Article Figures & Data
Figures
- Fig. 1.
Electrically evoked intracellular calcium concentration in the cell body of DRG neurons measured by calcium imaging with fura-2. Responses to 1, 6, 12, 18, 36, 54, and 90 action potentials at 10 Hz are shown, with the total calcium concentration–time integral indicated for each stimulus in brackets (nm min; A–G). A rapid increase in intracellular calcium and slower recovery after the stimulus is stopped are evident for all stimulus durations. A single action potential produces a small but detectable increase in [Ca2+]i in the cell body. The kinetics of calcium increase are well fit by a power function, using nonlinear regression (H–N) (see Table 1). The kinetics of recovery after the stimulus is terminated are well fit by a single exponential equation for stimulus durations <54 impulses (Q–U), but the recovery deviates from this function for longer bursts by having a longer sustained increase in intracellular calcium at ∼10 sec poststimulus (Table 1;arrow in insets to O,P). A double exponential function is required to adequately fit this slower recovery in the following 54 (P) and 90 (O) action potential bursts (see Table 1).
- Fig. 2.
Summary of pooled data from measurements on several neurons (mean and SEM) on the relation between action potential pattern and intracellular calcium transients. A, The peak concentration of intracellular calcium increases proportionately less for longer stimulus durations. This is consistent with the kinetics of calcium increase measured in single neurons, which follows a power function relation between action potential duration and calcium increase (Fig.1H–N) (n = 48 neurons, 6 neurons for each stimulus pattern). B, The calcium–time integral during one stimulus burst of 1–90 impulses. The integral calcium includes the increase in calcium from stimulation to recovery after the stimulus burst. The magnitude of the calcium–time integral for different burst durations is well fit by a linear regression to the duration of the burst (solid line= linear regression fit; dotted line = 95% confidence interval; n = 38 neurons).C, The total increase in calcium experienced during 30 min of stimulation. The comparison shown is for an equal number of impulses (540 impulses total during the 30 min stimulus) but delivered in 10 Hz bursts of 1, 6, 12, 18, 36, 54, and90 impulses and repeated at regular intervals (e.g., from one action potential every 3.3 sec to six bursts of 90 action potentials repeated every 5 min). Note that the total elevation in calcium experienced by the cell during the 30 min stimulus period (calcium–time integral) is relatively similar for stimulus bursts longer than ∼18 impulses (n = 38 neurons).
- Fig. 3.
Regulation of c-fos expression in response to different patterns of neural impulses. A, The effect of pulse train stimulation on c-fosexpression was studied by delivering the same number of impulses (540) in a 30 min period, but grouped into repeated bursts (10 Hz) of 1.8, 3.6, 5.4, and 9.0 sec, separated by 1, 2, 3, and 5 min interburst intervals, respectively. B, Expression ofc-fos mRNA was related inversely to the interval between bursts. This correlation held despite the countervailing increase in the duration of stimulus bursts (n = 31 cultures).C, Consistent with differences in c-fosmRNA in electrically stimulated cultures, Fos-β-galactosidase in DRG cultures from transgenic mice carrying the fos/LacZreporter gene decreases with increasing interburst intervals. The mean number of β-gal-positive neurons per microscope field is plotted for each culture. Means were determined from counts of neurons in 10 microscope fields in each culture, and differences in expression among the cultures stimulated with the four different impulse patterns were analyzed by ANOVA. The results indicate a highly significant difference among stimulus groups (p < 0.001;n = 15 cultures).
- Fig. 4.
Relation between electrically induced calcium transients and c-fos expression in response to different patterns of electrical stimulation. DRG neurons were stimulated with a total of 540 action potentials delivered in four different patterns for 30 min, as in Figure 3A, and the intracellular calcium transient was measured in the cell body by using ratiometric fluorescence imaging of cells loaded with the calcium indicator fura-2.A, D, G, J, The average calcium response of several neurons (n = 27, 35, 42, and 52 neurons) is shown in response to a single burst of stimulation at 10 Hz for different durations (1.8–9 sec). Higher peak calcium levels are reached after longer duration stimulus bursts, but the differences are small relative to the increase produced by a 1.8 sec burst (A vsJ). B, E, H, K, Long-term calcium recordings showing the average intracellular calcium levels in response to stimulus bursts repeated at different intervals (1–5 min). Note the full recovery of calcium to prestimulus levels after all stimulus patterns. C, F, I, L, Expression ofc-fos, measured by semiquantitative PCR (n = 31 cultures), does not correlate with the amplitude of the calcium transient but does correlate with the interval between stimulus bursts. This relation holds despite the countervailing differences in peak calcium produced by longer duration bursts.
- Fig. 5.
No differences in nuclear versus cytoplasmic calcium concentrations are seen by using confocal microscopy in response to action potential stimulation in mouse DRG neurons. Neurons were stimulated in multicompartment preparations with different burst durations (0.1–5 sec at 10 Hz) to stimulate calcium influx over a wide range of concentrations. The change in intracellular calcium was monitored by confocal microscopy along a line from the plasma membrane to the nucleus of each neuron, using ratiometric confocal microscopy in single line-scan mode in neurons loaded with the calcium indicator Indo-1. The intracellular calcium concentration in a region of interest adjacent to the plasma membrane and in the center of the nucleus is plotted. The points are well fit by linear regression with a slope of 1 (r2= 0.99;n = 14 test stimuli in six neurons from six cultures).
- Fig. 6.
CREB phosphorylation at Ser-133 in response to electrical stimulation of different patterns. Phosphorylation was determined by nuclear staining using an antibody that recognizes CREB phosphorylated at Ser-133 (P-CREB). The intensity of immunocytochemical staining was quantified in the nucleus of stimulated cells by densitometry of digitized images on a scale of 0–255. All values were normalized to the mean intensity of nuclear staining in unstimulated cells (A). A 10 min incubation in 60 mm KCl caused a large increase in the number and intensity of nuclei staining for P-CREB (F), which is evident by the rightward shift in the histogram of nuclear staining intensities. After electrical stimulation, localization of P-CREB in the nucleus varies with different stimulus patterns (B–E). The highest levels of nuclear staining were produced by short bursts repeated frequently (1.8 sec at 10 Hz, every minute) (B) or longer duration bursts repeated infrequently (9 sec at 10 Hz, every 5 min) (E). The intermediate patterns of stimulation produced less CREB phosphorylation at Ser-133 (C, D). No change in nuclear staining was evident after any stimulus when an antibody that recognizes both the phosphorylated and dephosphorylated forms of the protein was used (A–E).
- Fig. 7.
Western blot analysis of phosphorylation and activation of CREB. The antibody used for immunocytochemical studies stained a single band on immunoblots, consistent with the molecular weight of CREB. After electrical stimulation (10 Hz for 10 min), an increased amount of P-CREB was detected, as compared with unstimulated controls (Cnt.). Stimulation did not change the total amount of CREB (detected with an antibody that recognized both phosphorylated and nonphosphorylated CREB).
- Fig. 8.
The kinetics of changes in phosphorylated CREB in the nucleus of DRG neurons after electrical stimulation and comparison to MAPK. A, Stimulation at 1 or 10 Hz for 10 min caused a significant increase in CREB phosphorylation at Ser-133 (p < 0.001), with a greater increase in phosphorylation produced by higher stimulation frequency. Depolarization with 60 mm KCl induced a comparable increase in staining intensity (n = 589 neurons).B, The kinetics of CREB phosphorylation at Ser-133 are relatively rapid, with a significant increase detected after only 1 min of 10 Hz stimulation (p < 0.001). Near-maximal levels of CREB phosphorylation at Ser-133 are seen after 10 min of 10 Hz stimulation (n = 789).C, Dephosphorylation of CREB at Ser-133 followed slower kinetics than phosphorylation. Cells were stimulated at 10 Hz for 5 min to induce phosphorylation (poststimulus time = 0) and then fixed (P-CREB) or lysed (MAPK) between 1 and 25 min after the stimulus was stopped. Dephosphorylation was much more rapid for MAPK than for CREB. No significant dephosphorylation of CREB could be detected 1 min after the stimulus was terminated, but levels of MAPK phosphorylation were reduced by ∼50%. A small but sustained increase in P-CREB persisted 25 min after the 5 min stimulus was terminated (p < 0.001 relative to control; no significant difference between 5 and 10 min or 5 and 25 min; n = 1392 neurons) (p < 0.01 by ANOVA for MAPK;n = 17 dishes).
- Fig. 9.
Relation among c-fos expression, activation of CREB, and MAPK in response to different patterns of electrical stimulation. Expression of c-fos(A) did not correlate well with phosphorylation of CREB at Ser-133 (B). Phosphorylation of CREB was increased significantly after 30 min of 1.8 sec bursts repeated at 1 min intervals or 9 sec bursts repeated at 5 min, butc-fos expression was not increased in response to the latter stimulus. No increase in MAPK activation was observed in response to the stimulus that failed to induce expression ofc-fos (90/5). Phosphorylated CREB levels are summarized as the mean intensity of nuclear staining in DRG neurons ± SEM after electrical stimulation and normalized with respect to controls (n = 1547 neurons). C, In vivoactivation of the MAP kinase ERK1 was measured by Western immunoblotting, using the DAB detection method. Results are integrated intensity values normalized with respect to controls (mean ± SEM;n = 20 cultures).
- Fig. 10.
Time course of increase in CREB phosphorylation and MAPK activation in response to action potential bursts of 1.8 sec duration (10 Hz) repeated at 1 min intervals (18/1;filled symbols) and 9.0 sec duration repeated at 5 min intervals (90/5; open symbols) for 30 min. Neurons were analyzed just before the stimulus burst to estimate the residual increase in activation (t = 14.5 and 29.5 min) and compared with unstimulated controls (t = 0). A, Levels of CREB activation increase to near-maximal values in <14.5 min of stimulation with either stimulus pattern (18/1 = filled circles;90/5 = open triangles).B, A similar increase in MAPK activation with stimulus time is seen in response to stimulus bursts repeated at 1 min intervals (18/1 = filled circles). A single burst of stimulation at the end of the 30 min period of stimulation with these pulse patterns produces no further increase in CREB or MAPK activation (squares). A significant increase in P-CREB is produced from a single 10 sec burst of action potentials (arrow). A 9 sec burst increases MAPK activation significantly (open square; p < 0.05). Activation of MAPK was not sustained over 5 min interburst intervals, as shown by the failure to summate during the 30 min stimulus period (open circles; differences not significant comparing 0, 14, and 29.5 min). n = 1306 neurons in A and 42 cultures inB.
- Fig. 11.
Electrically induced expression ofc-fos is blocked by inhibitors of MAPK and CaM kinase. DRG neurons pretreated with 30 μm KN-62 (CaM kinase inhibitor), 50 μm PD098059 (MEK1 inhibitor), or both at the indicated concentrations for 1 hr showed no significant increase inc-fos mRNA levels after 30 min of electrical stimulation, as compared with controls (p = 0.53 by ANOVA). Stimulation was delivered in 1.8 sec bursts at 10 Hz, repeated at 1 min intervals (n = 22 dishes; *p < 0.003).
- Fig. 12.
MAPK activation in response to patterned action potentials. Activation of the ERK1 (p44) and ERK2 (p42) MAPK was determined in DRG neurons after eliciting a total of 540 impulses at 10 Hz in 1.8, 3.6, 5.4, and 9 sec bursts separated by 1, 2, 3, and 5 min intervals, respectively. The Western immunoblotting analysis was performed for activated MAPK after normalizing for total ERK1 by quantitative immunoblotting.
Tables
- Table 1.
Parameter fits using nonlinear regression to electrically induced calcium transients in the cell bodies of DRG neurons after action potential stimulation at 10 Hz for various durations
Number of action potentials Fitted parameter values a b c d e r2 Calcium increase y = a + b tc 1 0.608 38.1 1.02 – – 0.603 6 −1.00 132 0.686 – – 0.978 12 3.63 189 0.798 – – 0.973 18 −15.1 214 0.764 – – 0.976 36 −16.6 215 0.752 – – 0.980 54 −47.1 204 0.560 – – 0.975 90 −91.8 246 0.459 – – 0.955 Calcium recovery y = a + b exp (−t/c) 1 −12.0 36.0 3.60 – – 0.484 6 −1.03 119 1.69 – – 0.966 12 31.0 251 1.73 – – 0.986 18 28.5 300 1.96 – – 0.984 36 29.3 452 2.24 – – 0.989 54 23.2 460 2.71 – – 0.9791-160 90 51.2 504 2.99 – – 0.9681-160 Calcium recovery y = a + b exp (−ct) + d exp (−et) 54 24.4 450 0.439 27.5 0.103 0.981 90 27.6 447 0.545 130 0.0789 0.987 y, Intracellular calcium; t, time; r2, regression coefficient.
↵F1-160 A poor fit at approximately t = 10 sec because of sustained calcium increase.
