Study Of Relationship Between Long Splice Variant N Type Calcium Channel And CASK

The N type Ca channel (CaV2.2) is known to play a critical role in the gatingof transmitter release at presynaptic nerve terminals. Physiological, morphologicaland ultrastructural findings suggest that these channels are clustered in closeapposition to the transmitter release apparatus. CaV2.2 is known to exist in both short and long splice variant forms and thelong variant, termed here CaV2.2-L, has an extended C terminal region with anadditional 17 amino acids. Based on the targeting of short N terminal peptideconstructs or hemagglutinin-tagged full length channel constructs in hippocampalcultures, it has been reported that CaV2.2-L is targeted preferentially to thepresynaptic terminal. This is particularly interesting as the extended tail regionincludes binding sites for two 'modular adaptor' proteins, CASK and MINT.Mutational deletion of these sites reduced, but did not eliminate, the presence ofCaV2.2 channels at synapses in the neuron cultures. These studies providescompelling evidence that these two modular adaptor proteins play an importantrole in the targeting of N type Ca channels to the presynaptic terminal. The above findings have also been used to argue that that MINT and CASKplay a critical role in the anchoring of CaV2.2-L at its destination.The evidencefor channel attachment by these proteins is not, however, as compelling as that fortargeting. CASK and MINT can bind the CaV2.2-L and are present at presynapticnerve terminals .In addition, suppression of CASK and MINT by expressionconstructs markedly interferes with transmitter release.As assessed by reviews onthe subject, an anchoring role for the modular adaptor proteins has gained wideacceptance. However, a direct association of these proteins at the release site hasnot been confirmed while interference in channel transport alone could account forthe functional consequences on transmitter release.We set out to test two main hypotheses: first, that CaV2.2-L is targeted to thetransmitter release sites at a native and intact synapse and, second, to test if CASKand MINT are candidate CaV2.2-L anchors at the transmitter release face. Forthis we developed two new high-affinity antibodies, L4569 and L4570, directedagainst a unique region within the long splice C terminal of CaV2.2-L. We usedimmunocytochemistry to localize the channels at the chick ciliary ganglioncalyx-type presynaptic terminal. We compared the location of immunostainedproteins by a novel quantitative coimmunostaining analysis method, IntensityCorrelation Analysis (ICA), that can be used to test not only if the two targetproteins are colocalized but if their staining intensities vary in synchrony, asexpected for two proteins that are components of the same complex or sub-cellularstructure. We report that CaV2.2-L is, as predicted, localized to the transmitterrelease face and, based on costaining with the marker RIM, is associated with thetransmitter release sites.Reslut: A Western blot of chick brain lysate was probed with the two newantibodies, L4569 and L4570. Both antibodies identified a-250 kD band whereasthe control pre-immune-rabbit IgG band was negative. Further, Ab571, a wellcharacterized CaV2.2 antibody which is targeted to a very different region on theII-III loop, identified a similar-250 kD band on a Western blot of proteinsimmunoprecipitated by L4569 or L4570 from chick brain lysate. This confirmsthat these antibodies share a common target of the appropriate molecular weightfor CaV2.2. We conclude that both L4569 and L4570 are directed against thelong-splice region of CaV2.2 and both can be used for Western blot orimmunoprecipitation experiments.Calyx terminals were costained withanti-CaV2.2-L (L4569) and anti-pan-CaV2.2 (Ab571) using the poly-poly methodfor double rabbit polyclonal antibody staining. We used the standard coloredimage overlay method, generating a mixed color where the two psuedostainedimages dyes coexist, to provide an initial impression of colocalization. The ICAplots for both CaV2.2-L and the pan-CaV2.2 stains were very strongly skewed tothe right. The mean ICQ was 0.28±0.03 (p=0<0.002), which is strongly positivewithin the theoretical -0.5 to +0.5 range. Thus, the staining intensities forCaV2.2-L and pan-CaV2.2 vary in close synchrony, consistent with two antibodiesthat are bound to the same protein We next tested whether the CaV2.2-L punctaare associated with the transmitter release site by costaining with RIM, one of thefew proteins used as a specific release site marker. CaV2.2-L staining colocalizedwith RIM as assessed both dye overlay and ICQ analysis.CASK has previously been reported to be present at the presynaptic terminalCASK coprecipitated with MINT from chick brain lysates. We next tested for thiscomplex at the chick presynaptic calyx terminals. The protein was present mostlyin patches, both within the cytoplasm and at the surface membrane. Terminalsstained for MINT with a similar broad punctuate distribution. An overlay ofMINT and CASK staining gave bright costained puncta as well as large areaswhere the stains were segregated. This staining distribution generated complexscatter plots in the ICA analysis. The ICQ value, which effectively averages thevarious relationships, was low but highly significantly. consistent with thepresence of a MINT/CASK complex within the nerve terminal.Immunoprecipitation of native CaV2.2-L using a battery of specific antibodiesconsistently co-precipitated CASK, confirming that these two proteins can exist ina complex. However, immunoprecipitation analysis only demonstrates that twoproteins can form a complex not, in fact, whether they do so in situ nor where thiscomplex might be located. In order to test if the two proteins are associated at thepresynaptic transmitter release face we repeated the coimmunostaining analysis.The mean ICQ values were not significantly different from 0 for either of the twoCASK antibodies (Chemicon anti-CASK: 0.05±0.02, p=0>0.1;N=7;BDanti-CASK: 0.04±0.04, p=0>0.1, N=5). Thus, the ICA analysis indicates that theintensity for CaV2.2-L staining co-varies very weakly, if at all, with CASK at thetransmitter release face. The similar symmetrical ICA plots for both proteins argueagainst the possibility of a one-sided relationship. That CASK is widelydistributed with and without CaV2.2-L whereas CaV2.2-L is associated with thisprotein. Thus our findings suggest that CASK and CaV2.2-L are not colocalized atthe nerve terminal transmitter release face. We did, however, note prominentyellow-colored puncta within the nerve terminal cytoplasm, These puncta mayreflect colocalized CASK and CaV2.2-L associated with the same transportvesicles, we costained calyx terminals for CASK and the release site marker, RIM.The results were consistent with the Ca channel results above.A scaffold protein responsible for long term maintenance of a target proteinmust satisfy two minimal criteria: first, it must interact directly or indirectly withthe target protein in a molecular complex and, second, the two proteins must becolocalized at the cellular region of interest. Thus, to test whether CASK andMINT are candidate CaV2.2-L release site anchors we first carried out bindingstudies to see if these proteins associate with CaV2.2-L. We then used quantitativeimmunocytochemistry to test if the two proteins colocalize with CaV2.2 and/orRIM at the nerve terminal transmitter release face.Our results support the firsthypothesis since immunoprecipitation of CaV2.2-L coprecipitated both CASK andMINT. However, we failed to support the second hypothesis since neither proteincolocalized with either the channel or RIM. The latter finding argue against thehypothesis that the modular adaptor proteins serve as N type Ca channel anchors atthe transmitter release site.