A culture system of the postnatal rat retina was established to invest
igate Ca2+ currents and synaptic transmission in identified neurons. M
ethods are described that allowed us to select retinal ganglion neuron
s (RGNs) in short term cultures (up to 48 h in vitro) and in long-term
cultures (3 to 21 days in vitro). The specific aim of the present stu
dy was to identify channel specific components in whore-cell Ca2+ curr
ents of RGNs and to clarify the potential use of the lanthanide Gd3+ a
s a selective Ca2+ channel blocker. About one third of freshly dissoci
ated RGNs generated both low voltage activated Ca2+ currents (I-Ca(LVA
)) and high voltage activated Ca2+ currents (I-Ca(HVA)). The remaining
2/3 of RGNs in short term culture and most RGNs in long-term culture
displayed only I-Ca(HVA). The latter comprised at least three differen
t components that were functionally rather similar, but could be separ
ated pharmacologically. A significant portion (about 40%) of I-Ca(HVA)
was irreversibly blocked by the N channel antagonist omega-CgTx(5 mu
M). The L channel antagonist nifedipine (10 mu M) eliminated about 25%
of I-Ca(HVA). Thus, about 1/3 of the HVA Ca2+ or Ba2+ current remaine
d unaffected by either omega-CgTx or nifedipine. omega-AgaTx (200 nM)
completely failed to block HVA Ca2+ or Ba2+ currents in RGNs. Gd3+ exe
rted contrasting actions on LVA and HVA Ca2+ currents. While I-Ca(LVA)
consistently increased in the presence of Gd3+ (0.32-3.2 mu M), I-Ca(
HVA)) always decreased, especially when using higher concentrations of
Gd3+ (10-32 mu M) The blocking action of Gd3+ was not restricted to t
he omega-CgTx-sensitive HVA current component, but also concerned omeg
a-CgTx- and nifedipine-resistant components. The decay of Ca2+ current
s was accelerated in the presence of Gd3+. Even in RGNs lacking I-Ca(L
VA), application of 3.2 mu M Gd3+ significantly reduced the time const
ant of decay from an average of 64 ms to 36 ms (voltage steps from -90
to O mV; 10 mM [Ca2+](o);26 degrees C). This is in contrast to what h
ad to be expected if an N-type HVA current component was selectively s
uppressed by Gd3+. Gd3+ diminished glutamatergic spontaneous synaptic
activity in retinal cultures tested during the 3rd week in vitro. Both
frequency and amplitude were reduced. Occasionally, the application w
as followed by a rebound increase of EPSC frequency. A stimulatory eff
ect during application of Gd3+ has never been observed. These experime
nts indicate that RGNs express at least 4 different types of Ca2+ curr
ents, that resemble in some aspects T, N and L channel currents. A sig
nificant component of the HVA Ca2+ current was resistant to the availa
ble HVA channel blockers suggesting the presence of a pharmacologicall
y distinct type of HVA Ca2+ channel type in RGNs. Our experiments also
show that Gd3+ is not suitable for isolation of HVA subcomponents in
RGNs, but it can be used to distinguish between LVA and HVA Ca2+ curre
nts, as these currents reacted to Gd3+ in an opposite way. The purely
depressive effect of this lanthanide on spontaneous synaptic activity
is consistent with the assumption that in retinal neurons LVA Ca2+ cha
nnels are not involved in the regulation of glutamate release.