Surface L-type Ca2+ channel expression levels are increased in aged hippocampus

Age-related increase in L-type Ca2+ channel (LTCC) expression in hippocampal pyramidal neurons has been hypothesized to underlie the increased Ca2+ influx and subsequent reduced intrinsic neuronal excitability of these neurons that lead to age-related cognitive deficits. Here, using specific antibodies against Cav1.2 and Cav1.3 subunits of LTCCs, we systematically re-examined the expression of these proteins in the hippocampus from young (3 to 4 month old) and aged (30 to 32 month old) F344xBN rats. Western blot analysis of the total expression levels revealed significant reductions in both Cav1.2 and Cav1.3 subunits from all three major hippocampal regions of aged rats. Despite the decreases in total expression levels, surface biotinylation experiments revealed significantly higher proportion of expression on the plasma membrane of Cav1.2 in the CA1 and CA3 regions and of Cav1.3 in the CA3 region from aged rats. Furthermore, the surface biotinylation results were supported by immunohistochemical analysis that revealed significant increases in Cav1.2 immunoreactivity in the CA1 and CA3 regions of aged hippocampal pyramidal neurons. In addition, we found a significant increase in the level of phosphorylated Cav1.2 on the plasma membrane in the dentate gyrus of aged rats. Taken together, our present findings strongly suggest that age-related cognitive deficits cannot be attributed to a global change in L-type channel expression nor to the level of phosphorylation of Cav1.2 on the plasma membrane of hippocampal neurons. Rather, increased expression and density of LTCCs on the plasma membrane may underlie the age-related increase in L-type Ca2+ channel activity in CA1 pyramidal neurons.


Introduction
The calcium hypothesis of aging (Khachaturian, 1987;Landfield, 1987) posits that age-related cognitive deficits are mainly due to changes in cellular mechanisms that maintain and regulate intracellular Ca 2+ homeostasis. Among them, change in Ca 2+ channel number and/or function has been suggested to be a key factor (Khachaturian, 1994). While age-related increase in function of L-type Ca 2+ channels (LTCCs) in CA1 pyramidal neurons (Moyer & Disterhoft, 1994;Thibault & Landfield, 1996) and rescue of normal aging-and Alzheimer's disease-related cognitive deficits with LTCC antagonists have been demonstrated (Deyo et al., 1989;Ban et al., 1990), there is conflicting evidence for altered number of LTCCs with aging. Increased (Herman et al., 1998;Chen et al., 2000;Veng & Browning, 2002), no change (Blalock et al., 2003;Kadish et al., 2009), and reduced (Rowe et al., 2007) expression levels of the central pore-forming a 1 -subunits of the L-type Ca 2+ channels Ca v 1.2 and Ca v 1.3 in hippocampus from aged animals have been reported. These apparently conflicting findings may be due to the level of analysis conducted: from single cell to whole hippocampus. Therefore, we systematically examined the expression levels of Ca v 1.2 and Ca v 1.3 using Western blot, immunohistochemistry, and real-time quantitative PCR analysis in the three major hippocampal regions of young and aged rats.
While these results demonstrate that higher levels of LTCCs are present on the cell membrane, they did not provide the location of the increased surface expression. Therefore, we performed immunohistochemical analysis to identify the locus of the increase.
Ca v 1.2 immunoreactivity (Ca v 1.2-IR) is increased in somatic region of aged CA1 and CA3 neurons CA1 pyramidal neurons from aged subjects have been shown to have increased LTCC activity (Thibault & Landfield, 1996) and enhanced calcium action potentials (Moyer & Disterhoft, 1994). Therefore, we postulated that increases in LTCC subunit expression would be observed mostly in the somatic region of aged CA1 pyramidal neurons.
We observed Ca v 1.2-IR within the hippocampal formation and hippocampal cell layers similar to previous reports (Hell et al., 1993;Clark et al., 2003;Hall et al., 2013). Significant increases in Ca v 1.2-IR were observed in the somatic regions of aged CA1 (Fig. 4) and CA3 (Fig. 5) pyramidal neurons. No change in Ca v 1.2-IR and expression was observed in DG granule cells (Fig. 6). Furthermore, no significant changes in Ca v 1.2 subunit expression were observed in stratum radiatum of CA1 (Fig. 4C) or CA3 (Fig. 5C) hippocampal region.
In addition, as LTCCs are found in glial cells and its expression has been documented to change with astrocyte activation after brain injury, trauma, and aging (Wisniewski & Terry, 1973;Vaughan & Peters, 1974;MacVicar, 1984;Westenbroek et al., 1998;Chung et al., 2001;Djamshidian et al., 2002;Finch, 2003;Xu et al., 2007), we also examined whether the observed plasma membrane increases in LTCC subunits with aging were of glial/astrocytic origin. No detectable expression/colocalization of LTCC subunits was observed in glial cells with our antibodies .
Parallel to Ca v 1.2, immunohistochemical experiments were conducted to assess changes in Ca v 1.3 protein expression at the cellular level. However, the presence of various nonspecific, high-intensity bands in our Ca v 1.3 Western blots (Fig. S2B,D) prevents us from confidently reporting our Ca v 1.3 immunohistochemical findings, as the obtained Ca v 1.3 immunoreactivity might be the results of nonspecific binding of our current antibody in brain tissue. Hence, it remains to be determined whether immunohistochemical analyses of Ca v 1.3 expression at the cellular level are consistent with our Ca v 1.3 biotinylation experiments. Fig. 1 Characterization of antibody specificity for Ca v 1.2 and Ca v 1.3 proteins. Hippocampal lysates from wild-type (WT) and L-type-deficient (KO) mice were resolved by SDS-PAGE and immunoblotted with either CNC1 (J.H: Johannes W. Hell), ab144 (A.L: Amy Lee), commercially available anti-Ca v 1.2 (Alo: Alomone Labs, ACC-003; NM: Neuromab Antibodies Inc. L57/46,) or commercially available anti-Ca v 1.3 (Alo: Alomone Labs, ACC-005; NM: Neuromab Antibodies Inc. N38/8) antibodies. Blots were developed using Amersham ECL Plus and Hyperfilm ECL. Both anti-Ca v 1.2 and anti-Ca v 1.3 antibodies from commercial sources revealed nonspecific bands in hippocampal lysates from KO tissue. CNC1 and ab144 showed no cross-reactivity with either Ca v 1.3 or Ca v 1.2 proteins in hippocampal lysates. Note that this example figure is assembled from multiple blots with similar exposure time that have been aligned for illustrative purposes only. See Fig. S1 for immunoblots as loaded in gel.

Phosphorylation of surface-expressed Ca v 1.2 is increased in the DG of aged rats
Increased LTCC activity by 4-6-fold has been reported when the Ca v 1.2 subunits are phosphorylated (Sculptoreanu et al., 1993;Kavalali et al., 1997). In addition, Serine 1928 (S1928) (Davare & Hell, 2003) and Serine 1700 (S1700) (Fuller et al., 2010) of Ca v 1.2 can be phosphorylated, but only S1928 phosphorylation has been shown to be increased with normal aging (Davare & Hell, 2003). Therefore, we further explored the possibility that more S1928 in Ca v 1.2 might be phosphorylated in the hippocampal regions from aged rats. Biotinylated plasma membrane proteins were isolated and immunoblotted with anti-CH1923-1932P (p1928), an antibody designed to specifically detect Ca v 1.2 when phosphorylated at S1928 (De Jongh et al., 1996;Davare & Hell, 2003). We found a significant 1.18-fold increase in phosphorylated Ca v 1.2 in DG of aged rats (Fig. 7B). No significant age-related difference in phosphorylated Ca v 1.2 was detected in either CA1 or CA3.
Ca v 1.2 and Ca v 1.3 mRNA levels are unchanged with aging A positive correlation between Ca v 1.3 mRNA and LTCC activity has been previously demonstrated in CA1 pyramidal neurons using single cell reverse transcription PCR experiments . Therefore, we performed a systematic assay to assess Ca v 1.2 and Ca v 1.3 mRNA levels in each hippocampal region using the real-time quantitative PCR method to determine whether the mRNA levels were altered with normal aging and/or in a specific hippocampal region(s). We found no significant age-related changes in mRNA levels for both Ca v 1.2 and Ca v 1.3 (Table 1, Fig. S4).

Discussion
The present study is the first to demonstrate that the pore-forming a 1 subunits for the L-type voltage-gated Ca 2+ channels (Ca v 1.2 and Ca v 1.3) are significantly reduced in whole tissue lysates from all three major hippocampal regions of aged rats (Fig. 2). However, the biotinylation and immunohistochemical data demonstrate that age-related increases in Ca v 1.2 are observed in CA1 and CA3 regions. Notably, the increase in the Ca v 1.2 subunit was in the somatic region of CA1 and CA3 pyramidal neurons (Figs 4 and 5). In addition, no detectable expression/colocalization of Ca v 1.2 subunits was observed in glial cells with our antibodies (Figs 4-6). Therefore, the present results support the 'calcium hypothesis of aging' (Khachaturian, 1987;Landfield, 1987) in that the age-related increase in surface expression of L-type voltage-gated Ca 2+ channels (Ca v 1.2 and Ca v 1.3) in hippocampal pyramidal neurons we demonstrate may play an important role in the cognitive deficits observed in normal aging subjects. Phosphorylation of Ca v 1.2 LTCC has been previously shown to enhance Ca 2+ influx (Sculptoreanu et al., 1993;Kavalali et al., 1997). Ca v 1.2 a 1 subunit can be phosphorylated at Serine 1928 (S1928) (Davare & Hell, 2003) and at Serine 1700 (S1700) (Fuller et al., 2010), and it has been suggested that S1700 phosphorylation plays a greater modulatory role than S1928 phosphorylation (Brandmayr et al., 2012). However, only S1928 phosphorylation has been shown to be increased in the hippocampus with normal aging (Davare & Hell, 2003). Similarly, we also found significant age-related increase in S1928 phosphorylation, but only in the DG with no apparent changes in other hippocampal regions (Fig. 7). The difference with the previous report (Davare & Hell, 2003) may be due to our focus on the phosphorylation of cell surface channels in each hippocampal region of the dorsal hippocampus, whereas the previous report examined S1928 phosphorylation in the entire hippocampus. In addition, we found high level of Ca v 1.2 protein in the DG as compared with CA1 and CA3 from whole tissue lysate (Fig. S5). Furthermore, no age-related changes in the Ca 2 + -dependent postburst afterhyperpolarization have been observed in DG granule cells (Baskys et al., 1987;Niesen et al., 1988;Reynolds & Carlen, 1989). Therefore, S1928 phosphorylation cannot account for increased calcium influx and reduced neuronal excitability with normal Age-related increases in the ratio of Ca V 1.2 and Ca V 1.3 proteins found on the plasma membrane as compared with the total protein levels found in specific regions of the hippocampus. Eight 250 lm thick acute hippocampal slices from 9 young and 9 aged rats were randomly selected and exposed to Sulfo-NHS-SS-biotin-labeling reagent before cell surface proteins were isolated using streptavidin magnetic beads. Control assays with biotinylated lysate proteins were also carried out to verify successful isolation of plasma membrane proteins in all three regions. We found little to no detection of our internal control protein, GAPDH, on surface fractions (See aging observed in CA1 hippocampal pyramidal neurons (Landfield & Pitler, 1984;Moyer et al., 1992;Moyer & Disterhoft, 1994;Thibault & Landfield, 1996).
We found no age-related changes in Ca v 1.2 or Ca v 1.3 mRNA expression in the present study. Previous reports examining the mRNA or LTCC a-subunits in hippocampus from aged animals have been inconsistent; with groups reporting an increase (Herman et al., 1998;Chen et al., 2000;Veng & Browning, 2002), no change (Blalock et al., 2003;Kadish et al., 2009), or reductions (Rowe et al., 2007) in mRNA or a-subunit expression with aging. The discrepancy between the findings may be due to the anatomical specificity and/or method of analysis used in the studies (i.e. differences in splice variants amplified by the different primers of different groups). At single cell resolution, a positive correlation between Ca v 1.3 mRNA levels and functional channel density in the adult and aged CA1 pyramidal neurons has been reported: that is, the more Ca v 1.3 mRNA, the greater the LTCC activity . Using whole hippocampus, a report using semi-quantitative RNAse protection analysis revealed that Ca v 1.2 and Ca v 1.3 mRNA expression levels are increased in aged rats (Herman et al., 1998), whereas no change in either mRNA (Blalock et al., 2003;Kadish et al., 2009) or a reduction in Ca v 1.2 mRNA (Rowe et al., 2007) were reported using microarray methods. Our present real-time quantitative PCR data support the lack of change in Ca v 1.2 or Ca v 1.3 mRNA expression level with normal aging in the subregions of the hippocampus (Table 1). Notably, we also observed that the level of Ca v 1.2 and Ca v 1.3 mRNA expressions was different than that for the protein levels: for protein, DG > CA3 ≥ CA1 (Fig. S5): for mRNA, CA3 ≥ CA1 > DG (Table 1). Furthermore, a 4-fold increase in Ca v 1.2 mRNA relative to Ca v 1.3 mRNA was observed in the CA1 region, which supports the previous reports that approximately 80% of LTCCs are Ca v 1.2 channels (Hell et al., 1993;Clark et al., 2003).
The exact mechanism by which LTCC activity is increased with aging in CA1 neurons is as yet unclear. While our findings provide a better understanding of the processes that take place during aging, we cannot exclude the role that other processes may have in channel function, activity, and intracellular Ca 2+ concentrations. For example, the activity of the pore-forming LTCC a-subunits can be regulated by with relation to total cell surface-expressed Ca v 1.2 protein in young and aged rats. We found no significant age-related differences in phosphorylation of Ca V 1.2 on CA1 or CA3 region. However, a significant increase in phosphorylation was observed on DG region of aged rats. All results were confirmed by repeating the experiments and analysis twice. Unpaired t-test: **P < 0.005. Data reported as the mean AE SEM.
protein-protein interactions, a number of which can enhance L-type channel function (Catterall, 2000;Calin-Jageman & Lee, 2008). Second, calcineurin expression levels and activity have also been shown to be increased in hippocampus from aged animals (Foster et al., 2001;Norris et al., 2005;Eto et al., 2008) and inhibiting it reduces LTCC activity (Norris et al., 2002(Norris et al., , 2010 and the Ca 2+ -dependent postburst afterhyperpolarization (AHP) (Vogalis et al., 2004). Third, post-translational modifications of the Ca v 1.2 LTCC protein by proteolytic cleavage of the C-terminus region can significantly impact voltage-dependent activation and activity of the channel (Wei et al., 1994;Hulme et al., 2006). Finally, Ca v 1.2 and Ca v 1.3 channels are regulated by proteosomal degradation (Altier et al., 2011;Gregory et al., 2011) and oxidative stress through the action of reactive oxygen species, which can lead to increased accumulation of Ca 2+ inside hippocampal neurons. (Kourie, 1998;Fusi et al., 2001). Age-related increase in LTCC function, specifically in CA1 pyramidal neurons, has been a popular hypothesis to explain age-related cognitive deficits (Disterhoft & Oh, 2006;Foster, 2007;Thibault et al., 2007). Previous reports have demonstrated that Ca 2+ influx into neurons is significantly increased in CA1 pyramidal neurons from aged animals (Moyer & Disterhoft, 1994;Thibault & Landfield, 1996) due to increased density of LTCCs (Thibault & Landfield, 1996), which leads to reduced intrinsic excitability (Landfield & Pitler, 1984;Thompson et al., 1990;Moyer et al., 1992) and synaptic plasticity (Norris et al., 1998). Rescue of age-related cognitive deficits (Deyo et al., 1989) and restoration of intrinsic neuronal excitability (Moyer et al., 1992;Norris et al., 1998) and synaptic plasticity (Norris et al., 1998) with the use of LTCC blockers (e.g. nimodipine and nifedipine) have further provided support for the previously held viewpoint. In addition to the enhanced Ca 2+ influx via LTCCs, Ca 2+ released from the endoplasmic reticulum through ryano-dine receptors via the Ca 2+ -induced Ca 2+ -release (CICR) mechanism has been suggested and shown to greatly reduce excitability (i.e., increase postburst AHP) of CA1 pyramidal neurons from aged rats (Kumar & Foster, 2004;Gant et al., 2006;Kim et al., 2007;Thibault et al., 2007). Therefore, while increased LTCC function is a significant source of the age-related cognitive deficits, other sources of Ca 2+ that are changed with normal aging should also be systematically characterized.
The increased LTCC surface expression was not observed in all of the aged rats examined (Figs 3-5). This heterogeneity was expected as we have previously reported that nearly half of the aged cohort of 27-31month-old F344xBN rats are learning-impaired (Knuttinen et al., 2001;Matthews et al., 2009). Similar age-related heterogeneity has been reported for other strains of rats (Gallagher et al., 1993;Tombaugh et al., 2005) and in rabbits (Thompson et al., 1996;Moyer et al., 2000). More importantly, it has been demonstrated that learning-impaired aged animals have CA1 pyramidal neurons with enlarged postburst AHP (Moyer et al., 2000;Tombaugh et al., 2005;Matthews et al., 2009). Therefore, we hypothesize that only those aged subjects with increased surface expression of functional LTCCs are likely to be cognitively impaired given the contribution of Ca 2+ influx via L-type Ca 2+ channels to the AHP. Additional studies are required to determine whether such correlation exists.
In summary, our results suggest that age-related cognitive deficits cannot be attributed to a global change in L-type Ca 2+ channel expression or to the level of phosphorylation of Ca v 1.2 channels in the plasma membrane of aged hippocampal neurons. Rather, we provide evidence that age-related increases in plasma membrane expression and/ or distribution of L-type Ca 2+ channels in the somatic regions of CA1 and CA3 pyramidal neurons may underlie the reported changes in neuronal excitability and activity observed with normal aging.  (Bronson, 1990;Lipman et al., 1996). Every effort was made to ensure that only healthy rats were included in the experiments. Rats with palpable tumors, skin ulcerations, infections, or difficulty moving were excluded from the studies. All experimental procedures were approved by the Northwestern University Animal Care and Use Committee and conformed to NIH standards . All efforts were made to minimize animals' discomfort and the number of animals used. For CNC1 antibody specificity, brains and cerebellar tissue from conditional knockout mice with a forebrain-specific (hippocampus and forebrain) deletion of Ca v 1.2 were used White et al., 2008). For ab144 antibody specificity, brains from knockout mice with a global deletion of the gene encoding Ca v 1.3 were used (Platzer et al., 2000;Clark et al., 2003;McKinney & Murphy, 2006). Tissue from knockout mice and their wild-type littermates were generously provided by Dr. Geoffrey G. Murphy (University of Michigan, Ann Arbor, MI, USA).
For quantification of total L-type Ca 2+ channel expression in the three major hippocampal regions of young and aged rats, samples containing equal amounts of proteins (20 lg) were resolved by SDS-PAGE and analyzed by immunoblotting with either anti-CNC1 or Ab144. Fresh blots were used for each channel of interest, and each blot was reprobed with GAPDH for normalization and to control for variability during sample loading. Immunoreactive bands were visualized using a Chem-iDoc XRS+ Molecular Imager System with Image Lab TM Software (Bio-Rad Laboratories, Hercules, CA, USA), and only signals doubling with increasing exposure times were used for quantification and analysis. All immunoblots were measured and quantified by densitometry using NIH IMAGEJ image analysis software (rsb.info.nih.gov/ij/).

Whole slice cell surface biotinylation
To assess the relative cell surface expression of L-type Ca 2+ channels in hippocampal pyramidal neurons, cell surface biotinylation experiments were conducted on acute hippocampal slices (n = 8, 250 lm slices per rat) from nine young and nine aged rats. This technique has been successfully shown to reach all layers of acute hippocampal slices of up to 400 lm in thickness using Sulfo-NHS-SS-biotin as labeling reagent with very low to no labeling of intracellular proteins (Thomas-Crusells et al., 2003). Alternate slices from the dorsal half of both left and right hippocampi were exposed to 1 mg mL À1 Sulfo-NHS-SS-biotin-labeling reagent (Pierce) for 30 min before separating each hippocampal region for processing and isolation of surface proteins using streptavidin magnetic beads (Pierce).
Immunoblotting with p1928, CNC1, and Ab144 antibody was performed after SDS-PAGE separation of total region lysates (input) and biotinylated (surface) fraction proteins. GAPDH was used as both loading and internal protein control to confirm the success of the biotinylation procedures (Fig. S3).

L-type Ca 2+ channel surface expression and phosphorylation of Ca v 1.2 measurements
Following Western blot analysis, optical density values for total lysate (input) and biotinylated (surface) fractions were obtained and the relative ratio of surface expression for each subunit was determined by normalizing the surface optical density value to its corresponding input band density value. The level of phosphorylation of Serine 1928 on Ca v 1.2 protein was determined by normalizing the obtained p1928 band to the total CNC1 intensity for the loaded surface fraction. To quantify the level of p1928 phosphorylation of surface-expressed Ca v 1.2 channels, blots were initially probed with p1928, stripped, and reprobed with CNC1 antibody.
RNA isolation and cDNA synthesis CA1, CA3, and DG regions were isolated from young and aged rats in pairs, homogenized in RPLT-Plus Lysis Buffer (Qiagen, Valencia, CA, USA) and stored at À80°C until RNA isolation. Samples were further dissociated with QiaShredder columns, and the total RNA was isolated via Qiagen RNEasy Plus Kit according to manufacturer's directions. RNA was dissolved into 60 ll RNAse-free water, stored on ice, and the yield was determined with a nanodrop spectrophotometer (Thermo-Scientific, Rockford, IL, USA). 650 ng (CA1, CA3) or 240 ng (DG) of total RNA was converted into cDNA with reverse transcriptase and multiple primers (qScript TM cDNA SuperMix; Quanta Biosciences, Gaithersburg, MD, USA) on a thermocycler per manufacturer's instructions and stored at À20°C until use. Synthesis of cDNA was commenced within 2 h of RNA isolation.

Immunohistochemistry
Young (3-4 month old) and aged (30-32 month old) male F344xBN rats were anesthetized and intracardially perfused with ice-cold sodium phosphate buffer (0.1 M PB [pH 7.4]) supplemented with several protease and phosphatase inhibitors (PPI; cOmplete and PhosSTOP Inhibitor Cocktail Tablets; Roche, Indianapolis, IN, USA) and followed by ice-cold 4% paraformaldehyde in PB (supplemented with PPI). Brains were removed, postfixed (overnight), and cryoprotected by successively sinking in 10% (w/v) and 30% (w/v) sucrose in PB at 4°C for 72 h (Marshall et al., 2011). Forty-micrometer-thick coronal sections containing the hippocampus were made, hippocampus dissected out, and stored at À20°C in cryoprotectant solution (0.1 M PBS, pH 7.4, 30% (w/ v) sucrose, 30% (v/v) ethylene glycol and 1% (w/v) polyvinylpyrrolidone) (Watson et al., 1986) until processed for immunohistochemistry. Immunohistochemistry (IHC) was performed as previously described (Ferraguti et al., 2004;Wu et al., 2008) with some modifications. The tissue processing and data collection and analysis were performed blind to the age of the animals. Five hippocampal slices from each animal were systemically and randomly selected for double immunolabeling with antibodies against glial fibrillary acidic protein (GFAP) and Ca v 1.2 or Ca v 1.3. The free-floating sections were rinsed in 0.1 M Dulbecco's phosphate-buffered saline (DPBS) for 30 min to remove cryoprotectant, incubated with 0.05% boric acid in DPBS for 10 min, rinsed in DPBS containing 0.1% Tween-20 (DPBS-T) for 30 min, and blocked in DPBS containing 5% normal donkey serum, 1% immunoglobulin-and protease-free bovine serum albumin, and 0.3% Triton X-100 for 2 h. The sections were then incubated in chicken anti-GFAP (diluted 1:500) and anti-Ca v 1.2 (CNC1; diluted 1:500) or anti-Ca v 1.3 (ab144; diluted 1:2000) overnight at 4°C, rinsed in DPBS-T for 30 min, incubated in FITC-conjugated secondary antibody against rabbit IgG (diluted 1:500) and Texas Red-conjugated secondary antibody against chicken IgG (diluted 1:500) (Jackson immunoresearch, West Grove, PA) for 1 h at room temperature, rinsed in DPBS-T for 1 h, rinsed in DPBS for 10 min, and finally mounted and coverslipped using ProLong â Gold antifade reagent with DAPI (Molecular Probes, Eugene, OR, USA). Control hippocampal sections from young and aged rats were also included in the assay and treated the same way, but either the primary, secondary, or both antibodies were excluded from the incubating solution. All slices from young and aged animals were simultaneously processed during the experiment in order to minimize the effects of potential inter-batch staining variability.

Image analysis
Confocal images from CA1 region of hippocampus were obtained at a magnification of 409 using a Nikon Eclipse C1si Spectral Imaging Confocal Microscope System at the Nikon Imaging Center and Cell Imaging Facility at Northwestern University. Exposure parameters for Ca v 1.2, Ca v 1.3, and GFAP were standardized across all captured images and maintained throughout image acquisition for both young and aged hippocampal slides. Images were analyzed using MetaMorph â imaging software (Molecular Devices, Sunnyvale, CA, USA), and statistical analyses were performed using StatView software. To study age-related changes in L-type subunit expression in CA1 and CA3, data were collected from raw, unmodified, images by drawing three rectangular regions of interest (ROI) with equal dimensions in both the stratum pyramidale and the stratum radiatum layers of CA1 and CA3 regions. For DG, ROIs were placed atop the granular cell layer. The fluorescence intensity for each ROI was then averaged to calculate the integrated fluorescent intensity for each hippocampal slice. Averaged values from all 5 hippocampal slices from each animal were then averaged to collect the animal's integrated fluorescent intensity used for plotting and statistical analysis. Significant group differences in protein expression were evaluated using analysis of variance (ANOVA) with statistical significance set to P < 0.05. All data are reported as the means AE SEM.

Real-time PCR of mRNA levels
Primer sequences compatible with rat and mouse were a gift from C. Savio Chan (Northwestern University). The primers were designed to bridge exons of cDNA to eliminate concern of genomic DNA contamination and previously tested for comparable efficiency during PCR. Primers include Cacna1c, bridging exons 7-8 [sense primer-GGCATCACCAACTTCGACA, antisense Primer-TACACCCAGGGCAACTCATA], Cacna1d, bridging exons 41-42 [sense primer-TGACATTGGGCCAGAAATCC, antisense primer-GGTGGTATTGGTCTGCTGAA], and GAPDH, bridging exons 3-4 [sense primer-GCTGAGTATGTCGTGGAGTCTA, antisense primer-TTCTCGTGGTTCACACCCAT]. For real-time quantitative PCR, 1 pair of young and aged cDNA (1 lL) triplicate samples from each of CA1, CA3, and DG were measured in parallel for threshold fluorescence accumulation (C T ) of each gene target (Cacna1c, Cacna1d, and GAPDH) in a 96-well tray with a Step One Plus TM Applied Biosystems QPCR Machine using SYBR Green as a reporter and ROX TM dye as a passive reference control. After PCR, a melt curve analysis was done for each sample to ensure that the primers were specific. The threshold C T was manually set to be 1.1 DRn (reporter-reference control baseline fluorescence) for all targets and fell within the exponential phase of amplification. The comparative DDC T method described by Livak and Schmittgen (Livak & Schmittgen, 2001) was used to compare young and aged samples with GAPDH, which was used as internal housekeeping control. DC T for each sample target gene was calculated as follows: DC T = (mean C T Targetmean C T Gapdh ). To obtain fold change due to age, DDC T was calculated with the following equation: (DC T AGED -DC T Young ) and presented in Table 1.

Statistical analysis
All statistical analyses were performed using STATVIEW analysis software, and significant group differences in protein expression were evaluated using analysis of variance (ANOVA) with statistical significance set to P < 0.05. All data are reported as the mean AE SEM. Duplicates were performed on all reported phosphorylation and surface expression assays.
tissue; Dr C. Savio Chan and Vivian Hernandez for technical assistance, reagents, and use of real-time PCR equipment; Drs. Johannes W. Hell and Amy Lee for their contributions to conceptual discussions and practical input; Drs. Geoffrey T.

Supporting Information
Additional Supporting Information may be found in the online version of this article at the publisher's web-site.

Fig. S1
Depiction of Immunoblots from CNC1 and ab144 antibody specificity on hippocampal and cerebellar tissue lysates.