +/- S.D.): Purkinje cells, 22 +/- 3; Golgi cells 17 +/- 3; granule cells, 6 +/- 4; stellate and basket cells, 6 +/- 1; protoplasmic astrocytes, 11 +/- 1; Bergmann glia, 10 +/- 1; oligodendrocytes, 6 +/- 1. The total number of nuclear pores per nucleus varied from 18,451 +/- 2,336 (Purkinje cells) to 621 +/- 394 (granule cells) among neurons, and from 1,782 +/- 162 (protoplasmic astrocytes) to 402 +/- 67 (oligodendrocytes) among glial cells. The number of nuclear pores per unit nuclear volume (number/micron3), a parameter related to nucleocytoplasmic transport capacity, varied from 15 +/- 2 in Purkinje cells to 6 +/- 4 in granule cells. The proportion of nuclear membrane free of condensed chromatin was significantly (P less than 0.01) correlated to pore numerical density and total number of pores per nucleus. Some nuclear pores were associated in clusters of two or more pores. The amount of pore clustering was measured by counting the proportion of pores associated in clusters. This proportion varied among the different cell types from 82% in Purkinje cells to 44% in stellate and basket cells. The amount of pore clustering showed a positive linear correlation to pore numerical density and pore number per nucleus. However, the proportion of pores in clusters was not significantly correlated with the amount of condensed chromatin applied against the inner nuclear membrane.