BioSystems 57 2000 95 – 107
Canalization as a non-genetic source of adaptiveness during morphogenesis: experimental evidence from analysis of
reproductive development in Sorghum bicolor
G. Nissim Amzallag
1 Department of Plant Sciences, The Hebrew Uni6ersity of Jerusalem, Jerusalem
91904
, Israel Received 6 June 2000; accepted 8 June 2000
Abstract
In Sorghum bicolor, perturbations in reproductive development observed following salt-treatment also influence progeny grown in the absence of NaCl. However, a developmental reversion of these modifications may be observed
throughout two successive generations. This response, termed canalization, does not spontaneously occur following growth in the absence of NaCl, but is triggered by the level of perturbation in parental expression of reproductive
characters. Moreover, canalization is not specific to the perturbed character, but it includes modifications in reproductive development as a whole. A decrease in developmental variability coincides with amplitude of the
developmental reversion. This phenomenon is interpreted as an evidence for orientation of the developmental process towards the lowest free-energy state of the ‘epigenetic landscape’. Involvement of this phenomenon of canalization in
developmental stability, adaptiveness, and evolution is discussed. Moreover, these results point to the need for a posteriori methods of investigations in order to analyze self-organized transformations in biological systems. © 2000
Elsevier Science Ireland Ltd. All rights reserved.
Keywords
:
Adaptiveness; Canalization; Developmental plasticity; Evolvability; Non-genetic information; Self-organization www.elsevier.comlocatebiosystems
ally understood as the expression of a develop- mental homeostasis, a capacity of the developing
organism to reach a final, defined state. It has been be related to a series of cybernetic regula-
tions in metabolism and protein synthesis, directly or indirectly controlled by the pre-existing genetic
information. However, this ‘teleonomic property’ of development seems more complex than as-
sumed for at least three reasons.
The first is that, as emphasized by Waddington 1957, p. 121, ‘homozygotes seem usually to be
1. Introduction
The self-buffering capacity of development has been recognized for a long time. It was termed
equifinality von Bertalanffy, 1950 , correcti6e
pleiotropy Warburton, 1955, or canalization Waddington, 1957. This phenomenon is gener-
1
Present address: The Judea Center for Research and De- velopment, Carmel 90404, Israel. Fax: + 972-2-9960061.
0303-264700 - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 3 - 2 6 4 7 0 0 0 0 0 9 3 - 9
more variable in an inconstant environment than the heterozygotes are’, a situation incompatible
with a complete genetic control of canalization. The second is that characters are generally nor-
mally-distributed in wild populations in spite of a strong genetic and environmental variability Bar-
ton and Turelli, 1989. The third is that genetic information is not completely controlling develop-
ment. There is rather a temporal alternance be- tween self-organized and direct genetic control,
while self-organization is largely involved in the achievement of major developmental and physio-
logical events Sachs, 1994; Amzallag, 2000a,b.
Beyond these
considerations, canalization
strongly differs from developmental homeostasis in the capacity of development to tend towards
the optimal end-result according to the internal or environmental constraint Waddington, 1957, pp.
43 – 44. This adaptive aspect of canalization was clearly observed in plants Bradshaw, 1965;
Moran et al., 1981; Sultan, 1992 and was termed adapti6e determinism by Seligmann and Amzallag
1995.
Existence of a ‘developmental buffer’ seems very important for integration of complex physio-
logical regulations, and evolution would require extremely rare combinations of simultaneous mu-
tations in its absence Warburton, 1955. In spite of the evidence, canalization was ‘unfairly under-
played in evolutionary discussions’ Alberch, 1980. It was also excluded from developmental
and physiological considerations. This obscure sit- uation does not only result from a deterministic
approach of development centered on expression of a pre-existing information, but we must also
recognize that the phenomenon of canalization confounds investigation. The high velocity and
efficiency of this process prevents any opportunity to identify transitory phases towards the ‘develop-
mental resolution’. Fortunately, this process be- comes visible when it takes place over an extended
period. For example, Oono 1985 described a dwarf mutation in some individuals regenerated
from cultured cells of rice. Although this trait is due to an homozygous mutation, a ‘chimeric re-
version’ towards normal stem height was observed in progeny of regenerated plants, and it remained
stable for at least three successive generations. This change was obviously directed, because a
normal-to-dwarf transition has never been ob- served after plant regeneration Oono, 1985. The
requirement for two sexual generations suggests the occurrence of intermediate states in the devel-
opmental reversion, which may be analyzed.
A similar opportunity exists in Sorghum bicolor. In this species, a 3-week exposure to a moderate
concentration of NaCl induces an ability to grow and set seeds at an otherwise lethal salinity. This
response was defined as salt-adaptation Amzallag et al., 1993. All the individuals of the salt-treated
population were able to grow at the NaCl concen- tration lethal for non-treated plants, but induction
of salt-adaptation was accompanied by a consid- erable increase in phenotypic diversity Amzallag
et al., 1995; Amzallag, 1999a. Salt-adapted plants displayed many perturbations in reproductive de-
velopment Amzallag, 1996, 1998, suggesting that canalization was disrupted by expression of salt-
adaptation. The reproductive development was also disturbed in progeny of salt-adapted plants
Amzallag, 1996, even for individuals grown in absence of NaCl Amzallag et al., 1998. More-
over, the evidence for a ‘developmental reversion’ in reproductive development of progeny of salt-
adapted plants was already suggested by previous observations Amzallag and Seligmann, 1998.
However, the latter analysis was complicated by re-exposure of the successive generations to a
salt-adaptation treatment. The purpose of this study is to analyze the phenomenon of ‘develop-
mental reversion’ in offspring of one or two suc- cessive generations of salt-adapted Sorghum that
are grown in absence of NaCl.
2. Materials and methods
