Vol. 281, Issue 3, R706-R715, September 2001
Temperature-sensitive properties of rat suprachiasmatic
nucleus neurons
Penny W.
Burgoon and
Jack
A.
Boulant
Department of Physiology and Cell Biology, College of Medicine,
The Ohio State University, Columbus, Ohio 43210
 |
ABSTRACT |
The hypothalamic suprachiasmatic nucleus (SCN) contains
a heterogeneous population of neurons, some of which are temperature sensitive in their firing rate activity. Neuronal thermosensitivity may
provide cues that synchronize the circadian clock. In addition, through
synaptic inhibition on nearby cells, thermosensitive neurons may
provide temperature compensation to other SCN neurons, enabling postsynaptic neurons to maintain a constant firing rate despite changes
in temperature. To identify mechanisms of neuronal thermosensitivity, whole cell patch recordings monitored resting and transient potentials of SCN neurons in rat hypothalamic tissue slices during changes in
temperature. Firing rate temperature sensitivity is not due to
thermally dependent changes in the resting membrane potential, action
potential threshold, or amplitude of the fast afterhyperpolarizing potential (AHP). The primary mechanism of neuronal thermosensitivity resides in the depolarizing prepotential, which is the slow
depolarization that occurs prior to the membrane potential reaching
threshold. In thermosensitive neurons, warming increases the
prepotential's rate of depolarization, such that threshold is reached
sooner. This shortens the interspike interval and increases the firing rate. In some SCN neurons, the slow component of the AHP provides an
additional mechanism for thermosensitivity. In these neurons, warming
causes the slow AHP to begin at a more depolarized level, and this, in
turn, shortens the interspike interval to increase firing rate.
circadian rhythms; threshold; intracellular recording
| |
INTRODUCTION |
CIRCADIAN
RHYTHMS IN MAMMALS are generated by the suprachiasmatic nucleus
(SCN) in the rostral hypothalamus. The SCN contains an internal
timekeeper that modulates several regulatory systems, making them
oscillate with a near 24-h rhythm. Daily changes in light and body
temperature are two cues that help synchronize the circadian clock
(2, 10). Light information from the eyes is directly
relayed to the SCN by the retinohypothalamic tract (6).
Information about body temperature may be sensed by the SCN itself,
because some SCN neurons are thermosensitive (11, 21, 24).
In vitro electrophysiological studies find that more than 10% of SCN
neurons are warm sensitive, showing significant increases in their
firing rates during increases in temperature (5, 11). Inasmuch as body temperature is highest during an animal's active period and lowest during the inactive period, warm-sensitive neurons could provide temporal information to synchronize the SCN clock. In
addition, some thermosensitive neurons synaptically inhibit other SCN
neurons, causing the inhibited neurons to be less sensitive to thermal
changes (5). This synaptic inhibition may be one mechanism
for temperature compensation in the biological clock.
Previous studies have not addressed the inherent mechanisms by which
suprachiasmatic neurons sense changes in temperature. Other
investigations, however, have studied mechanisms of thermosensitivity in neurons in the nearby preoptic region, and two different hypotheses have been proposed. Temperature could affect steady-state currents that
determine the resting membrane potential, such that warming depolarizes
the membrane potential, resulting in an increased firing rate
(19, 20). This hypothesis is opposed by research that
indicates that resting membrane potential is not an important factor in
preoptic neuronal thermosensitivity. Instead, these studies find that
temperature affects transient potentials that depend on preceding
action potentials (9, 13, 14). Spontaneously firing
neurons, for example, often display depolarizing prepotentials that
reach threshold to produce action potentials. Intracellular recordings
of preoptic thermosensitive neurons show that warming increases the
prepotential's rate of depolarization. This, in turn, shortens the
interval between consecutive action potentials, causing the firing rate
to increase.
The purpose of the present study was to determine the cellular
mechanisms responsible for neuronal thermosensitivity in the rat SCN.
These experiments employed whole cell intracellular recordings of
warm-sensitive and temperature-insensitive SCN neurons. For different
neuronal types, comparisons were made of thermal effects on resting
membrane potentials and transient potentials, including depolarizing
prepotentials as well as fast and slow afterhyperpolarizing potentials (AHPs).
| |
METHODS |
Preparation of tissue slices for electrophysiological recording.
As described previously (11), male Sprague-Dawley rats
(172-405 g) were housed in a temperature-controlled vivarium under a 12:12-h light-dark cycle for at least 2 wk before use. Rats were
decapitated in accordance with procedures approved by the National
Institutes of Health and the Ohio State University Laboratory Animal
Care and Use Committee. All decapitations occurred during the
subjective daytime to prevent phase shifting of circadian rhythms
(12). A block of hypothalamic tissue was cut and sectioned to thicknesses of 350-450 µm using a vibrating tissue slicer
(Vibratome). Coronal tissue slices containing the SCN were transferred
to a humidified, oxygenated (95% O2-5% CO2)
recording chamber and were constantly perifused at 1 ml/min with a 300 mOsm/kgH2O artificial cerebrospinal fluid
consisting of (in mM) 124 NaCl, 26 NaHCO3, 5 KCl, 2.4 CaC12, 1.3 MgSO4, 1.24 KH2PO4, and 10 glucose. This fluid was gas
saturated with 95% O2-5% CO2 and heated to
36-37°C using a thermoelectric Peltier assembly
(18). The thermoelectric assembly also allowed the tissue
slice to be periodically warmed and cooled. Tissue temperature was
monitored by a thermocouple placed in the perfusion medium directly
below the slices.
Recording intracellular activity.
After the slices had incubated for 2 h at 36-37°C, SCN
neurons were intracellularly recorded in the current clamp mode.
Neurons were identified using a blind-patch approach (3),
and recordings were made with 1.5- to 2-µm-tip glass microelectrodes
filled with a solution consisting of (in mM) 130 potassium gluconate,
10 EGTA, 10 HEPES, 2 ATP, 1 CaCl2, 1 MgCl2, and
5 NaCl. This solution was adjusted to 295 mOsm/kgH2O and pH
of 7.3. To minimize thermally induced changes in electrode tip
potentials, the ground electrode was maintained at a constant
temperature in an outer bath connected to the inner recording bath by a
filter paper bridge, as previously described (13). The
liquid junction potential for this solution has been experimentally
determined to be 12 mV (13), and this value was subtracted
from all reported values of recorded potentials. Temperature's effect
on liquid junction potential was not measured; however, a recent report
on neurons in rat hypothalamic slices determined that liquid junction
potential varied by <1 mV during temperature changes from 32 to 40°C
(16). All recordings were made using an Axon Instruments
200A amplifier. When determining spontaneous neuronal activity and
thermosensitivity, no holding current was applied to neurons.
Stimulation protocols to measure input resistance were conducted using
pCLAMP software from Axon Instruments.
Many whole cell recordings were conducted using a perforated patch
clamp mode to prevent washout of the intracellular contents (17). For these experiments, pipette tips were first
backfilled with an antibiotic-free electrode solution (identical to the
above pipette electrode solution but containing no ATP). The electrode was then filled with electrode solution containing 200-240 µg/ml nystatin or amphotericin B (Sigma). Electrodes were lowered onto neuron
membranes, and negative pressure was applied to establish gigaohm
seals. Antibiotics generally established electrical access into the
neuron within 5-15 min.
Membrane and action potentials were displayed on an oscilloscope and
recorded on a digital VCR tape recorder for computer analysis. In
addition to recording fast transient potentials, a recording also was
made of changes in the resting membrane potential (RMP). As previously
described (13), RMP was recorded by filtering out all
rapidly changing potentials (including action potentials). This
filtering was accomplished by a Grass 7DA driver amplifier adjusted to
give 0.5 Hz half-amplitude response. This RMP signal was recorded on a
polygraph and VCR tape recorder.
During each experiment, integrated firing rate, RMP, and tissue
temperature were monitored on a chart recorder. Neurons were also noted
for location within the dorsomedial SCN (dmSCN) and ventrolateral SCN
(vlSCN) (11). Cells were noted as being in an intermediate
(iSCN) region when electrode placement was in an area between the vlSCN
and dmSCN. Neuronal activity was transferred to an analog-to-digital
converter and stored on videotape for later analysis. Each neuron was
recorded for 2-5 min to determine its normothermic firing rate at
36-37°C. The neuron was then tested for temperature sensitivity
with a temperature cycle (duration 5-10 min) that covered a range
of at least 33-39°C but not exceeding 31-40°C.
Criteria for acceptable recordings were 1) action potential
amplitudes 
55 mV as measured from threshold voltage (which usually was 10-20 mV more positive than RMP) and 2) a stable
RMP (that did not change more than ±3 mV) during recordings of
spontaneous activity at 36-37°C.
Criteria for classifying SCN neuronal thermosensitivity were similar to
numerous investigations of preoptic temperature sensitivity (4). Thermosensitivity
(impulses · s
1 · °Cl) was
defined by the linear regression slope (or thermal coefficient) of
firing rate plotted as a function of temperature. This plot was
determined over a (minimal 3°C) temperature range in which a neuron
was most sensitive. With the use of the same criteria as previous
studies, warm-sensitive neurons had thermal coefficients of 0.8 impulses · s1 · °Cl or
greater, and temperature-insensitive neurons had lesser thermal coefficients. The temperature-insensitive neurons were further divided
into two subpopulations (11). Low-slope
temperature-insensitive neurons were almost completely unresponsive
to changes in temperature, and the absolute values of their thermal
coefficients were <0.2 impulses · s1 · °Cl.
Moderate-slope temperature-insensitive neurons exhibited modest changes
in their firing rates during changes in temperature, and their thermal
coefficients were 0.2 but <0.8
impulses · s1 · °Cl. In
addition, silent neurons were classified as cells that did not generate
spontaneous action potentials, although spike activity could be evoked
by application of depolarizing current pulses.
Data analysis.
Neuron firing rates and membrane properties were analyzed with
respect to temperature and SCN region. Membrane potential was plotted
as a function of temperature, and the slope of this plot defined
membrane potential thermosensitivity (mV/°C). Action potentials were
collected (minimum of 10) and signal averaged at each temperature for
examination of action potential and resting membrane properties. Figure
1 illustrates the components of transient
potentials that were analyzed as a function of temperature. All
measurements of potential amplitudes were taken relative to the
threshold voltage. Threshold voltage was determined by a set of
procedures explained in Estimation of action potential
threshold. The rate of rise of the depolarizing prepotential was
calculated from the slope of the membrane potential during the 4- to
20-ms period immediately preceding the action potential. The slow AHP
was determined from an averaged 5-ms window of data collected 8 ms
after the peak depolarizing afterpotential (DAP) value.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 1.
Components of transient membrane potentials. Before the
action potential, the depolarizing prepotential approaches threshold.
The rate of rise of the prepotential is measured during the 4- to 20-ms
interval before threshold. All amplitudes are measured from the
threshold voltage. The slow afterhyperpolarizing potential is the mean
of the 5-ms interval occurring 8 ms after the depolarizing
afterpotential (DAP).
|
|
Values are expressed as means ± SE unless otherwise indicated.
Multivariate analysis of variance with repeated measures compared neuronal responses as a factor of cell classification, temperature (cool = 32-33°C, neutral = 36-37°C,
warm = 39-40°C), and location (vlSCN, iSCN,
dmSCN). When appropriate, post hoc comparisons
(Scheffé's adjustment) identified where statistical differences
were located. A 
2 analysis was used to compare
the proportions of cell types by SCN region. Statistical significance
was defined as P 
0.05.
Estimation of action potential threshold.
It is possible that temperature may alter a neuron's action potential
threshold and, thereby, affect firing rate. There are no established
procedures for determining action potential threshold in spontaneously
firing neurons. Therefore, in this study, criteria for estimating
threshold were developed based on changes in the variability of the
membrane potential before and during the generation of action
potentials. Figure 2 illustrates that a
standard deviation (SD) plot can be generated when several action
potentials are superimposed (Fig. 2A) and signal averaged
(Fig. 2B). Figure 2A shows that the membrane
potential SD decreased as the membrane potential approached the action
potential. The estimation of threshold is based on the assumption that
(at a constant temperature, during successive action potentials)
threshold should remain relatively constant compared with changes in
the membrane potential. Therefore, the threshold potential would
correspond to the minimum impulse-to-impulse variability (i.e., SD)
immediately preceding the action potential. The minimum point in the SD
was selected (Fig. 2B, point a) that immediately
preceded the first large SD peak associated with the upstroke of the
action potential. To ensure that this point was a local minimum, two
criteria had to be met. First, all subsequent SD values after
point a had to be of greater value. This comparison was made
until it was clear that comparisons were being made well into the
action potential. Second, point a could not exceed the values in the preceding 0.5 ms by >1 SD of the SD values. With the use
of a 32-kHz sampling rate, this 0.5-ms window provided ~17 points for
comparison. If the value at point a exceeded the SD values
of the previous 0.5 ms by >1 SD, then it was assumed that point
a was not the minimum and another minimum point was selected for
threshold and rechecked for conformation to the above criteria. Figure
2B shows that the value at point a (i.e., SD = 0.52) fell within 1 SD (SD range: 0.46-0.54) of the values in the preceding 0.5 ms. When an SD value conformed to the above criteria,
the corresponding membrane potential at that same point in time was
defined as the estimated threshold voltage (point b). In
this case, the threshold voltage was 
39.1 mV.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2.
Estimation of threshold voltage in a moderate-slope
temperature-insensitive neuron (thermal coefficient: 0.4 impulses · s1 · °Cl;
resting membrane potential: 56 mV). A: action potentials
are superimposed (thin traces, n = 30) with the
standard deviation values (SD; thick line) associated with their
average. SD values decrease as the membrane potential approaches
threshold voltage. B: average action potential (thin line)
and SD for the neuron shown in A. A minimum point in the SD
values (point a, SD = 0.52) was selected using criteria
explained in METHODS. The membrane potential at the
corresponding time point (point b, 39.1 mV) is the
estimated threshold voltage.
|
|
| |
RESULTS |
Firing rate and temperature sensitivity.
The examples shown in Fig. 3 illustrate
the effect of temperature on three different types of spontaneously
firing SCN neurons. All of the neurons in Fig. 3 displayed depolarizing
prepotentials that reached threshold to produce action potentials. The
low-slope temperature-insensitive neuron had many postsynaptic
potentials and a low spontaneous firing rate that remained nearly
constant during changes in temperature (thermal coefficient: 0.1 impulses · s1 · °Cl). In
Fig. 3, the moderate-slope temperature-insensitive neuron (thermal
coefficient: 0.3 impulses · s1 · °Cl) had
a higher spontaneous firing rate and exhibited depolarizing prepotentials before action potentials. Similarly, the warm-sensitive neuron in Fig. 3 had a high spontaneous firing rate at 36-37°C and was very responsive to changes in temperature (thermal coefficient: 1.1 impulses · s1 · °Cl).
Seventy-one SCN neurons met the criteria for acceptable neurons defined
in METHODS and were characterized for firing rate and temperature sensitivity. These 71 neurons consisted of 17 low-slope temperature-insensitive neurons (24%), 37 moderate-slope
temperature-insensitive neurons (52%), 13 warm-sensitive neurons
(18%), and 4 silent neurons (6%). Eighteen neurons were recorded in
membrane-ruptured patch clamp, and 53 neurons were recorded in
perforated patch clamp. There were no significant differences in the
proportions of cell types recorded with these two techniques.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 3.
Effect of temperature on different types of suprachiasmatic nuclei
(SCN) neurons. Resting membrane potential and thermal coefficient:
low-slope temperature-insensitive neuron, 54 mV, 0.1 impulses · s1 · °Cl;
moderate-slope temperature-insensitive neuron, 54 mV, 0.3 impulses · s1 · °Cl;
warm-sensitive neuron, 52 mV, 1.1 impulses · s1 · °Cl.
|
|
Of the 71 recorded neurons, 25 were in the dmSCN, 37 were in the vlSCN,
and the remaining 9 were in the iSCN. Because there are morphological
and physiological differences between dmSCN and vlSCN neurons (6,
25), comparisons were made on the proportions of neuronal types
in these two regions. In the dmSCN, 12% of the neurons were warm
sensitive, 48% were moderate slope temperature insensitive, and 40%
were low slope temperature insensitive. The vlSCN contained more
thermosensitive neurons, with 22% warm sensitive, 62% moderate slope
temperature insensitive, and 16% low slope temperature insensitive.
2 analysis showed that these regional differences
reached statistical significance (P = 0.05).
As suggested in Fig. 3, Table 1
indicates that of the spontaneously firing neurons, low-slope
temperature-insensitive neurons had the lowest firing rates, and
warm-sensitive neurons had the highest firing rates. There were no
significant differences in the RMPs of the different cell types in
Table 1. Accordingly, membrane potential does not appear to be a
primary reason for the differences in spontaneous firing rate.
Membrane potential and temperature sensitivity.
To determine the role of membrane potential in neuronal
thermosensitivity, measurements of RMP were made during changes in tissue temperature. Figure 4 shows the
effect of temperature on membrane potential and firing rate for the
warm-sensitive neuron in Fig. 3. This neuron's firing rate increased
during warming and decreased during cooling; however, RMP did not
contribute to these firing rate changes, because (as noted in Fig.
4C) the membrane potential thermosensitivity was very low
and negative (i.e., 0.1 mV/°C). Had RMP been a contributing factor
to this neuron's firing rate response, membrane potential
thermosensitivity would be positive in value, with cooling producing
membrane hyperpolarization and heating producing membrane
depolarization. Table 1 also shows that a thermally induced change in
membrane potential was not an important determinant of neuronal
thermosensitivity. There were no significant differences in membrane
potential thermosensitivity among the four different neuronal types.
Moreover, there was no correlation between firing rate
thermosensitivity and membrane potential thermosensitivity. For
example, membrane potential thermosensitivity was virtually identical
in the warm-sensitive neurons and low-slope temperature-insensitive
neurons; and membrane potential thermosensitivity was slightly greater
in the low-slope temperature-insensitive neurons, compared with the
moderate-slope temperature-insensitive neurons. In addition, the silent
neurons (having the lowest firing rate thermosensitivity) had the
greatest membrane potential thermosensitivity.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 4.
Effect of temperature on membrane properties of the SCN
warm-sensitive neuron shown in Fig. 3. A: firing rate
(impulses/s) and membrane potential (mV) during changes in tissue
temperature. Firing rate increased with warming and decreased with
cooling, but membrane potential remained relatively constant during the
temperature changes. B: firing rate is plotted as a function
of temperature. C: membrane potential is plotted as a
function of temperature. D: membrane potentials during
hyperpolarizing current injections. Input resistance is the slope of
each current-voltage plot. Resistance increased with cooling and
decreased with warming.
|
|
Effects of temperature on membrane input resistance.
As shown in Fig. 4D, input resistance was measured at three
different temperatures. At each temperature, the neuron received 10 (210-250 ms) hyperpolarizing current pulses (ranging from 20 to
110 pA). When the resulting membrane potential (mV) was plotted as a
function of current, input resistance was determined by the regression
coefficient. Resistance decreased with warming and increased with
cooling, and this response was observed in all neurons, regardless of
their thermosensitivity. For each neuron tested, there was a
significant difference between the input resistances measured in
hypothermic and hyperthermic ranges.
Input resistance of 43 neurons was measured at cool, neutral, and warm
temperatures. Table 2 shows the mean
input resistances for each neuronal population at the three temperature
ranges. The average resistance at thermoneutral temperatures was
327 ± 6 M
, and the range was 151-659 M. For the entire
population, resistance significantly increased with cooling and
significantly decreased with warming. In addition, at 36-37°C,
resistances were different when comparing different cell classes.
Silent neurons had significantly lower resistances compared with the
other cell types, and low-slope temperature-insensitive neurons had
lower resistances than the moderate-slope temperature-insensitive and warm-sensitive neurons. There were no significant differences in the
resistances of the moderate-slope temperature-insensitive neurons and
warm-sensitive neurons.
Thermal effects on action potential properties.
Figure 1 describes the parameters used to measure the transient
potential properties. Measurements of threshold, action potential amplitude, half-amplitude duration, fast AHP amplitude, DAP amplitude, slow AHP, and rate of rise of the depolarizing prepotential were collected from 44 neurons at all three temperatures (9 low-slope temperature-insensitive neurons, 25 moderate-slope
temperature-insensitive neurons, and 10 warm-sensitive neurons). Silent
neurons were not included in the analysis because they did not generate
spontaneous action potentials. Action potential amplitudes during
warming were significantly smaller compared with amplitudes at cold and neutral temperature (cool = 59.05 ± 0.64 mV, neutral = 60.01 ± 0.64 mV, warm = 50.96 ± 0.64 mV,
P < 0.01). Cooling prolonged the half-amplitude
duration (cold = 1.05 ± 0.01 ms, neutral = 0.95 ± 0.01 ms, warm = 0.93 ± 0.01 ms, P < 0.01) , and DAPs were significantly different when comparing DAP amplitudes at
warm and cool temperatures (cold = 5.64 ± 0.26 mV,
neutral = 5.02 ± 0.26 mV, warm = 4.73 ± 0.26 mV,
P < 0.05). Although these variables were significantly
affected by temperature, there were no differences between the cell types.
Effects of temperature on threshold.
Firing rate thermosensitivity was not affected by temperature-dependent
changes in the threshold voltage. Figure
5A illustrates the
signal-averaged traces at three different temperatures for the same
warm-sensitive neuron shown in Fig. 4. Threshold voltage did not
markedly change at different temperatures (Fig. 5B).
Additionally, despite a significantly higher firing rate at 39°C, the
threshold voltages at 36 and 39°C were nearly identical. Table
3 presents the effect of temperature on
the average threshold for the different types of spontaneously firing
neurons. These data illustrate that action potential threshold is not a
reliable predictor of either spontaneous firing rate or neuronal
thermosensitivity. Of the different neuronal types, for example,
warm-sensitive neurons had the highest firing rates at 36-37°C,
but their thresholds occurred at more depolarized levels compared with
the other cell types. If threshold was an important determinant of
firing rate, one might have predicted that high firing neurons have
thresholds occurring at more hyperpolarized levels. Moreover, by
definition, warm-sensitive neurons showed the greatest increases in
firing rate during warming, and yet their thresholds at 39-40°C
occurred at more depolarized levels (i.e., 29.8 mV) compared with
their thresholds at 36-37°C (i.e., 31.5 mV). Thus there were
no consistent changes in threshold that would explain the differences
in spontaneous firing rate or the thermal-induced changes in firing
rate.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 5.
Temperature effects on action potential, afterhyperpolarizing
potential (AHP), and threshold in the warm-sensitive neurons shown in
Figs. 3 and 4. A: superimposed signal-averaged action
potentials at 3 different temperatures: 33°C (n = 51), 36°C (n = 70), 39°C (n = 95).
Action potential and AHP amplitudes and durations increased with
cooling and decreased with warming. These same averaged action
potentials are shown in B, except the records are expanded,
truncated, and not superimposed. Arrows, threshold voltages (as
described in METHODS). Temperature did not affect threshold
voltages: 33°C, 27.2 mV; 36°C, 28.1 mV; and 39°C, 28.0 mV.
Resting membrane potential was 52 mV.
|
|
Effects of temperature on AHPs.
As suggested in Fig. 5A, warming tended to decrease the
amplitudes of action potentials and fast AHPs. Figure
6 indicates that there was a trend for
warming to decrease fast AHP amplitudes in all three neuron types;
however, only warm-sensitive neurons showed significantly smaller
amplitudes at 39°C.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 6.
Thermal effects on the fast AHP amplitudes in different
types of SCN neurons. Although warming decreased fast AHP amplitudes in
each group, the only significant reductions occurred in warm-sensitive
neurons (*P < 0.05).
|
|
Another example of thermal effects on the AHP is shown in the
warm-sensitive neuron in Fig. 7, where
warming not only reduced the fast AHP amplitude but also caused this
fast AHP to terminate at a more depolarized level. Moreover, in Fig. 7,
both the DAP and slow AHP were more depolarized at 39 than at 33 or
36°C. This contributed to the neuron's thermosensitivity, because it
allowed the neuron to reach threshold faster at 39°C. However, this
trend was not statistically significant when slow AHPs of
warm-sensitive neurons were compared at different temperatures,
nor were there significant thermal effects on the slow AHPs of
moderate-slope temperature-insensitive neurons.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 7.
Thermal effects on AHPs and membrane potential of a SCN
warm-sensitive neuron (thermal coefficient = 0.8 impulses · s1 · °Cl).
A: superimposed single records of action potentials at 3 different temperatures. B: superimposed signal averages of
fast and slow AHPs at 3 temperatures: 33°C (n = 58),
36°C (n = 69), 39°C (n = 82). Slow
AHP levels: 33°C = 59.8 mV; 36°C = 60.7 mV;
39°C = 55.9 mV. Warming from 33 to 36°C did not
substantially affect fast and slow AHPs; however, the depolarizing
prepotential started sooner at 36°C, thus shortening the interspike
interval. Warming to 39°C reduced the fast AHP and caused the slow
AHP to occur at a more depolarized level. This further shortened the
interspike interval. C: effect of temperature on the
recorded resting membrane potential (0.1 mV/°C).
|
|
Effects of temperature on depolarizing prepotential.
Most SCN neurons displayed depolarizing prepotentials that brought the
membrane potential to threshold to produce action potentials. As noted
in Fig. 1, the prepotential rate of rise was measured during the 4- to
20-ms interval before threshold. Figure 8
suggests that temperature has different effects on the prepotentials of different types of neurons. Temperature had little or no effect on the
prepotential rates of depolarization in the low-slope and moderate-slope temperature-insensitive neurons in Fig. 8A.
As a result, the interspike intervals of these two neurons remained relatively constant over a 32-39°C temperature range (Fig.
8B). This was not the case in the warm-sensitive neuron in
Fig. 8, in which warming to 39°C increased the prepotential's rate
of depolarization. This shortened the interspike interval and increased the firing rate. Conversely, cooling to 32°C decreased the
prepotential's rate of depolarization, which, consequently, lengthened
the interspike interval and decreased firing rate.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 8.
Effect of temperature on the depolarizing prepotential in 3 different types of SCN neurons. A: superimposed
averaged prepotentials. B: superimposed single records of
action potentials at 3 different temperatures. Temperature had little
or no effect on the prepotential in the low-slope
temperature-insensitive neuron. In the moderate-slope
temperature-insensitive neuron, cooling to 32°C reduced the
prepotential's rate of rise, which lengthened the interspike interval.
In the warm-sensitive neuron, cooling reduced the prepotential's rate
of rise and lengthened the interspike interval, whereas warming
increased the prepotential's rate of rise and shortened the
interspike interval.
|
|
Figure 9 describes the effect of
temperature on the prepotentials of each SCN neuronal population. When
comparing prepotentials by general cell type, rates of rise were
smallest in low-slope temperature-insensitive neurons, which had the
lowest spontaneous firing rates. Rates of rise were largest in the
warm-sensitive neurons, which had the highest firing rates
(P < 0.05). Accordingly, the rate of rise of the
depolarizing prepotential appears to be an important factor that
determines the spontaneous firing rates of SCN neurons.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 9.
Thermal effects on the depolarizing prepotential's rate
of rise in different types of SCN neurons. When comparing all neuronal
types at all temperatures, prepotential rates of rise were greatest in
warm-sensitive neurons and smallest in low-slope
temperature-insensitive neurons (P < 0.05). In the
warm-sensitive neurons, prepotential rates of rise were significantly
higher at 39°C compared with 36°C (P < 0.05).
|
|
Similar to the examples in Fig. 8, Fig. 9 suggests that prepotential
differences may account for thermosensitive characteristics in
different SCN populations. Temperature had no consistent effect on the
prepotential rates of rise in low-slope temperature-insensitive neurons. In moderate-slope temperature-insensitive neurons, there was a
trend toward increasing rates of rise with increasing temperatures; however, there were no significant differences between the three temperatures. In contrast, the population of warm-sensitive neurons had
prepotentials with significantly higher rates of depolarization at
39°C compared with 36°C (P < 0.05). Thus there is
evidence that the ionic currents determining the prepotential are an
important determinant of neuronal warm sensitivity.
| |
DISCUSSION |
Our previous study (5) identified mechanisms by which
synaptic activity contributed to the thermosensitivity of SCNs. The present study is the first investigation to examine SCN membrane properties to identify a mechanism of thermosensitivity. In this manner, inherently temperature-sensitive neurons within the SCN could
serve as local temperature sensors, capable of sending efferent thermal
information to nearby neurons. We demonstrated in a previous report
(5) that inhibitory synaptic output from these neurons can
also serve as an intercellular temperature compensatory mechanism to
other SCN neurons.
Temperature-dependent changes in neuronal firing rate could
theoretically be the product of thermosensitive changes in RMP or
action potential threshold. Therefore, in the case of a warm-sensitive neuron, elevations of temperature would depolarize membrane potential or lower the threshold voltage, resulting in an increased firing rate.
As shown in Figs. 4 and 6, the present study on SCN neurons and another
investigation on preoptic and anterior hypothalamic neurons
(13) revealed that there is little relationship between temperature and membrane potential in hypothalamic neurons. It is
possible that the relative insensitivity of membrane potential to
temperature may be a property of hypothalamic neurons in general. By
contrast, a strong temperature dependence in RMP has been observed in
cat spinal motoneurons (23) and in rat visual cortical
cells (26). Additionally, the present study also revealed
that threshold voltage remained stable during changes in temperature in
both temperature-sensitive and insensitive SCN neurons, although
temperature altered the size and duration of other components of the
action potential. Stable spike thresholds have also been reported in a
cooling study examining rat visual cortical cells (26). By excluding threshold as a determinant of firing rate temperature sensitivity, efforts were focused on the transient potentials within
the interspike interval.
The AHP and prepotential are both part of the interspike interval, and
both of these membrane properties displayed temperature-dependent changes in warm-sensitive neurons. However, slow AHP values for warm-sensitive neurons were not statistically different from
temperature-insensitive neurons. This negated the idea that warming
caused smaller amplitudes in fast AHPs, thereby producing a more
depolarized slow AHP and a decreased time to reach threshold. Instead,
the main contributor to SCN inherent temperature sensitivity is in the
depolarization of the prepotential. Figure 8 illustrated temperature
dependence in the rate of depolarization of the prepotential that was
characteristic of warm-sensitive neurons and, to a lesser extent,
moderate-slope temperature-insensitive neurons. Previous studies of SCN
neuron electrophysiology have examined the contribution of the
prepotential to spontaneous firing rate activity (1, 22).
The current study shows that the prepotentials of some neurons have the
additional role of determining temperature sensitivity. In
warm-sensitive neurons, the rate of rise of the prepotential increased
during warming, causing shortening of the interspike interval and
increased firing rate. This response was consistent in warm-sensitive
neurons and appears to be the primary mechanism for neuronal
thermosensitivity. Prepotentials are present in temperature-insensitive
SCN neurons, as well; however, temperature was less effective in
influencing their rate of rise to threshold.
Future investigations to study the ionic basis for the
prepotential should prove to be complex. In SCN neurons, Akasu and colleagues (1) attributed the prepotential to an
inwardly rectifying, nonspecific cation current
(IH) and a low-threshold calcium current. Pennartz et al. (22), however, reported that prepotentials
were not altered in the presence of IH current
antagonists. They, instead, proposed that prepotentials were the
product of a slowly inactivating sodium current. In nearby preoptic and
anterior hypothalamic neurons, the temperature-dependent inactivation
of a potassium A current is considered to be an important contributor
to the rate of rise of the prepotential (14). This
hyperpolarizing current is activated during the interspike interval and
is an important determinant of firing rate (7, 8, 15).
Warming causes more rapid inactivation of this hyperpolarizing current,
presumably allowing membrane potential to depolarize at a faster rate
toward threshold, i.e., the steeper prepotential seen in
temperature-sensitive neurons (14). It is possible that
while one type of current may be responsible for production of the
prepotential, another current could serve to modulate the prepotential
response. Given the heterogeneity of neurons within the SCN,
experiments will have to be carefully developed to understand the ionic
basis of SCN temperature sensitivity and temperature compensation.
| |
ACKNOWLEDGEMENTS |
The authors are grateful to D. A. Burgoon for assistance with
analysis and interpretation of the data.
| |
FOOTNOTES |
J. A. Boulant is the Hitchcock Professor of Environmental
Physiology at the Ohio State University College of Medicine.
This research was supported by National Institutes of Health (NIH)
Grant NS-14644 and by an NIH Neural Development, Plasticity and
Regeneration training grant (NS-07291).
Address for reprint requests and other correspondence: J. A. Boulant, Dept. of Physiology & Cell Biology, College of Medicine, Ohio State Univ., 1645 Neil Ave., Columbus, OH 43210 (E-mail: boulant.1{at}osu.edu).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 16 January 2001; accepted in final form 18 April 2001.
| |
REFERENCES |
1.
Akasu, T,
Shingo S,
and
Hasuo H.
Inward rectifier and low-threshold calcium currents contribute to the spontaneous firing mechanism in neurons of the rat suprachiasmatic nucleus.
Pflügers Arch
425:
109-116,
1993[Web of Science][Medline].
2.
Barrett, KR,
and
Takahashi JS.
Temperature compensation and temperature entrainment of the chick pineal cell circadian clock.
J Neurosci
15:
5681-5692,
1995[Abstract].
3.
Blanton, MG,
Lo Turco JJ,
and
Kriegstein AR.
Whole cell recording from neurons in slices of reptilian and mammalian cerebral cortex.
J Neurosci Methods
30:
203-210,
1989[Web of Science][Medline].
4.
Boulant, JA.
Hypothalamic neurons regulating body temperature.
In: Handbook of Physiology. Environmental Physiology. Bethesda, MD: Am Physiol Soc, 1996, sect. 4, vol. 1, chapt. 6, p. 105-126.
5.
Burgoon, PW,
and
Boulant JA.
Synaptic inhibition: its role in suprachiasmatic nucleus neuronal thermosensitivity and temperature compensation in the rat.
J Physiol (Lond)
512:
793-807,
1998[Abstract/Free Full Text].
6.
Card, JP,
and
Moore RY.
The organization of visual circuits influencing the circadian activity of the suprachiasmatic nucleus.
In: Suprachiasmatic Nucleus: The Mind's Clock, edited by Klein DC,
Moore RY,
and Reppert SM.. New York: Oxford University Press, 1991, p. 51-76.
7.
Connor, JA,
and
Stevens CF.
Voltage clamp studies of a transient outward membrane current in gastropod neural somata.
J Physiol (Lond)
213:
21-30,
1971[Abstract/Free Full Text].
8.
Connor, JA,
and
Stevens CF.
Prediction of repetitive firing behaviour from voltage clamp data on an isolated neurone soma.
J Physiol (Lond)
213:
31-53,
1971[Abstract/Free Full Text].
9.
Curras, MC,
Kelso SR,
and
Boulant JA.
Intracellular analysis of inherent and synaptic activity in hypothalamic thermosensitive neurones in the rat.
J Physiol (Lond)
440:
257-271,
1991[Abstract/Free Full Text].
10.
DeCoursey, PJ.
Function of a light response rhythm in hamsters.
J Comp Cell Physiol
63:
189-196,
1964.
11.
Derambure, PS,
and
Boulant JA.
Circadian thermosensitive characteristics of suprachiasmatic neurons in vitro.
Am J Physiol Regulatory Integrative Comp Physiol
266:
R1876-R1884,
1994[Abstract/Free Full Text].
12.
Gillette, MU.
The suprachiasmatic nuclei: circadian phase shifts induced at the time of hypothalamic slice preparation are preserved in vitro.
Brain Res
379:
176-181,
1986[Web of Science][Medline].
13.
Griffin, JD,
and
Boulant JA.
Temperature effects on membrane potential and input resistance in rat hypothalamic neurones.
J Physiol (Lond)
488:
407-418,
1995[Abstract/Free Full Text].
14.
Griffin, JD,
Kaple ML,
Chow AR,
and
Boulant JA.
Cellular mechanisms for neuronal thermosensitivity in the rat hypothalamus.
J Physiol (Lond)
492:
231-242,
1996[Abstract/Free Full Text].
15.
Hille, B.
Ionic Channels of Excitable Membranes. Sunderland, MA: Sinauer Associates, 1992, p. 116-121.
16.
Hori, A,
Minato K,
and
Kobayashi S.
Warming-activated channels of warm-sensitive neurons in rat hypothalamic slices.
Neurosci Lett
275:
93-96,
1999[Web of Science][Medline].
17.
Horn, R,
and
Korn SJ.
Prevention of rundown in electrophysiological recording.
In: Ion Channels, edited by Rudy B,
and Iverson LE.. New York: Academic, 1992, p. 149-155.
18.
Kelso, SR,
Nelson DO,
Silva NL,
and
Boulant JA.
A slice chamber for intracellular and extracellular recording during continuous perfusion.
Brain Res Bull
10:
853-857,
1983[Web of Science][Medline].
19.
Kiyohara, T,
Hirata M,
Hori T,
and
Akaike N.
Hypothalamic warm-sensitive neurons possess a tetrodotoxin-sensitive channel with a high Q10.
Neurosci Res
8:
48-53,
1990[Web of Science][Medline].
20.
Kobayashi, S,
and
Takahashi T.
Whole-cell properties of temperature-sensitive neurons in rat hypothalamic slices.
Proc R Soc Lond B Biol Sci
251:
89-94,
1993[Medline].
21.
Miller, JD,
Cao VH,
and
Heller HC.
Thermal effects on neuronal activity in suprachiasmatic nuclei of hibernators and nonhibernators.
Am J Physiol Regulatory Integrative Comp Physiol
266:
R1259-R1266,
1994[Abstract/Free Full Text].
22.
Pennartz, CMA,
Bierlaagh MA,
and
Geurtsen AMS
Cellular mechanism underlying spontaneous firing in rat suprachiasmatic nucleus: involvement of a slowly inactivating component of sodium current.
J Neurophysiol
78:
1811-1825,
1997[Abstract/Free Full Text].
23.
Pierau, FK,
Klee MR,
and
Klussmannn FW.
Effects of local hypo- and hyperthermia on mammalian spinal motoneurones.
Fed Proc
28:
1006-1010,
1969[Web of Science][Medline].
24.
Ruby, NF,
and
Heller HC.
Temperature sensitivity of the suprachiasmatic nucleus of ground squirrels and rats in vitro.
J Biol Rhythms
11:
126-136,
1996[Abstract/Free Full Text].
25.
Van den Pol, AN.
The suprachiasmatic nucleus: morphological and cytochemical substrates for cellular interaction.
In: Suprachiasmatic Nucleus: the Mind's Clock, edited by Klein DC,
Moore RY,
and Reppert SM.. New York: Oxford University Press, 1991, p. 17-50.
26.
Volgushev, M,
Vidyasagar TR,
Chistiakova M,
Yousef T,
and
Eysel UT.
Membrane properties and spike generation in rat visual cortical cells during reversible cooling.
J Physiol (Lond)
522:
59-76,
2000[Abstract/Free Full Text].
Am J Physiol Regul Integr Comp Physiol 281(3):R706-R715
0363-6119/01 $5.00
Copyright © 2001 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
