The Journal of Neuroscience, June 7, 2006, ():
Insulin-Like Growth Factor-1 Modulation of CaV1.3 Calcium Channels Depends on Ca2+ Release from IP3-Sensitive Stores and Calcium/Calmodulin Kinase II Phosphorylation of the 
1 Subunit EF Hand
J. Neurosci. Gao et al.
26: 6259
Supplemental data
Files in this Data Supplement:
- supplemental material
-
Supplemental Figure 1. RNAi knock-down of CaV1.3 strongly reduces CaV1.3 without affecting other calcium channel subtypes.
A,B, Total CaV1.3 levels (surface + intracellular) in cortical pyramidal neurons transfected with RNAi to CaV1.3 (arrowheads, "GFP+ neuron") were strongly reduced as compared to non-transfectants ("GFP- neuron") on the same coverslip. Specifically, neurons transfected with RNAi were identified by GFP co-expression (left panels), and an antibody against an intracellular epitope and that does not distinguish between CaV1.3 isoforms was used to assess total CaV1.3 levels immunocytochemically (center, right panels). We tested a short hairpin RNAi specific for the long isoform, CaV1.3a (A, "RNAi-1.3a") and a pan-specific RNAi expected to decrease expression of both long and short isoforms (B, "RNAi-1.3pan"). Both RNAi clearly reduce total CaV1.3 levels. N= 5 independent experiments, each performed in triplicate.
C, CaV1.3 currents in the soma of cortical neurons were evident in the absence of RNAi (C1), but undetectable in neurons expressing RNAi-1.3pan (C2). Recordings were performed in the presence of 1uM nimodipine to inhibit CaV1.2 activity as well as a cocktail of additional inhibitors to block other voltage-sensitive channels (see Materials and Methods). N= 17 neurons tested for control CaV1.3 currents levels (C1), and 8 neurons tested for CaV1.3 currents levels in the presence of RNAi-1.3pan (C2). Shown here are currents evoked by depolarization to 10mV because CaV1.3 currents are expected to be maximal at approximately this potential: However, in RNAi-1.3pan expressing neurons, no CaV1.3 activity can be detected.
D, Identical results are obtained when SH-SY5Y neuroblastoma cells are transfected with CaV1.3a in the absence or presence of RNAi-1.3a. Test potential, 10mV. Shown here are the results from 2 different cells, with the red arrow indicating the control CaV1.3a current level ("No RNAi") and the blue-green arrow indicating the near absence of CaV1.3a current in a cell co-expressing CaV1.3a with RNAi-1.3a. Cumulative results: the control CaV1.3a current level was 108+10 pA/pF, while the CaV1.3a level in the presence of RNAi-1.3a was almost 100-fold lower (1.4+0.4 pA/pF). Indistinguishable results were obtained when CaV1.3b was expressed with and without RNAi-1.3pan: 224+23 pA/pF (control) vs. 4.0+1.5 pA/pF (+RNAi-1.3pan). N> 8 cells for each group.
E, RNAi to CaV1.3 does not prevent expression of functional CaV1.2 channels in cortical neurons (see "1"). Furthermore, as expected from our earlier studies (Blair and Marshall, 1997; Blair et al., 1999; Bence-Hanulec et al., 2000), IGF 1 is able to rapidly potentiate the CaV1.2 activity (see "2"). Moreover, subsequent addition of 1uM nimodipine, a concentration that blocks CaV1.2 with almost no effect on CaV1.3 (Lipscombe et al., 2004), inhibited all detectable activity (see "3"). Shown here are CaV1.2 currents from a single neuron, recorded at 10mV before ("1"), 30 seconds after adding 20ng/ml IGF 1 ("2"), and 2 minutes after adding 1uM nimodipine ("3"). N= 6 neurons tested.
F, Similarly, CaV1.2 channels express well in SH-SY5Y neuroblastoma cells regardless of the presence of RNAi to CaV1.3. Shown here are the results from 2 different cells, with the red arrow indicating the control CaV1.2 current level ("No RNAi") and the blue-green arrow indicating that co-expressing CaV1.2 with RNAi-1.3a yields almost identical CaV1.2 currents. Cumulative results: CaV1.2 current level (no RNAi) was 138+40 pA/pF, and the CaV1.2 level in the presence of RNAi-1.3a was 147+29 pA/pF; test potential, +10mV. N= 5 cells for each group.
G, Large N-type calcium channel currents are detected in cortical neurons transfected with RNAi to CaV1.3. Test potential, +10mV. Here, N currents were recorded in the presence of 1uM nimodipine to block CaV1.2 activity and in the absence of ω conotoxin GVIA. Other voltage-gated channels were inhibited as described in the Materials and Methods.
- supplemental material
-
Supplemental Figure 2. Epitope-tagged CaV1.3a and S1486A-CaV1.3a expression in pyramidal neurons.
Upper panels, CaV1.3-transfected cortical pyramidal neurons from sister cultures were identified by GFP co-expression.
Middle panels, Surface CaV1.3 levels in transfectants were assessed immunocytochemically using an antibody against the extracellular HA epitope.
Lower panels, Overlay of CaV1.3a staining onto the GFP-identified neurons.
A, Wild-type (wt) HA-tagged CaV1.3a is found at the cell surface in somata and dendrites with relatively high levels in the dendritic shafts and spines.
B, HA-tagged S1486A-CaV1.3a is also on the surface of somata and dendrites.
- supplemental material
-
Supplemental Figure 3. Co-localization of surface CaV1.3a with Shank-1a requires the C-terminus of CaV1.3a and an interaction with the Shank SH3 domain.
HA-tagged CaV1.3a (wild-type or with the carboxyl terminus deleted) was co-transfected into COS-1 cells with Shank-1a (wild-type or with a W520A point mutation in the SH3 domain) and, to identify transfectants, BFP. Two days post-transfection, the extent of co-localization of Shank-1a with CaV1.3a expressed on the cell surface was assessed.
A, Surface wild-type (wt) CaV1.3a strongly co-localizes with wild-type (wt) Shank-1a. Overlaying the surface wt-CaV1.3a (red) and total wt-Shank-1a (green) immunolabelling yields a large number of small and large overlapping (yellow) puncta.
B, In contrast, deletion of the CaV1.3a C-terminus (expected target of the Shank-1a PDZ domain) eliminates co-localization of wild-type Shank-1a with surface CaV1.3a-DEL. Puncta of CaV1.3a-DEL fail to overlap with wt-Shank-la staining.
C, Disruption of the Shank-1a SH3 domain also disrupts co-localization: Co-expressing W520A-Shank-1a with wild-type CaV1.3a again results in poor co-localization.
