Calcium-dependent inactivation of calcium channels in the medial striatum increases at eye opening

Abstract

Influx of calcium through voltage-gated calcium channels (VGCCs) is essential for striatal function and plasticity. VGCCs expressed in striatal neurons have varying kinetics, voltage dependences, and densities resulting in heterogeneous subcellular calcium dynamics. One factor that determines the calcium dynamics in striatal medium spiny neurons is inactivation of VGCCs. Aside from voltage-dependent inactivation, VGCCs undergo calcium-dependent inactivation (CDI): inactivating in response to an influx of calcium. CDI is a negative feedback control mechanism; however, its contribution to striatal neuron function is unknown. Furthermore, although the density of VGCC expression changes with development, it is unclear whether CDI changes with development. Because calcium influx through L-type calcium channels is required for striatal synaptic depression, a change in CDI could contribute to age-dependent changes in striatal synaptic plasticity. Here we use whole cell voltage clamp to characterize CDI over developmental stages and across striatal regions. We find that CDI increases at the age of eye opening in the medial striatum but not the lateral striatum. The developmental increase in CDI mostly involves L-type channels, although calcium influx through non-L-type channels contributes to the CDI in both age groups. Agents that enhance protein kinase A (PKA) phosphorylation of calcium channels reduce the magnitude of CDI after eye opening, suggesting that the developmental increase in CDI may be related to a reduction in the phosphorylation state of the L-type calcium channel. These results are the first to show that modifications in striatal neuron properties correlate with changes to sensory input.

the dorsal striatum is a brain region important for habit and skill learning as well as motor control (Knowlton et al. 1996). To serve these functions, the dorsal striatum receives information from all parts of the cerebral cortex, including the sensory cortices and thalamus. To properly integrate signals from these diverse inputs, striatal neurons precisely control their response to these signals through synaptic plasticity and dendritic excitability. Voltage-gated calcium channels (VGCCs) play a critical role in both striatal synaptic plasticity (Fino et al. 2010) and the dendritic generation of striatal upstates (Plotkin et al. 2011). The duration of the VGCC current and consequently the timing of the dendritic calcium elevation are of particular importance for both synaptic plasticity and dendritic excitability. It has been hypothesized that the duration of the calcium elevation controls the direction of plasticity (reviewed in Evans and Blackwell 2015) and the duration of the VGCC current can affect the calcium influx during the upstates (Evans et al. 2013). The duration of the VGCC current is regulated by both voltage-dependent inactivation and calcium-dependent inactivation (CDI).

Although VGCCs play a role in striatal function, differences in VGCC characteristics between striatal regions have been largely ignored. There are known differences in cortical input pattern (Reig and Silberberg 2014; Voorn et al. 2004), receptor expression (Chapman et al. 2003; Herkenham et al. 1991), learning strategies (Pauli et al. 2012; Yin et al. 2009), and synaptic plasticity (Partridge et al. 2000; Smith et al. 2001) between the dorsomedial (DM) and dorsolateral (DL) striatum. However, it is not clear whether the VGCC composition or inactivation properties contribute to the difference between these regions.

Similarly, the composition of VGCCs and their inactivation properties can be modulated throughout an animal's life. This modulation occurs through an increase or decrease in the expression of various VGCC types (L, R, T, N, and P/Q) (Martella et al. 2008). In addition, inactivation properties of VGCC may be altered through phosphorylation or dephosphorylation (Dittmer et al. 2014) over development. Indeed, striatal neurons from aged rats show less CDI than neurons from younger adult rats (Dunia et al. 1996); however, VGCC characteristics during early striatal development have not been investigated.

Eye opening, occurring around 12–13 days of age, is a critical period in development that has been studied extensively in the cortex but not the striatum. An increase in synaptic input to visual cortex produces significant changes on the molecular, cellular, and network level (Karmarkar and Dan 2006). The DM striatum receives dense inputs from visual cortex (Reig and Silberberg 2014), suggesting that eye opening may trigger intracellular changes in DM neurons that will alter the way they integrate synaptic information.

Here we characterize CDI across striatal regions and at three time points during development: before eye opening (p11-12), after eye opening (p13-15), and after weaning (p22-24). We show that CDI in the DM striatum increases sharply between the ages of p11-12 and p13-15, which correlates with eye opening. We also show that this increase is dependent on L-type calcium channels and channel phosphorylation state.

METHODS

Electrophysiology.

All animal handling and procedures were in accordance with the National Institutes of Health animal welfare guidelines and were approved by George Mason University's Institutional Animal Care and Use Committee. Every effort was made to minimize anxiety and pain in the animals. We used C57Bl6 male and female mice (Charles River) of ages p11-12 (eyes closed), p13-15 (eyes open), and p22-24 (weaned). Mice of age p11-12 with eyes open were not used nor were mice of age p13-15 with eyes still closed. Mice were rapidly anesthetized with isoflurane and euthanized by decapitation. Brains were extracted and cut 350-μm thick using a vibratome (Leica VT 1000S) in ice-cold sucrose slicing solution (in mM: 2.8 KCl, 10 dextrose, 26.2 NaHCO₃, 1.25 NaH₂PO₄, 0.5 CaCl₂, 7 Mg₂SO₄, and 210 sucrose). Slices were immediately placed in an incubation chamber containing artificial cerebrospinal fluid (aCSF; in mM: 126 NaCl, 1.25 NaH₂PO₄, 2.8 KCl, 2 CaCl₂, 1 Mg₂SO₄, 26.2 NaHCO₃, and 11 dextrose) for 30 min at 33°C and then removed to room temperature (22–24°C) for at least 90 more minutes before use. Recording aCSF was modified from the incubation aCSF by replacing Mg₂SO₄ with MgCl₂, using either 2 mM CaCl₂ or 2 mM BaCl₂, and including the potassium channel blocker 4-AP (4 mM). Slices were placed in a submersion chamber (ALA Scientific) at room temperature, and cells were visualized using differential interference contrast imaging (Zeiss Axioskop2 FS plus). Pipettes (2–4 MΩ) were pulled from borosilicate glass on a laser pipette puller (Sutter P-2000), coated with candle wax approximately 3–5 mm along the pipette tip (to reduce pipette capacitance) and fire-polished (Narshige MF-830). Pipettes were filled with a cesium-based internal solution (in mM: 85 Cs-gluconate, 10 Cs₃-citrate, 1 KCl, 10 NaCl, 10 HEPES, 1.1 EGTA, 0.1 CaCl₂, 0.25 MgCl₂, 15 TEA-Cl, 3.56 Mg-ATP, and 0.38 Na-GTP) at pH 7.26 (Rankovic et al. 2011). Voltage-clamp recordings were obtained using the HEKA EPC-10 and sampled at 20 kHz. Data were acquired in Patchmaster (HEKA Elektronik) and filtered with three-pole (10 kHz) and four-pole (2.9 kHz) Bessel filters. Series resistance (4–15 MΩ) was compensated 70–90%.

Drugs and drug application.

Unless otherwise stated salts were from Fisher Scientific. Mg-ATP, Na-GTP, gluconic acid, citric acid, 4-AP, CsOH, TEA-Cl, papaverine, nickel chloride, and nifedipine were obtained from Sigma-Aldrich. Tetrodotoxin (TTX) and ω-conotoxin GVIA were purchased from Tocris Bioscience; ω-conotoxin MVIIC was obtained from Alomone Laboratories; forskolin was obtained from Ascent Scientific; and FK506 was obtained from LC Laboratories. FK506, forskolin, and nifedipine were dissolved in DMSO for a final DMSO concentration of 0.04–0.05%. ω-Conotoxin GVIA was dissolved in 20% acetonitrile for a final concentration of 0.02%. Nifedipine was protected from light and made fresh the day of use. TTX, calcium channel blockers, and the cocktail of FK506, forskolin, and papaverine were applied via a custom made Y-tube local perfusion system (Murase et al. 1989). Application of TTX (0.5 μM) through the Y-tube abolished action potentials in <1 min, demonstrating rapid and spatially adequate perfusion of the patched neuron (data not shown). Slices were not reused after application of barium-containing bath solution or being exposed to a drug treatment.

Data analysis.

Pharmacologically isolated, high-voltage-activated (HVA) calcium currents were obtained during a pair of 200-ms depolarizing pulses from −40 to +10 mV, separated by 50 ms. Currents were obtained every 60 s and leak subtracted. Inactivation was measured from the mean current at the end of the 200-ms pulse divided by the peak amplitude of the current (r200; Fig. 1A). An alternative inactivation measure, the peak amplitude of the second pulse divided by the peak amplitude of the first pulse (Rankovic et al. 2011), gave similar results. The difference between inactivation in barium (average of 5 points as indicated in Fig. 1C, beginning ∼4 min after switching to barium-aCSF in the bath and ∼2 min after applying barium-aCSF with the Y-tube) and mean inactivation in calcium (average of 5 baseline points indicated in Fig. 1C) was taken as the measure of CDI (Ba-Ca Diff, Fig. 1D). For experiments using calcium channel inhibitors or phosphorylation enhancing drugs, after the control response in calcium was collected, the drugs were applied using the Y-tube. Inactivation in drug + calcium was calculated as the average of responses between 2 and 7 min after drug application. Then, similar to the time course of the barium switch illustrated in Fig. 1C, barium was applied, both in the bath and in the Y-tube, with drug added to the barium-aCSF in the Y-tube. When specified, ramp depolarizations from −80 to +50 mV (0.5 mV/ms; Fig. 1E) were interleaved with the square pulses, also every 60 s. Mean current amplitude was measured as greatest negative amplitude of the leak-subtracted current during the ramp. Data were not corrected for junction potential, which was calculated as −12 mV. Data were processed using python 2.7. Statistical tests were performed using the general linear model procedure (GLM) or nonparametric tests (NPAR1WAY) in SAS 9.3 (SAS Institute).

Fig. 1.

Striatal calcium-dependent inactivation (CDI) depends on age and region. A: example trace of calcium currents produced by depolarizing pulses to 10 mV. A1: inactivation was measured as ratio of mean current at end of first pulse to peak current. A2: example trace showing reduced inactivation in barium. B: schematic illustrating medial (DM) and lateral (DL) regions of the striatum where recordings were made. C: typical inactivation time course in barium and calcium. Average of 5 points labeled as Ca avg were subtracted from average of 5 points labeled Ba avg to calculate the Ba-Ca Diff. D: summary of CDI (Ba-Ca Diff) for each age group by striatal region (n is specified in each bar; error bars are ± SE). **Slope vs. age is significant at P < 0.005. E: example trace of calcium currents produced by ramp pulse of 0.5 mV/ms from −80 to + 50 mV. F: summary of total current amplitude in calcium for each age group by striatal region. ***Slope vs. age is significant at P < 0.0005. G: summary of current density (amplitude divided by cell capacitance) for each age group by striatal region. For both F and G, n is specified in each bar; error bars are ± SE.

RESULTS

Isolating CDI in HVA channels.

Voltage-gated calcium currents are isolated by blocking voltage-gated potassium currents (with 15 mM TEA, 115 mM cesium, and 4 mM 4-AP) and sodium currents (with 0.5 μM TTX). HVA channels are isolated by holding striatal neurons at −40 mV, which inactivates the low voltage-activated calcium currents (such as T-type channels). Inactivation (r200) is calculated from the mean current at the end of the 200-ms pulse divided by the peak amplitude of the current (Fig. 1A). The ratio of the second peak to the first peak (Rankovic et al. 2011) is an alternative measure of inactivation and is highly correlated to r200 (mean corr = 0.97). CDI is the difference between voltage-dependent inactivation (assessed from currents measured in barium) and total inactivation (assessed from currents measured in calcium; Fig. 1C).

CDI increases with age in a region-dependent manner.

Because there are well-documented differences across striatal regions and over the course of striatal development, we tested whether CDI varied with region and age. Using either calcium or barium as the charge carrier, we measured currents through HVA calcium channels from the DM and DL striatum (Fig. 1B) at three different murine developmental ages: p11-12 (before eye opening), p13-15 (after eye opening), and p22-24 (after weaning). Figure 1C shows the r200 over time for one experiment from each region of one mouse (p14) and for an experiment in which the barium solution was omitted. This demonstrates that the measure of CDI is stable over time and that the current inactivation (Fig. 1A2) and r200 ratio is lower in calcium than in barium. Figure 1D illustrates the mean CDI for all three age groups and both striatal regions, and demonstrates that CDI increases with age (GLM for age × region, F = 7.08, P = 0.0024) in the DM striatum (slope: 0.008 ± 0.003, P = 0.002) but not the DL striatum (slope: 0.005 ± 0.003, P = 0.085), showing that this developmental change is region specific. CDI did not differ by sex in any age group or region (data not shown), as expected with low levels of sex hormones in juvenile mice (Garris 1985; Overpeck et al. 1978; Vandenberg et al. 2006).

The total calcium currents are quantified from the ramp depolarization (Surmeier et al. 1995) (Fig. 1E). Analysis shows that the amplitude of the calcium current increases with age (GLM age × region, F = 21.06, P < 0.0001; Fig. 1F), but the increase is similar in both regions (DM: slope = 0.102 ± 0.018, P < 0.0001; DL: slope = 0.123 ± 0.019, P < 0.0001). The change in current amplitude is unlikely to be responsible for the increase in CDI because the largest increase in current amplitude medially occurs between open and weaned conditions, rather than between closed and open conditions, and the current density (current amplitude divided by capacitance) does not significantly increase with age for either region (GLM age × region, F = 2.35, P = 0.109; Fig. 1F). Furthermore, the correlation between either current amplitude or current density and CDI is not significant (P = 0.624 for amplitude, P = 0.078 for density). In summary, the DM-specific increase in CDI with development is not solely due to a region-specific change in total current amplitude.

L-type calcium channels contribute to the increase in medial CDI that occurs at eye opening.

Because L-type calcium channels are known to have strong CDI (Liang et al. 2003) and contribute significantly to the calcium current in medium spiny neurons (MSNs) (Martella et al. 2008), we tested CDI in DM and DL neurons in the presence of 20 μM nifedipine, a selective L-type calcium channel blocker, and separately in the presence of a cocktail designed to isolate L-type channels, consisting of calcium channel inhibitors: 1 μM ω-conotoxin GVIA, 1 μM ω-conotoxin MVIIC, and 30 μM nickel (blocking N-, P/Q-, and R-type channels, respectively). Nifedipine prevented the difference in CDI between closed and open groups. All regions and age groups had the same CDI in the nifedipine condition (GLM age × region, F = 0.85, P = 0.442; Fig. 2A), arguing that L-type currents are important for the age-dependent change in CDI in DM striatum. Conversely, in dorsomedial striatum, isolating L-type channels reduced CDI for both age groups by similar amounts [GLM age drug, F = 10.53, P = 0.0004, P(age) = 0.019, P(drug) = 0.0006]. Although this seems counterintuitive, it may mean that inactivation of the L-type calcium channels is partially sensitive to the calcium from other nearby HVA calcium channels. As in the control condition, CDI in these groups did not differ by animal sex (data not shown).

Fig. 2.

L-type calcium channels contribute to the increase in CDI that occurs at eye opening, but other channels contribute to the magnitude of CDI. A: summary of CDI (Ba-Ca Diff) before and after eye opening for nifedipine (20 μM) treated (Nif: striped bars) or ω-conotoxin GVIA (1 μM), ω-conotoxin MVIIC (1 μM), and nickel (30 μM) (IsoL: checked bars). Gray bars in the background show the control data for comparison; n is specified in each bar. B: fraction of calcium current remaining, measured using ramp protocol, after application of calcium channel inhibitors; n is specified in each bar. *Significant reduction compared with the other nifedipine treated groups. Dashed line shows the reduction in calcium current over the same time frame in the absence of drug application (n = 7). Error bars are ± SE.

These data show that L-type calcium channels are important for the increase in CDI seen in the dorsomedial striatum at eye opening. This can be caused by several mechanisms. There could be an increase in the density or proportion of L-type channels in DM striatum at eye opening. Alternatively, there could be a similar density of L-type calcium channels that undergo more CDI at eye opening. To test the first possibility, we measured the amplitude of current produced by ramp depolarization in control conditions and in the presence of nifedipine for both regions before and after eye opening. Nifedipine decreased the amplitude of the calcium current by 20–25% for each group (Fig. 2B). Interestingly, nifedipine blocks more of the current from DM in the eye open group (Fig. 2B; GLM age × region, F = 3.64, P = 0.028, post hoc contrast DM nifedipine vs. others, F = 10.6, P = 0.0036). This suggests that the increase in CDI at eye opening may be due to an increase in L-type channel density. However, isolating L-type currents reduces the DM currents to a similar degree in both age groups (Wilcoxon two-sample test, P = 0.125). In summary, these data collectively suggest that both current density and the CDI of L-type currents increases in DM striatum at eye opening.

Protein kinase A phosphorylation reduces CDI.

An alternative mechanism for the change in CDI we observe after eye opening is a change in calcium channel properties. Studies have shown that certain mutations (Barrett and Tsien 2008) and phosphorylation states (Dittmer et al. 2014; Oliveria et al. 2012; Rankovic et al. 2011) can alter the inactivation properties of L-type channels. Specifically, L-type channels are phosphorylated by protein kinase A (PKA) and dephosphorylated by calcineurin which decreases and increases CDI, respectively (Dittmer et al. 2014; Oliveria et al. 2012). To test whether the phosphorylation state of HVA calcium channels is the same before and after eye opening, we applied the calcineurin antagonist FK506 (50 μM), the adenylyl cyclase activator forskolin (25 μM), and the phosphodiesterase inhibitor papaverine (10 μM) to enhance PKA via increased cAMP and then measured the response beginning 2–3 min after drug application. Previous studies have demonstrated changes in phosphorylation using these drugs within 2–3 min of application, both in cultured neurons (Dittmer et al. 2014; Oliveria et al. 2012) and in brain slice (Castro et al. 2013; Nishi et al. 2008). We confirmed that the drugs were affecting the cell within this time frame by the reduction in ramp current amplitude, which was 70% of control for both groups (Fig. 3). This ramp current reduction is not due to rundown as the current maintains 95% of its level over the same time period in the absence of drug application (dashed line in Fig. 3B). Maximizing the phosphorylation of calcium channels significantly reduces CDI for the eyes open group (Fig. 3A; Wilcoxon, DM open drug vs. no drug, P = 0.021), yielding CDI that is not significantly different before and after eye opening (Wilcoxon, DM drug open vs. closed, P = 0.405). These results suggest that after eye opening, the calcium channels in the DM striatum are in a weakly phosphorylated state resulting in enhanced CDI.

Fig. 3.

Increased PKA phosphorylation of calcium channels reduces calcium current and CDI. A: enhanced PKA phosphorylation reduces CDI more after eye opening, resulting in a similar CDI for both age groups. B: PKA phosphorylation reduces current amplitude. n is specified in each bar. Error bars are ± SE.

DISCUSSION

Region and age-dependent variation in calcium currents.

Here we have shown that dorsal striatal calcium currents change during specific developmental stages in a region-dependent manner. Specifically, we have shown that the amplitude of the total calcium current increases during development in both the medial and lateral striatum, whereas the CDI increases at eye opening in the medial striatum only. This finding contributes to previous observations showing that several key characteristics of striatal neurons change over this time period of development. Both morphology, such as density of spines on dendrites, and inward rectification due to K_ir channels increase during development (Belleau and Warren 2000; Tepper et al. 1998). N-methyl-d-aspartate (NMDA) receptor expression also increases during development (Portera-Cailliau et al. 1996; Wenzel et al. 1997), with both GluN2A and GluN2B increasing between p10 and p21/26, the age range studied here. While those studies did not specifically investigate developmental differences in the dorsomedial and dorsolateral striatum, Tepper et al. (1998) mentions a general trend toward the medial and caudal neurons becoming electrically and morphologically mature sooner than the lateral and rostral ones.

Although this is the first study to test CDI in striatal neurons during development, CDI has been recorded in striatal neurons from adult animals. One previous study recorded calcium plateaus from adult and aged rats and found that the calcium plateaus were longer in duration for adult rats than aged, which may suggest less CDI in adult rats compared with aged (Dunia et al. 1996). However, switching between barium and calcium as the charge carrier revealed that this difference in plateau duration was not due to a difference in CDI. Instead, Dunia et al. demonstrate that the aged rats have less CDI than adult rats. Taken together, the results indicate that there is not a monotonic change in CDI with age. The functional significance of this finding is not yet clear, but the increase in CDI may alter the excitability of the medial neurons during development and young adulthood and could affect their ability to undergo synaptic plasticity.

We also found that calcium current amplitude in striatal neurons increases with age in a region-independent manner. Specifically, we observed a large increase in current amplitude between eye opening and weaning in both regions. A previous study showed that the amplitude of total calcium currents from acutely dissociated striatal neurons does not change with age (Martella et al. 2008). The increase in current amplitude observed between eye opening and weaning in the present study may be due to an increase in dendritic calcium channels, which would not be apparent when measuring from acutely dissociated neurons. L-type channels are located in dendrites and spines (Carter and Sabatini 2004); thus the observed increase in calcium current may be partly due to the previously observed increase in dendritic spine density (Tepper et al. 1998).

Mechanisms underlying the change in CDI.

Blockade of L-type calcium channels prevents the increase in CDI observed at eye opening in the medial striatum, which suggests the increase in CDI could be due to two mechanisms. First, L-type channel density increases after eye opening in DM striatum, as blocking L-type channels reduces the total current amplitude to a greater degree after eye opening than before eye opening. Presumably, an increased density of channels results in greater calcium current and enhanced calcium-dependent processes such as CDI. Secondly, a change in inactivation properties occurs through different phosphorylation states of L-type channels (Budde et al. 2002). Several studies have shown that activation of PKA and inhibition of calcineurin alters CDI in hippocampal neurons (Dittmer et al. 2014) and thalamic neurons (Rankovic et al. 2011) and that L-type channels have increased phosphorylation with age (Davare and Hell 2003). Our results showing that enhanced PKA phosphorylation reduces CDI predominantly after eye opening are consistent with these prior studies and further suggest that the enhanced CDI in DM striatum with eye opening involves modulation of PKA phosphorylation sites. Further studies using immunohistochemistry or in situ hybridization could reveal any differences in phosphorylation state of L-type channels immediately after eye opening in the medial striatum.

L-type calcium channels have long been known to undergo strong CDI, while the other VGCCs particularly N- and R-type channels have more recently been shown to undergo CDI also (Liang et al. 2003). L-type channels inactivate in response to elevations in small calcium microdomains, while N- and R-type channels inactivate in response to global calcium (Liang et al. 2003). In this study we did not use EGTA or any calcium buffer in our internal solution, allowing for optimal CDI of all channels (Liang et al. 2003; Norris et al. 2002). Because CDI is still present even when L-type calcium channels are blocked, our results indicate that in both medial and lateral striatal neurons, the non-L-type channels also undergo CDI. Surprisingly, our data show that both blocking L-type calcium channels with nifedipine and isolating L-type calcium channels with nickel and ω-conotoxin GVIA reduce the total CDI. This result seems counterintuitive because it is thought that the L-type calcium channels have the strongest CDI and sense only their own calcium. However, because we are not using calcium buffers in our internal solution, calcium from N- and R-type calcium channels may reach the calcium binding domain of the L-type calcium channels, enhancing its CDI. Thus blocking N- and R-type channels may reduce the total CDI simply because it is reducing global calcium. Further work is needed to parse out the specific contributions of each calcium channel to the total CDI in these neurons.

Our results show a lower contribution of L-type calcium channels (20–25%) than a previous study measuring the proportion of calcium current carried by each VGCC in dissociated striatal cells (Martella et al. 2008). They showed that L-type calcium channels made up ∼50% of the total current at p14 and that the proportion of current carried by L-types decreased with age. The differences could be due to a difference in slice preparation, which preserves dendrites and spines, compared with acutely dissociated cells, which only contain soma and proximal dendrites. On the other hand, our results are similar to a separate study of acutely isolated neurons from the ventral striatum (nucleus accumbens) of mice p5-11, which reported that L-type calcium channels account for 18% of the total current (Churchill and Macvicar 1998). In addition, the change in peak current produced by enhanced PKA phosphorylation is similar to results seen in cultured striatal neurons (Surmeier et al. 1995). In conclusion, our measurements of calcium channel amplitude are consistent with most prior studies.

The significance of eye opening.

Our data show that CDI was increased immediately after eye opening only in the medial striatum. This finding is of particular interest in light of anatomical tracing studies linking the visual cortex to the medial striatum in rodents and new in vivo experiments demonstrating that medial striatal neurons respond to visual stimulation (Reig and Silberberg 2014). In rodents, developmental eye opening corresponds with enhanced synaptogenesis (Blue and Parnavelas 1983), an increase in synaptic events (Desai et al. 2002), and increases in synaptic strength or spine density (Konur and Yuste 2004) in the visual cortex. These changes would likely be reflected in the cortical output to other brain regions such as the medial striatum or hippocampus. Indeed, eye opening has been shown to affect hippocampal development, with early eye opening accelerating the maturation of synaptic strength (Dumas 2004). Accordingly, the change in CDI at eye opening may be accompanied by changes in synaptic plasticity of dorsomedial striatum. Our study is the first to suggest that sensory input can alter calcium dynamics in striatal neurons. We have demonstrated a correlation between eye opening and a selective increase in CDI in dorsomedial neurons; however, further work will be required to demonstrate a causative link. For example, using monocular deprivation or early eye opening will be necessary to determine whether visual input directly causes electrophysiological changes in medial striatal neurons.

Functional implications.

Developmental changes in ion channel expression and characteristics produce diverse effects on MSN electrical properties. NMDA receptor contribution to synaptic current doubles between the first and third postnatal weeks (Hurst et al. 2001), which may contribute to the characteristic upstates (and downstates) that develop in vivo during that time (Tepper et al. 1998). Opposing this increase in excitability, the increase in inward rectifier limits cell excitability by accentuating the hyperpolarized potential during the downstate. It is possible that the occurrence of upstates, observed earlier in medial than lateral striatum (Tepper et al. 1998), triggers the change in CDI during development or that the developmental increase in calcium current is essential for upstates to occur. Future work is necessary to test whether the development of CDI and upstates occur in the same neuron at the same time.

The increase in CDI during development may reflect a greater trend toward tighter temporal control of calcium influx. Our previous computational work suggests that CDI is a key mechanism governing the spike-timing-dependent calcium influx during upstates (Evans et al. 2013). Thus CDI serves to increase the temporal specificity of the calcium transient during upstates, limiting large calcium transients to spikes occurring early in the upstate and therefore limiting the subset of inputs able to undergo synaptic plasticity. Calcium through NMDA receptors also becomes more tightly controlled temporally during striatal development, as the decay time constant of NMDA receptors becomes faster with age (Hurst et al. 2001). A subunit shift from the slower GluN2B to the faster GluN2A during development (Chapman et al. 2003) can cause a narrowing of the timing window within which synaptic plasticity can occur (Evans et al. 2012).

The change in channel density and CDI of L-type calcium channels may contribute to an age-dependent change in striatal synaptic plasticity. L-type calcium channels are required for striatal long-term depression (Adermark and Lovinger 2007), whereas NMDA receptor channels are required for striatal long term potentiation (Calabresi et al. 1992). An increase in L-type calcium channels at eye opening may shift the balance of synaptic plasticity from potentiation to depression. On the other hand, striatal long-term potentiation requires dopamine (Centonze et al. 1999), which activates PKA and enhances the phosphorylation state of both NMDA receptors (Snyder et al. 1998) and L-type calcium channels (Dittmer et al. 2014; Rankovic et al. 2011). PKA phosphorylation of NMDA receptors enhances their calcium permeability (Murphy et al. 2014; Skeberdis et al. 2006). Here we have shown that PKA phosphorylation of HVA calcium channels reduces CDI, increasing the calcium influx. Because the effect of PKA phosphorylation is greater after eye opening, and dopamine leads to PKA activity in D1 MSNs, this developmental change may increase the sensitivity of bidirectional synaptic plasticity to dopamine through PKA phosphorylation of both NMDA receptors and HVA calcium channels.

Conclusions.

We have shown that CDI increases in the medial, but not lateral, striatum at eye opening and have determined that the increase in CDI depends on L-type calcium channels and is dependent on their phosphorylation state. Future studies are needed to determine whether visual input directly causes this CDI increase, and to examine the effects of striatal CDI enhancement on behavior.

GRANTS

This work was supported by Office of Naval Research Grant MURI N00014-10-1-0198 and National Insitute on Drug Abuse Grant R01-DA-038890. R. C. Evans was supported by National Institute of Neurological Disorders and Stroke Grant NRSA1F31-NS-066645.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: R.C.E. and K.T.B. conception and design of research; R.C.E., G.A.H., and S.L.H. performed experiments; R.C.E., G.A.H., and K.T.B. analyzed data; R.C.E., G.A.H., and K.T.B. interpreted results of experiments; R.C.E., G.A.H., and K.T.B. prepared figures; R.C.E. and K.T.B. drafted manuscript; R.C.E., G.A.H., S.L.H., and K.T.B. edited and revised manuscript; R.C.E., G.A.H., S.L.H., and K.T.B. approved final version of manuscript.