Behavioural Processes, 28 (1993) 145-164 0 1993 Elsevier Science Publishers B.V. AIf rights reserved 0376-6357/93/$06.00
Effects of aging on habituation Caenorhabditis C.D.O. Department
in the nematode
Beck and C.H. Rankin
of Psychology, University of British Columbia, British Columbia, Canada
(Accepted 6 August 1992)
The effects of aging on spontaneous locomotor behavior and habituation in a mechanosensory reflex were examined in the nematode Caenorhabditis elegans. Worms were tested at 4 days (at the peak of egg laying), at 7 days (when egg laying ends) and at 12 days post-hatching. Both spontaneous and reflexive movements were smaller in older worms than in younger worms. In addition the magnitude of these movements was related to life span; the shorter an animal’s life span, the smaller its reversal movements while still young. Worms at all ages expressed habituation and dishabituation at a 10 s interstimulus interval (151); thus even aged worms were capable of non-associative learning. However, older worms showed greater habituation than did 4-day-old worms to stimuli delivered at a 60 s ISI. There was also an age-related change in the recovery from habituation. At days 4 and 7, worms had recovered from habituation by 30 min after training; however, responses of day 12 worms were still significantly smaller than baseline at 30 min after training. Further behavioral tests with normal and mutant worms may help elucidate the nature of the age-related changes in the learning and memory processes of C. elegans and the genetic mechanisms which underlie them.
Correspondence to: C.H. Rankin, Department of Psychology, 2136 West Mall, Columbia,
Introduction The effects of aging on the relationship
between behavior and biology can be studied
through the use of model systems. One such system is the marine mollusc Aplysia californica, in which the neurophysiological function of identified neurons in circuits controlling the behaviors of gill and siphon withdrawal are known. Studies of aging in Aplysia have shown that these simple animals display losses in learning and memory similar to those found in vertebrates. Rattan and Peretz (1981) found that the threshold for a behavioral response to gill stimulation was significantly higher in old Aplysia than in young, mature animals. Furthermore they found that old Aplysia habituated more rapidly than the younger animals. In addition, these old animals did not exhibit dishabituation when a neuron (L7), identified as one that produces dishabituation in younger animals, was stimulated. The studies with Aplysia have advanced our understanding of how age-related neurophysiological changes in identified neurons may be reflected in behavior. However, if we are to understand the role of the genome in behavioral aging, we should use a model system in which both the nervous system and the genetics may be readily understood. These experiments were designed to study the effects of aging on behavioral plasticity in such a model system, the nematode Caenorhabditis elegans. This work will provide a basis for investigations of the role of the genome in the effects of aging on the nervous system of this simple organism. Caenorhabditis elegans, a small free-living soil nematode, has been widely used as a genetic model in the study of anatomy, development and behavior. A map of over 90% of the C. elegans genome (which is half the size of the genome of Drosophila) has been established (Coulson et al., 1986; Hodgkin et al., 1988). The short reproductive cycle (about 3 days) makes C. elegans a convenient subject for genetic analysis. The genetics of C. elegans have been extensively studied and many mutant strains have been isolated. Because the neuroanatomy of C. elegans is simple, containing only 302 neurons, and the cell lineage of each somatic cell is known, the relationship between gene, neuron and behavior may be narrowly determined. The neuroanatomical map of this nematode has been described completely (White et al., 1986; Chalfie, 1984). Furthermore the functional elements of neural circuits underlying behaviors such as touch withdrawal circuit have been demonstrated, establishing the link between behavior and anatomy (Chalfie et al., 1985; Chalfie and Au, 1989). Recent work on the behavior of this organism has shown that C. elegans has a rich repertoire of behavioral plasticity (Rankin and Chiba, 1988; Rankin et al., 1990; Kumar et al., 1989). Rankin et al. (1990) showed that the tap withdrawal reflex in C. elegans exhibits the major forms of non-associative learning: habituation, dishabituation and sensitization. An advantage of working with the tap withdrawal reflex is that the neural circuit for touch withdrawal, described by Chalfie et al. (1985), has been shown to underlie the tap withdrawal reflex as well (Rankin and Chalfie, 1989; Wicks and Rankin, 1991). Thus C. elegans possesses a range of behavioral plasticity, making it a promising candidate for the investigation of the biological mechanisms underlying age-related changes in behavioral plasticity. The short life cycle of C. elegans makes studies of behavior across life span convenient. In this paper, we have focused on changes in behavior during post-reproductive development. The average life span of an individual worm is 14 days under the conditions maintained in our laboratory. In these experiments worms were
tested at four days of age which is at the peak of egg laying, at seven days, when egg laying has ended, and at 12 days, just prior to the end of the 14day
Subjects Mature hermaphroditic
were used throughout these studies. Subjects
were maintained on solid media at 20°C. Under these conditions the average life span observed was 14 days post-hatching. Tests of behavior were performed at 4, 7 and 12 days post-hatching.
Materials Worms were maintained on Nematode Growth Medium (NCM; Brenner, 1974) agarfilled Petri plates (5 cm diameter) and fed Escherichia co/i (strain OP50). Behavioral observations were made through a stereomicroscope with attached videorecording equipment. Mechanical and electrical stimulation were controlled by a Grass S88 stimulator. Vibrational stimuli were produced by a mechanical tapper controlled by an electromagnetic relay. The mechanical tapper tapped the side of the Petri plate holding the worm. Electrical shocks were produced using a spanning electrode with the two wires placed one on either side of the animal on the surface of the agar (approximately 2 mm apart; individual C. elegans are about 0.05 to 0.1 mm in diameter and 1 mm in length). Each shock stimulus consisted of a 600 ms train of 6-l 0 ms pulses delivered at IO Hz.
that were used in these studies were hatched syn-
chronously. To ensure synchrony, mature egg-laying worms were selected from the general population and were placed on an agar plate streaked with E. co/i. These worms were permitted to lay eggs for 3-4 h and were then removed from the plate. The eggs hatched in 9-l 1 h. At 3 days post-hatching (72-84 h) the maturing worms were placed individually on numbered agar plates spotted with E. co/i. In Experiment 1, in which individual worms were followed throughout post-reproductive development, the worms were tested at intervals 4, 7 and 12 days after hatching. Tests were performed on plates without food. After each test, the worms were placed onto new plates with fresh E. co/i to ensure a consistent food source and to prevent confusing the subjects with their offspring. In Experiments 2 to 4, in which worms were tested at only one of the test ages, worms were plated individually and raised to that test age. Worms were replated approximately every two days to maintain isolation. Behavioral responses were scored in several ways depending on the experiment. The reversal response (swimming backward) to vibrational stimuli (taps> was the chief behavioral response considered in these experiments. In order to be considered a reversal to tap, the response must have occurred within 1 s after the stimulus. Both the frequency and the
magnitude of these reversals were scored. The magnitude of the reversal response was quantified by tracing the path of the reversal onto an acetate sheet. These tracings were then digitized and measured using a digitizing tablet, MacMeasure software and a Macintosh computer. Statistical analyses of repeated-measures proportion data were performed with Cochran Q tests. Magnitude data were analyzed using ANOVAs with Fisher’s post-hoc comparisons when statistical significance was achieved. The level of significance was set at LY = 0.05
unless the performance of multiple tests made it necessary to adjust the (Y level
In order to determine whether there were changes in the baseline activity during aging that might affect the expression of plasticity by that reflex, the level of spontaneous locomotor activity of the worms during late (post-reproductive) development was measured by recording the time spent active over IO min, and the number and magnitude of any spontaneous reversals (swimming backward) that occurred within the 10 min observation period. In addition to spontaneous reversals, worms also show reflexive reversals in response to a variety of stimuli. In previous experiments we have shown that, in 4-day-old adults, a tap to the dish elicits a backward swimming withdrawal response called a reversal (Rankin et al., 1990). In this experiment, age-related changes in the tap withdrawal reflex were examined. The number of worms responding to taps with reversals, and the magnitude of the reversal responses to tap were measured. In addition the occurrence of other types of response to tap (accelerations forward and pauses) were scored. Finally, each worm’s response to a light touch to the head with a hair was tested in order to determine whether the worm was capable of swimming backward (as all healthy worms reverse to head-touch; Chalfie et al., 1985). If a worm did not respond to a head-touch with a reversal, it was assumed to be incapacitated and was not included in the results. The magnitude of the reversal to head-touch was not scored because the strength of the hand-delivered headtouch could not be accurately controlled.
Subjects The same 21 hermaphroditic
worms were tested for spontaneous activity three times,
once at each of the test ages: 4, 7 and 12 days after hatching. Since some worms died before 12 days of age, 45 worms had to be tested at day 4 (35 at day 7) to obtain data for 21 animals on day 12. These data permitted an investigation of behavioral differences between worms that died young (between day 4 and day 7 or between day 7 and day 12) and those that survived 12 days. Following the recording of spontaneous activity, 20 worms at each age were tested for their response to tap and head-touch.
Procedure Spontaneous locomotor activity was observed over a IO min period during which the time active was determined by measuring the amount of time the worm spent swimming
149 (either forward or backward). The number of spontaneous reversals (swimming
in the absence of an obvious external stimulus) was scored over the IO min period. The magnitude (mm) of these spontaneous reversals was determined by video stop-frame analysis and computer-aided digitizing of reversal length. A single tap and a head-touch were administered immediately following the observation period for spontaneous activity with a 3 min interval between tap and head touch. Head-touch was administered with a fine hair to the region of the worm’s pharynx. The magnitudes of the reversals to tap were determined computer-aided digitizing of reversal length.
by video stop-frame
Results and Discussion Some of the baseline behavioral measures changed with
did not. No
sjgnificant differences were found in time spent active among the test ages (day 4: X k S.E. = 96.048 + 3.152; day 7: 99.889 f 0.084; day 12: 91.563 IE 2.085; F(2,38) = 2.67, n.s.). The number of spontaneous reversals during the 10 min observation period did not change with age (F(2,38) = 1.63, n.s.). However, as seen in Fig. IA, the mean magnitude (mm) of spontaneous reversals at day 12 was significantly smaller than at day 4 or 7 (f(2,38) = 15.43, P = 0.0001). To examine the worms’ response to tap across the three test ages both the occurrence and the magnitude of reversal responses to tap were scored. When the occurrence of reversal responses was analyzed with a Cochran Q test, no change with age was evident (Q(2) = 2.8, P > 0.05). However, as seen in Fig. 1 B, the mean magnitude (mm) of reversal responses to tap was significantly smaller at day 12 than at day 4 (F(2,38) = 7.48, P = 0.002). This is reminiscent of the observation that spontaneous reversals were smaller at day 12. In both cases the magnitude of the reversals changed while the occurrence of reversals did not. The decrease in the magnitude of spontaneous and reflexive reversals at day 12 may have been related to these worms’ approaching deaths. To determine whether life span affected the spontaneous and reflexive activity of the worms, the behavior at day 4 of worms that survived to 12 days was compared with the behavior of those that died between day 4 and day 7 and the behavior of those that died between day 7 and day 12. In addition, the behavior at day 7 of worms that survived to 12 days was compared with the behavior of those that died between day 7 and day 12. During the course of the experiment 22% of the worms run died between day 4 and 7 and 31% died between day 7 and 12. Only those behaviors that changed with age (the mean magnitude of spontaneous and reflexive reversals to tap) were examined. In both cases, the magnitude of the swimming reversals was smaller in worms that were to die earlier. As seen in Fig. IC, the mean magnitude of spontaneous reversals measured on day 7 was related to lifespan while it was not when measured on day 4 (day 4: F(2,42) = 0.302, ns.; day 7: t(33) = 2.352, P = 0.025). The magnitude of the swimming reversal response to tap (Fig. 1 D) was related to lifespan when measured on day 4 but not when measured on day 7 (day 4: F(2,42) = 3.69, P = 0.0336; day 7: t(33) = 0.51, n.s.). Thus, the magnitude of spontaneous and reflexive reversals may reflect the worms’ lifespan. Further investigations of this sort may help to define this relationship. A possible explanation for the decrease in response magnitude in 12 day old worms is that there is a change in sensory ability with increased age. An examination of the magnitude of the reversal reflex in response to a graded series of taps (beginning with a very weak stimulus) at each of the three test ages was used to test this hypothesis.
mean magnitude of spontaneous
magnitude of spontaneous stimuli)
during the 10 min observation
in old worms.
backward in the absence of any obvious external
period at each of the three test ages, 4, 7 and 12 days
post-hatching (mean number of spontaneous
by each worm f S.E.M.
Day 4: 34.7 k 3.2;
Day 7: 29.4 i 1.5; Day 12: 34.9 F 2.1). (B) The magnitude of the reversal response to tap (mm) was smallest
responded with a reversal;
magnitude of reversals
number of worms
to tap (including
reversing Day 4: 20;
Day 7: 17;
Life-span was retdted to the mean magnitude of spontaneous reversals at day 7. The mean magnitude of spontaneous
reversals of worms that survived until day 12
that of worms
that died between day 4 and day 7 (D4-D7;
n = 14).
velocity at day 4; D7 TEST=
( > D12; R = 21) was compared with n = IO)
and day 7 and day 12
velocity at day 7. (D)
related to the magnitude of reversal responses to tap at day 4. The magnitude of reversals of worms that survived until day 12 ( > D12;
n = 21) was compared with that of worms that died between day
4 and day 7 (D44D7;
Experiment In this were during
n = IO) and ddy 7 and day 12 (D7-D12;
n = 14).
2: Graded response
the effects tap intensities induced
of aging on the responses chosen
to taps of different and strong)
were ones which,
total of 60 worms. before
were tested with
taps of three different
were tested at day 12 to examine
at each age for a the effect of delay
Procedure Worms were placed individually on test plates with a small amount of E. co/i (strain OP50) 24 h before testing. Tap intensity was altered by changing the voltage from the Grass S-88 stimulator to the electromagnetic relay which controlled the tapper. The lowest voltage (38 V) produced a weak tap, while the highest voltage (60 V> produced a strong tap and was the intensity used in all other experiments. An intermediate voltage of 40 V produced a tap of moderate strength. The taps were administered individually to worms at IO-min intervals. In studies of recovery from habituation, Rankin and Broster (1992) showed that taps administered at IO-min intervals did not produce significant response decrement in worms that were 4 days old. Of the 20 worms tested at each age (4, 7 and 12 days post-hatching), IO worms received the stimuli in ascending order of intensity (weak, moderate intensity (strong,
and strong) and 10 worms received the stimuli moderate and weak). To test for an interaction
in descending order between the order
presentation and the response to stimuli of different intensities at each age, the responses of worms that received the stimuli in ascending and descending orders were compared with a two-factor repeated-measures ANOVA. The magnitude (mm) of the reversals at each age for different intensities were analyzed with a repeated-measures ANOVA. Non-reversals (pauses, accelerations forward and no response), which made up approximately 5% of the data, were excluded from the analysis of reversal magnitude and were replaced with the mean of the group for the purposes of the statistical analysis. At each test age, 4, 7 and 12 days post-hatching, the numbers were analyzed with a Cochran Q test.
of reversals to taps of different
To determine whether the data from the groups which received the stimuli in different orders could be pooled together, the mean response magnitudes from worms that received the stimuli in ascending order were compared with the mean response magnitudes from those that received the stimuli in descending order across the three stimulus intensities at each alpha (three tests,
test age (4, 7 and 12 days post-hatching). To control experiment-wise error level (0.05) was divided by the number of tests performed on the data from tests: one initial comparison of the orders of stimulation, and possibly two one for each stimulus order; alpha = 0.05/3 = 0.016). At day 4 and day 7,
rate, the each age follow-up there was
an effect of stimulus intensity, but no effect of stimulus order nor any interaction between stimulus order and intensity (day 4: Order: F(1,18) = 2.16, ns.; Intensity: F(2,36) = 5.51, P = 0.008; Order x Intensity: F(2,36) = 0.66, ns.; day 7: Order: F(1,18) = 0.14, ns.; Intensity: f(2,36) = 5.39, P = 0.009; Order X Intensity: F(2,36) = 0.091, n.s.1. However, at 12 days of age the order of stimulation interacted with stimulus intensity in addition to a significant effect of intensity (day 12: Order: F(1,18) = 5.759, P = ns.; intensity: F(2,36) = 11.78, P = 0.0001; Order X Intensity: f(2,36) = 10.26, P = 0.0003). Therefore, while data from worms run in ascending and descending orders from day 4 and 7 worms could be pooled, the data from day 12 could not. Response occurrence At days 4 and 7, stimulus intensity did not affect the number of worms responding (day 7: Q(2) = 2.4; n.s.>. At day 12, the data from worms that received the stimuli in ascending
Fig. 2. The reversal response to tap (mm) was graded as a function of stimulus tested, The magnitude of reversal responses+S.E.M.
days post-hatching (n = 20 at each age). At day 4 (A) and day 7 (B), received the stimuli
in ascending order (ASCENDING;
worms that received the stimuli in descending order (DESCENDING;
at all ages
at 4, 7 and 12
the data from worms
in ascending and descending orders were pooled, while at day 12 (C) the data
from worms that received the stimuli
to taps of different intensities
n = 10) and the data from
n = IO) are shown separately.
Response magnitude As shown weak (Fig.
worms or weak
did not show
day 4 worms a moderate
tap than to either
significantly = 5.66,
2C) that received
responded or strong
P = 0.007).
tap (F(2,38) reversals
At day 7
tap than either
the data from
P = 0.007).
to a strong
At day 12,
day 12 worms
and weak) responded with significantly larger reversals to the strong tap than to either the weak or moderate taps (F(2,18)= 13.79, P = 0.0002). Furthermore, when the first stimuli of the ascending and descending order series were compared, the first tap of the ascending series (weak intensity) evoked a reversal response that was significantly smaller than the response to the first stimulus of the descending series (strong intensity) (two-tailed t(l8) = - 3.61; P = 0.002). This finding shows that day 12 worms, like day 4 and 7 worms, are capable of responding to stimuli of different intensities with graded responses, but that unresponsiveness produced by repeated stimulation (habituation) even when the stimuli are administered at IO-min intervals.
mask the effect
An alternate explanation for the unresponsiveness of day 12 worms to stimuli in the ascending order of presentation is that treatments, other than stimulation, such as vibrational stimulation from placement on the microscope stage and the light and warmth on the microscope stage, may have caused fatigue in these older worms during the 20-min testing period. Thus the responses of an additional 20 day-12 worms were tested after a delay of 0 or 20 min (IO worms with a 0-min delay and IO worms with a 20-min delay) on the microscope stage. A two-tailed t-test showed that there was no significant difference in the reversal
to strong tap between
the tap immediately
after being placed on the microscope stage and worms that received the tap after a 20.min delay (0-min delay: 2 = 0.377 + 0.165; 20 min delay: x = 0.601 f 0.305; t(l1) = -0.61, ns.). Simply being on the microscope stage did not appear to affect the day 12 worms’ responses to tap. This finding suggests that indeed the day 12 worms responded with small reversals during the tests with taps of graded intensities because of habituation to stimuli delivered 10 min apart. It is evident that even the worms tested on day 12 were sensitive to changes in stimulus intensity. Thus there is no reason to attribute the decreased responsiveness of the older worms to a loss of sensory ability. These results suggest that older worms are more susceptible to habituation training than younger worms. If so, day 12 worms might be expected to exhibit a faster rate of response decrement during habituation training.
The effects of aging on habituation
to 60 stimuli
at short (IO s) or long (60 s)
interstimulus intervals in C. elegans were examined at each of the three test ages, 4 days, 7 days and 12 days post-hatching. Four-day-old worms show more rapid and greater habituation to stimuli at a 10 s ISIthan to stimuli at a 60 s ISI (Rankin and Broster, 1992). In addition the effect of aging on dishabituation was tested following habituation with 60 stimuli at a 10 s ISI.
Subjects Twenty naive worms were used at each of the three test ages (4, 7 and 12 days) for each of the two ISIS (IO and 60 s) for a total of 180 worms.
Procedure Trains of taps (1 train = 6 taps at 8.5 Hz) were the stimuli used in all of the following habituation experiments because this stimulus produces larger responses than do single
154 taps (Chiba plasticity tap
at a 10 s or a 60 s interstimulus
after the last habituating
60 V train of shocks
of taps were
range in which
at day 4 the mean reversal
a large response
of taps were
at 10 Hz)
For the 10 s ISI group
train of shocks that
In the present
In this experiment, interval
a 60 V
is a stimulus
at a 10 s ISI to test for
IO s ISI
Response occurrence The
4, 7 and 12 days of age (see Figs. 10%
of the responses)
Q tests initial Q(2)
P < 0.05;
data. Cochran across
P < 0.02;
as seen in Fig. 3, the
did appear different
at day 7. At 4 and 12
began responding with
in the habituation
again as the habituation
in a change in the proportion
At 7 days of age, the worms
P < 0.001).
from the analysis
and C, and 5A).
days of age, the number
that at each age there
at each age confirmed response,
that both habituation
at day 7 than at day 4 or day 12 (see Table
P = 0.0001).
Response magnitude In order
of a response
habituated P < 0.0001;
as a score
be compared of zero
by the group
and in addition
as a percent
on each of the three ages. To control
at all ages showed
and condition Worms
was set at 1000/o) for each worm.
the dishabituated = 786.63,
B and C) and dishabituation
At all ages the
P < 0.0001;
P < 0.0001).
rate, the alpha level
day 7: were
Fig. 3. During decreased habituation
s ISI and 60 s ISI), the proportion
at all ages tested. The proportion
of worms responding to stimuli with reversals during
training to trains of taps (IO s ISI and 60 s ISI, 30 stim; n = 20 at each age, day 4 (A), day 7 (B) and day 12 (C)j.
in the amount of dishabituation across the three ages as measured by difference scores produced by subtracting each animal‘s response magnitude for the first response
following shock from its initial response magnitude (see Table 1 for means; F(2,57) = 0.66, ns.). There was no significant difference in the rate of response decrement during habituation training. The rate of decrement was analyzed by calculating the slope of the regression lines for reversal responses to the first ten stimuli in the habituation training for each worm and then taking the mean of these slopes for each age (see Table 1 for means; F(2,56) = 0.36, n.s.). Both the frequency and response magnitude data demonstrate that worms of all ages show habituation and dishabituation. The significant dishabituation demonstrates that the response decrement was indeed habituation and not simply the result of sensory adapta-
3: the mean magnitude of reversal responsesf5E.M.
during habituation training. Initial
response set at 100%
ISI 1 OS
Age D4 D7
27.07 2.45 25.01 3.15 24.48 3.35
6.62 1.72 12.25 2.45 9.65 2.50
22.98 7.30 21.48 10.64
71 .I 5 9.98 60.46 a.99 44.48 5.00
0.67 ~ 7.61 1.24 - 7.08
1.61 - 7.00 1.60 - 5.60 0.93
tion or motor fatigue (Groves and Thompson,
3.77 I I .a0 2.85
The only difference across the ages
tested, was in response frequency during habituation. At day 7, worms initially decreased in their response frequency, but then increased again and continued a relatively high response rate throughout habituation training; however, their responses were very small. This pattern of responding was not as pronounced in a later experiment (Experiment 4) on recovery from habituation. Further work on the patterns of response occurrence would help to clarify the significance of these results. Based on the results of the graded response experiment (Experiment 2) we hypothesized that day 12 worms habituate to stimuli delivered at longer interstimulus intervals than do younger animals. Although habituation training at a IO s ISI revealed no age-related difference in response magnitude, the 10 s ISI may have been so short that differences in rate or degree of response decrement were lost. Thus, a 60 s ISI was used to test whether, with longer intervals, older animals might show greater habituation than younger animals.
60 s ISI
Response occurrence As seen in Fig. 3A, B and C, the proportion of worms responding with reversals appeared to decrease more in older worms than in younger ones. Chi-square tests at each age comparing the proportion of worms responding on the initial stimulus with the proportion of worms responding on the final stimulus of the habituation training showed that the proportion training at day 4 habituation training chi-square = 11.25, during habituation habituation training
of worms responding did not change significantly with habituation (chi-square (df I) = 1.8, n.s.) but did decrease significantly with at day 7 and day 12 (day 7: chi-square = 5.0, P < 0.05; day 12: P < 0.05). Furthermore the mean proportion of worms responding training changed significantly with age (see Table I). Throughout day 4 worms responded significantly more frequently with reversals
Fig. 4. During habituation training (IO
s IS1 and 60 s ISI), the mean magnitude of reversal responses
decreased at all ages tested. The mean magnitude5S.E.M.
of reversal responses during habituation
training with trains of taps (IO s ISI and 60 s ISI, 30 stimuli;
n = 20 at each age; day 4 (A),
day 7 (B)
and day 12 (C)). The magnitude of reversal responses was expressed as a percent of each worm’s response to the initial stimulus
than did day 12 worms worms
of reversal B and 0.
P = 0.0001).
as much with
the proportion training
as the proportion
did not decrease
in habituation training.
rate of habituation,
to the first
at each age with
by the regression
did not change with
at a 60 s ISI changed
age (see Table
at all ages as (d4: P=
for the reversal
Fig. 5. Habituation and dishabituation were evident in both the proportion of worms responding and the magnitude of reversal responses at all ages tested. Reversal responses before and after habituation training with trains of taps (IO s ISI, 60 stimuli)
and after dishabituation
with electric shock at 4, 7
and 12 days post-hatching (n = 20 at each age). INIT = initial response of the habituation training; HAB = final response of the habituation training; lus. (A) Proportion habituation
DIS = first response after the dishabituating stimu-
responding to stimuli
with reversals before habituation training,
at a 10
of the reversal
reduced: interstimulus magnitude
at a 60
60 s ISI habituation
each test B and C;
at each age, the mean to trials The
both the proportion reversal
to a greater extent
of each worm
at a 60 s ISI.
at 10 s and 60 s ISI’s within
of the reversal
to the last ten stimuli
day 4 worms.
of each worm
of the effects of habituation
I>. At each age, the mean magnitude
at a 60 s ISI than
in habituation training.
than day 7 or day 12 worms
Day 7 and day 12 worms
of taps administered
in day 4 worms
P = 0.0001).
Mean magnitude of reversal
before habituation training, after habituation training and following the dishabituat-
Magnitude of reversal responses was expressed as a percent of each worm’s response to the initial stimulus
age. At greater
at a 10 s ISI were 21-30 alpha of the day
of each level effects
10 s ISI habituation P = 0.0001;
P = 0.0001).
7 the mean
reversals at a 60 s ISI was greater than at a 10 s ISI in early training but not late training (trials I-IO t(38) = 3.72, P = 0.0006; trials 21-30 t(38) = 1.63, r1.s.). At day 12 the mean magnitude of reversals at a 60 s ISI was greater than at a IO s ISI in early training but not late training (trials I-IO t(38) = 3.32, P= 0.002; trials 21-30 t(38) = 0.57, n.s.>. Thus at each age there was a difference between the mean magnitude of reversals to the first ten stimuli at 10 s and 60 s ISl’s. However, at day 4 the responses to stimuli at a 60 s ISI remained significantly greater than those at a 10 s ISI in late habituation training, while at day 7 and 12 the responses at 10 s and 60 s ISl’s were habituation training.
In addition we can compare the rate of habituation at 10 s and 60 s ISl’s at each age by comparing the slopes of the regression lines of the first ten stimuli. At day 4 the rate of habituation was significantly greater at a 10 s ISI than at a 60 s ISI (see Table 1; t(38) = -3.007, P = 0.0047). At day 7 and 12 the rate of habituation training was not affected by the duration of the interstimulus interval (see Table 1; t(38) = -0.30, n.s.; t(38) = - 1.04, n.s.>. These comparisons of 10 s and 60 s ISI habituation at each testing age reveal two age-related differences. First, at day 4 the mean magnitude of reversals during 60 s ISI habituation training was significantly greater than that during 10 s ISI training in both early and late trials. However, at day 7 and 12, interstimulus interval only had an effect in early training. This finding confirms that worms at all ages are sensitive to changes in interstimuIus interval. In addition, it suggests that there are age-related differences in the effects of habituation training as stimulus number increases. Second, at day 4 the rate of habituation was slower to a 60 s than a 10 s ISI, but at day 7 and day 12 the rate of habituation was not significantly slower at the longer ISI. This finding suggests an age-related difference in the early part of habituation. Thus, with a longer ISI, older animals appeared to habituate earlier and to a greater degree than younger animals. No such aging effect was seen at the short interstimulus interval.
4: Recovery from habituation
Recovery from habituation may be memory. The stronger the memory, Thus if the processes underlying the recovery from habituation might also
thought of as a measure of the strength of short-term the slower the recovery from habituation might be. formation of short-term memories change with age, change with age.
were tested at each of the test ages 4, 7 and 12 days post-hatching.
Procedure Habituation was established by delivering 60 trains of taps at a 10 s ISI; at the end of habituation training, single trains of taps at 30 s, 10 min, 20 min and 30 min were
Fig. 6. (A) The
to trains of taps with
habituation training and subsequent recovery tests at all ages tested. Proportion
of worms responding
with reversals during recovery following habituation training (10 s ISI, 60 stimuli;
n = 20 at each age)
shown. Tests of recovery were given 30 s, 10 min, 20 min, and 30 min post-habituation. response to the first habituation stimulus; The magnitude of reversal responses
HAB = the response to the last habituation stimulus.
at the end of habituation training and during recovery from
habituation at day 4. Magnitude of reversal responses was expressed as a percent of each worm’s reversal
to the initial
magnitude of reversal
responses at the end of habituation training and during recovery from habituation at day 7. (D) The magnitude of reversal
to test for recovery
4, it produced
habituation at day 12.
from habituation. A 10 s ISI was used because, in Experiment the same degree of habituation at all ages. This procedure has been used to recovery from habituation 20 to 30 min after the last habituating stimulus in
In the analysis of the recovery of the magnitude of the reversal response after habituation training, reversal magnitude was standardized as percent initial response as in the previous experiment. Only the frequency and magnitude data from the 30 min recovery tests were statistically analyzed because the variance in the intermediate tests (at 30 s, 10 min and 20 min) was too great to be usefully included. The magnitude data from all tests are depicted in Fig. 6B, C and D.
Response occurrence The occurrence of reversals, scored as the proportion of worms responding to each stimulus with a reversal, showed that both habituation and recovery from habituation were evident at 4, 7 and 12 days of age (see Fig. 6A). Cochran Q tests at each age comparing the initial, habituated and recovered (30 min post-habituation) response occurrence confirmed that at each age there were significant changes with the treatments (day 4: Q(2) = 17.231,
P < 0.001;
Q(2) = 13.0,
P < 0.01;
Q(2) = 19.0,
P < 0.001 I Response magnitude A different pattern is seen in the analysis of the magnitude of reversal responses. As shown in Fig. 66 and C at 4 days and 7 days post-hatching worms showed significant recovery of response magnitude over habituated levels by 30 min following habituation training; in contrast, worms tested at 12 days post-hatching (Fig. 6D) did not show significant recovery over habituated levels by 30 min post-habituation (day 4: F(2,38) = 23.56, P< .OOOOl; day 7: F(2,38)= 16.95, P< 0.0001; day 12: F(2,38)= 3.43, P < 0.0001). The slower
in day 12 worms
may reflect a change
in the way
short-term memory for habituation is expressed in older worms. It is not clear whether this longer time to recover is related to the greater sensitivity to habituation to long ISIS seen in the day 12 worms. Further investigations into this issue are needed before conclusions can be drawn.
Examination of the spontaneous and reflexive behavior of the worms indicated that there was a decrease in some measures of motor activity with age. The magnitude of spontaneous reversals and the reversal response to tap diminished with age. The mean magnitude of spontaneous reversals and the magnitude of response to single taps was related to life span. In future studies it may be useful to probe the relationship in C. elegans between an individual worm’s life span, their physiological state, the quality of their activity and the effects of that activity on behavioral plasticity. To test the sensitivity of the aged worms to changes in stimulus intensity, above-threshold stimuli of different intensities (weak, moderate and strong) were administered at a 10 min ISI in either ascending (weak, moderate, strong) or descending (strong, moderate, weak) orders of intensity. At day 4, 7 and 12 worms showed graded responses to stimuli of graded intensities. However, at day 12 there was an interaction between the order of presentation and the effect of stimulus intensity. In the ascending order of presentation, the responses to moderate and strong stimuli were of the same magnitude as the small responses seen to the weak stimulus. This may have been due to habituation to the tap stimulus. This finding was unexpected as Rankin and Broster (1992) found no evidence of response decrement when they administered four taps of equal intensities at a 10 min ISI to day 4 worms. Exposure to the environmental conditions of the graded response’experiment alone did not affect reversal magnitude of day 12 worms. In addition, since 12-day-old worms showed an average of 34.857 + 2.145 spontaneous reversals during a IO-min observation period, it is unlikely that three evoked reversals would lead to exhaustion. Thus, the unresponsiveness observed in day 12 worms in the graded response experiment when
were administered in order of ascending intensity may have indeed been caused by
habituation to tap. This finding supports the hypothesis that processes underlying tion in C. elegans change with age.
Habituation at a 10 s ISI and dishabituation to shock were present in worms at all ages and did not appear to change greatly during post-reproductive development. Although there were no apparent differences in the rate or amount of habituation to a 10 s ISI between age groups, habituation training using a longer ISI (60 s) revealed that older worms showed greater habituation at long interstimulus intervals than day 4 worms as reflected by both the proportion of worms responding and the magnitude of reversal response. These findings are similar to those of Rattan and Peretz (1981) who found that older Aplysia habituated more quickly than young, mature Aplysia. However, unlike Rattan and Peretz’s observations in Aplysia, we observed dishabituation at all ages of C. elegans tested. Worms tested at 4 and 7 days showed recovery from habituation (10 s ISI) of both the frequency and response magnitude 30 min after habituation training, but worms tested at 12 days did not. This deficit in recovery suggests a persistence of habituation in the older worms. This persistence is unlikely to be explained in full by a factor such as more rapid physical exhaustion since day 12 worms with the same habituation training showed dishabituation (facilitation of the habituated response) immediately after a mild electric shock. Perhaps this persistence reflects a change in short-term memory for habituation training in the older worms. Further studies on the dynamics of habituation and recovery from habituation and on long-term memory in aged worms may help the understanding of mechanisms underlying the age-related changes in habituation. In summary, older worms show smaller spontaneous reversals and smaller reversals to tap than younger worms. In addition older worms show greater habituation to stimuli delivered at long interstimulus intervals and recover from habituation more slowly than do younger worms. These changes appear to be independent of fatigue. The studies described here provide groundwork for further research on the effects of aging on learning and memory in C. elegans. Future research should explore the behavioral, neural and genetic mechanisms that underlie changes in learning and memory with age. First, behavioral experiments on the parameters of habituation can further define the effects of aging on non-associative learning. Second, laser ablation studies and work with mutants missing particular neurons will aid in understanding the roles of individual neurons in the aging nervous system. Finally mutants with specific changes in the aging process may be used to examine the genetic mechanisms underlying changes in learning and memory with age. For example, Johnson (1987) isolated a single gene mutation that produces a 60-I 10% increase in life span; attempts are underway to clone this gene. By integrating the findings of these various lines of research a better understanding of the effects of aging on behavioral plasticity in C. elegans
may be achieved.
Acknowledgements We thank Cathy Chiba for her editorial input on an earlier version of this manuscript, and Brett Broster, Shannon Cerniuk, Bill Mah, Theresa Marion and Steve Wicks for useful discussions. This work was supported by a BCHCRF grant to C.H.R.
The genetics of Caenorhabditis
Chalfie, M., 1984.
Chalfie, M. and Au, M., 1989. receptor neurons. Chalfie, M., S&ton,
Glass, G.V. and Hopkins, Groves,
7: 197-202. elegans touch
J.N. and Brenner, S., 1985.
elegans. J. Neurosci.,
elegans. J. Neurobiol.,
Coulson, A., Sulston, J., Brenner, S. and Karn, J., 1986. nematode Caenorhabditis
of the Caenorhabditis
I.E., White, J.G., Southgate, E., Thomson,
reversals in the nematode Caenorhabditis
Genetic control of differentiation
circuit for touch sensitivity Chiba, C.M.
elegans. Genetics, 77: 71-94.
Neuronal development in Caenorhabditis
5: 956-964. of spontaneous
Towards a physical map of the genome of the
elegans. Proc. Natl. Acad. Sci. USA, 83: 7821-7825.
Statistical Methods in Education and Psychology (2nd Edn.).
Englewood Cliffs, 436 pp.
41 g-450. Hodgkin, J., Edgely, M., Riddle, D.L. and Albertson, The Nematode Caenorhabditis
Genetics. In: W.B.
elegans. Cold Spring Harbor Laboratory, Cold Spring Harbor,
491-586. Johnson, T.E.,
Aging can be genetically dissected into component processes using long-lived
lines of Caenorhabditis Kumar, N., Williams,
elegans. Proc. Natl. Acad. Sci. USA, 84: 3777-3781.
M., Culotti, J. and van der Kooy, D., 1989.
the nematode C. elegans. Sot. Neurosci. Abstr., Rankin, C.H.,
and Chiba, C.M.,
Evidence for associative learning in
elegans: A new model system
for the study of learning and memory. Behav. Brain Res., 37: 89-92. Rankin, C.H. and Broster,
C. elegans. Sot. Neurosci. Rankin, C.H. and Broster,
the nematode Caenorhabditis Rankin,
Factors affecting habituation and recovery from habituation in
elegans. Behav. Neurosci.,
and Chalfie, M., 1989.
circuit mutations. Sot. Neurosci. Abstr.,
Rankin, C.H. and Chiba, C.M., 1988. Neurosci.
Factors affecting habituation and recovery from habituation in 16: 626. 239-249. learning in C. elegans:
Short and long term learning in the nematode C. elegans. Sot.
Rattan, K.S. and Peretz, B., 1981.
Age-dependent behavioral changes and physiological changes in
identified neurons in Aplysia californica. J. Neurobiol., White, J.G., Southgate, E., Thomson, of Caenorhabditis
J.N. and Brenner,
elegans. Phil. Trans.
Wicks, S.R. and Rankin, C.H.,
in C. elegans. Sot. Neurosci.
469-478. The structure of the nervous system
Lond. B Biol. Sci., 314:
Circuit analysis of interactions between two antagonistic reflexes