olfactory plasticity in Caenorhabditis elegans: a separation of adaptation and habituation
NIRIT BERNHARD
A thesis submitted in conformity with the requirements for the degree of Master of Science
Graduate Department of Anatomy and Ce11 Biology University of Toronto
O Copyright by Nirit Bernhard 1999
National Library (*m of Canada Bibliothèque nationale du Canada
Acquisitions and Acquisitions et Bib liographic Services services bibliographiques 395 Wellington Street 395, rue Wellington ût!awaON K1A ON4 Ottawa ON K I A ON4 Canada Canada
The author has granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sel1 copies of ths thesis in microform, paper or electronic formats.
L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/film, de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.
OLFACTORY PLASTICITY IN CAENORHABDrrIS ELEGANS: A SEPARATION OF ADAPTATION AND HABITUATION
Master of Science 1999
NIRIT BERNHARD
Department of Anatomy and Ce11 Biology University of Toronto
Abstract
Continuous presentation of an oifactory stimulus causes a chemotaxis response
decrement in the nematode Cuenorhabditis eelegans, but the differences between the
leaming process of habituation (a reversible decrease in behavioural response) and other
olfactory plasticity such as adaptation (a decrement in response due to sensory/motor
fatigue, which c m o t be dishabituated) have not been addressed. Using the volatile
odorant diacetyl (DA) 1 assessed the distinct processes of olfactory adaptation and
habituation. Pre-exposing and testing worms to 100% DA caused a chemotaxis
decrement which was not reversible despite the presentation of potentially
dishabituating stimuli. This DA adaptation is abolished in odr-10 mutants but remains
intact in odr-1 mutants. Although DA pre-exposure to intermediate concentrations of DA
(0.01% and 25%) produced no chemotaxic response decrement, pre-exposure to low DA
concentrations (0.001%) produced either a dishabituable response decrement
(habituation) or sensitization. The distinct behavioural effects o b s e ~ e d after DA pre-
exposure highlight a concentrationdependent dissociation between adaptation and
habituation.
1 would to thank rny supervisor, Derek van der Kooy, for al1 his tirne, patience and advice over the past couple of years. From you 1 have gained a new appreciation for the critical scientific mind, leamed about the advantages of thinking laterdy and being BRIEF. Thank you for expanding my ability to use the words: shockhg, pathetic, pitiful and dandy. 1 would &O like to thank Mike Wiley for king an inspiring teacher and helpful graduate student cooridinator and the members of my supervisory committee, Marc Perry, h Shettleworth, and especiaily Marla Sokolowski for all the staüstics help.
Throughout these advenhirous two years in the windowless basement of MSB (a.k.a. the Zombie Zone) I have seen many people corne and go (some permanently). 1 thank al1 the lab rats who made every day memorable. Firstly, I would like to thank the excellent technical support that has been the backbone of the worm project: Sue - you the woman! Thanks for your wit and good humour, and for listering to me as I cut the parafilm. Rachel, Nilo and Jim, thanks for all yow help. I'd also like to thank Nadine Livaya and Inge Marge for their contributions to UUs project. nianks to Bill for the interesting worm discussions with a molecular twist. Glenn, thanks for putting up with all of my stupid questions - your help has been invaluable, and your patience inspirational. Of course, I am also indebted to: Brenda - for the smiles and stupid e- mails; Danka - for the smiles and the stories; Hance - for king so unorganized and teaching me the significance of the Benjamins; Catherine - for providing constant background stories; Steve - for being a source of cwiosity; Dave-Bob - for being out, about and in-my-face; Raewyn - for being yourself; Cindi - for bringing in the sunshine, for your advice and understanding about what a schlep life can be sometimes; Colleen - for being so sweet, understanding and a constant source of exercise and culinary knowledge and Vince - for everything. Cuesta bella persona - the wind at my back - every moment has k e n memorable. Thank you so much for the stimulating discussions, support, advice, incredible patience and understanding, and for being there.
Lastly, and most importantly, thank-you Mom, Dad, Roy, Eric, Igael, Beni and family, Bali Dora and all of the rest of my family, extended family and wonderfd friends (especially H.N., R.R. and A.B.) - you have been so supportive, understanding, caring and patient. Thankyou justdoesn'tsayenough, but it'sall Icancomeupwith. It's been a long ride, prepare for part II of the adventure!!
Abs tract
Acknowledgments
Figure Index
Introduction
Methods
Results
Discussion
Bibliography
TABLE OF CONTENTS
Page
iii
LIST OF FIGURES
Figure 1: Habituation/ Adap tation Apparatus
Figure 2: DA Dose Respowe Curve
Figure 3: tncreasing exposure time, but not volume, affects degree of approach decrement after pre-exposure to 100% DA.
Figure 4: Lack of dishabituation after pre-exposure to 100% DA.
Figure 5: No response decrement seen after exposure to intermediate DA concentrations (25%, 0.01 %)
Figure 6: Nonassociative learning occurs after pre-exposure to low concenhations of DA (0.001%)
Figure 7: Adaptation is odr-10 dependent, but odr-1 independent.
Figure 8: Adaptation, Habituation and Sensitization are separate forms of olfactory plasticity.
Page
13
17
20
INTRODUCTION
Introduction
Our understanding of the mechanisms underlying some forms of learning and
memory have been greatly facilitated by the use of invertebrate mode1 systems.
Extensive studies on associative and nonassociative learning have been camed out in
molluscs, such as Aplysia californica (Castelucci et al. 1970; Carew and Sahley 1986;
Walter et al. 1979, 1981) and Hennissenda crassicomiss (Rogers and Matzel, 1995), as
well as in other invertebrate species such as Drosophila melanogaster (Corfas and Dudai
1988; Qullui et al. 1974; Tully and Quinn 1985) and Apis melli@ra camica (Braun and
Bicker 1992; Hammer 1997; Hellerstern et al. 1998). More recently, the nematode
Caenorhabditis elegans has been identified as a useful organism for the study of
learning and memory (Hedgecock and Russell 1975; Morrison et al. 1999; Rankin et
al. 1990; Wen et al. 1997). C. elegans has a relatively simple nervous system,
composed of 302 neurow whose synapses have been fully mapped out at the electron
microscopic level (Hali and Russel 1991; White et al. 1986). Combined with its
detailed physical map and a nearly completely sequenced genome (C. elegans
Sequencing Consortium 1998; Sulston et al. 1992), C. elegans is well suited for
ident@ing specific molecular pathways and genes that play a role in learning,
memory and other forms of behavioural plasticity.
The search for molecular and genetic components of learning and memory in
any organisrn must first begin with a clear definition of the different types of
learning being investigated and distinguishing between leaming and other forms of
plasticity . Different temporal relationships between stimuli during training d o w s
for classification of two elementary types of leaniing: nonassociative and associative
3
learning (Byrne et al. 1987). Whereas associative learning involves a change in
behaviour due io specific temporal contingencies between stimuli or between a
stimulus and a behavioural response, nonassociative learning involves modification
of a behaviour due to presentation of a single cue (Brown 1998; Carew and Sahley
1986). Types of nonassociative learning include sensitization, habituation and
dishabituation. Sensitization refers to the enhancement of a behavioural response
caused by a strong and often noxious stimdus (Groves and Thompson 1970).
Habituation is the decrement in response (independent of sensory/motor fatigue or
adaptation) due to repeated or continuous presentations of a single stimulus (Groves
and Thompson 1970; Harris 1943; Thompson and Spencer 1966). On the other hand,
dishabituation is the restoration from an habituated respowe following presentation
of a novel or noxious sthnulus, which has been hypothesized to be a superimposition
of sewitization on the habituated response (Groves and Thompson 1970) or may
involve different processes from sensitization (Cohen et al. 1997). The re-
establishment of baseline response levels due to a dishabituating stimulus is distinct
from a spontaneous recovery of the habituated respowe, which occurs without a
dishabituating stimulus after a prolonged period of time (Thompson and Spencer
1966).
Harris (1943) and 'Thompson and Spencer (1966) suggest that continuous or
repeated stimulus presentation leading to sensory adaptation can be caused by
decreased receptor activity (receptor adaptation) or limitations of effector response
(efictor fatigue). Respowe decrements that can be accounted for by either of these
mechanisms are not referred to as habituation, since they indicate an inability to
respond to the stimulus as opposed to an active learning associated modulation in
4
behaviour. There are examples of newoadaptation within the central nervous system
(Ochoa et al. 1990), as well as in sensory systems such as the mechanosensory (Corfas
and Dudai 1990), visual (Dizhoor et al. 1991; Kawamura and Murakami 1991) auditory
(Rauschecker and Korte 1993) and olfactory systems (Chen and Yau 1994; Dawson et
al. 1993; Kramer and Siegelbaum 1992; Kurahashi and Menini 1997; Leinders-Zufall et
al. 1999). While the olfactory studies f o w mainly on the cellular changes that lead to
modulations in sensory transduction, they do not assess the behavioural processes
that underlie the changes in olfactory perception leading to olfactory behavioural
plas ticity .
The ability of C. elegans to sense, approach and discriminate between volatile
odorants has been shown to be manipulable at the cellular and genetic levels
(Bargrnann et al. 1990; Bargmann et al. 1993; Colbert and Bargmann 1997). Laser
ablation studies have identified the AWA and AWC primary chemosensory neurons
as responsible for mediating the olfaction to specific odorants, while ethyl
methanesulfonate mutagenesis of wild-type Worms has identified novel genes
involved in olfaction (Bargmann et al. 1993; Colbert and Bargmann 1995; Sengupta et
al. 1996). The AWA neuron has been shown to mediate approach to pyrazine as weU
as lower concentrations of diacetyl (Bargmann et al. 1993), and the odr-10 gene which
encodes a seven trammembrane domain diacetyl receptor is expressed in AWA
(Sengupta et al. 1996; Troemel et al. 1997; Zhang et al. 1997). In addition, odr-3 (a Ga-
protein subunit proposed to be involved in dowwtream intracellular signaling) and
osm-9 (a putative ion channel with some homology to the Drosophila transient
receptor potential protein) have been identified and behaviourally characterized as
affecting AWA olfactory h c t i o n (Colbert et al. 1997; Roayaie et al. 1998). Laser and
5
genetic ablations of the AWC neuion indicate that this neuron mediates responses to
odorants such as benzaldehyde, butanone and isoamyl alcohol (Bargmann et al. 1993).
In addition to the high-affinity odr-IO diacetyl receptor on AWA, there appears to be a
putative low-affinity diacetyl receptor on AWC which mediates response to high
concentrations (100%) of diacetyl and 2,3-pentanedione (Chou et al. 1996). This low-
affinity DA receptor may signal to a downstream G-protein coupled receptor via the
candidate receptor guanylyl cyclase, odr-1 (N. L'Etoile and C. Bargmann 1997, penonal
communication) and lead to the opening of a cyclic nucleotide gated cation channel,
such as the tax-2/tax4 channels (Coburn and Bargmann 1996).
In C. elegans, habituation to tactile st imul i (Chalfie and Sulston 1981; Croll
1975; Rankin et al. 1990) and chemosensory stimuli (Wen et al. 1997) as well as
associative leaming (Wen et al. 1997; Morrison et al. 1999) and olfactory habituation
(Nuttley and van der Kooy submitted; Morrison et al. 1999) have been demonstrated.
In addition, Colbert and Bargmann (1995; 1997) have characterized olfactory
adaptation. While genetic components specifically underlying nonassociative
learning have yet te be identified, there has been extensive work at the cellular level
in idenhfying pathways in the tapwithdrawal response, a form of mechanosensory
habituation (Wicks and Rankin 1996,1997).
In the present study, I asked if both olfactory adaptation and habituation could
be obsemed in the same paradigm, and if these two processes could be differentiated.
After pre-exposure to very high concentrations (100%) of diacetyl @A), Worms
exhibited a chernotaxis decrement to a point source of DA which did not r e m to
baseline levels, with the presentation of various potentially dishabituating stimuli.
The adaptation of this response was dependent on proper odr-10 function (although
6
baseline chemotaxis to 100% DA did not), but odr-l was not required for DA
adaptation. h contrast to 1 0 % DA pre-exposure, after pre-exposure to low DA
concentrations (0.001%), the habituated response could be reinstated to naive
chemotaxis levels after presentation of a dishabituating stimulus (centrifugation at
250g). Moreover, at these low DA concentrations the baseline naive chemotaxic index
(CI) determined whether habituation or sensitization would occur such that High
initial Responders demowtrated habituation and Low initial Responders exhibikd
sensitization. Surprisingly, exposure to intermediate concentrations of DA (0.01 %
and 25%) did not c a w worms to exhibit a response decrement at all. Thus, there are
at least two distinct processes that underlie the observed behavioural response
decrement after continuous presentation of an olfactory stimulus, and these two
processes, adaptation and habituation, are disassociable in a concentration-dependent
METHODS
Materials and Methods
SmAINs
The Wild-type C. elegans Bristol strain (N2), the odr-10 (ky255) mutant s h a h
(Sengupta et al. 1996) and the odr-l (n 1936) mutant strain (Bargmann et al. 1993) were
utilized (mutant strains were provided by the Caenorhnbditis elegans Genetic Center).
The general culturing techniques w d are described by Sulston and Hodgkin (1988).
Adult worms were tested at approximately 4 days post hatching (3 day old from the
L1 stage). The population of worms was synchronized using the following method.
Populations of unsynchronized non-dauer plates containing many l a r d stage 1 and
2 worms were washed off a plate with distilled water (dHzO) into a 15 ml conical
centrifuge tube and spun down at 40g for 1 minute. This procedure causes the older
(heavier) worms to form a pellet, leaving the younger woms in the supernatant. The
supernatant was then transferred to a second 15 ml conical centrifuge tube and spun
down at lSOg for 1 minute. This technique of sequential centrifugation results in a
pellet of young larvae that could then be transferred to agar plates seeded with E. coli
(strain OP50) as a food source.
MATERLALS
Conditioning and testing were camed out using various different DA
concentrations, as specified. Lower concentration DA solutions (0.01 % and 0.001 %)
were prepared the day of the experiment from a stock solution of 1% diacetyl (in
ethanol) that was made the day @or. High concentration diacetyl and benzaldehyde
were aliquoted (100%) or diluted (1%, .1%, 25%) on the day of the experiment. All
attractants were obtained from Aldrich Chernicals and were always diluted in
9
anhydrous ethyl alcohol. Chemotaxic media (CD<) was prepared using an
autoclaved mixture of DIFCO Purified Agar with 100 ml/l of a 0.1 MOPS buffer (pH
7.2 with NhOH). Once the mixture was autoclaved and aliowed to cool, 2.5 ml/l
Tween 20 was added and test plates were made by pouring 6 ml of CTX into 10 ml
petri dishes (Fisher Scientific). Agar was evenly distributed throughout the plate and
aliowed to air dry for 20 minutes before capping with the iids; plates were generally
poured the day prior to the experiment. A plexiglass grid was used under the testing
plate to determine the location of placement of the odorant spots and the placement
origin of the worms (Fig. 1). A 1 pl spot of 1 M sodium azide OIJaN3) was placed on
test plates 15 minutes prior to testing in order to anesthetize animals as soon as they
chose to approach a specific attradant and to prevent the effects of odorant exposure
on leaming during the actual test period.
BEHAVIOURAL ASSAYS
Worms were placed in 15 ml conical centrifuge tubes, washed twice with dH20
and given two gentle 1 minute spins at 40g. The resulting peilet of adult worms was
carefully transferred to the conditioning CTX plate. Worms were gently dried with a
Kimwipe, and diacetyl was pipetted ont0 agar plugs piaced on the lids of the petri
dishes. The standard amount of diacetyl used was 5 4 distributed among 5 agar
plugs (variations from this amount are noted in a specific experiment). Control
animals were placed on CTX plates with no diacetyl added to the agar plugs. The
duration of odorant exposure ranged between 15 minutes for low concentration
experiments (0.001% and 0.01%) to 60 and 120 minutes for higher concentration
experiments (25%, 100%). The duration of pre-exposure was varied with different DA
10
concentrations since 15 minutes was sufficient to induce habituation at low DA
concentrations, whereas 60 to 120 minutes of pre-exposure were required to see a
similar approadi decrement at higher DA concentrations. After diacetyl pre-
exposure, dH20 was w d to wash habituated worms off of conditioning plates and
into a 15 ml conical centrifuge tube. Sufficient was added to the tubes to bring
the final volume up to 15 ml, and the worm suspension was allowed to settle for 5
minutes. The supernatant was removed and a 5 pl spot of Worms (containing as many
as two hundred worms) was aliquoted to the centre of the test plate at a point
equidistant from the aitractant and control spot using an Eppendorf micropipettor
with the tip cut short to prevent damage to the animals. The spot of worms was then
dned using a Kimwipe, and lpl spots of diacetyl or control solution (ethanol) were
placed on opposite sides of the plate. Plates were parafilmed, and left on the counter
for a 60 minute test duration, a standard tirne point (Bargrnaru et al. 1993). Each plate
consisted of 50-200 of worms and constitutes an n = 1.
Dishabituated worms were pre-exposed to diacetyl using the same procedure
as above, but instead of allowing to settle for 5 minutes in a 15 ml conical centrifuge
tube, the animals were spun down once at 250g for 1 minute, the supernatant was
removed, and more dHQ was added to a final volume of 15 ml prior to centrifuging
for a second time at 250g for 2 minutes. The speed of centrifugation varied across the
different pre-exposure experiments, and the different speeds used are noted for those
instances. Worms were placed on the test plates in an identical manner as described
above for the habituated groups. Naive Worms were placed on blank CïX plates for
the duration of exposure time (i.e. 15 minutes for low concentrations, 60 minutes for
high concentrations), and given the same strong centrifuga1 spins as the dishabituated
11
groups before testing. There was no difference in chemotaxic approach between
unmanipulated naive animais and naive worms exposed only to the dishabituation
treatments (data not shown), but as a control Naive groups were always given the
dishabituating treatment without DA presxposure. The order of groups tested were
randomized within an experiment.
SCORING AND ANALYSE
After the 60 minute test period, plates were inverted and placed at 4OC for at
least 30 minutes prior to counting. Any worms within 20 mm of the spot were
considered to have chosen that odorant (Fig. 1). A chemotaxic index (CI) adapted
from Bargmann et al. (1993) was calculated based on the numbers of animals at the
attractant :
CI = # at test spot - # at control spot Totd on plate
The CI can Vary from + 1.0, indicating pure attraction, to -1.0 indicating pure
aversion. In general, less than 10% of the w o n s do not leave the origin, and since
this sniali aggregate of animals in the middle of the plate is likely due to possible
damage inmred during the handling procedures, such clumps of worms were not
included in the CI cakulation. The % Response was w d as a measure of change from
Naive baseline CI and was calculated as:
% Response = Mean of Pre-Exposed Treatments iOO% Mean Naive Treatments
where the Mean of each treatment represents the Mean for one experiment with n = 4
plates. Mean values, standard error of the mean (SEM), regression analysis, t-tests,
analyses of variance (ANOVAs) and post- hoc analyses (Neuman-Keuls) were
calculated using the Statistica software program (Macintosh StatSoft software version
Figure 1: Habituation/Adaptation Apparatus. Conditionhg agar plates were used for exposing worms to the various concentrations of diacetyl during training. Circles with hatched bars represent agar plugs saturated with 1 pl of diacetyl, and short wavy lines represent aggregates of worms. Following treatment, hundreds of worms were placed at the origin point on a test plate with a 1 y1 control spot (ethanol, E1OH) and test spot (DA) placed on either end. Plates were treated with 1 pl sodium azide (NaN3) at control and test spots to capture the animalsr initial respowes. After a 60 minute test period woms were placed at 4 ' ~ , and later counted to establish a chemotaxic index per plate. Each plate comisting of hundreds of worms was considered an n =l; each treatment group had an average of n = 4 per experiment.
Control spot Test spot of DA
Conditioning plate Test plate
CI = # at test m o t - # at control spot Total on plate
RESULTS
Results
C. ELEGANS APPROACHES DIFFERENT CONCENTRATIONS OF DIACETYL IN A DOSE-DEPENDENT MANNER
In order to assess behavioural plasticity in C. elegnns using diacetyl (DA) as an
olfactory stimulus, the naive baseline approach to a variety of different DA
concentrations was established. The animals approached DA in a dose dependent
manner, such that the highest chemotaxic response was elicited at the highest
concentrations and as the concentration of DA became more dilute, the CI decreased
(Fig. 2). This corresponds with previous data which show that DA is a significant
olfactory attractant over a broad range of concentrations (Bargmann and Horvitz,
1991). In addition, it was also noted that the variability was highest at the lowest
concentrations of DA.
PRE-EXPOÇURE TO HIGH CONCENTRATIONS OF DIACFPiIL (100%) LEADS TO A DECREMENT IN APPROACH THAT IS TIME-DEPJ3lDENT BUT V O L W - INDEPENDENT
Ln order to test whether worms are capable of showing nonassociative
learning at very high concentrations of diacetyl, populations of woms were pre-
exposed to 5 pl of 100% DA for varying exposure times and their CI was scored after
the standard 60 minute test period. Naive animals showed very high approaches
towards a test spot of 100% DA. A Naive Starved group that was starved for 2 hours
(there was no noted difference in naive DA approach for animals starved for 30
minutes, 1 hour or 2 hours, data not shown) was used as a control for effeds of
Figure 2: Dose Response Curve. C. elegans chemotax to a broad range of diacetyl concentrations. Approach was elicited to a variety of different concentrations of diacetyl, with the highest approach to diacetyl seen at the highest dilutions. n = 4 for all groups.
18
starvation during pre-exposure t h e , however there was no significant difference
between Naive or Naive starved groups (t(7) = 1.26, P > 0.05). Within 1 hour, animals
showed a signihcant decrease (Neuman-Keuls test, P < 0.05) in approach to DA, and
this approach decreased even further after a 2 hour exposure (Fig. 3A). A one-way
ANOVA comparing Naive or Naive Starved animals to Pre-exposed groups with
varying exposure times revealed a signihcant effect of pre-exposure time (F(5,18) =
26.0, P < 0.05). Neuman-Keuls post-hoc tests demowtrated that there was no
significant effect (P > 0.05) following a 30 minute pre-exposure, but the 1 hour and 2
hour pre-exposure treatments yielded Ch that were significantly different from
Naive starved levels and from each other (P < 0.05). A 3 hour pre-exposure time was
also tested, but despite the substantial decrement in diacetyl approach (data not
shown), this lack of response was likely due to damage to the animals as seen by
general tack of movement and gross clumping at the site of origin on the test plate.
Based on these results, the 2 hour exposure period was chosen to examine the effects
of varying the volume of DA on approach. A dose-independent decrease in diacetyl
approach was seen, such that regardless of the pre-exposure volume, the decrement in
CI remained constant compared to Naive levels (Fig. 38). A one-way ANOVA
comparing Naive treatment to Re-exposed treatments with vary hg exposure
volumes revealed a significant effect of pre-exposure (F(4,15) = 7.7, P < 0.05).
Neuman-Keuls post-hoc tests showed that followuig pre-exposure, ail groups were
significantly different from Naive (P < 0.05) and were not signihcantly different from
one another (P > 0.05). Thus, t h e of exposure but not exposure volume affects the
decrement in diacetyl approach.
Figure 3: Increasing exposure time, but not volume, affects degree of approach decrement after pre-exposure to 100% DA. (A) Worms pre-exposed to 5 pl of 100% diacetyl for 30 minutes to 2 hours had significantly lower mean diacetyl chemotawis. The unstarved naive group versus the naive group that was starved for 2 hours were not significantly different. After 1 hour of diacetyl exposure, worms showed a signihcant decrease in approach, but the effed was most pronounced after the 2 hour exposure. n = 4 for all groups. (B) A 2 hour pre-exposure time to diacetyl with vasring volumes of odorant shows that the approach decrement decreased to approximately 40% of Naive levels for all groups, regardless of pre-exposure volume. n = 4 for all groups.
- 1 E
0.9 PRE-EXPOSED
0.8 c 8 0.7 - 0-6 5 8 0'5 a g 0.4
$ 0.3 O O ri 0.2 O - 0.1 U
n V
Naive Naive 30 min Starved
PRE-EXPOSURE TIME
PRE-EXPOSURE VOLUME
EXPOSURE TO HIGH CONCENTRATIONS OF DIACETYL E L K E ADAPTATION, BUT NOT EIABITUATION
Worms demonstrate a very high CI to a test spot of 100% DA and can show a
decrement in approach foliowing various pre-exposure h e s and odorant volumes.
Is this decrement in response adaptation, cawd by receptor or effector fatigue, or is
it due to nonassociative leaming which is subject to reversal? By definition,
habituation requires a behavioural demonstration of dishabituation. In order to
determine whether the observed decrease in approach behaviour represented
adaptation or habituation, worms were pre-exposed to 3 pl of 100% DA for 2 hours,
and the dishabituating stimuli varied from centrifuga1 spins of lOOg to 1000g.
Regardless of the magnitude of the spin, the stimuli were not sufficient to induce
dishabituation of the observed response decrement after pre-exposure (Fig. 4A). A
one-way ANOVA comparing Naive, Re-exposed and Pre-exposed + Spin treatments
indicated a signihcant effect of treatment (F(5,17) = 8.2, P < 0.05). In a Neuman-Keuls
post-hoc analysis, the CIs of the Pre-exposed and Re-exposed + Spins groups were
significantly different from the Naive group (P c 0.05), but not from each other (P >
0.05). In a subsequent test 1 asked if a stronger dishabituating stimulus couid c a w
the response decrement to retm to naive levels. After pre-exposure to 5 pl of
diacetyl for 1 hou, two different groups of wonns were subjected to two different
dishabituating treatments: dishabituation group 1 was given two centrifuga1 spins at
2000g and 30 seconds of vortexing, and dishabituation group 2 was spun twice at lOOg
and Gven a 15 minute cold shock at 4 ' ~ (Fig. 4B). A one-way ANOVA revealed that
animals decreased their approach response after pre-exposure to 100% DA (F(5,52) =
229, P< 0.05). Neuman-Keuls post-hoc analysis showed that foiiowing
Figure 4: Lack of dishabituation after pre-exposure to 100% DA. (A) Pre-exposure to 3 pl of 100% diacetyl for 2 hows caused Worms to exhibit a response decrement to 100% DA which could not be reversed despite being presented with dishabituating stimuli ranging from lOOg to lOOOg spins. The Naive group was given lOOOg treatment without pre-exposure to DA. Ail groups given dishabituation stimuli were significantly different from the mean Naive, but not different from the pre-exposed group without dishabituation stimuli. 100g= a 1 minute spin at 100g followed by a 2 minute spin at 100g 300g= a 1 minute spin at 300g followed by a 2 minute spin at 300% 500g= a 1 minute spin at 500g followed by a 2 minute spin at 500g. 1000g= a 1 minute spin at lOOOg followed by a 2 minute spin at 1000g. n = 4 for all groups, except n = 3 for Naive. (B) Worms exposed to 5 y1 of 100% diacetyl for one hour do not show a dishabituated response to 100% DA following different dishabituating treatments. Pre-expose = pre-exposed without dishabituating treatments, Group 1 = pre-exposure + 2 x 2000g spins (3 minutes total) + 30 sec vortex, Group 2 = pre-exposure + cold shock treatment of 4°C for 15 minutes, followed by a 1 minute spin at 100g Spont. Recovery = Spontaneous recovery d e r 3 hours. n = 20,18, 4,8,4 respectively from left to right.
NAlVE
0 PRE-EXPOSED
El PRE-EXPOSED + DISHABITUATION TREATMENT
Naive Pre- lOOg 300g 500g lOOOg expose only
DISHABITUATION spins at various g
' ~ a i v e ' Pre- Group Group Spont. expose
only 1 2 Recovery
TREATMENTS
24
conditioning, the Pre-exposed group showed a deaeased CI that was approximately
50% of the naive CI (P c 0.05), and that regardless of attempts at dishabituation, this
decrement was not reinstated to naive levels. Moreover, the decrement in response
seen in the two groups given the dishabituating treatments was not caused by any
potentially damaging effects of the treatments, since the Naive groups (combined)
were given equivalent treatments without DA pre-exposure and these treatments did
not affect the high naive baseline CI. When left for 3 hours after an equivalent DA
pre-exposure to spontaneously recover prior to testing, worms exhibited chemotaxic
indices that were not significantly different from naive levels (Fig. 48, P > 0.05).
Thus pre-exposure to 100% DA results in adaptation, since dishabituation is not
revealed despite the potency of the dishabituating stimulus.
EXPOSURE TO INTERMEDIATE CONCENTRATIONS OF DIACETn (25% AND 0.01 %) LEADS TO NO RESPONSE DECREMENT IN DA APPROACH
Since exposure to 100% diacetyl leads to adaptation, might UA concentrations
lower than 100% elicit a respome decrement that is capable of dishabituation? Pre-
exposure to 25% DA was not suffident to cause a decrement in approach to the same
25% DA stimulus regardless of exposure time or amount of diacetyl present (Fig. 5A).
The dishabituating stimulus of two strong centrifuga1 spins at 500g did not affect
approach to diacetyl in either the Naive or Pre-exposed + Spin groups. A one-way
ANOVA examining a 90 minute pre-exposure to 2 ~ 1 of 25% DA in Naive, Pre-exposed
and Pre-exposed + Spin groups revealed no signihcant effect of treatments (F(2,8) =
2.3, P > 0.05). Similarly, a 2 hour pre-exposure to 3p1 of 25% DA yielded no significant
effect of treatment (F(2,9)=2.1, P > 0.05). Since approach to 25% DA was so high, the
effects of pre-exposure to a concentration of diacetyl whch was lower than 25% DA
by several orders of magnitude (0.01 % DA) was examined. Pre-exposure to 5 p l 0.01%
DA for 15 minutes was not sufficient induce a decrement in CI to 0.01% DA, A one-
way ANOVA comparing Naive, Pre-exposed and Pre-exposed + Spin groups revealed
no signihcant effects of treatment (F(2,15) = 0.2 P > 0.05) (Fig. 58). Pre-exposure to
the same concentration of diacetyl for a longer 60 minute duration of, however,
showed a trend toward a deaease in CI, but a one-way ANOVA revealed no
significant difference between these two groups (F(2,lI) = 2.63 P > 0.05). In
subsequent attempts to see if an even longer pre-exposure time of 90 minutes would
drive the CI down even further, there was no significant reduction from Naive CI
(F(2,12) = 1.97 P > 0.05, data not shown). Despite variations in length and volume of
pre-exposure to the odorant, intermediate concentrations of diacetyl (<IO098 and 2
0.01 %) do not produce decrementd respowes.
PR.-EXPOSURE TO LOW CONCENTRATIONS OF DIACETYL, (0.001%) FAVOURÇ NONASSOCIATIVE LEARNING
The effects of pre-exposure to low concentrations of DA (0.001%) were
observed and compared with pre-exposure to intermediate DA (25% and 0.01%)
concentrations to determine if there would be a similar lack of response decrement as
seen after 25% or 0.01% DA pre-exposure. At these low odorant concentrations, naive
response to 0.001% DA varied from 0.2 to 0.8 on any given day @ut was consistent
within days). A one-way ANOVA comparing Day x Naive CI demonstrated there
Figure 5: No response decrement seen &ter exposure to intermediate DA concentrations ( 2 5 O 0 , 0.01%) (A) Pre-exposures to 2 pl of 25% DA for 90 minutes or 3 pl of 25% DA for 2 hours were not sufficient to elicit a response decrement to 25% DA in wild-type animals, regardless of conditioning treatments. The dishabituating stimulus was a 1 minute centrifugation at 500g, followed by a 2 minute 500g spin. n = 4 for all, except n = 3 for Naive 2 pL/90 minute exposure group. (B) A 5 pl pre-exposure to 0.01% diacetyl did not cause a decreased chemotaxis approach to 0.01% diacetyl despite varying the time of exposure from 15 to 60 minutes. The dishabituating stimulus was a 1 minute centrifugation at 500g, followed by a 2 minute 500g spin. n = 4 for all 15 minute group, and n = 3 for au60 minute groups.
A Exposure to 25% Diacetyl
Naive Pre-
2 pl exposure for 90 3p1 exposure for 2 hr min
B Exposure to 0.01% Diacehrl
F Pre-
Expose ' ~ r e - ~ x ~ ;se
+ Spin
5p1 exposure for 15 min
5p1 exposure for 60 min
28
was a significant effect of Day (F(2S173) = 3.0, P < 0.05). In an analysis of ail
experiments (including ail ranges of Naive CI), Pre-exposure treatment (habituation)
was compared to Pre-exposure + Spin treatment (dishabituation). Worms exhibited a
25% decrease in approach to 0.001% DA after 0.001% DA pre-exposure, and after the
Dishabituation treatment their CI retumed to Naive levels (Fig. 6A). If left to recover
for two hours after exposure, animals spontaneously recovered back to naive DA
approach levels . A one-way ANOVA corn paring Naive, Habitua ted, Dishabituated
and Spontaneous Recovery treatments showed that there was a significant effect of
treatment (F(2,282) = 10.0 P < 0.05) and post-hoc tests revealed that the Habituation
treatment was significantly different h m the Naive, Dishabituation, and
Spontaneous Recovery treatments (P < 0.05), which were not significantly different
from each other (P > 0.05).
Investigating the day to day variability, 1 noticed that when the Naive baseline
CI was higher, the habituation tended to be greater than the average 25% response
decrement. 1 therefore hypothesized that as the Naive CI increased, the % Response
after habituation treatment would decrease. A regression analysis of the % Response
to the Mean Naive CI (for each day's experiment) demonstrated a correlation of r = - 0.59 (F(1,26) = 13.6, P < 0.05) (Fig. 68). There were two noted outliers ( > 2 s.d. away
from mean CI and mean % Response). When each of these points were excluded from
the analysis, the correlation stiîl remained significant (F(1,24) = 6.1, P < 0.05, r = -0.45).
Based on this regression analysis, the experiments were divided into two groups: the
Sensitized Group, which exhibited an increase in % Response after 0.001 % DA pre-
exposure, and the Habituated Group, which demonstrated a decrease in % Response
after 0.001% DA pre-exposure. A closer analysis of the Çewitized group revealed that
29
the mean Naive CI was 0.34 f 0.02 with a mean % response of 51.9% f 7.7% (n = 33)
after DA pre-exposure, while the mean Naive CI of all scores in the Habituated group
was 0.48 î 0.01 with a mean % response of -51.5% I 4.3% (n = 67) after DA pre-
exposure. A one-way ANOVA comparing Group to Naive CI and % Response
revealed a main effect of Group x Naive CI (F(1,98) = 32.9, P < 0.05) and % Respowe
(F(1,98) = 162.3, P < 0.05). Neuman-Keuls post-hoc analyses demonstrated that both
Naive CI and % Response were significantly different between the Habituated and
the Çensitized groups (P < 0.05).
Based these results and previous expeninents which suggest that Low initial
Responders (in this case, those with low Naive CI) tend to demonstrate sensitization
and High initial Responders (those with a High Naive CI) demonstrate habituation
(Eisenstein et al. 1991; Eisenstein 1997), all the experiments were divided into two
groups: the Habituated group with % Response less than zero as the High initial
Responders (naive CI values > O.%), and the Sensitized group with % Response
greater than zero as the Low initial Responders (naive CI < 0.34). An analysis of Low
initial Responders and High initial Responders in a two-way ANOVA comparing
Group x Treatment demonstrated a signihcant effect of Group alone (F(1,279) = 33.7, P
< O.OS), and a signihcant interaction between Group and Treatment (F(2,279) = 17.9, P
< 0.05). Worms in the High Responder group (naive CI > 0.34) that were pre-exposed
to 5 111 DA for 15 minutes showed a 40% decrement in approach to the odorant in the
standard chernotaxis test (Fig. 6C). In addition, this decrement could be dishabituated
upon presentation of a strong stimulus of two centrifugai spins of 250g. A one-way
ANOVA comparing Naive, Habituated, Dishabituated and Spontaneous Recovery
treatments in the High initial Responder group revealed a signihcant effect of
30
sigruficant effect of treatment (F(2,214) = 27.4, P < 0.05). Post-hoc tests indicated that
following the habituation treatment, the CI to 0.001% DA was significantly different
h m naive and dishabituatecl leveis (P < 0.05). Following the dishabituation
treatment, the CIs increased to a level that was significantly different from
Habituated (P < 0.05), but not statistically different from naive levels (P > 0.05).
Interestingly, under the exact same Pre-exposure and Pre-exposure + Spin
treatments, pre-exposure in the Low initial Responder group (naive CI < 0.34)
showed sensitization, as seen by a 25% increase above baseline levels (Fig. 6D). A
one-way ANOVA comparing Naive, Pre-exposed and Preexposed + Spin
(dishabituation treatment) groups demonstrated a significant effect of treatment
(F(2,89) = 3.6, P < 0.05) and post-hoc analysis indicated that the pre-exposed group was
significantly different from Naive (P < O.OS), while the Pre-exposed + Spin group was
not significantly different from either Naive or Pre-exposed groups. Thus, when
baseline approach to DA is high (> O%), wild type Worms exhibit habituation and
when baseline CI is low (< 0.34) they display sewitization.
Figure 6: Nonassociative leamhg occurs after pre-exposure to low concentrations of DA (O.OOlO/o) (A) Pre-exposure to 0.001% DA leads to habituation. Worms pre-exposed for 15 minutes to 5 pl of 0.001% DA showed a 25% response decrement when tested to 0.001% DA compared to naive values. Dishabituated groups centrifuged once at 250g for 1 minute and again at 250g for 2 minutes exhibited a recovery of chemotkwic approach to 0.001% DA that was not significantly different from the Naive or Spontaneous Recovery groups. Al1 naive wonns were given the dishabituahg treatment without diacetyl pre-exposure. n = 100,106, 103,10, respectively from leh to right. (B) Linear regression of % Response versus Mean Naive CI. Analysis revealed a correlation of r = -0.59 (represented by the slope of the diagonal line) between % Response (% difference between Mean Naive CI and Mean Pre-Exposed CI) and Mean Naive CI. Mean Naive CI is calculated as the mean of 4 plates within one day's experiment. n = 28 individual experiments. The dashed horizontal line represents no change in behavioural response (0% Response) after pre-exposure treatment; all points above the line reflect the experiments where sewitization was obsewed (Sewitized group) and those below the 0% Response represent the Habituated group. (C) High initial Responders to 0.001% DA show habituation after 0.001% DA pre-exposure. Naive animals with baseline CIs of greater thm 0.34 were classified as the High initial Responder group. Worms exposed for 15 minutes to 5 pl of 0.001% DA showed a 40% response decrement to 0.001% DA compared to naive values. Dishabituated groups exhibited a recovery of chemotaxic approach to DA that was not significantly different from the Naive or Spontaneous recovery groups. 11 = 69, 77/71, respectively from left to right. (D) Low Naive Baseline CI to 0.001% DA favours sensitization. in Low initial Responders where wiid-type naive CIs are less than or equal to 0.34, pre-exposure to 5 pl 0.001% DA caused a 25% increase above baseline levels. n = 31,29,32, respectively from left to right.
0.55
0.5
E: W 0.45 rn +l 0.4 4 O S
; 03 0.25 s Sensitizcd Cmup s" 0.2
$ 0.15 Hahiiuatcd Gmup CI t; 0.1
0.05
O 9C Naive Habituatcd Dishabituateci Sporitcncous O .1 .2 .3 -4 .5 .6 .7 Rt?Covcry
Habituafed Group - % Rrsponsc < O Sensitizd Group - % Response > O
High Initial Respondcrs: Mean Naive CI>0.34
Low [ni tial Responders: Mean Naivc. CIc0.34
33
ADAPTATION AND NONASSOCIATIVE LEARNING ARE DISTINCT PROCESSES
THAT ARE SEPARABLE IN A CONCENTRATION-DEPENDENT MANIER
Re-exposure to both 100% DA and O.ûûl% DA elicited a response decrement,
however the decreased response after pre-exposure to 0.001% DA (in High
Responders) could be dishabituated whereas the decreased response after 100% DA
pre-exposure could not. In addition, pre-exposure to intermediate concentratiow of
DA did not elicit any signihcant response decrement. In order to test if there is a
signihcant difference between the various concentrations (100%, 2596, 0.01% and
0.001 % DA) and treatments (Naive, Pre-exposed/ Habituation, and Pre-exposed +
Spin/ Dishabituation), a two-way ANOVA was carried out. This showed a signihcant
interaction of Concentration x Treatment (F(6,432) = 5.27, P < 0.05), demonstrating
that the effects of the treatments are different at differing DA concentrations.
DIA- ADAPTATION REQUIRES odr-10, BUT NOT odr-l
Wild-type N2 worms demonstrate adaptation after being pre-exposed to 100%
DA, however the receptor which mediates th& response is unknown. Approaches to
high concentrations of diacetyl are likely mediated by a low-affinity DA receptor on
the AWC neuron, since animals lacking the odr-20 gene or function of the AWA
neuron s till approach high DA concentrations (Chou et al. 1997; Sengupta et al. 1996),
but animals with deficiencies in AWC function show impaired chernotaxis to high
DA concentrations (Bargmann et al. 1993). Thus, mutations that affect the hinction of
the two putative DA sensing primary chernosensory neurons (odr-IO which is the
34
high-affinity receptor in AWA (Sengupta et al. 1996) and odr-2 which affects AWC
function (Bargmann et al. 1993)) were tested for their adaptation respowes to 100%
DA. Wild-type, odr-10 and odr-l strains were pre-exposed for 60 minutes to 5 pl of
100% DA, and tested for their responses to 100% DA, 0.1% DA and 1% benzaldehyde
(BZ) (Fig. 7). As expeded, Md-type worms pre-exposed and tested to 100% DA
demonstrated an -50% decrease in CI compared to wild-type Naive. Wild-type
worms pre-exposed to 100% DA and tested to 0.1% showed an even larger decrement
in approach (-75% decrease). Pre-exposed wild-type anirnals tested to either 100%
DA or 0.01% DA did not display any signs of dishabituation after being spun for three
minutes at 500g (Pre-exposure + Spin treatment). Approach to BZ was high and while
there did appear to be a small yet signihcant decrease after DA pre-exposure
(Neuman-Keuls post hoc tests, P < 0.05), this slight decrease was not different from
Pre-exposed + Spin CI levelç which were not significantly different from Naive (P >
0.05). A two-way ANOVA revealed a signihcant interaction of treatments with
olfactory stimuli (F(4,125) = 9.84, P < 0.05) for wild-type and post-hoc analyses
confirm that 100% DA pre-exposure caused a significant reduction in DA approach to
both 100% DA and 0.1% DA @<0.05) that could not be dishabituated.
Worms of the odr-20 strain exhibited high naive chernotaxis to 100% DA and
1% BZ, with approach to 0.1% DA almost completely eliminated (shown by Sengupta
et al. 1996). Intereshgly, testing to 100% DA revealed that unlike wild-type, odr-10
worms did not dernonstrate any signihcant decrement in response to DA after a 60
minute pre-exposure to 100% DA. A three-way ANOVA of odr-20 and wild-type
worms comparing Strain x Treatment x Olfactory Stimulus showed signihcant
interactions (F(4,197)=7.69 p<0.05) and a closer cornparison of Strain x Treatment in
35
response to 100% DA using a two-way ANOVA showed a significant interaction
(F(2,68) = 5.87 pd.05). Post-hoc cornpansons of wild-type and odr-10 groups pre-
exposed and tested to 100% DA yielded signihcant differences between wild-type
Naive and Pre-exposed or Re-exposed + Spin groups (P < 0.05) (with no difference
between Pre-exposed and Presxposed + spin (P 1 0.05)), while no significant
differences between odr-20 Naive and Pre-exposed or Pre-exposed + Spin groups (P >
0.05) were observed. Thus, unlike wild-type, odr-20 worms did not adapt.
The odr-1 worms displayed a compromised naive CI to 100% and 0.1% DA
@<0.05, significantly different from the comparable wild-type values). Despite this
lower baseline, however, after 100% DA pre-exposure the odr-1 worms still
demonstrated a -50% and -80% decreased CI when tested to 100% DA, and 0.1% DA,
respectively. This decrement in the odr-1 worms could not be dishabituated with a
diçhabituating stimulus (3 minute sp.h at 500g). An almost complete lack of BZ
approach in Naive odr-l worms confirms the prior classification of the odr-2 stralli as
particularly deficient in AWC function (Bargmann et al. 1993). A three-way ANOVA
comparing Strain x Treatment x Olfactory Stimulus interactions between wild-type
and odr-2 strains revealed signihcant interactions (F(4,200) = 4.03, P < 0.05). Within
the wild-type and odr-1 strains, a simple cornparison of response to 0.1% DA (the
concentration where wild-type and odr-1 woms had naive CI values that were most
similar) in a two-way ANOVA investigating Strain x Treatment interactions showed
a signihcant interaction (F(2,78) = 5.20, P < 0.05). Excluding the previously noted
differences in BZ approach, the interactions in the above cornparisons could be
accounted for entirely by the differences in Naive CI between the wild-type and odr-1
strains, since the odr-1 Naive CI to both 100% DA and 0.1% DA were significantly
36
different from wild-type Naive CI (Neuman-Keuis, P < 0.05). Both straiw
demonstrate a signihcant decrease in CI after 100% DA pre-exposure tested to either
100% DA and 0.1% DA (P < 0.05). Neither the wild-type Pre-exposed nor the odr-1
Pre-exposed groups differed significantly from each other, although unlike wild-type
woms, there appeared to be a trend towards dishabituation in the odr-l Pre-exposed
+ Spin group. Closer analyses of this trend using a two-way ANOVA to investigate
interactions of Strain x Pre-exposure or Pre-exposure + Spin treatments to 100% DA
and testing to 0.01% DA, revealed no significant interaction between Strain or
Treatment (F(1,53) = 3.7, P > 0.05) and that the differences between odr-1 Pre-exposed
or Pre-exposed + Spin treatments were not signihicantly different from each other or
control values (P > 0.05). The overall low baseline approach to DA and BZ in odr-1
may suggest a broad inability to sense volatile odorants or to chemotax normally.
To test this possibility, 1 camed out an additional experiment to ask if the odr-1 strain
had any generalized motor or sensory deficits by obsewing the chernotaxis approach
to the AWA sensed odorant, pyrazine (Bargmann et al. 1991). Baseline (naive)
approach to 2 mg/ml of pyrazine was tested and 1 observed that odr-2 naive CI did
not differ signüicantly from wild-type naive CI (wild-type Naive CI = 0.47 f 0.05, odr-
1 Naive CI = 0.61 f 0.12, n = 4, 4, respectively; P > 0.05). Overall, these experiments
demonstrate that wonns are stili capable of exhibiting normal adaptation to DA
(through an odr-2-independent pathway), and that lack of odr-10 prevents adaptation.
Figure 7: Adaptation is odr-10 dependent, but odr-2 independent. Wild-type, odr-IO and odr-2 worms were pre-exposed to 5 pl of 100% DA for 60 minutes and tested to 100 % DA, 0.1% DA or 1% BZ. Wild-type worms displayed a 50 % decrease in CI to 100 % DA and a 75% decrease in CI to 0.1 % DA (compared to their respective Naive values), but the CI to BZ did not show a similar decrement after 100% DA pre- exposure. Pre-exposed group = pre-exposure with a gentle wash and settled in a tube; Pre-exposed + Spin = presxposed and spun at 500g for 1 minute followed by another 500g spin for 2 minutes. Within each strain, there was no significant difference behveen Pre-exposed or Pre-exposed + Spin treatments. odr-IO animals, which showed a nearly zero response to 0.1% DA, did not demonstrate DA adaptation after 100% DA pre-exposure and testing, and did not show any difference in BZ approach compared to wild-type. odr-1 animals diçplayed lower baseline CIs to 100% DA and 0.1 % DA and no response to BZ, yet still demonstrated DA adaptation to 100% and 0.1% DA after 100% DA pre-exposure. Respectively from left to right: for Wild-type 100% DA, n = 17,16,15; Wild-type 0.1% DA, n = 17,18,15; 1% BZ, n = 12,11,14. For odr-10 100% DA, n = 9,9,8; 0.1% DA, n = 9,9,10: 1% BZ n = 9 for all groups. For odr-1 100% DA,n=7,6,8,O.l% DA,n = I l , 12,12andl% B Z n =11,9,8.
DISCUSSION
DISCUSSION
The decrement in chemotaxic approach subsequent to continuous presentation
of the volatile attractant diacetyl c m be accounted for by hvo distinct foms of
olfactory plasticity depending on the odorant concentration. After pre-exposure to
high DA concentrations (100%), the obsewed chemotaxis decrement could not be
reversed to naive approach levels despite strong dishabituating stimuli, and thus the
cause of such a decrement is likely due to some form of sensory or motor fatigue
yielding behavioural adaptation of the response. M e r pre-exposure to intermediate
DA concentrations (0.01% to 25%) there was no observable decrease in DA
chemotaxis. After pre-exposure to low concentrations of DA (0.001%) the behavioural
plasticity observed was due to nonassociative learning: the decreased response was a
result of habituation in High initial Responders (naive CP.34) since a strong stimulus
was sufficient to induce dishabituation, and an increased CL above baseline levels in
tow initial Responders was indicative of sensitization. On a mechanistic level, lack
of the ODR-10 high-affinity DA receptor on the AWA neuron prevented DA
adaptation without affecting high baseline naive CI to 100% DA, and knocking out
AWC primary chemosensory function by eliminating the odr-2 gene reduced baseline
naive CI, but did not prevent normal DA adaptation.
In order to account for the concentration specific effects of DA pre-exposure the
following three processes are postulated to underlie the observed behavioural
responses: adaptation (receptor fatigue), sensitization and habituation (Fig. 8).
Although 100% DA is sensed by both low-affinity and high-affinity receptors, naive
approach to 100% DA requires the low-affinity DA receptor through odr-1 signahg
while odr-10 may account for only half of 100% DA approach in the absence of odr-1
41
function (Fig. BA, and Fig. 7). After pre-exposure to 100% DA, excessive activation of
the high-affinity ODR-IO DA receptor causes ODR-10 downstream targets to interact
with the AWC low-affinity DA receptor pathway leading to its down-regulation and
causing adaptation of the approach to 100% DA. This interaction between the two
primary chemosensory neurons may be through common downstream targets such
as the AIY or A U interneurons. Evidence of such cross-talk between AWA and AWC
is supported by studies which demonstrate that butanone (an AWC-sensed odorant)
sensitivity is slightly increased when odr-10 is knocked out (Sengupta et al. 1996).
After pre-exposure to 100% DA, approach to this concentration of odorant can then
only be reinstated to naive levels after sufficient time for upregulûtion of the AWC-
mediated approach pathway. Despite the strong intewity of the stimulus,
sensitization is not favoured to occur at this concentration since cellular fatigue
prevents any modulation due to nonassociative learning from occurring.
At intermediate DA concentrations (0.01% to 25%) there is less activation of the
AWC low-affinity receptor as well as strong activation of the high-affinity ODR-10
receptor (Fig. 8B). While the stimulation of the ODR-10 receptor is still strong, it is
not as great as with 100% DA pre-exposure so that adaptation is less and /or
sensitization is greater. Thus, there is less cellular fatigue/ adaptation and now both
sensitization and adaptation are equally favoured, which leads to a cornpetition
between the two processes. While sensitization is favoured due to the strength of the
stimulus (high stimulus intensity), adaptation is also occurring. Any potential
decrement in response that would be caused by adaptation is opposed by the
sensitization which is attempting to facilitate the respowe. Hence, the overau
obsewed behavioural response afkr DA preexposure to such intermediate
42
concentratiow is a failure to decrease the chernotaxis approach. It muçt be noted,
however, that only two different concentrations of DA that were tested within the
intermediate DA concentratiow. It is therefore possible that the ideal conditions to
observe a net behavioural response lies somewhere in between the tested
concentrations. While this remains a possibility, it may be uniikely since the
assessed concentrations (25% and 0.01% DA) covered several orders of magnitude
without any obsewable difference of treatments.
At the lowest DA concentrations (0.001%) only the ODR-IO receptor is
stimulated without activation of the AWC low-affinity DA receptor, which elhinates
any possible adaptation and d o w s the nonassociative learning processes of
sensitization or habituation to occur (Fig. SC). At low stimulus intensities, Groves
and Thompson (1970) would predict that habituation should occur; this is observed in
the High Responders (but not the Low Responders, see below) which demonstrate an
habituated response after DA pre-exposure that can be dishabituated with a
sufficiently strong stimulus. The data here suggest that on a day when the animal
does not find the diacetyl particularly appetitive, it will not habituate as well as on a
day when the cue is particularly salient. What then accounts for sensitization seen in
Low Responders? This trend of habituation in High initial Responders and
sewitization in Low initial Responders has been reported previously in the ciliate
protozoa, Spirostomurn and in the galvanic skin response of humans after being given
a shock stimulus (Eisenstein et al. 1991; Eisenstein 1997). However, the dual process
theory (Groves and Thompson 1970) would predict that a weaker stimulus is more
conducive to eliciting an habituated response while a stronger stimulus would elicit
Figure 8: Adaptation, Habituation and Sensitization are separate processes -
which underlie the changes in behavioural response after pre-exposure to diacetyl. (A) Under baseline conditions the putative low-affinity DA receptor on AWC mediates a portion of the naive approach to 100% DA via an odr-1-dependent pathway. Although the ODR-IO receptor on AWA is not necessary for naive approach to DA, it can account for half of the response to 100% DA in the absence of AWC hinction. Pre-exposure to 100% DA causes excessive activation of the high-affinity odr-10 receptor on AWA which leads to a down regulation of the AWC mediated approach pathway causing adaptation. Thiç interaction between AWC and AWA may be via direct connections between the two primary chemosensory neurons or tluough a common downstream target. Lack of the putative receptor guanylyl cyclase in AWC encoded by odr-l partially eliminates baseline approach to 100% DA, but does not prevent the odr-20 mediated adaptation. Any sensitization that may be occurring in AWA is overshadowed by the adaptation. (8) Pre-exposwe to intermediate DA concentrations causes less stimulation of the low-affinity DA receptor on AWC and strong stimulation of the ODR-IO DA receptor @ut less stimulation than with 100% DA so that adaptation is less and/or sensitization is greater). This leads to the processes of adaptation and sensitization to be equivalently favowed and the cornpetition between these two opposing processes results in no net behavioural decrement of DA response. (C) Pre-exposure to low DA concentrations causes stimulation of only the odr-10 receptor (no activation of the low-affinity DA receptor), eliminating any possible adaptation. This results in either of two nonassoaative learning processes: habituation or sensitization, depending on the baseline initial response to the odorant. Gray = inactive neurons; Blue = activated AWA neuron; Red = activated AWC neuron.
3 A High DA Concentrations , B intermediate DA Concentrations I C Low DA Concentrations -
(1 W/O) I BASELINE RESPONSE I
ADAPTATION > SENSITIZATION
DECREMENT IN D A RWPONSE, NO RETURN TO BASELINE WLTH DISHABïIWG STIMULI
ADAPTATION NO DECREICIENT IN DA RESPONSE
NON-ASSOCIATIVE LEARNING
45
sensiti~ation~ therefore questionhg why there should be any sensitization in Low
Responder animals exposed to a weak DA stimulus (0.001%). in this case, the
demonstration of sensitization in responders with low initial CI to DA is likely due
to the fact that once the animals' baseline response has reached su& a low output
response, the probability of the response to be driven down any hrrther is less likely
and with a less prominent habituation, the cornpethg process of sensitization is
unmasked.
This does not address the curious phenornenon of why there is such great
variability in the approach to low concentrations of diacetyl on different
experimental days. Environmental factors, such as temperature and relative
humidity are known to affect the response of C. elegans in a variety of behavioural
paradigms (Hedgecock and Russell, 1975; Gannon and Rankin, 1995). 1 attempted to
maintain these variables as consistent as possible hom day to day, however, formal
assessmentç on the degree to which these factors affect nematode leaming are still
lacking. In addition to these environmental signals, food availability and density of
worms on growth plates change dauer signals and have been shown to affect
chernotaxis and benzaidehyde adaptation (Colbert and Bargmann, 1997). Although
ali woms are prepared in the same way, it is possible that variation in the amount of
E. coli present on the NGM plates from one experiment to another may lead to
differences in nematode development that a f k t the worms abiiity to withstand the
effects of starvation at a later stage, such as during the odorant exposure period.
Another developmental consideration might be the absolute age of the worm; the
stage at which the worms are tested is on the cusp of adulthood, straddling the
reproductive peak. If it is the case that age plays such an important role, then a
46
simple experiment comparing naive chemotaxic approach to low concentratiow of
diacetyl in different aged Worms should address this issue. Attempts at tlus have
yielded inconclusive results, since woms that are younger or older in age than those
presently used have movement deficits which affect their chemotaxis to volatile
odorants (unpublished obseruntions). The multitude of environmental factors which
may affect the sensitive response to low DA concentratiow do not appear to have
such a great influence on more intense stimuli such as the higher DA concentrations,
since cues such as 100% DA are so strong and may be potentially noxious so that it
would be evolutionarily disadvantageous for the animal to have a variable response
to such a salient cue.
Other experiments in drosophila have demonstrated that the mutants rutabaga
(rut) and dunce (dm) can show normal adaptation despite their inability to habituate
normally (Corfas and Dudai, 1990). Within studies investigaüng C. rlegans olfactory
plasticity, however, there has been a lack of interest in systematically categorizing
these independent processes. For example, Colbert and Bargmann (1995) have
demonstrated benzaldehyde 'adaptation' by showing that following a 90 minute pre-
exposure to benzaldehyde (sensed by the AWC neuron), worms exhibit a diminished
chemotaxis towards a test spot of this odorant. This decrement in approach was not
reversed after washing and centrifugation three times in a buffer, and animals were
only able to restore their initial attraction to the odorant after a waiting period of
three hours (spontaneous recovery). The animals' ability to restore their olfaction to
the odorant atiests that there was no permanent damage done to the worms.
However, a similar experiment (Nuttley and van der Kooy, subrnitted) has
demomtrated that dishabituation can be achieved after BZ pre-exposue if worms are
47
washed in dH20 and given a stronger centrifuga1 spin. Thus, a phenomenon
previously identified as olfactory adaptation has all the hallmark characteristics of
nonassociative learning, suggesting this type of odorant pre-exposure termed
adaptation may be a misclassifica tion and shouid be considered habituation. Here, 1
demonstrate that a decrement in DA chernotaxis after pre-exposure to DA can be
caused by both of these processes, and it is the ability of the worms to exhibit a return
to naive approach subsequent to a dishabituating stimulus that defines whether
adaptation or habituation is occurring. Although habituation has been traditionaily
defined as a decrement in response after repeated presentation of a single stimulus
(Harris, 1943; Thompson and Spencer, 1966), 1 consider this equivalent to a continuous
presentation of a stimulus for a given period of tirne, and my controls demonstrate
that the criteria for both nonassociative learning and adaptation c m be h M e d by
changing only the odorant concentrations (stimulus intensities) and the tirne of
exposure. The longer duration (60 minutes) necessary for animals to display a 50%
response decrement after pre-exposure to 100% DA (as opposed to the 15 minutes
needed to see similar habituation after 0.001 % DA pre-exposure) is likely a reflection
of the differences in molecular timing of the two distinct processes. Whereas
olfactory habituation may involve readily reversible processes such as receptor
phoshphory lation or similar manipulations in downstream secondary messenger
pathways, the longer the-course of adaptation may involve changes in gene
transcription leading to downregulation of olfactory receptors or upregulation of
other unidentified proteins which may act to inhibit chernotaxis via an adaptation-
dependent process. Such differences in timing also serve to explain why habituation
can be easily reversed with a dishabituating stimulus, whereaç adaptation requites a
48
two to three hour spontaneous recovery period before animals can demonstrate a
non-decremented respowe.
Experirnents with the odr-10 and odr-l mutants have allowed genetic dissection
of some of the molecular pathways involved in adaptation. Initial baseline approach
can be dissociated from adaptation, since odr-20 animals exhibit high naive CIs to
100% DA but do not adapt to this odorant, while odr-1 animals have a compromised
baseline CI to higher concentrations of DA (100% and 0.1%) but still demonstrate
adaptation. 1 conclude that the interactions of the hgh-affinity ODR-10 receptor
pathway in AWA with some downstream targets of the proposed low-affinity
receptor on AWC, either within AWC via a direct connection or through a common
downstream interneuron, such as AIZ, are essential for adaptation and that despite
the importance of odr-1 for mediating naive approach to 100% DA, it is not involved
in DA adaptation. In addition, the ODR-10 receptor alone also mediates responses
that lead to olfactory habituation and sensitization, however, the current lack of
knowledge regarding many downstream targets in the odr-10 pathway combined
with the fact that odr-10 animais do not approach low DA concentrations presents a
challenge to investigations of the role of the odr-10 gene in nonassociative learning.
The low naive approach to DA that was observed in odr-1 mutants is not
surprising given the postulated presence of a low-affinity receptor on AWC. It is
possible that this AWC receptor is signaling via the ODR-1 guanylyl cyclase and loss
of odr-1 leads to loss of the signaling such that only ODR-10 activity in AWA can
account for the residual approach to DA seen in odr-1 mutants. Although previously
published data suggest that approach to 0.1% DA is not irnpaired in odr-l mutants
(Bargmann et al. 1993), 1 find that the CI to 0.1% DA in odr-2 Worms is - 20% lower
49
than wüd-type. This deficit, however, was specific to DA because their approach to
another AWA-sewed odorant, pyrazine, was the same as wild-type controls.
Interesüngly, the observation that a compromised baseline approach to DA in odr-2
mutants did not prevent a 50% response decrement after 100% DA presxposure
highlights yet another difference between adaptation and habituation, since it was
noted that an habituation-related response decrement is less likely to occur when the
baseline approach to low DA concentrations is diminished. However, the trend of
habituation and dishabituation seen in odr-l animals pre-exposed to 1 0 % DA and
tested to 0.01% DA suggests that as the baseline to DA decreases and the effect of the
low-affinity receptor as part of a stimulus for adaptation is diminished,
nonassociative learning processes may be revealed.
These findings allow distinction of three separate behavioural effects of
diacetyl pre-exposure. The first type of behavioural plasticity caused by odorant pre-
exposure is habituation, as characterized by a decrement in diacetyl approach which
can be dishabituated back to naive levels. The second process underlying a
behavioural decrement in response is adaptation, which is likely caused by receptor
down-regdation or fatigue since the baseline response can return over tirne, but c m
not be dishabituated. The third form of plasticity, which has not been demonstrated
before in C. ekgans olfaction, is sensitization. This form of plasticity facilitates the
behavioural response causing an increase above baseline DA approach, or competing
with adaptation to prevent a decrement in DA approach. The development of these
paradigms in C. elegans have allowed the exploration of behavioural adaptation at a
genetic level. 1 propose that DA adaptation may rely on interactions between an
AWC-dependent (but odr-1-independent) process that relies on ODR-10 receptor
50
function in AWA. Further analyses at the molecular level will help elucidate the
underlying relationship between nonassociative learning and other forms of
olfactory plasticity by idenhfying both common and distinct genetic and cellular
pathways.
BIBLIOGRAPHY
52
Ref erences
Bargmann, C.L. E Hartwieg. and H.R. Horvitz. 1993.Odorant-selective genes and neurow rnediate olfaction in C. elegans. Cell, 74: 515-527.
Bargmann, C.I. and H.R Horvitz. 1991. Chemosewory neurons with overlapping functiow direct chernotaxis to multiple chernicals in C. elegans. Neuron, 7: 729-742.
Bargmann, C.I., J.H. Thomas, and H.R. Horvitz. 1990. Chemosensory Ce11 Function in the Behavior and Development of Gzenorhnbditis elegnns. Cold Spring Harbor Symposia on Quantitative Biology, LW 529-538.
Braun, G. and G. Bicker. 1992. Habituation of an appetitive reflex in the honeybee. l. Neurophysiol, 67: 588-598.
Brown, G.D. 1998. Operational terminology for stimulus exposure (SE) conditioning. B e h Brain Res, 95: 143-150.
Byrne, J.H. 1987. Cellular analysis of associative learning. In Phvsiol Reu @p. 329439).
C. elegans kquencing Consortium (1998). Genome sequence of the nematode C. elegans: a platform for investigating biology. Science, 282 2012-2018.
Carew, T.J. and C.L Sahiey. (1986). Invertebrate learning and memory: from behavior to molecules. Ann Rev Neurosci, 9: 435487.
Castellucci, V., H. Pinsker, 1. Kupferman, and E.R. Kandel. 1970. Neuronal mechanisms of habituation and dishabituation if the gill-withdrawal reflex in Aplysia. Science, 167: 1745-1748.
Chen, T.Y. and K.W. Yau. 1994. Direct modulation by Ca*+-calmodulin of cyclic nucleotide-activated channel of rat olfactory receptor neurons. Na hcre, 368: 545- 548.
Chou, J.H., E.R. Troemel, P. Sengupta, L. Colbert, L. Tong, D.M. Tobin, K. Roayaie, J.G. Crump, N D . Dwyer, and C.I. Bargmann. 1996. Olfactory recognition and discrimination in Caenorhabditis elegans. Cold Spring Harbor Symposia on Quantitative Biology., LXI: 157-164.
Chalfie, M. and J. Sulston. 1981. Developmental genetics of the mechanosensory neurons of C. elegans. Developmental Biology, 82: 35û-370.
Cohen, T.E., S.W. Kaplan, E.R. Kandel, and R.D. Hawkins. 1997. A simplified preparation for relating cellular events to behavior: Mechanisms contributing to habituation, dishabituation, and sensitization of the ApIysia @-withdrawal reflex. 1. Neurosci, 17: 2886-2899.
Colbert, H.A., T.L. Smith, and C.I. Bargrnam. 1997. osm-9, a novel protein with structural simiiarity to charnels, is required for olfaction, mechanosensation and olfactory adaptation in Coenorhabditis elegans. 1. Neurosci, 17: 825908269,
53 Colbert, H. A. and C.I. Bargmann. 1997. Environmental Signais modulate olfactory
acuity, disaimination, and memory in Caenorhabditis elegans. Leaming b Memo y, 4: 179-191.
Colbert, H. A. and C.I. Bargmann. 1995.Odorant-specific adaptation pathways generate olfactory plasticity in C. elegans. Neuron, 14: 803-812.
Corfas, G. and Y. Dudai. 1988. Habituation and Dishabituation of a cleaning reflex in normal and mutant Drosophila. 1 Neurosci, 9: 56-62.
Corfas, G. and Y. Dudai. 1990. Adaptation and fatigue of a mechanosensory neuron in wild-type drosophila and in memory mutants. 1 Neurosci, lO(2): 491499.
Coburn, C. and C.I. Bargmann. 1996. A putative cyclic nucleotide-gated channel is required for sensory development and function in C. elegans. Neuron, 17: 695-706.
Croll, N. 1975. Indoleakylamines in the cooridination of nematode behavioral acitivities. Cm 1. Zool., 53: 894-903.
Dawson, T., J. Arriza, D. Jaworski, F. Borisy, H. A m a d a l , R. Lefkowitz, and G. Ronnet 1993. f3-adrenergic receptor kinase-2 and fl-amestin-2 as mediators of odorant-ind uced desensitiza tion. Science, 259: 825-829.
Dizhoor, A.M., S. Ray, S. Kuma. , G. Niemi, M. Spencer, D. Brolley, K.A. Walsh, P.P. Philpov, J.B. Huriey, and L. Stryer. 1991. Recoverin: a calcium sensitive activator of retinal rod guanylate cyclase. Science, 251: 915-918.
Eisenstein, E.M., D.G. Brunder, and H.J. Blair. 1991. Initial Habituation and Sensitization of the GSR depends on magnitude of first response. Physiol Behnv, 49: 211-215.
Eisenstein, E.M. 1997. Selecthg a mode1 system for neurobiological studies of leaming and memory. Behav Brain Res, 82: 121-132.
Gannon, T.N. and C.H. Rankin. 1995. Methods of studying behavioral plasticity in Caenorhabditis elegans. In H. F. Epstein and D. C. Shakes (Eds.), Cnenorhabditis elenans: Modem Bio loaical Anal vsis of an Ornanim. @p. 205-223). San Diego: Academic Press.
Groves, P.M. and R.F. Thompson. 1970. Habituation: A dual process theory. h j c h Rmiew, 77: 419-450.
Hall, D.H. and L.R. Rwel . 1991. The posterior nervous system of the nematode Caenorhabditis ekgans: serial reconstruction of identified neurow and complete pattern of synaptic interactions. \ Neurosci, 11: 1-22.
Hammer, M. 1997. The neural bais of associative reward learning in honeybees. Trends in Neurosci, 20: 245-252.
Harris, J.D. 1943. Habituatory respowe decrement in the intact organism. Psydiological Bulletin, 40(6): 385-422.
Hedgecock, E.M. and R.L. Rwell. 1975. Normal and mutant thennotaxis in the nematode, Gienorhahditis elegans. Proc Natl Acad Sci, 72: 40611065.
Hellerstern, F., R. Malaka, and M. Hammer. 1998. Backward inhibitory learning in honeybees: A behavioral analysis of reinforcement processing. Leaming and Memoy, 4: 429444.
Kramer, R.H. and S.A. Siegelbaum. 1992. Intracellular Ca2+ regulates the sensitivity of cyclic nucleotide gated channels in olfactory receptor neurow. Neuron, 9: 897-906.
Kurahashi, T. and A. Menini. 1997. Mechanism of odorant adaptation in the olfactory receptor cell. Nature, 385: 725-729.
Kawamwa, S. and M. Murakami. 1991. Wum-dependent regulation of cyclic GMP phosphodiesterase by a protein from frog retinal rods. Nature, 349: 420-423.
Leinders-Zufall, T., M. Ma, and F. Zufd. (1999). Impaired odor adaptation in olfactory receptor neurow after inhibition of Ca2+/Calmoduh Kinase II. 1. Neurosci, 19:RC19,1-6.
Morrison, G.E., J.Y.M Wen, S. Runciman, and D. van der Kooy. 1999. Olfactory associative learning in Caenorhabditis elegans is impaired in lm-1 and lrn-2 mutants. Behav Neuroaci, 113(2): 358-367.
Nuttley W.M., Harbinder S., and van der Kooy, D. 1999. Distinct attractive and aversive mechanisms contribute to benzaldehyde responsive chemotaxic behaviour in Caenor habditis elegams. 1. Neu rosa', submitted.
Ochoa, E., L. Li, and M. McNamee. 1990. Desensitization of central cholinergie mechanisms and neuroadaptation to nicotine. Mol Neurobiol, 4: 251-287.
Quinn, W.G., W.A. Harris, and S. Benzer. 1974. Conditioned behavior in Drosophiln melanogaster. Proc Natl Acad Sci, 7l: 708-712.
R a n h , C.H., C.D.O. Beck, and C.M. Chiba. 1990. Cuenorhabditis elegans: a new mode1 system for the study of leaming and memory. Behav Brain Res, 37: 89-92.
Rauschecker, K. and M. Korte. 1993. Auditory Compensation for early blindness in cat cerebral cortex. [ Neuroscz, 13 453M548.
Rogers, R.F. and L.D. Matzel. (1995). Higher-order associative processing in Hennissenda suggests multiple sites of neuronal modulation. Learning and Memo y, 2(6): 279-298.
Roayaie, K., J.G. Cnunp, A. Sagasti, and C.I. Bargmann. 1998. The Ga Protein ODR-3 mediates olfactory and nociceptive h c t i o n and controls SLim morphogenesis in C. elegam olfactory neurons. Neuron, 20: 5567.
55 Sengupta, P., J.H. Chou, and C.I. Bargmann. 1996. odr-10 encodes a seven
transmembrane domain olfactory receptor required for responses to the odorant diacetyl. Cell, &Q: 899-909.
Sulston, J. and J. Hodgkin. (1988). Methods. In W. B. Woods (Ed.), The nematode Crrrnorhabditis ele9an.s @p. 587-606). Cold Spring Harbor, NY.: Cold Spring Harbor Laboratory Press.
Sulston, J., 2. DU, K. Thomas, R. Wilson, L. Hilier, et al. 1992. The C. elegans genorne sequencing project: a beginning. Nature, 356: 3741.
Thompson, R.F. and W.A. Spencer. 1966. Habituation: A mode1 phenornenon for the study of neuronal substrates of behavior. Psych Rw, 173: 1643.
Troemel, E.R., B.E. Kimmel, and C.I. Bargmann. 1997. Reprogramming chernotaxis responses: sensory neurons define oifactory preferences in C. elegans. Cell,91: 161- 169.
Tully, T. and W.G. Quinn. 1985. Classical conditioning and retention in normal and mutant Drosophiln rnelanogster. J Comp Physiol, 157: 263-277.
Walter, E.T., T. J. Carew, and E.R. Kandel. 1979. Classical Conditionhg in Aplysia Californicn. Proc Na tl A cad Sci, 76: 6675-6679.
Walter, E.T., T.J. Carew, and E.R. Kandel. 1981. Associative learning in Aplysia: Evidence for conditioned fear in invertebrates. Science, Zïl: 504-506.
Wen, J., N. Kumar, G.E. Morrison, S. Runciman, G. Rambaldini, J. Rousseau, and D. van der Kooy. 1997. Mutations that prevent associative learning in C. elegans. Behav Neurosci, 111(2) 342-353.
White, J.G., E. Çouthgate, J.N. Thompson, and S. Brenner. 1986. The structure of the nervous system in the nematode îaenorhabditis elegans. Phil. Tram. R. Soc. Lund. B. Biol. Sci., 314: 1-34.
Wicks, S.R. and C.H. Rankin, C. H. 1996. The integration of antagoristic reflexes revealed by laser ablation of identified neurons determines habituation kinetics of the Gienorhabditis elegans tap withdrawal response. ] Comp Physîol, A176 675- 685.
Wicks, S R . and C.H. Rankin. 1997. Effects of tap withdrawal respowe habituation on other withdrawal behaviors: the localization of habituation in the nematode Caenorha bditis elegans. Behav Ne urosci, ll(2): 354-368.
Zhang, Y.. J.H. Chou, J. Bradley, C.1. Bargmann, and K. Zinn. 1997. The Caenorhabditis elegans seven-hansmembrane protein ODR-IO functions as an odorant receptor in mammmalian cells. Pmc Nat1 Acad Sci, 94: 1216242167.