Magazine

R877

Stable inheritance would have an improved attraction

to olfactory cues if its parent was
Oliver Hobert and I [2] have
previously reported that olfactory
of an acquired imprinted. Intriguingly, this appears
to be the case: as shown in
imprinting involves at least
two classes of neurons, the
behavior in Figure 1A, a parental imprint was chemosensory neurons AWC and

Caenorhabditis inherited by F1 worms, but not

transmitted to F2 worms.
the interneurons AIY. Imprinting
inheritance suggests that neuronal
elegans
A
Jean-Jacques Remy Imprinted
4 Non-imprinted F1

Migration index (mean +/- s.e.)

Sensory imprinting produces life-long Non-imprinted F2
3.5
attachment to environmental features 3
experienced during a critical period
2.5
of early development. Imprinting
of this kind is highly conserved in 2
evolution and is an important form 1.5
of adaptive behavioral plasticity 1
[1]. The nematode Caenorhabditis 0.5
elegans undergoes such adaptation 0
to new environments through BA 1/900 BA 1/500 BA 1/300
imprinting: attractive odorants, odor
when present during the first larval
stage, produce life-long olfactory
imprints that enhance attraction
and egg-laying rates in the adults B
Number of generations grown in the presence of odours
[2]. Here I report evidence that the
Naives
olfactory imprint can be transmitted 1 2 3 4 5 6 7 8 9
N2 worms
to the next generation. If the imprint Modified sensory
is generated successively over environment
more than four generations, it is
not just transmitted through one
further generation, but rather, it
is stably inherited through many
First generation grown in Original odour-free
following generations. While the
the absence of odours environment
transient nature of the inheritance
suggests the existence of resetting
mechanisms, stable trans-
generational inheritance of the kind Second generation
reported here raises the possibility grown in the absence
that a behavioral alteration produced of odours
by an environmental change might
be genetically assimilated after a
limited number of generations.
All following generations
Behavioral plasticity is the ability
grown in the absence
of an animal to modify its behavior
of odours
so as to respond to an environmental Current Biology
change. C. elegans adapts to novel
olfactory environments by imprinting: Figure 1. Inheritance of an acquired behavior in C. elegans.
worms exposed to odorants during (A) Olfactory imprints are passed to the F1, but not to the F2 generation. N2 worms were im-
the critical first larval stage of printed independently with three dilutions of benzaldehyde: BA 1/900, BA 1/500, or BA 1/300
development (L1) keep life-long (Imprinted). F1 and F2 from imprinted parents were grown into the regular odor-free environ-
imprints that shape their behavior ment (Non-imprinted). The chemotaxis behavior of F1 and F2 were compared to that of imprint-
ed and naïve animals. Mean migration indices of non-imprinted F1 were significantly different
when they encounter the olfactory
from naïve (**, p < 0.01, n = 10), as determined by unpaired two tailed Student t tests using the
cues again as adults. In C. elegans, Kaleidagraph statistical analysis software. Chemotaxis of non-imprinted F2 was not different
olfactory imprinting increases from naïve animals (n = 10). (B) Transient or stable evolution of the C. elegans chemotaxis be-
attraction to imprinted odorants havior after iterated olfactory imprinting over generations. N2 wild-type worms were imprinted
and enhances odorant-induced independently with the five odors described in Table 1. One part of each progeny was grown in
egg-laying rates [2]. Such long-term the presence of the same odor (i.e. imprinted), another part was grown in the absence of any
odor for at least 10 generations. Population chemotaxis assays were performed at the adult
memory of favorable conditions
stage for every generation. Whatever their own history or history of their progenitors, worms
might be useful to subsequent display only one of the two levels of responses: while all ‘black’ generations behave as naïve
generations, so we therefore tested animals, all ‘red’ generations behave as imprinted. The same pattern of inheritance was found
whether a non-imprinted animal for the five odorants, imprinted independently (Table S1 in the Supplemental information).
Current Biology Vol 20 No 20
R878
information is translated into a
meiotically stable germ-line form.
An environmental change that
leaves a lasting impact on a worm
may last longer than a single
generation. We therefore investigated
how imprinting the same olfactory
cues generation after generation
could influence adaptation of worm
populations to new chemosensory
environments. We cultured up to
nine generations of worms in five
different sensory environments by
adding five different dilutions of
attractive odorants (benzaldehyde
1/900, 1/500 or 1/300, or citronellol
1/500 or 1/300) to the environment
during the critical early period of
worm development. Part of the
progeny from every imprinted
generation was imprinted again with
the same odorant while the other
part was grown back into the original
regular odor-free environment
(Figure 1B).
The behavior of each Figure 1B
generation was compared to that
of naïve worms. Every generation
of worms has a different ‘history’,
being itself imprinted or descending
from imprinted or non-imprinted
ancestors. We found that,
independently of their history, worms
display only one of two levels of
responses: either a response of naïve
(black), or a significantly enhanced
response (red). Even after nine
successive generations of imprinting,
chemotaxis did not surpass the
migration index of the first imprinted
generation, and no intermediate
levels between naïve and imprinted
were observed (see Supplemental
Table S1 in the Supplemental
Information).
Olfactory imprinting is inherited
in the absence of the triggering
odors. It can be passed either only
to the F1 generation or to the F1,
F2 and all subsequent generations.
Imprinting inheritance after F1 only
depends on the number of imprinted
ancestral generations. For the five
odor dilutions tested independently,
we observed stable inheritance of the
imprinting-modified behavior when
at least five ancestral generations
had been consecutively imprinted,
suggesting that a switch between
reversible and stable imprinting
occurred in all cases after four
consecutive generations have been
imprinted (Figure 1B and Table S1).
As shown in Supplemental Figure
S1 (see Supplemental Information),
stable imprinting was odor-specific
and maintained over the course of
at least 40 generations. One cannot,
however, exclude the possibility that
resetting of the imprinted behavior
will occur eventually.
Once parental imprints have been
erased, worms recover their initial
plasticity and can be imprinted again
with the same cue, but when they
have become innately expressed,
worms have lost plasticity of
response to the imprinted cues. In
that sense, it is a form of behavior
canalization, yet different from the
classically described phenotypic
canalization [3], in which different
genotypes produce the same
phenotype. Importantly, imprinting
inheritance in C. elegans seems not
to be based on selection of individual
variations, as all members of a given
generation behave only according
to its own or to its ancestor’s
experience (not shown).
It has been shown that learned
behaviors can be transmitted in
many animal species by cultural
means; through natural selection,
newly acquired behaviors that give
selective advantages eventually
come to be expressed by all
members of a population. Because
the mechanism of inheritance of
the olfactory imprinting behavior is
not based on cultural transmission
nor on selection of individual
genetic variations, it might involve
epigenetic modifications: imprinting
induces the same unique behavioral
switch in all individuals, which is
transiently or stably inherited by all
the progeny in the absence of the
initial conditions that caused the
switch.
The process of genetic assimilation
usually involves fixation of an
environmentally induced novel
phenotype which is made possible
by plasticity. Epigenetic mechanisms
may greatly reduce the number
of generations required to fix an
adaptive phenotypic change within a
population, and thus have significant
evolutionary consequences [4,5].
A lot of controversy still remains
regarding the concepts of innateness
and heritability, and the mechanisms
by which genetic assimilation
of external cues may influence
evolution [6–9]. More specifically, the
contribution of behavior plasticity to
animal speciation has been highly
discussed [10]. Our findings of
trans-generational inheritance of
olfactory imprinting in C. elegans
provide a simple experimental
paradigm for studying the evolution
of adaptive behavior by assimilation
of an external change. Further
research on this system should
help to uncover the molecular
mechanisms by which a reiterated
sensory experience over a fixed
limited number of generations can
be assimilated and stably alter
the innate behavior of an animal
population.
Supplemental Information
Supplemental Information includes one
figure and one table and can be found with
this article online at
doi:10.1016/j.cub.2010.08.013.
Acknowledgements
Thanks to INRA-Phase for funding, M.N. for
support, and O. Hobert for comments and
discussion on the manuscript. The author
has no conflicts of interest related to this
work.
References
1. Lorenz, K. (1970). Studies in Animal and
Human Behavior (Cambridge, USA: Harvard
University Press).
2. Remy, J.J., and Hobert, O. (2005). An
interneuronal chemoreceptor required for
olfactory imprinting in C. elegans. Science
309, 787-790.
3. Waddington, C.H. (1953). Genetic assimilation
of an acquired character. Evolution 7,
118-126.
4. Turner, B.M. (2009). Epigenetic responses to
environmental change and their evolutionary
implications. Phil. Trans. R. Soc. B 364,
3403-3418.
5. Pigliucci, M., and Murren, C.J. (2003). Genetic
assimilation and a possible evolutionary
paradox: can macroevolution sometimes be
so fast as to pass us by? Evolution 57,
1455-1464.
6. Mameli, M., and Bateson, P. P. G. (2006).
Innateness and the sciences. Biol. Philosoph.
22, 155–188.
7. Griffiths, P.E. (2009). The Distinction Between
Innate and Acquired Characteristics. The
Stanford Encyclopedia of Philosophy, (E.N.
Zalta editor).
8. Price, T.D., Qvarnstrom, A., and Irwin, D.E.
(2003). The role of phenotypic plasticity in
driving genetic evolution. Proc. R. Soc. Lond.
270, 1433-1440.
9. Pigliucci, M., Murren C.J., and Schlichting,
C.D. (2006). Phenotypic plasticity and
evolution by genetic assimilation. J. Exp. B.
209, 2362-2367.
10. Avital, E., and Jablonka, E. (2000). Animal
Traditions: Behavioral Inheritance in Evolution
(Cambridge, UK: Cambridge University
Press).
Neurobiology of Cellular Interactions and
Neurophysiopathology, UMR 6184 CNRS,
Mediterranée University, USC INRA-Phase,
Marseilles, France.
E-mail: jean-jacques.remy@univmed.fr