- Email: [email protected]
Behavioral kinetics: Dynamics of the unconditioned response to footshock
Physiology & Behavior, Vol. 29, pp. 957-960. Pergamon Press, 1982. Printed in the U.S.A.
Behavioral Kinetics: Dynamics of the Unconditioned Response to Footshock F. D A R E L L T U R N E R A N D F R E D H. G A G E s
Chemistry of Behavior Program, Texas Christian University, Fort Worth, TX 76129 Received 24 April 1982 TURNER, F. D. AND F. H. GAGE. Behavioral kinetics: Dynamics of the unconditioned response to footshock. PHYSIOL. BEHAV. 29(5)957-960, 1982.--The purpose of this paper is to present an instrument and behavioral assay for monitoring the unconditioned response to footshock in the rat. The instnunent provides measures of the total force exerted by a subject in response to a footshock, the duration of the subject's response, and the subject's latency to respond to the stimulus. The assay is developed to minimize the error variance inherent in behavioral measurements. In the second part of this paper, the relationships between the response and the physical parameters of the stimulus are investigated. The total force exerted by a subject, and the duration of the subjects response are linearly related to the multiplieative intea'actionof stimulus current and stimulus duration. The subject's latency to respond is curvilinearly related to stimulus current alone. The form of this relationship is hyperbolic. These results are discussed in the context of the measurement of behavioral and psychophysieal processes in psychobiological research. Behavioral measurement
Footshock
Psychophysics
IN recent years, electrophysiological, biochemical and anatomical assays have been developed which enable researchers to reliably monitor small alterations in the physiological functioning of the nervous system. The development of behavioral assays, however, has not paralleled the development of physiological assays. When a behavioral response is to be related to underlying physiological function, the validity of the conclusions drawn will depend on the reliability and sensitivity of the behavioral assay which is used. Because of this crucial role of behavioral measurements in psychobiological research, the development of behavioral measurements requires extensive consideration. Several behavioral assays have been developed for measuring the unconditioned response to footshock. Recently developed behavioral assays [2, 4, 6, 8] employ instruments which measure the amplitude of the largest movement made by a subject in response to the stimulus. This type of instrument has been used successfully to investigate the behavioral effects of several physiological manipulations [2, 4, 6], and the measurements obtained appear to be reliable and reasonably sensitive to behavioral changes. The instrument to be presented in this paper is an elaboration of this type of instrument. In general, when a subject is presented with a footshock stimulus, a number of movements are made. Thus, if the only measurement of a subject's response is the amplitude of the largest o f these movemeats, the measurement does not permit a ~ asl,~sment of the aOxtal response. The instntment which will be discussed in this paper provides a
Sensory reactivity
Unconditionedresponse
more complete assessment of the subject's behavior by incorporating several measures of the response to footshock. When the physical parameters of a stimulus are held constant, there are two primary sources of variance which contribute to the measured variance in the responses of subject's to that stimulus. One source of variance is attributable to differences in the expected response to the stimulus of individual subjects. The second source of variance is associated with differences within the responses of individual subjects to a presentation of the stimulus. If there is no systematic alteration in a subject's response due to repeated stimulus presentations, the mean of single responses to several presentations of a stimulus may be used as an approximation of the mean of the possible responses to a single presentation of a stimulus, thus reducing the between subject variance. Pilot studies in our laboratory indicate that when more than five trials are used to calculate the subject's mean response, the reductions in the between subject variance are not appreciably greater than when five trials are used. Therefore, in the current study we have chosen to use no more than five stimulus trials for any sub~ct in any test session. A footshock stimulus has two parameters which may affect the response o f the subject. The effect of stimulus current intensity (I) on the response to footsboc~ has been investigated in a nua0~r of studies using a variety ,of tech, niques [2, 4, 6, 8]. The effects of the duration (t) of a footshock stimulus have not been investigated with respect to the unconditioned response to footshock. A study by
tThis research was partially wppeo, zd by N ~ H research grant MH34609, and the Texas Christian UaiversRy Research Fotmdation. 2Requests for ~ s should be addressed to F. H. Gage at the above address.
Copyright © 1982 Pergamon Press~0031-9384/82/llO957-(M$03.00/0
958
TURNER AND GAGE
A
B i
I 10 1K
,
lOOK 1or
"ft*~ 1
J--1 A
In] C ~TII~IBL
1 SECOND
AJ
IC1
1/2
LF353N
I,~,~ r --
r
o
q
_.J
E
.
.
.
.
.
.
~
F , ~
H
J
IC3
I__
I I _,
L. ~c' r 1,,~ ~M
÷ ¢,8
--]
I
,
1/2 LF353N
,o . . . . . . . . . . tO5
LM4066
ic6
LMS55CN
IC7
1/4 LM324N
ic8
LM556N
IC9
LM555CN
let0
1/2 LF353N
ICll
1/2 LF353N
tc12
112 UA1458
tC13
1/2 UA1458
~--~
~
....
IOOK~IOT 15K ~l
~o~
,
I
,~.~ 1~
F
t7
FIG. 1 (A) Schematic circuit diagram and key. (B) Schematic representation of circuit response. Letters in B represent responses recorded from the correspondingly lettered points in A. For detailed wiring diagrams of IC6, IC8, and IC9 see [3].
Church, Raymond, and Beauchamp [1] demonstrated that the suppression of free operant responding by f o o t s h o c k is a function of the multiplicative interaction of stimulus current and duration. The purpose of this paper is twofold. First we will present a sensitive and reliable instrument for use in measuring the unconditioned response to footshock in the rat. We will then discuss the use of this assay to determine the relationship between the response of subjects to electrical footshock and the physical parameters (duration and intensity) of the stimulus. METHOD
Subjects The subjects used in this study were Sprague-Dawtey rats between 80 and 100 days old. Subjects were individually housed under a 12 hour light/dark cycle with free access to food and water. Testing was done in dim light 4 to 8 hours after the beginning of the dark phase of the cycle.
Apparatus The basic apparatus used in this study has previously been described in detail by Gage, Armstrong and Thompson [2]. The basic apparatus consists of a test box which is wired to present footshock, and balanced on a force transducer. The apparatus is designed so that forces exerted by a subject's movements in the test box are translated into voltages and recorded with a polygraph. The modified apparatus used in this study electronically monitors the output of the polygraph driver amplifier and provides three indices of a subject's response to shock. Latency to respond (LAT) is the elapsed time from shock onset until the momentary force exerted by a subject exceeds a preset
threshold level. Response duration (DUR) is the period of time a subject's responses occur with an amplitude greater than a preset threshold level and a frequency greater than one per second. The third measure (INT) is proportional to the dynamic forces exerted by the subject integrated over the previously defined response duration. The electronic circuitry used to monitor the output o f the polygraph amplifier is shown schematically in Fig. 1A. The input is AC coupled at the input to IC2 and the output of IC2 rectified, so that only the voltages due to the subject's movements in the chamber are recorded. Temporal measurements (LAT and DUR) are made using linear voltage ramp circuits, so that all three measures may be recorded with a digital voltmeter. Scrambled shock stimuli are provided by a Grason Stadler model 700 shock generator. The shock schedule is controlled by conventional relay equipment. Initiation of a shock stimulus causes the monostable timer IC6 to produce a 1 msec output pulse. The output pulse from IC6 triggers the retriggerabie one-shot IC8 (see D and G Fig. IB). The output of the Tetriggcrable one-shot IC8 opens channel 1 of the analog s~,itch IC5 thus permitting the input from the polygraph to the monitoring circuitry (see B Fig. IB). The input from the polygraph is intc~ratod by IC3 to produce the integrated force measure (INT). The polygraph input also provides additional triggers to the re~riggerable one-shot IC8 via the output of the voltage comparator I(27. If a period greater than one second occurs during which the retriggcrablc one-shot IC8 receives no further input tri.ggcr pulses, its output will go low, thereby blocking the i l ~ t from the polygraph (see D Fig. 113). During the period in which the output of the retriggerable one-shot IC8 is high a constant 15 volts is applied to the input of the linear voltage ramp generator IC12. At the end of the response the output of this linear voltage ramp generator is read as the uncorrected
F O O T S H O C K R E S P O N S E DYNAMICS response duration (DUR) (see Fig. IB). The output of the R/S flip IC9 flop (which is set high by the initiation of a shock stimulus and reset by the first output from the voltage comparator IC7) provides a constant 15 V input to the linear voltage ramp generator IC10. At the end of the response, the output of the linear voltage ramp generator is read as latency to respond (LAT). Shorting across the feedback capacitors of the three integrating circuits (IC3, IC10, and IC12) will clear the device in preparation for another trial. The uncorrected duration of response is recorded from the time of shock onset until one second beyond the last response of the period of time wherein responses occur with a frequency of greater than one per second. To correct the recorded DUR it is necessary to subtract the recorded LAT plus one second from the recorded DUR. The corrected DUR is thus the period of time wherein responses of an amplitude greater than a threshold value occur with a frequency of greater than one per second (see Fig. 1B).
959
a
S
Q
Z
8
8 -2, o •
8 ,O0
,~
.tO
.IS
,20 .2S ,30 C[,IR~NT x [XIRflTION
.3S
.40
g
b
Calibration of Monitoring Circuitry The two temporal measurements (DUR and LAT) are calibrated by presenting a shock stimulus with no subject in the chamber. Under these circumstances no response can be recorded and DUR should read 1.00 V(1.00 see) and LAT should read 10.00 V (1000 msecs). These values were chosen for optimal range, resolution, and ease of reading. The outputs are adjusted by changing the settings on the ten turn potentiometers which control the variable gain of amplifiers IC 11 and IC 13. The polygraph amplifier is set such that when a 300 g mass is placed in the center of the chamber there is a 280 mV change in the static voltage at the input to IC1. The integrated force measure (INT) is calibrated by applying 15 volts to the input of IC1 with the monitoring circuitry held open. The change in voltage due to the charging of the capacitor at the input of IC2 is then integrated at IC3 and the output adjusted to read 2.00 V by setting the ten turn potentiometer on the variable gain amplifier IC4.
.so
,4S
eseoe oo 8 ,00
".0~
'.10
'.IS
'.20
'.2S '.~ CLIq~ENT x 0U~qTION
".~
'.40
'.4S
C
Separation of Responses from Spontaneous Movements Reduction of measurement error is an approach used to increase the sensitivity of a measurement. When measuring a subject's response to a sensory stimulus, one way to reduce measurement error is to insure that the response to the stimulus is clearly distinguishable from non-stimulus linked behaviors. This reduces the variance produced when the instrument measures irrelevant responses in addition to the stimulus evoked responses. The movements made in response to footshock are clearly distinguishable from movements which are not in response to footshock because of the larger amplitude of the subject's movements in response to footshock. Ten subjects were placed in the experimental chamber, and for each subject, the monitoring circuitry was opened ten times for one second on a VT 30 second schedule; however shock was not delivered. The maximum voltage occurring due to the spontaneous movements of the animal was then recorded during each one second period by monitoring the circuitry at point B (Fig. 1A) with a storage osciUiscope. The mean and standard deviation of the maximal voltages produced by the spontaneous movements of the subjects were calculated across animals and trials. The one-tailed critical value for a normalized score with p <0.001 was obtained from a table of normal probabilities. Using this critical value, and the calculated mean and standard deviation of the volt-
•
.00
".is
'.3o
".4s
'.ea
'.Ts
1
'.go
i.os
i.ao
i'.3s
1.540
COl[RENT
FIG. 2. (a) Integrated force (INT) of subject's response plotted relative to the multiplicative interaction of stimulus current and stimulus duration. (b) Duration of subject's response plotted relative to the multiplicative interaction of stimulus current and stimulus duration. (c) Latency to respond to the stimulus plotted relative to stimulus current. See text for a description of the units employed•
ages produced by the spontaneous movements o f the subjects, a threshold value was calculated. With this threshold, the probability of a voltage greater than threshold occurring due to the spontaneous movements of an average subject is less than 0.001.
960
T U R N E R AND G A G E
Procedure
In order to assess the effects of stimulus current (I) and duration (t) on the measured response, 18 subjects were each presented with one of the 18 possible combinations of the following stimulus amplitudes: 0.4, 0.5, 0.6, 0.8, 1.0, 1.3 mA; and durations: I00, 200, 300 msec. Each subject was presented with 5 stimuli on the first day of testing, and 5 stimuli on the second day. Presentations of stimuli were made on a VT 30 second schedule. Scatter plots were made for each of the three measures as a function of stimulus current, duration, and the multiplicative interaction of stimulus current and duration. These plots were examined for both linear and curvilinear trends. RESULTS The best predictor of a subject's response to a shock stimulus, as measured by I N T and DUR, was found to be the multiplicative interaction between stimulus intensity and stimulus duration. The relationship, shown in Fig. 2A and 2B, was found to be linear by a least squares criterion, (r=.94 for INT and r = . 8 8 for DUR). In nearly every case, a subject's latency to respond to the stimulus was less than the duration of the stimulus. It is therefore unlikely that the duration o f the stimulus has a significant effect on the subject's latency to respond to the stimulus. A scatterplot o f stimulus current (I) versus L A T revealed a curvilinear relationship between I and L A T which suggested a hyperbolic relationship between I and LAT. The linear transformation of a hyperbolic relationship is a double reciprocal plot [5]. This transformation produced a significant least squares solution ( r = - . 7 9 ) . The hyperbolic relationship between L A T and I is shown in Fig. 2C. DISCUSSION The results shown in Fig. 2A and 2B demonstrate that stimulus current and duration have equivalent effects on the subject's unconditioned response to the stimulus as measured by I N T and DUR. This implies that the effects o f the electrical charge passed across the subject on the subject's response summate along the temporal duration of the stimulus. L A T is the period of time from the onset of the stimulus until the subject's first response. The subject's response as measured by I N T an.d D U R demonstrates that the effects of the electrical charge passed across the subject on the subj e c t ' s response summates along the temporal duration of the stimulus. A hyperbolic relationship between I and L A T then suggests that LAT is a measure of the temporal summation of
electrical charge necessary to evoke a response from the subject (i.e., stimulus current multiplied by LAT is equivalent to the total charge passed across the subject at the time of the subject's first response). The presence of a Y-asymptote in the relation between I and L A T suggests that there is a minumum latency (Lm) to respond such that regardless of the amplitude of I, the subjects cannot respond more rapidly than Lm. Lm is representative of physical constraints on the subject, that is, the mimihai~idtime necessary for the subject to perceive the stimulus, and produce a coordinated movement. The value obtained for Lm in this study was 45 msec. Lm is actually the expected minimal latency to respond for an average subject, therefore, variance above and below this value would be expected among individual subjects. The presence of the X-asymptote in the I-LAT relationship suggests that there is some level of current (In,) below which the effects of I do not temporally summate so that a response cannot be evoked. Im then is the predicted threshold current for a stimulus of long duration. The value obtained for Im in this study was 0.29 mA. These results suggest that in experiments designed to investigate shock response thresholds, the current threshold will in large part depend upon the duration of the stimulus employed. If a long duration stimulus is employed, the recorded threshold will correspond to Ira. If a short duration stimulus is employed, the current threshold will reflect only the amplitude of current necessary for the summation of the electrical charge over the temporal duration of the stimulus to be large enough to evoke a response. In summary, we have demonstrated: (l) the unconditioned response to footshock is in response to the summated effects of electrical charge passed across the subject by the stimulus over the duration of the stimulus; and, (2) there appear to be two response thresholds to footshock, one to the summated effect of electrical charge over time, and the other is a current threshold, below which the effects of electrical charge do not summate to produce a response. We are convinced that the development of sensitive and reliable behavioral measures like the one described in this report are essential for an accurate and meaningful understanding of the physiological mechanisms underlying behavior. ACKNOWLEDGEMENTS We thank J. E. Springer and G. R. Stewart for assistance with data collection, and C. O. Larson for assistance in drawing figures. We also thank R. C. Bay, P. E. Moes, J. P. Kesslak, andR. G. Thompson for critical reading of this manuscript.
REFERENCES
1. Church, R. M., G. A. Raymond and R. D. l~auchamp. Response suppression as a function of intensity and duration of a punishment. J. comp. physiol. PsychoL 63:39 A.A.,1967. 2. Gage, F. H., D. R. Armstrong and R. G. Thompson. Behavioral kinetics: A method for deriving qualitative and quantitative changes in sensory responsiveness following septal nuclei damage. Physiol. Behav. 23: 479-484, 1980.
3. Jung, W. G. IC Timer Cookbook. Indianapolis, IN: Howard W. Sams and CO., Inc., 1977. 4. Leiter, D. S., A. S. Powers and H. S. Hoffman. The neural substrate of the startle response. Physiol. Behav. 25: 291-297, 1980.
5. Lewis D. Quantitative Methods in Psychology. New York: McGraw-Hill, 1960. 6. Lubar, L F., J. M. Brener, J. H. Deagle, R. Numan and W. J. Clemens. Effect of septal tesions on detection threshold and unconditioned response to shock. Physiol. Behav. 5: 459--463, 1970. 7. Nunnally, J. C. Psychometric Theory. N e w York: McGraw-Hill. 1978. 8. Smith, R. F. Scopolamine does not affect footshock sensitivity m the rat. Pharmac. Biochem. Behav. 8: 31-34, 1978.
Behavioral Kinetics: Dynamics of the Unconditioned Response to Footshock F. D A R E L L T U R N E R A N D F R E D H. G A G E s
Chemistry of Behavior Program, Texas Christian University, Fort Worth, TX 76129 Received 24 April 1982 TURNER, F. D. AND F. H. GAGE. Behavioral kinetics: Dynamics of the unconditioned response to footshock. PHYSIOL. BEHAV. 29(5)957-960, 1982.--The purpose of this paper is to present an instrument and behavioral assay for monitoring the unconditioned response to footshock in the rat. The instnunent provides measures of the total force exerted by a subject in response to a footshock, the duration of the subject's response, and the subject's latency to respond to the stimulus. The assay is developed to minimize the error variance inherent in behavioral measurements. In the second part of this paper, the relationships between the response and the physical parameters of the stimulus are investigated. The total force exerted by a subject, and the duration of the subjects response are linearly related to the multiplieative intea'actionof stimulus current and stimulus duration. The subject's latency to respond is curvilinearly related to stimulus current alone. The form of this relationship is hyperbolic. These results are discussed in the context of the measurement of behavioral and psychophysieal processes in psychobiological research. Behavioral measurement
Footshock
Psychophysics
IN recent years, electrophysiological, biochemical and anatomical assays have been developed which enable researchers to reliably monitor small alterations in the physiological functioning of the nervous system. The development of behavioral assays, however, has not paralleled the development of physiological assays. When a behavioral response is to be related to underlying physiological function, the validity of the conclusions drawn will depend on the reliability and sensitivity of the behavioral assay which is used. Because of this crucial role of behavioral measurements in psychobiological research, the development of behavioral measurements requires extensive consideration. Several behavioral assays have been developed for measuring the unconditioned response to footshock. Recently developed behavioral assays [2, 4, 6, 8] employ instruments which measure the amplitude of the largest movement made by a subject in response to the stimulus. This type of instrument has been used successfully to investigate the behavioral effects of several physiological manipulations [2, 4, 6], and the measurements obtained appear to be reliable and reasonably sensitive to behavioral changes. The instrument to be presented in this paper is an elaboration of this type of instrument. In general, when a subject is presented with a footshock stimulus, a number of movements are made. Thus, if the only measurement of a subject's response is the amplitude of the largest o f these movemeats, the measurement does not permit a ~ asl,~sment of the aOxtal response. The instntment which will be discussed in this paper provides a
Sensory reactivity
Unconditionedresponse
more complete assessment of the subject's behavior by incorporating several measures of the response to footshock. When the physical parameters of a stimulus are held constant, there are two primary sources of variance which contribute to the measured variance in the responses of subject's to that stimulus. One source of variance is attributable to differences in the expected response to the stimulus of individual subjects. The second source of variance is associated with differences within the responses of individual subjects to a presentation of the stimulus. If there is no systematic alteration in a subject's response due to repeated stimulus presentations, the mean of single responses to several presentations of a stimulus may be used as an approximation of the mean of the possible responses to a single presentation of a stimulus, thus reducing the between subject variance. Pilot studies in our laboratory indicate that when more than five trials are used to calculate the subject's mean response, the reductions in the between subject variance are not appreciably greater than when five trials are used. Therefore, in the current study we have chosen to use no more than five stimulus trials for any sub~ct in any test session. A footshock stimulus has two parameters which may affect the response o f the subject. The effect of stimulus current intensity (I) on the response to footsboc~ has been investigated in a nua0~r of studies using a variety ,of tech, niques [2, 4, 6, 8]. The effects of the duration (t) of a footshock stimulus have not been investigated with respect to the unconditioned response to footshock. A study by
tThis research was partially wppeo, zd by N ~ H research grant MH34609, and the Texas Christian UaiversRy Research Fotmdation. 2Requests for ~ s should be addressed to F. H. Gage at the above address.
Copyright © 1982 Pergamon Press~0031-9384/82/llO957-(M$03.00/0
958
TURNER AND GAGE
A
B i
I 10 1K
,
lOOK 1or
"ft*~ 1
J--1 A
In] C ~TII~IBL
1 SECOND
AJ
IC1
1/2
LF353N
I,~,~ r --
r
o
q
_.J
E
.
.
.
.
.
.
~
F , ~
H
J
IC3
I__
I I _,
L. ~c' r 1,,~ ~M
÷ ¢,8
--]
I
,
1/2 LF353N
,o . . . . . . . . . . tO5
LM4066
ic6
LMS55CN
IC7
1/4 LM324N
ic8
LM556N
IC9
LM555CN
let0
1/2 LF353N
ICll
1/2 LF353N
tc12
112 UA1458
tC13
1/2 UA1458
~--~
~
....
IOOK~IOT 15K ~l
~o~
,
I
,~.~ 1~
F
t7
FIG. 1 (A) Schematic circuit diagram and key. (B) Schematic representation of circuit response. Letters in B represent responses recorded from the correspondingly lettered points in A. For detailed wiring diagrams of IC6, IC8, and IC9 see [3].
Church, Raymond, and Beauchamp [1] demonstrated that the suppression of free operant responding by f o o t s h o c k is a function of the multiplicative interaction of stimulus current and duration. The purpose of this paper is twofold. First we will present a sensitive and reliable instrument for use in measuring the unconditioned response to footshock in the rat. We will then discuss the use of this assay to determine the relationship between the response of subjects to electrical footshock and the physical parameters (duration and intensity) of the stimulus. METHOD
Subjects The subjects used in this study were Sprague-Dawtey rats between 80 and 100 days old. Subjects were individually housed under a 12 hour light/dark cycle with free access to food and water. Testing was done in dim light 4 to 8 hours after the beginning of the dark phase of the cycle.
Apparatus The basic apparatus used in this study has previously been described in detail by Gage, Armstrong and Thompson [2]. The basic apparatus consists of a test box which is wired to present footshock, and balanced on a force transducer. The apparatus is designed so that forces exerted by a subject's movements in the test box are translated into voltages and recorded with a polygraph. The modified apparatus used in this study electronically monitors the output of the polygraph driver amplifier and provides three indices of a subject's response to shock. Latency to respond (LAT) is the elapsed time from shock onset until the momentary force exerted by a subject exceeds a preset
threshold level. Response duration (DUR) is the period of time a subject's responses occur with an amplitude greater than a preset threshold level and a frequency greater than one per second. The third measure (INT) is proportional to the dynamic forces exerted by the subject integrated over the previously defined response duration. The electronic circuitry used to monitor the output o f the polygraph amplifier is shown schematically in Fig. 1A. The input is AC coupled at the input to IC2 and the output of IC2 rectified, so that only the voltages due to the subject's movements in the chamber are recorded. Temporal measurements (LAT and DUR) are made using linear voltage ramp circuits, so that all three measures may be recorded with a digital voltmeter. Scrambled shock stimuli are provided by a Grason Stadler model 700 shock generator. The shock schedule is controlled by conventional relay equipment. Initiation of a shock stimulus causes the monostable timer IC6 to produce a 1 msec output pulse. The output pulse from IC6 triggers the retriggerabie one-shot IC8 (see D and G Fig. IB). The output of the Tetriggcrable one-shot IC8 opens channel 1 of the analog s~,itch IC5 thus permitting the input from the polygraph to the monitoring circuitry (see B Fig. IB). The input from the polygraph is intc~ratod by IC3 to produce the integrated force measure (INT). The polygraph input also provides additional triggers to the re~riggerable one-shot IC8 via the output of the voltage comparator I(27. If a period greater than one second occurs during which the retriggcrablc one-shot IC8 receives no further input tri.ggcr pulses, its output will go low, thereby blocking the i l ~ t from the polygraph (see D Fig. 113). During the period in which the output of the retriggerable one-shot IC8 is high a constant 15 volts is applied to the input of the linear voltage ramp generator IC12. At the end of the response the output of this linear voltage ramp generator is read as the uncorrected
F O O T S H O C K R E S P O N S E DYNAMICS response duration (DUR) (see Fig. IB). The output of the R/S flip IC9 flop (which is set high by the initiation of a shock stimulus and reset by the first output from the voltage comparator IC7) provides a constant 15 V input to the linear voltage ramp generator IC10. At the end of the response, the output of the linear voltage ramp generator is read as latency to respond (LAT). Shorting across the feedback capacitors of the three integrating circuits (IC3, IC10, and IC12) will clear the device in preparation for another trial. The uncorrected duration of response is recorded from the time of shock onset until one second beyond the last response of the period of time wherein responses occur with a frequency of greater than one per second. To correct the recorded DUR it is necessary to subtract the recorded LAT plus one second from the recorded DUR. The corrected DUR is thus the period of time wherein responses of an amplitude greater than a threshold value occur with a frequency of greater than one per second (see Fig. 1B).
959
a
S
Q
Z
8
8 -2, o •
8 ,O0
,~
.tO
.IS
,20 .2S ,30 C[,IR~NT x [XIRflTION
.3S
.40
g
b
Calibration of Monitoring Circuitry The two temporal measurements (DUR and LAT) are calibrated by presenting a shock stimulus with no subject in the chamber. Under these circumstances no response can be recorded and DUR should read 1.00 V(1.00 see) and LAT should read 10.00 V (1000 msecs). These values were chosen for optimal range, resolution, and ease of reading. The outputs are adjusted by changing the settings on the ten turn potentiometers which control the variable gain of amplifiers IC 11 and IC 13. The polygraph amplifier is set such that when a 300 g mass is placed in the center of the chamber there is a 280 mV change in the static voltage at the input to IC1. The integrated force measure (INT) is calibrated by applying 15 volts to the input of IC1 with the monitoring circuitry held open. The change in voltage due to the charging of the capacitor at the input of IC2 is then integrated at IC3 and the output adjusted to read 2.00 V by setting the ten turn potentiometer on the variable gain amplifier IC4.
.so
,4S
eseoe oo 8 ,00
".0~
'.10
'.IS
'.20
'.2S '.~ CLIq~ENT x 0U~qTION
".~
'.40
'.4S
C
Separation of Responses from Spontaneous Movements Reduction of measurement error is an approach used to increase the sensitivity of a measurement. When measuring a subject's response to a sensory stimulus, one way to reduce measurement error is to insure that the response to the stimulus is clearly distinguishable from non-stimulus linked behaviors. This reduces the variance produced when the instrument measures irrelevant responses in addition to the stimulus evoked responses. The movements made in response to footshock are clearly distinguishable from movements which are not in response to footshock because of the larger amplitude of the subject's movements in response to footshock. Ten subjects were placed in the experimental chamber, and for each subject, the monitoring circuitry was opened ten times for one second on a VT 30 second schedule; however shock was not delivered. The maximum voltage occurring due to the spontaneous movements of the animal was then recorded during each one second period by monitoring the circuitry at point B (Fig. 1A) with a storage osciUiscope. The mean and standard deviation of the maximal voltages produced by the spontaneous movements of the subjects were calculated across animals and trials. The one-tailed critical value for a normalized score with p <0.001 was obtained from a table of normal probabilities. Using this critical value, and the calculated mean and standard deviation of the volt-
•
.00
".is
'.3o
".4s
'.ea
'.Ts
1
'.go
i.os
i.ao
i'.3s
1.540
COl[RENT
FIG. 2. (a) Integrated force (INT) of subject's response plotted relative to the multiplicative interaction of stimulus current and stimulus duration. (b) Duration of subject's response plotted relative to the multiplicative interaction of stimulus current and stimulus duration. (c) Latency to respond to the stimulus plotted relative to stimulus current. See text for a description of the units employed•
ages produced by the spontaneous movements o f the subjects, a threshold value was calculated. With this threshold, the probability of a voltage greater than threshold occurring due to the spontaneous movements of an average subject is less than 0.001.
960
T U R N E R AND G A G E
Procedure
In order to assess the effects of stimulus current (I) and duration (t) on the measured response, 18 subjects were each presented with one of the 18 possible combinations of the following stimulus amplitudes: 0.4, 0.5, 0.6, 0.8, 1.0, 1.3 mA; and durations: I00, 200, 300 msec. Each subject was presented with 5 stimuli on the first day of testing, and 5 stimuli on the second day. Presentations of stimuli were made on a VT 30 second schedule. Scatter plots were made for each of the three measures as a function of stimulus current, duration, and the multiplicative interaction of stimulus current and duration. These plots were examined for both linear and curvilinear trends. RESULTS The best predictor of a subject's response to a shock stimulus, as measured by I N T and DUR, was found to be the multiplicative interaction between stimulus intensity and stimulus duration. The relationship, shown in Fig. 2A and 2B, was found to be linear by a least squares criterion, (r=.94 for INT and r = . 8 8 for DUR). In nearly every case, a subject's latency to respond to the stimulus was less than the duration of the stimulus. It is therefore unlikely that the duration o f the stimulus has a significant effect on the subject's latency to respond to the stimulus. A scatterplot o f stimulus current (I) versus L A T revealed a curvilinear relationship between I and L A T which suggested a hyperbolic relationship between I and LAT. The linear transformation of a hyperbolic relationship is a double reciprocal plot [5]. This transformation produced a significant least squares solution ( r = - . 7 9 ) . The hyperbolic relationship between L A T and I is shown in Fig. 2C. DISCUSSION The results shown in Fig. 2A and 2B demonstrate that stimulus current and duration have equivalent effects on the subject's unconditioned response to the stimulus as measured by I N T and DUR. This implies that the effects o f the electrical charge passed across the subject on the subject's response summate along the temporal duration of the stimulus. L A T is the period of time from the onset of the stimulus until the subject's first response. The subject's response as measured by I N T an.d D U R demonstrates that the effects of the electrical charge passed across the subject on the subj e c t ' s response summates along the temporal duration of the stimulus. A hyperbolic relationship between I and L A T then suggests that LAT is a measure of the temporal summation of
electrical charge necessary to evoke a response from the subject (i.e., stimulus current multiplied by LAT is equivalent to the total charge passed across the subject at the time of the subject's first response). The presence of a Y-asymptote in the relation between I and L A T suggests that there is a minumum latency (Lm) to respond such that regardless of the amplitude of I, the subjects cannot respond more rapidly than Lm. Lm is representative of physical constraints on the subject, that is, the mimihai~idtime necessary for the subject to perceive the stimulus, and produce a coordinated movement. The value obtained for Lm in this study was 45 msec. Lm is actually the expected minimal latency to respond for an average subject, therefore, variance above and below this value would be expected among individual subjects. The presence of the X-asymptote in the I-LAT relationship suggests that there is some level of current (In,) below which the effects of I do not temporally summate so that a response cannot be evoked. Im then is the predicted threshold current for a stimulus of long duration. The value obtained for Im in this study was 0.29 mA. These results suggest that in experiments designed to investigate shock response thresholds, the current threshold will in large part depend upon the duration of the stimulus employed. If a long duration stimulus is employed, the recorded threshold will correspond to Ira. If a short duration stimulus is employed, the current threshold will reflect only the amplitude of current necessary for the summation of the electrical charge over the temporal duration of the stimulus to be large enough to evoke a response. In summary, we have demonstrated: (l) the unconditioned response to footshock is in response to the summated effects of electrical charge passed across the subject by the stimulus over the duration of the stimulus; and, (2) there appear to be two response thresholds to footshock, one to the summated effect of electrical charge over time, and the other is a current threshold, below which the effects of electrical charge do not summate to produce a response. We are convinced that the development of sensitive and reliable behavioral measures like the one described in this report are essential for an accurate and meaningful understanding of the physiological mechanisms underlying behavior. ACKNOWLEDGEMENTS We thank J. E. Springer and G. R. Stewart for assistance with data collection, and C. O. Larson for assistance in drawing figures. We also thank R. C. Bay, P. E. Moes, J. P. Kesslak, andR. G. Thompson for critical reading of this manuscript.
REFERENCES
1. Church, R. M., G. A. Raymond and R. D. l~auchamp. Response suppression as a function of intensity and duration of a punishment. J. comp. physiol. PsychoL 63:39 A.A.,1967. 2. Gage, F. H., D. R. Armstrong and R. G. Thompson. Behavioral kinetics: A method for deriving qualitative and quantitative changes in sensory responsiveness following septal nuclei damage. Physiol. Behav. 23: 479-484, 1980.
3. Jung, W. G. IC Timer Cookbook. Indianapolis, IN: Howard W. Sams and CO., Inc., 1977. 4. Leiter, D. S., A. S. Powers and H. S. Hoffman. The neural substrate of the startle response. Physiol. Behav. 25: 291-297, 1980.
5. Lewis D. Quantitative Methods in Psychology. New York: McGraw-Hill, 1960. 6. Lubar, L F., J. M. Brener, J. H. Deagle, R. Numan and W. J. Clemens. Effect of septal tesions on detection threshold and unconditioned response to shock. Physiol. Behav. 5: 459--463, 1970. 7. Nunnally, J. C. Psychometric Theory. N e w York: McGraw-Hill. 1978. 8. Smith, R. F. Scopolamine does not affect footshock sensitivity m the rat. Pharmac. Biochem. Behav. 8: 31-34, 1978.