Strength of signaling to CREB in SCG neurons: dependence on activation of L-type Ca²⁺ channels and a CaM kinase pathway. (A) Nuclear pCREB levels after a 3-min, 40-mM K⁺ or FPL 64176 stimulation. Drugs (KN-93, CaMK inhibitor; KN-92, inactive congener of KN-93; U0126, MEK inhibitor; KT 5720, PKA inhibitor; Bis [bisindolylmaleimide I], PKC inhibitor) were added 1 min before stimulation. Some cultures were preloaded with 100 μM BAPTA-AM for 30 min at 37°C, stimulated in 0 mM of extracellular Ca²⁺ and EGTA (0 Ca²⁺) or 100 μM Ca²⁺ and no EGTA (low Ca²⁺), and/or stimulated in 200 μM La³⁺ or Cd²⁺ after preincubation for 1 s. Data are from two or more experiments of >40 neurons each. (B) Cumulative histograms of pCREB levels in individual neurons from a representative experiment. Cells stimulated for 3 min as indicated. Infecting cells with adenovirus expressing dihydropyridine-insensitive L-type Ca²⁺ channel (DHPi) rescued signaling to CREB. (C–E) Data from a representative experiment. Cells were either mock stimulated in 5-mM K⁺ Tyrode's (t = 0) or stimulated with solutions containing 20, 30, or 40 mM K⁺ for the time indicated, placed back in 5 mM K⁺ solution, and then fixed at 3 min. (C) 5–d in vitro SCG neurons stained for pCREB and MAP2; nuclei counterstained with DAPI. Bar, 50 μm. (D) Cumulative histograms of pCREB levels from individual neurons. (E) Mean pCREB levels plotted against stimulation time; [K⁺] as indicated. Dashed lines are linear fits of initial data points; slope used as index of CREB signal strength. (F) Isochronic nuclear pCREB levels measured at 3 min (•); CREB signal strength (□), measured as in E. Isochronic data from three independent experiments, except the 10- and 15-mM K⁺ data are the mean of >15 cells from a single duplicate experiment. Error bars represent SEM.

CREB signal strength is steeply dependent on channel P_o but not Ca²⁺ flux per se. (A) Log–log plot of signal strength versus Ca²⁺ flux for the [K⁺]_o indicated. Solid line shows the linear fit of the data. (B) Whole-cell Ca²⁺ flux (as in Fig. 2 D), in the presence of 0, 20, or 200 μM Cd²⁺. (C) CREB signal strength for 40 K⁺ with 0, 20, or 200 μM Cd²⁺ measured as described in Fig. 1 and Materials and methods; n = 3–6 experiments. The solid line is a linear fit of the Cd²⁺ data. For comparison, the dashed line and gray data points are reproduced from A. Error bars represent SEM.

L-type channel activity produces phosphoCaMKII puncta near the cell surface. (A) SCG neurons stimulated with 40 mM K⁺ for 0, 10, or 60 s and immediately fixed and stained for Map2 (red) and pCaMKII (green). Arrows point to pCaMKII puncta (magnified in the inset). Background pCaMKII staining was subtracted to highlight punctate staining (see Materials and methods). Bar, 20 μm (B) pCaMKII puncta weight from experiments as in A. (C) Nimodipine blocks pCaMKII puncta formation. (D) pCaMKII puncta weight from cells treated for 60 s as shown on a log–log plot against the rise in bulk Ca²⁺ (Δ[Ca²⁺]_i are taken from Fig. 4). For all, n = 3 independent experiments. Error bars represent SEM.

Knockdown of α and βCaMKII inhibits L-type channel signaling to CREB. (A) pCREB levels from neurons stimulated for 2.5 s with 5- or 40-mM K⁺ Tyrode's and transferred to a 5 mM K⁺ solution for 45 s before fixation. Cultures were infected with GFP alone, α and βCaMKII, or nonsilencing lentiviruses. Data are from three to six independent experiments (15–30 cells from duplicate coverslips in each). *, P < 5 × 10⁻¹⁷; **, P < 5 × 10⁻¹⁰. (B) Total nuclear CREB levels in neurons infected with GFP alone or α and βCaMKII lentiviruses. Data from three independent experiments, two done in parallel with pCREB experiments in A (46–87 neurons). (C) Peak Ca²⁺ elevations (Δ[Ca²⁺]_i) from sister cultures to those in B. Data are from seven separate experiments (31–49 cells). Error bars represent SEM.

CREB signal strength is a steep function of membrane potential and channel activity. (A) SCG neuron under current clamp in TTX. In 4 mM K⁺, the cell rested at −60 mV. Exposure to [K⁺] as indicated caused stepwise depolarizations. (B) Membrane potential plotted against [K⁺]_o. Slope of linear fit is 58.1 mV per 10-fold change in [K⁺]_o. (C) CREB signal strength, measured as in Fig. 1 E and in Materials and methods (five to eight experiments), plotted against voltage. 100% is the full change in pCREB from baseline to maximal levels. Linear fit shows that signal strength is steeply voltage dependent, as if governed by a gating particle with valence (z) between 4 and 5. (D) Voltage-clamp recordings with 2-mM Ca²⁺ Tyrode's as bathing solution. Pulse depolarizations to the voltages indicated. Traces are the mean of ten consecutive sweeps at each voltage. (E) Ca²⁺ flux (▪; as in D) and signal strength (○) plotted versus membrane potential. Solid lines are single exponential fits. Dashed lines indicate the voltages generated by each [K⁺]_o (n = 5–8 experiments). Error bars represent SEM.

Bulk [Ca²⁺]_i is neither necessary nor sufficient to dictate CREB signal strength. (A) Despite Cd²⁺ effects on Ca²⁺ flux, release from intracellular stores might maintain bulk [Ca²⁺]. (B) Fura-2 Ca²⁺ imaging experiment. Various [K⁺]_o with and without Cd²⁺ were applied as indicated. (C) Increased [Ca²⁺]_i plotted against Ca²⁺ flux (from Fig. 3). Δ[Ca²⁺]_i data from 10–23 cells and three to five experiments. Linear fit correlation coefficient equals 0.995 (P < 0.0005). (D) Log–log plot of CREB signal strength versus Δ[Ca²⁺]_i for the indicated conditions. Solid lines are linear fits through the data points collected with varying [K⁺]_o or [Cd²⁺]. (E) Possible gating-driven conformational changes working in concert with elevated bulk Ca²⁺. (top) Hypothetical conformational change repositions a local signaling determinant (circle). When combined with high local Ca²⁺ elevations, strong signaling to the nucleus is engaged (green). In the Ca²⁺-free solution (0 Ca²⁺), the conformational change remains, but the signaling determinant is not activated (red). (bottom) Can conformational change contribute to signaling in conjunction with increases in bulk Ca²⁺. (F) Nuclear pCREB levels from cells stimulated for 10 s in 2 mM Ca²⁺ (high local Ca²⁺) or for 15 s in the presence of caffeine with 0 Ca²⁺ (bulk Ca²⁺). Error bars represent SEM.

Knockdown of α and βCaMKII eliminates the formation of pCaMKII puncta. αCaMKII (A) and βCaMKII (B) mRNA levels in SCG neurons assessed by quantitative RT-PCR. Some cultures infected with lentiviruses expressing GFP alone, shRNAs against α and βCaMKII, or nonsilencing control shRNAs. Data normalized against three housekeeping genes and plotted relative to the GFP only (empty vector) control (see Materials and methods). n = 3–4 experiments done in triplicate. (C) SCG neurons infected as in A and B and stained with an antibody against βCaMKII; nuclei were counterstained with DAPI. Bar, 50 μm. (D) pCaMKII puncta weight from cells infected with α and βCaMKII or nonsilencing control lentiviruses, stimulated for 10 s with 40 mM K⁺ and immediately fixed. Data from two independent experiments done in duplicate (10–15 neurons each). Error bars represent SEM.

Model of local CaM saturation and frequency-dependent CaMKII activation. Schematic of the effects of three different types of stimulation on single channel kinetics and subsequent local effects on [Ca²⁺], Ca²⁺/CaM levels, and CaMKII activity that lead to CREB phosphorylation in the nucleus. (A) Individual L-type channels in the open state. Ca²⁺ flows through the channel (red arrow) and results in a plume of Ca²⁺ near the pore (red/yellow gradient). Pore block with Cd²⁺ markedly decreases flux (dashed red arrow), resulting in a proportionately lower [Ca²⁺] in the plume. Local CaM molecules (dots) are either in the apo-CaM form (gray) or Ca²⁺ bound (green). A local CaM target (presumably CaMKII) is depicted as a blue square. (B) Schematized record of L-type channels opening briefly and intermittently during depolarization to give rise to unitary current (black). Closed intervals are abbreviated by increasing depolarization. Open channel current amplitude is decreased by Cd²⁺. Local Ca²⁺ transients (red) mirror single channel currents, rising and falling quickly. During channel opening, the local Ca²⁺ levels saturate the local pool of apo-CaM, resulting in large elevations in Ca²⁺/CaM even when Ca²⁺ transients are diminished by Cd²⁺. Thus, transient elevations of Ca²⁺/CaM, shown as solid green bars, match the channel openings in frequency but are not greatly affected by the magnitude of the single channel current. (C) Local Ca²⁺/CaM pulses each have an incremental effect on local CaMKII activity (blue trace). With mild depolarization, pulses arrive at a low frequency and CaMKII activation decays between pulses. Upon stronger depolarization, interpulse intervals shorten and decay of activation is slowed by intersubunit phosphorylation, resulting in greater accumulation of CaMKII activation. This holds even when i_Ca is decreased with Cd²⁺. (D) Schematic representation of data on CREB signal strength. The seemingly paradoxical finding that potent block of Ca²⁺ flux (red) with Cd²⁺ has a relatively weak effect on signal strength (blue) can be explained by local CaM saturation, allowing for potent local CaMKII activation even when Ca²⁺ flux is greatly decreased.