Download The use of mutants to probe models of gravitropism

Journal of Experimental Botany, Vol. 51, No. 349, pp. 1323–1340, August 2000
REVIEW ARTICLE
The use of mutants to probe models of gravitropism
Richard D. Firn1, Carol Wagstaff and John Digby
Department of Biology, University of York, York YO10 5YW, UK
Received 10 January 2000; Accepted 3 May 2000
Abstract
It has been widely believed for more than 70 years
that auxin plays a central role in the induction of differential growth which causes gravitropic curvature.
However, this long-standing consensus about a role
for auxin in gravitropism has only been achieved by
allowing several mutually exclusive models to coexist.
Furthermore, because there is no detailed model
which is unchallenged by evidence, consensus is now
centred on ill-defined models which have a low predictive value, hence are harder to challenge experimentally. An increasing number of mutants with
abnormal gravitropic behaviour are becoming available. Such mutants should be very helpful in challenging existing models of gravitropism and in providing
new evidence on which to build improved, more precise models. However, to date, most studies of mutants
with abnormal gravitropism have been guided, experimentally and conceptually, by the old inadequate and
vague models. Consequently, the full potential of
modern molecular analysis in aiding our understanding
of gravitropism has yet to be realized.
Key words: Gravitropism, mutants, Cholodny, Went, auxin.
Introduction
It is 70 years since Cholodny and Went, working independently, proposed that plant tropisms were caused by
the lateral redistribution of a plant growth substance.
This work subsequently led to the isolation from plants
of auxin ( later identified as indolyl-3-acetic acid (IAA)),
the discovery of which motivated a generation of plant
physiologists to study the regulation of plant growth
and development by endogenous chemicals. The
Cholodny–Went model (as defined by Went and
Thimann, 1937), gained a wide acceptance and was to
dominate the study of plant tropisms for the rest of the
century. The Cholodny–Went model, which continues to
have both detractors and admirers, has evolved to produce many variants which coexist with the original. So
what is the current thinking about the role of auxin
in gravitropism? Which ideas associated with the
Cholodny–Went model still have an experimental or
conceptual validity? Do the increasing number of mutants
with abnormal gravitropic behaviour provide useful data
to inform us about the role of auxin in gravitropism or
to challenge the various old models? How effectively has
the availability of the new mutants been used as a driving
force to produce improved, detailed models?
What is ‘the Cholodny–Went model’?
Although it is usual to refer to ‘the Cholodny–Went
model’, there is no single coherent model. Instead, there
is a collection of several quite distinct models which share
one common element—they are based on the proposition
that the plant growth regulator auxin is involved in some
way in gravitropism (Firn and Myers, 1987; Firn, 1992).
The variants of the original model are sometimes more
specific and sometimes much less specific than the original.
Each version of the Cholodny–Went model produces its
own predictions hence there can be no valid experimental
test of ‘the model’, rather there can only be a test of one
form of the model. Consequently, it is important that
those working on the possible role of auxin in plant
tropisms understand the sometimes subtle, but often
fundamental, differences between the various versions of
the model and have some understanding of why the
different variants of the original model exist.
The ‘original’ Cholodny–Went model
The confusion as to the precise nature of the
Cholodny–Went model began at its birth. Cholodny and
Went were not collaborators, they apparently never met
and they never jointly defined the model associated with
1 To whom correspondence should be addressed. Fax: +44 1904 432860. E-mail: [email protected]
© Oxford University Press 2000
1324 Firn et al.
their names. They worked independently and each followed a path leading from the work conducted previously
by many scientists including Darwin, Rothert, Fitting,
Boysen-Jensen, Paál, Stark, and Söding (see Avery and
Burkholder, 1936; Went and Thimann, 1937; Hart, 1990;
for fuller historical accounts). The majority of those
studying plant tropisms in the 40 years before the discovery of auxin were guided by the conviction that there was
a spatial separation of the regions of perception and
response in organs that showed a tropistic response. It
was this belief in a spatial separation of the perception
and response zones that was central to the development
of the idea that a chemical messenger must be involved
in controlling shoot and root extension. Cholodny, and
subsequently Went, were both developing existing ideas
that the organ apex in some way controlled the growth
of the cells in the elongation zone. Cholodny’s contribution, made largely between 1918 and 1927, is best known
to the English speaking world through Went and
Thimann’s very influential book, Phytohormones. Quoting
from translations of Cholodny’s work, Went and
Thimann ( Went and Thimann, 1937, 155–156), noted
that Cholodny proposed that:
(1) ‘growth hormones play an essential role in the mechanism of the geotropic reaction’
(2) ‘in vertically placed stems and roots the growthregulating substances are equally distributed on all
sides’
(3) ‘as soon as these organs are placed in a horizontal
position, the normal diffusion of the growth hormones is disturbed; the upper and lower cortical cells
now obtain different amounts of these substances’
(4) the opposite signs of the reactions of roots and shoots
fit in with the fact that they react in opposite ways
to the growth hormones coming from the tips
(5) during phototropism ‘the cells of the coleoptile first
become polarized under the influence of the unilateral
illumination, and this causes the continuously produced growth hormones to diffuse from the light
towards the dark side more rapidly than in any other
direction’
Went, starting his research on phototropism after
Cholodny had been active in the area for some years,
developed the first practical bioassay in the search for the
substance that many others (including Cholodny) had
argued must flow from the tip of an organ to control
elongation. Using this Avena curvature bioassay Went
was able to quantify the amount of diffusible growth
promoting substance coming from isolated tips and he
produced experimental evidence which he interpreted
( Went and Thimann, 1937, 156) as showing that:
(1) during phototropic stimulation, auxin is moved laterally in the apex of a coleoptile, consequently the
supply of auxin to the elongating cells below is
changed such that cells on the shaded side grow faster
than cells on the illuminated side
(2) ‘geotropic perception is caused by a polar alteration
in the coleoptile cells . . . instead of moving rectilinearly the growth regulators are more strongly conveyed towards that side which under geotropic
stimulation was turned downwards.’
The combining of the ideas of the two workers to give
the now famous Cholodny–Went model was undertaken
not by Cholodny and Went but by Went and Thimann
who produced the most widely quoted version of the
model: ‘Growth curvatures, whether induced by internal
or by external factors, are due to an unequal distribution
of auxin between the two sides of a curving organ. In the
tropisms induced by light and gravity the unequal auxin
distribution is brought about by a transverse polarization
of the cells, which results in lateral transport of the auxin’
( Went and Thimann, 1937).
Remarkably, what is omitted from the Went and
Thimann definition of the model is the central role that
the organ apex was supposed to play. Both Cholodny
and Went, like most of the workers at that time, ascribed
a central function to the role of the organ tip. ( The word
‘tip’ is commonly used but it must always be recognized
that the apical regions of roots, coleoptiles and hypocotyls
are anatomically and functionally very different.) Since
the very early work (Ciesielski, 1872; Darwin, 1880), the
tip had been considered by many to be the site of stimulus
perception (especially in the case of gravity perception in
roots and phototropic perception in coleoptiles).
Following these 19th century workers, Rothert, Fitting,
Boysen-Jensen, Paál, Stark, and Söding among others,
considered that processes in the tip were central to
understanding the control of organ growth. When
Cholodny, and subsequently Went, carried out their
experiments, their studies were based on the assumption
that the tip played a pivotal role. For example, Cholodny
carried out many gravitropism experiments where the
gravitropic responses of organs were recorded after the
tips were excised and replaced. Likewise, Went’s classical
oat coleoptile curvature assay was based on the assumption that the tip supplied auxin to the coleoptile and
Went used this bioassay to measure the auxin flowing
from the tip after phototropic stimulation. It was Dolk
who later used the same bioassay to measure auxin
flowing from gravistimulated coleoptile tips (Dolk, 1936).
Thus, the failure of Went and Thimann to make any
reference to the tip in their version of the model left an
extraordinary central ambiguity at the heart of the definition that is most widely quoted. It is unclear whether
Went and Thimann’s failure to include a role for the tip
in their definition was due to the fact that the role of the
tip was taken as proven and self-evident (nearly all the
Models of gravitropism
experiments conducted by Cholodny and especially Went
only made sense if the tip played a unique role) or
whether it was because Went and Thimann had come to
accept that a large body of evidence already existed that
both phototropic perception and gravitropic perception
could occur in regions other than the tip in some organs.
The subsequently expressed view of Went, referring to
the Cholodny–Went model, supports the former explanation ( Went, 1956): ‘This theory is based on the fact that
the lower zones of the Avena grow under the influence of
the growth-promoting substance auxin, which is formed
in the coleoptile tip.’
Although Leopold correctly states that the tip is ‘peripheral’ to the Cholodny–Went theory as defined by Went
and Thimann, it cannot be argued that it was anything
other than central to the original studies of Cholodny or
Went (Leopold, 1992). The most modern enunciation of
the Cholodny–Went theory by Hart explicitly identifies
the tip as being the site of both auxin synthesis and of
tropistic perception and gives the tip a central role in the
model (Hart, 1990): ‘The growth of the plant organ is
dependent on auxin, which is synthesized in the organ
tip. A directional stimulus brings about an asymmetric
redistribution of auxin in the tip of the organ. The
asymmetry in auxin distribution is transported longitudinally along the organ to produce a response in another
region. The tropic response represents a growth response
to altered auxin levels, involving growth stimulation and
growth inhibition on opposite sides of the responding
organ. The opposite responses of shoots and roots to
gravity are a result of their different sensitivities to auxin.’
Furthermore, if the work of the predominant groups
in the 1950s, 1960s and early 1970s (see Hart, 1990, for
references to work by the groups of Thimann, Briggs,
Wilkins, and Pilet) is examined, it is clear from their
experimental rationale that they all subscribed to the idea
that the tip played a central role.
Thus the most widely quoted version of the
Cholodny–Went model ( Went and Thimann, 1937) can
be seen as the first stage in a process of weakening the
precision of the models that Cholodny and Went independently developed. This reinterpretation and generalizing of a key idea was a process that was set to continue.
As data began to appear which was inconsistent with the
some of the predictions of the original concepts of
Cholodny and Went (for instance the growing acceptance
that the tip played no role of gravity sensing in shoots),
the model often avoided rejection by becoming less precise. Although these variants of the original model have
been criticized in some detail previously ( Firn and Myers,
1987), a new hierarchical analysis of the many variants
of the models inspired by Cholodny and Went’s work
provides a more structured way of presenting the framework necessary to evaluate the current state of the models
1325
and to assess the new information about the models that
is becoming available from mutants.
A hierarchy of Cholodny–Went models
In Fig. 1, some alternative models which attempt to
explain the gravitropic behaviour of organs in terms of
some aspect of auxin physiology have been laid out in a
hierarchical way. This arrangement of the variants of the
Cholodny–Went model emphasizes some important
points:
$
$
$
$
a vague conceptual model can encompass many more
specific models;
that the historical development of the model does not
match the logical hierarchy;
that the broad conceptual models need not distinguish
between models that are specific to roots ( letter R
alongside the model number) and to shoots (S);
that models at the bottom of the hierarchy can be
challenged more easily than models at the top because
precise models make precise predictions (R or S
denotes models for roots and shoots, respectively, that
have been seriously challenged by experimental
evidence)
Broad conceptual models—a good starting point but a bad
end point?
Research normally progresses by devising and testing
increasingly precise models, each model being rejected
when data are obtained which cannot be accounted for
by the model. Consequently, a productive research area
normally shows a clear historical progression from broad
conceptual models to more precise specific models and
then to detailed mechanistic models. This progression was
indeed shown in the first three decades of the century
with regards to the control of organ elongation by growth
substances, especially in the study of phototropism. As
early as 1910, Boysen Jensen, following the broad conceptual ideas of Darwin, was attempting to explain phototropism in terms of an unknown growth-regulating
material emanating from the apex (a phototropism model
equivalent to model 2). In the following decade Paál,
Stark, and Seubert developed experimental approaches
to probe the role of the tip, using agar blocks and tip
replacements on decapitated coleoptiles. While BoysenJensen argued that phototropism was caused by a stimulation of elongation at the shaded side of the stimulated
organ, Paál presented evidence that a growth inhibition
of the illuminated flank was most important (which is
now known to be true). Went built on some of these
ideas and by devising a practical bioassay he was able to
demonstrate that phototropic stimulation of a coleoptile
tip did change the amounts of growth-regulating activity
diffusing from an isolated tip (note that Hasegawa et al.,
1989 have since shown that the changes in growth-
1326 Firn et al.
Fig. 1. The hierarchy of models related to the Cholodny–Went model.
promoting activity reported by Went cannot be ascribed
to changes in IAA concentration). Thus during this
period, Darwin’s broad conceptual framework was
developed, by logic and by experiment, to explain coleoptile phototropism in terms of the supply of a specific
growth-promoting substance, culminating in Went’s
experimental evidence in favour of the concepts being
developed by several people at that time. Meanwhile,
others such as Cholodny were extending these ideas,
producing a broader, unifying model that brought
together phototropism and gravitropism and which
attempted to explain both root and shoot gravitropism.
However, when the ideas of Cholodny and of Went were
combined by Went and Thimann into a simplified, unifying model ( Went and Thimann, 1937), the historical
progression of model development began to break down.
By making no explicit reference to the role of the tip in
their definition (model 2.2.2), Went and Thimann started
a process of making the model less precise. In the decades
that followed, some workers further extended the boundaries of what might be included in the Cholodny–Went
model, to the extent that ideas that were quite contrary
to those of Cholodny and Went began to be incorporated
into vague models that bore their names. The extent of
this retreat to broader models is clear in some recent
redefinitions of the model. For example, a ‘contemporary
generalization’ of the Cholodny–Went model by Pickard
(Pickard, 1985a, b), ascribes differential growth to a
change in content of one or more of auxin, gibberellin,
protons, and calcium and hence is akin to a version of
model 1. Such a model can incorporate all the models
lower in the hierarchy simply by being imprecise.
Likewise, Evans produced a definition ( Evans, 1992):
‘The Cholodny Went Hypothesis can be interpreted more
broadly to mean that differential growth results from a
gradient in auxin activity which may arise from gradients
in auxin concentration, auxin co-factors/inhibitors, auxin
sensitivity or some combination of these.’ which is also
equivalent to model 1. The addition of a second unknown
growth-controlling factor to the Cholodny–Went model
by Hertel (Hertel, 1992) results in a definition of the a
model which is also close to model 1. The scale of this
retreat to a vague model is clear from the fact that of 15
authors offering their views on the Cholodny–Went model
in 1992 ( Trewavas, 1992), seven subscribed to models
akin to model 1 (where auxin is simply a part of the
control mechanism together with some other unidentified
regulator) and none of the 15 participants defended the
‘original’ precise models (models 2.2.1 or 2.2.2). In other
words, only a very weak consensus could be reached and
Models of gravitropism
that was only achieved by weakening the definition of the
Cholodny–Went model to the extent that it includes many
features that were not part of Cholodny or of Went’s
thinking and encompasses concepts that should rightly
be ascribed to others. The Cholodny–Went model had in
effect been reduced to a belief that auxin was in some
way involved in plant tropisms.
Specific conceptual models—a necessary step en route to
mechanistic models?
A broad conceptual model can encompass several more
specific models. The specific conceptual models offer more
opportunities for experimental challenge than the broad
conceptual models hence are of greater value.
Sensitivity to a auxin changes as part of the tropistic
mechanism (model 2.1): The failure to find auxin concentration gradients across some tropistically stimulated
organs ( Firn and Digby, 1980; Mertens and Weiler, 1983)
and the increasingly fashionable idea that sensitivity
changes to hormones might be more important than had
previously been thought ( Trewavas, 1982), led to the idea
that sensitivity changes might be incorporated into modified versions of the Cholodny–Went model (Pickard,
1985a, b; Salisbury et al., 1986; Evans, 1991, 1992). Some
earlier work which discussed auxin sensitivity changes
(Brauner, 1966) was largely ignored at the time, maybe
because it appeared during a period when measuring
radioactive IAA movement was fashionable. However,
the concept that a cell may change its sensitivity to a
given regulator as it develops was nearly as old as the
Cholodny–Went model itself but there had been no place
for such thinking in the model previously, possibly
because such concepts undermined many of the basic
ideas which were so fundamental to the original model.
In general, a sensitivity change might be an appropriate
way for a locally perceived event modulating the effect of
a hormone, however, it is a less appropriate way of
explaining processes that involve a spatial separation of
perception and response. For instance, if it is argued that
the sensitivity to auxin is changed in cells in the elongation
zone of roots as part of a gravitropic response, how do
those cells obtain the instructions to carry out their
change in sensitivity if they are incapable of sensing
gravity themselves? If it is postulated that another chemical message is needed to control the sensitivity of a cell
to auxin, a model is produced which has another chemical,
not auxin, as the key regulator (a version of model 1).
In the absence of unambiguous ways of measuring the
elements that can control the sensitivity of cells to a
hormone (Firn, 1986; Weyers et al., 1987), no convincing
data have been published to support the idea that cells
taking part in any tropistic response change their sensitivity to a growth regulator as a necessary part of the
gravitropic response. The current methodology to investi-
1327
gate changes in ‘sensitivity’ largely rely on comparisons
of dose–response curves for cells which might differ in
their sensitivity. However, in the case of tropistic
responses, it has so far proved impossible to use populations of cells for such bioassays that have not already
shown (or are about to show) a tropistically-induced
change in growth rate. For example, if semicylinders of
hypocotyls are grown in various auxin concentrations,
the ‘lower’ halves of the organ show a larger response to
lower concentrations of auxin than the ‘upper’ halves
(Brauner, 1966; Firn and Digby, 1977), but the lower
halves grow more in the absence of any applied auxin
and the experiment cannot distinguish between the possibility that the cells in the lower half are now more
sensitive to endogenous auxin and the possibility that the
cells have started to grow faster for some reason unconnected with auxin.
In summary, model 2.1 seems more popularly incorporated as an optional extra into other models (for example
as part of model 1) than as an independent model. To
date, no specific or mechanistic models have been devised
or tested to explain at a cell level how gravity could
change the sensitivity of cells in the elongation zones of
gravitropically competent organs. This is a weakness
caused by the inherent difficulties of devising ways of
changing the sensitivity of a cell to auxin without admitting that another substance is the true regulator. The
mechanistic underpinning of proposed sensitivity changes
in organs showing complex spatial and temporal changes
in growth rate has never been seriously addressed by
advocates of such models yet such considerations are
necessary for a credible model.
Auxin concentration changes are responsible for differential
growth (model 2.2): The great majority of those studying
the role of auxin in plant tropisms between 1930 and
1980 were guided by the proposition that differential
growth across tropistically stimulated organs is caused by
an underlying auxin concentration gradient (model 2.2).
This model gives two clear predictions. Firstly, the model
predicts the establishment of an auxin gradient, of sufficient magnitude, in a gravistimulated organ at some time
during the lag phase before differential growth is evident.
Secondly, it is predicted that if the gradient of auxin
causing differential growth is not allowed to form or is
dissipated or is swamped, then differential growth should
be abolished.
Can one find appropriate auxin gradients in gravistimulated
organs? Despite decades of effort, there remains a serious,
unresolved conflict of data as to whether an auxin gradient
is always formed, in sufficient magnitude, early enough
to be the cause of differential elongation in all gravistimulated organs ( Firn and Digby, 1980). On the one hand
some very sensitive physicochemical and immunological
assays have failed to detect such auxin gradients in a
1328 Firn et al.
range of gravistimulated organs (Mertens and Weiler,
1983). But on the other hand, more recently several
reports have appeared which report a differential expression of an IAA-responsive promoter–reporter gene construct in gravistimulated organs of transgenic plants (Li
et al., 1991). There are ways of resolving this apparent
conflict and these need experimental study. It could be
argued that some subcompartment of auxin changes at
positions across a gravistimulated organ while the total
solvent extractable does not (equivalent to specific model
2.2.6). Alternatively, it could be argued that IAA-responsive promoter–reporter constructs cannot unambiguously
distinguish between auxin concentration changes and
changes in sensitivity to a constant amount of auxin
(taking us back to model 2.1).
Can one influence gravitropic curvature by influencing auxin
gradients? The treatment of organs with phytotropins,
such as NPA or one of the fluorenols, abolishes auxin
gradient formation and abolishes gravitropic curvature
hence this prediction is verified (Firn, 1968). However,
the prediction is challenged by the results of studies of
gravitropic curvature in organs that are grown in the
presence of high external auxin concentrations. If exogenous auxin is supplied at a concentration that is demonstrably having an effect on cell elongation, it must be
assumed that endogenous auxin pools have been overwhelmed and it is predicted that differential growth should
be impossible under those circumstances. However, many
organs have been shown to be capable of showing gravitropic curvature in the presence of exogenous auxin
supplied at concentrations clearly capable of swamping
endogenous auxin levels. Studies of hypocotyls and coleoptiles (Firn and Digby, 1977) and roots ( Katekar and
Geissler, 1992; Muday and Haworth, 1994; Ishikawa and
Evans, 1993) provide extensive evidence to falsify this
important prediction.
Model 2.2 is therefore supported by some evidence but
is quite seriously challenged by other data. Given that a
refutation of this model would demolish all the detailed
models below it in the hierarchy, it would be useful to
encourage further attempts at disproof or to encourage
more attempts at resolving the conflicting data. Simply
gathering more data of the same type, either in support
or opposition, would seem to be unproductive and what
is needed is a concentration on the key areas of dispute.
For instance, IAA-responsive promoter–reporter gene
construct expression in gravistimulated organs needs to
be quantified, spatially and temporally, with a precision
that matches the precision of the measurements of differential growth. Such measurements will help identify which
events are causal. It would also be interesting to see the
patterns of IAA-responsive promoter–reporter gene construct expression in gravistimulated organs in the presence
of exogenous auxin at a concentration thought to be
swamping any endogenous auxin gradients. If there was
an elevated expression of the reported gene, but no
gradient across the organ of a root which was showing
differential growth, then the generation of auxin gradients
would clearly not be essential for differential growth.
Detailed models
Where more than one model exists at any level in the
hierarchical diagram, those models should obviously not
be seen as likely to coexist in one organ. The expectation
should be that evolution will have provided only one
basic mechanism to control one piece of physiology.
However, one should always ponder whether quite different pieces of physiology might have unwittingly been
drawn together under a common name by physiologists.
Thus it might be expected that different mechanisms
would operate to drive gravitropism, for example, in
single cells and in multicellular organisms. Likewise, the
idea that roots and shoots share a common gravitropic
response mechanism is appealing but by no means certain.
However, a tolerance to the possibility that more than
one type of mechanism is used to control gravitropism in
quite different organs should not encourage tolerance to
the idea that different models of gravitropism might
coexist in very similar organs. If it is argued that more
than one model does exist in one type of organ, an onus
should be placed on the proponents of each model to
state explicitly under which circumstances their favoured
model does and does not apply. The evolutionary arguments as to how the two or more systems evolved and
how they interact also needs to be addressed.
Lateral auxin movement across the tip gives rise to auxin
concentration gradients in the elongation zone (model
2.2.1): It is not the purpose to review each of the broad
mechanistic models in detail. The main purpose is briefly
to outline how the models differ and to consider which
models have the most predictive value for the future.
The role of the tip in gravity perception is still a matter
of some confusion and debate. In roots there is considerable evidence that the root cap is the site of gravity
perception (Ciesielski, 1872), hence there would seem to
be a need to postulate a means of communication between
the cells that sense gravity and those that respond further
back in the elongation zone. However, in coleoptiles,
hypocotyls and epicotyls, gravity can be perceived along
the whole length of the responding region and there is no
such need to postulate any longitudinal transfer of a
message ( Firn and Digby, 1980). It must be borne in
mind that Cholodny, and more particularly Went, largely
derived their ideas for a role for the tip in shoot gravitropism from an analogy with shoot phototropism and root
gravitropism and not from experimental studies of shoot
gravitropism. Hence model 2.2.1 could be retained for
roots, but must be rejected for shoot gravitropism. Even
Models of gravitropism
in roots, the localization of the site of graviperception in
the tip has been questioned (Ishikawa and Evans, 1990)
and there are serious doubts as to whether lateral auxin
redistribution alone can account for the complex temporal
and spatial elongation rate changes that cause root gravitropism (Sievers and Zieschang, 1992; Konings, 1995;
Evans and Ishikawa, 1997). The inability of exogenously
supplied auxin to abolish root gravitropism also challenges this model ( Katekar and Geissler, 1992; Muday
and Haworth, 1994; Ishikawa and Evans, 1993).
Trans-organ lateral auxin movements across the elongation
zone gives rise to the auxin concentration gradients causing
differential growth (model 2.2.2): The fact that model
2.2.1 was inappropriate for shoots was maybe acknowledged in the existence of model 2.2.2. Yet it is quite
extraordinary that this model, the most widely quoted
version of the Cholodny–Went model as applied to shoots,
ever gained wide support. It had been known for decades
before the model was defined that longitudinally bisected
sections of hypocotyls, epicotyls or coleoptiles can show
an adequate gravitropic response. Indeed, Copeland specifically noted that the fact that a gravitropic response
could be seen in bisected hypocotyls demonstrated that
each half of an organ required no information from the
other half (Copeland, 1900). Such experiments showed
that if any lateral movement of auxin were needed to
cause differential growth, the movement had to be localized and not across the organ. The main reason why the
old semicylinder data were disregarded by supporters of
this version of the Cholodny–Went models was finally
shown to be invalid by experiment (Firn and Digby,
1977), sadly not before a large number of studies in the
1960s and 1970s measured radioactive auxin moving
across whole organs.
Auxin concentration in the epidermal layer control differential growth (model 2.2.3): This model for shoots evolved
from model 2.2.2. Classic studies in the last century
showed that the peripheral cell layers of hypocotyls or
coleoptiles played a particular structural role in elongating
organs and it was predicted that these cells would be
important in regulating organ extension (Sachs, 1887).
Iwami and Masuda incorporated such ideas into a version
of the Cholodny–Went model as applied to shoots (Iwami
and Masuda, 1974), suggesting that auxin from the tip
of a gravistimulated organ was directed into the epidermal
layers of the lower flank and kept from the epidermal
layers of the upper flank of the elongating zone. Sadly
the model carried with it two false assumptions derived
from the two previous versions of the Cholodny–Went
model. Firstly, it wrongly assumed the tip of the hypocotyl
was the site of gravity perception. Secondly, it presumed
that transorgan IAA movement was essential.
Auxin concentration in the epidermal and sub-epidermal
layers control differential growth (model 2.2.4):
1329
MacDonald and Hart (MacDonald and Hart, 1987)
advanced a modified version of model 2.2.3 which
addressed the false assumptions of Iwami and Masuda
(Iwami and Masuda, 1974). The revised model proposed
that gravity had a localized effect on the auxin movement
between the epidermis and the adjacent cell layers (see
also Jones, 1992, for a similar view that auxin movement
might be very localized ). Furthermore, it was suggested
that auxin promoted elongation in the epidermis but
inhibited the elongation of the sub-epidermal layer. By
this means, authors addressed the need ( long ignored) to
account for the cessation of elongation at the upper flank
of a gravistimulated hypocotyl as well as the acceleration
of elongation on the lower flank. However, an analysis
of the IAA concentration of epidermal layers before and
after periods of gravistimulation (Mertens and Weiler,
1983) reported that no changes in IAA concentration
were evident. Less direct evidence comes from a study of
IAA-responsive promoter–reporter gene construct expression in gravistimulated tobacco (Li et al., 1991) where it
was stated that the reporter gene was activated in the
epidermal layers of the lower flank and down-regulated
in the same cells of the upper flank (the ambiguity in
these data will be discussed again later).
Localized movement of auxin within an organ from a subpool of auxin (model 2.2.5): Although the shoot tip has
no role in gravity sensing it was thought by many to have
a role as the source of auxin in the basipetal transport
system, auxin which was supposed to control organ
extension. At the heart of the Cholodny–Went model was
the concept that auxin in the basipetal transport flow was
controlled to cause differential elongation. However, in
etiolated hypocotyls ( Tamimi and Firn, 1984) or coleoptiles (Parsons et al., 1986) the auxin supply via the
basipetal transport can be interrupted without reducing
the elongation of cells in the elongation zone over
time scales that are relevant to tropistic responses.
Consequently, Firn and Tamimi (1985) concluded that if
auxin were to be controlling elongation during gravitropism, it must come from a specific pool other than that in
the basipetal transport system. They proposed a model
for tropistic curvature which involved auxin movement
between a supply pool and a growth limiting pool, both
of which were isolated from the basipetal auxin transport
pool. If the movement between these pools involved a
phytotropin-sensitive step then the ability of those compounds to inhibit gravitropism could be explained.
However, until the existence of such pools can be demonstrated, the model remains speculative.
Unequal rates of longitudinal auxin transport alter the rates
of auxin supply to the different sides of the elongation zone
hence cause differential growth (model 2.2.6): This version
of the model had but a brief life and was soon challenged
by numerous studies in the 1960s and 1970s which failed
1330 Firn et al.
to show significant effects of gravitropic stimulation on
the longitudinal basipetal movement of IAA. Surprisingly,
there was little comment by those carrying out such
studies on the lack of effect of gravitropic stimulation on
the basipetal auxin movement (Cane and Wilkins, 1969)
given that Cholodny and Went had both implied that
tropistic responses resulted from a diversion of auxin flow
in the stimulated organs.
Differential rates of auxin synthesis give rise to auxin
gradients within organs and cause differential growth
(model 2.2.7): This model was a recurring minority view
(see Leopold and Kriedermann, 1975, p 211, for several
examples). It possibly lacked popularity because, like the
sensitivity model, it offered no explanation for any transmitted message which was so central to thinking for many
years. The model was also challenged by many measurements of the auxin content of organs, pre- and posttropistic stimulation, which were found to be unchanged
in most cases. An exception was found in grass node
gravitropism. In grass nodes, there was little perceived
need for any transmitted message and no trans-organ
lateral auxin redistribution was found. However, the
nodes were auxin-sensitive, so it was proposed that local
auxin synthesis was changed by gravistimulation to give
an auxin gradient across the organ ( Wright et al., 1978).
There are doubts, however, as to whether the changes in
auxin content that were reported were sufficient to
account for the magnitude of differential growth.
Differential rates of auxin conjugation at the two sides of
an organ give rise to an auxin concentration gradient and
cause differential growth (model 2.2.8): It has long been
known that hormone conjugates (the products of combining a hormone molecule with another low molecular
weight substance) can be found in plant extracts and roles
for these substances have been sought. There is no
inherently attractive feature of a model of tropistic regulation that is based on the production and reversible
breakdown of hormone conjugates, but attempts have
been made to incorporate such ideas under the
Cholodny–Went umbrella (Bandurski et al., 1990). The
auxin concentration changes reported to arise in coleoptiles from these processes seem very unimpressive (after
30 min gravistimulation 56% of the free auxin is found at
the lower side) and the model has not been developed
further, conceptually or experimentally.
Can models 2.2.1–2.2.8 coexist?
It is an unfortunate property of the hierarchical nature
of the Cholodny–Went models that tests of the mechanistic models (2.2.1–2.2.8) may apparently offer support
for the more conceptual models (2.1 and 2.2) which then
can give support to other mechanistic models merely by
association. For example, evidence that the tip of a maize
root may be the site of graviperception supports model
2.2.1, but such information actually provides no support
for models 2.2.2–2.2.8. Indeed, evidence that the tip plays
a role in gravitropic perception would actually provide
evidence that challenges models 2.2.2–2.2.8. Thus advocates of each of the models 2.2.1–2.2.8 should not be seen
as offering each other mutual support simply because
they all agree that model 2.2 is a good point of common
ground, quite the contrary. It would be possible for two
or more models to coexist if one abandons the concept
of a unifying model and if there are sound evolutionary
arguments which would support such a diversity of mechanisms. To argue that roots might use one variant of a
model and hypocotyls another would be plausible, but it
would be less plausible to argue that Zea roots use a
different mechanism from Arabidopsis roots.
Mechanistic cell-based models
Both Cholodny and Went specifically incorporated cellbased mechanisms into their whole organ models—they
both suggested that cells became polarized as a result of
a tropistic stimulation. There was a conceptual link
between the mechanism used for basipetal longitudinal
movement of auxin in cells (based on a cell polarity) and
the tropistic-driven lateral auxin movement. This link
seems to have been retained as a concept, but not
adequately developed conceptually and even less so
experimentally. Considerable attention has been paid by
workers as to how the longitudinal polarity of auxin
movement is generated—it was postulated that membrane-associated auxin carriers might be preferentially
located at the base of cells thence drive auxin efflux from
cells predominantly in a polar direction (Hertel and
Leopold, 1963). However, very little attention has been
paid to how gravity stimulation might drive lateral auxin
movement. The ‘valve hypothesis’ linked starch grain
movement to an impairment of substance movement via
the plasmodesmata (Juniper, 1976), but there are doubts
as to whether starch grains necessarily move close enough
to the plasmamembrane to allow such a direct association
and the evidence that plasmodesmata are the route
through which lateral auxin moves is controversial.
Summary of ‘the model’
The unity that the Went and Thimann brought to gravitropism research with their enunciation of the ‘CholdnyWent Model’ should now be seen as an illusion. The
conceptual unity the model gave to roots and shoot
gravitropism defied some anatomical, morphological and
physiological evidence that the organs showed significant
differences. Consequently, it should now be accepted that
no satisfactory unifying mechanistic model has yet been
proposed. The concepts of Cholodny survive best in the
model for root gravitropism (model 2.2.1), but little
Models of gravitropism
remains unchallenged of either Cholodny’s or Went’s
mechanistic ideas on shoot gravitropism. The broad conceptual models that survive for shoot gravitropism do so
largely because they give imprecise predictions and hence
do not encourage experimental falsification. The complex
spatial and temporal growth rate changes that are now
known to occur in gravistimulated organs (Digby and
Firn, 1979; Cosgrove, 1990; Evans and Ishikawa, 1997)
require much more sophisticated models of hormonal
control than were envisaged 70 years ago.
Mutants and the Cholodny–Went model
The study of mutants with abnormal sensory physiology
can yield spectacular results. This has been well illustrated
with studies of phytochrome mutants where old physiological confusions have been clarified and new levels of
understanding have been made possible. As yet, little
progress has been achieved by those studying and
exploiting gravitropism mutants, but there is an optimism
that these mutants will provide the key finally to unlock
the secrets of gravitropism. In order to confine the current
discussion, gravitropic mutants which are clearly gravitropic perception mutants will only be considered when
they provide information about the possible location of
the gravisensing cells. However, the exclusion from the
discussion of information about gravity perception is
artificial and possibly unwise because there must be a
clear link between the sensing mechanisms and the
response mechanisms and the nature of that link may be
a key into each of the processes themselves.
Mutants that showed abnormal gravitropic behaviour
were recognized well over 70 years ago and there were
attempts to relate the abnormal physiology to the prevailing models of gravitropism ( Kaiser, 1935; van
Overbeek, 1936). However, the study of gravitropism
mutants in order to probe the mechanisms involved in
the induction of differential growth only flourished when
the attractions of Arabidopsis thaliana as an experimental
system became evident. The many searches for mutants
with abnormal gravitropic responses has revealed one
fact quite clearly—that the majority of genes involved in
various aspects of root gravitropism must be distinct from
those involved in shoot gravitropism ( Tasaka et al., 1999).
Consequently it is convenient to group root mutants and
shoot mutants separately until a more rational grouping
suggests itself ( Tables 1, 2). Not all the mutants listed in
the tables will be discussed but they are included to
indicate the diversity of phenotypes available for study.
Root gravitropism mutants
Two strategies for the discovery of mutants with abnormal
root gravitropism have been widely used. The simplest is
to seek mutants which show an abnormal development
1331
of the response and then one can explore the basis of the
abnormality by experiment, guided by prevailing models.
Alternately one can seek a mutant with an altered ability
to make, move, degrade, sense or respond to an effector
molecule that is postulated to be involved in gravitropism
with the predicted phenotype as a guide. The former
strategy is complicated in the case of gravitropism because
the easiest event to measure, organ curvature or lack of
it, is the end result of a cascade of processes. For example,
an agravitropic root could be deficient in sensing, transduction (with at least three possible levels of transduction,
molecular, cell and organ) or response. Thus without
defining the nature of the abnormality of the gravitropism
in some detail, it is hard to utilize information from
studies on agravitropic mutants with confidence.
Consequently, the strategy of seeking mutants with abnormal hormonal physiology has so far proved most popular
and productive, but this approach suffers from being
guided in its approach by the very models one would like
to probe.
Abnormal gravitropism mutants
Evidence of an absence of or a deficiency in gravitropism
can take many forms:
$
$
$
$
$
no gravitropic curvature evident after prolonged gravitropic stimulation (caused by a deficiency in any one
of the multiple signal-transduction pathways involved
in the response)
a delayed gravitropic response—a long latent period
between the onset of stimulation and the onset of the
response
a normal latent period but a subsequent reduced
capacity for differential growth
—reduced magnitude of differential growth (e.g. the
capacity to elevate the elongation rate on the convex
flank is limited )
—reduced duration of differential growth (e.g. inability
to sustain the normal increased elongation on the
convex flank)
a changed ‘sensing deadzone’ (due to a sensing impairment) causing premature cessation of the period of
differential growth or showing no response to small
initial displacements
an altered GSA (gravitropic setpoint angle) change
following gravitropic stimulation causing premature
cessation of the period of differential growth (Digby
and Firn, 1995)
Consequently, to describe a mutant as showing ‘abnormal’ or ‘reduced’ gravitropism is ambiguous, although
the difficulty in fully characterizing any mutant makes
one sympathetic to the use of such terms. It would be
helpful if the standards of characterization in physiologi-
Mutant
How was
it found?
Root
gravitropism
Shoot
gravitropism
Root
elongation
Basipetal
auxin
transport
Auxin
‘sensitivity’
changed
Shoot
phototropism
What else is known?
aux1
Root growth insensitive
to applied IAA
and 2,4-D
×
m
Very small
stimulation in light,
none in dark
?
<IAA or 2,4-D
=NAA
m
axr1
Root growth insensitive
to 2,4-D
Reduced rate
of curvature
development
m
Enhanced
?
2–3 < 2,4-D/IAA
in roots and shoots
?
axr2
Root growth insensitive
to applied IAA
×
×
Slightly enhanced
?
m
?
axr3
Root growth insensitive
to ACC
Reduced
?
Reduced
?
m
?
axr4
Root growth insensitive
to applied IAA
and 2,4-D
?
Enhanced
?
<2,4-D and IAA
?
agr1
eir1
Atpin2
wav6–52
Root gravitropism
Abnormality (agr1)
Ethylene insensitivity
(eir1)
Reduced rate
of curvature
development
– maybe an
increased lag
time and lower
final angle?
×
Influx carrier. Also resistant to
ethylene and cytokinins. NAA can
restore gravitropism. Starch
sedimentation slow?
Resistance to all auxins, ethylene and
cytokinins. Encodes protein related to
ubiquitin-activating enzyme. A number
of morphological effects on shoots.
Auxin response lacking, judging by
SAUR assays. Inflorescence laterals
grow down or horizontally? Reduction
of hypocotyl elongation in dark but
less effect in light. No apical hook.
The product of AXR3 is IAA17.
Semi-dominant. Enhanced auxin
response.
Normal sensitivity to ACC, ABA and
cytokinins (but a bit variable?)
Inflorescence nearly normal. Slight
effect on leaves. Lateral root
development reduced.
but see
Utsuno
Reduced in
Atpin2
m
No/yes/maybe
m
agr2
Root gravitropism
abnormality
Reduced
Changed but
see Utsuno
?
?
=IAA
m
agr3
Root gravitropism
abnormality
Reduced
Appear normal
but see Utsuno
?
?
>2,4-D=IAA
but see Utsuno
m
rgr
Root gravitropism
abnormality
Screening for
abnormal gravitropic
response
Root gravitropism
abnormality
Less
directional
m
m
?
< IAA, 2,4-D,
NPA and TIBA
m
m
‘Reduced’
Slight reduction
in light only
‘reduced’
×
m
m
?
m(IAA) ×
(NAA)
?
arg1
agr—barley
?
?
Only expressed in root. Contradictory
evidence as to whether there is
reduced sensitivity to auxins. Some
resistance to ethylene. Recessive.
Allelic to agr2 and agr3.
Allelic to agr1 and agr3. Screened for
IAA insensitivity. Reduced sensitivity
to ACC. Recessive
Allelic to agr1 and agr2. Reduced
sensitivity to ACC. AGR may function
like a bacterial membrane transporter.
Root specific. Recessive
ARG1 codes for a DnaJ-like protein
potentially interacting with the
cytoskeleton.
IAA content normal
Key code: m=normal; ×=absent; ?=no information; >more sensitive; <less sensitive (partial resistance).
aux 1 references: Bennett et al., 1996; Yamamoto and Yamamoto, 1998; Maher and Martindale, 1980; Okada and Shimura, 1992; Roman et al., 1995; Timpte et al., 1995; Olsen et al., 1984;
Marchant et al., 1997; Chen et al., 1999; Evans et al, 1994.
axr mutants references: Estelle and Somerville, 1987, axr1; Lincoln et al., 1990, axr1; Leyser et al., 1993, axr1; Rouse et al., 1998, axr3; Wilson et al., 1990, axr2; Timpte et al., 1992, axr2;
Leyser et al., 1996, axr3; Hobbie and Estelle, 1995, axr4; Ishikawa and Evans, 1997, axr1; Evans et al., 1994, axr1, 2.
agr mutants: Maher and Bell, 1990, agr1, 2, 3; Bell and Maher, 1990, agr1, 2, 3; Luschnig et al., 1998, agr1=eir1; Utsuno et al., 1998, agr1, 2, 3; Muller et al., 1998, agr1=Atpin2; Sinclair
et al., 1996, agr3.
Other mutants: Sedbrook et al., 1999, arg mutant; Abe and Suge, 1993, lazy rice; Abe et al., 1994, lazy rice; Mullen et al., 1998, rgr; Tagliani et al., 1986, agravitropic barley.
1332 Firn et al.
Table 1. Root gravitropism mutants not thought to be caused by gravisensing lesions (all Arabidopsis unless otherwise stated)
Table 2. Shoot gravitropism mutants (all Arabidopsis unless otherwise stated)
Mutant
How was it found?
Shoot
gravitropism
Root
gravitropism
Basipetal auxin
transport
Auxin
‘sensitivity’
Shoot
phototropism
What else is known?
sgr1
Screened for nflorescence
stem gravitropism abnormality
Screened for nflorescence
stem gravitropism abnormality
Screened for nflorescence
stem gravitropism abnormality
Screened for nflorescence
stem gravitropism abnormality
Screened for inflorescence
stem gravitropism abnormality
Screened for inflorescence
stem gravitropism abnormality
Screened for inflorescence
stem gravitropism abnormality
Screening for abnormal
gravitropic response
× (i) (h)
m
?
m
(i) (h)m
Recessive. sgr1 is about 33% of wt in size.
× (i) (h)
m
?
m
(i) (h) m
Recessive
× (i) (h)
m
?
m
(i) (h) m
Recessive
×(i) (h)
m
?
m
(i) (h) m
Recessive
× (i) m(h)
m
?
m
(i) (h) m
Recessive
× (i) m(h)
m
?
m
(i) (h) m
Recessive
× (i) m(h)
m
?
m
(i) (h) m
Recessive
‘reduced’
‘reduced’
?
m
m
No pleiotropic effects reported. ARG1 codes for
a DnaJ-like protein potentially interacting with
the cytoskeleton.
sgr2
sgr4
sgr7
sgr3
sgr5
sgr6
arg1
dgt tomato
lazy-1 tomato
?
?
×
m
×
?
?
?
m
?
?
m
lazy-2 tomato
?
?
?
m
m
la maize
?
?
?
?
m
Pepper lazy
fruit stalk
lazy rice
?
m
(GSA change?)
m
(GSA change?)
m
?
?
?
m
?
?
?
?
?
m
(GSA change?)
Forms starch after 1AA treatment which restores
gravisensitivity
The phenotype is only evident in light-grown
plants
The phenotype is only evident in light-grown
plants? Recessive gene on chromosome 4.
The effect is only evident in the fruit stalks
Key code: m=normal ×=absent ?=no information>more sensitive<less sensitive (partial resistance) (i)=inflorescence (h)=hypocotyl
sgr references: Fukaki et al., 1996a, b, c, 1998; Yamauchi et al., 1997; Tasaka et al., 1999.
lazy references: dgt: Jackson, 1979; Kelly and Bradford, 1986. lazy 1 tomato: Roberts, 1984. lazy 2 tomato: Roberts, 1987; Roberts and Gilbert, 1992; Gaiser and Lomax, 1992. lazy
pepper: Kaiser, 1935. la maize: Jenkins and Gerhardt, 1931; Eyster, 1934; Emerson et al., 1935; van Overbeek, 1936. Lazy barley: Jones and Adair, 1938; Abe and Suge, 1993; Abe et al., 1994.
Models of gravitropism
In some lazy barley lines, the young seedling
sinitially grow vertical then bend down. A GSA
change as found in lazy tomato?
1333
1334 Firn et al.
cal terms matched those now demanded in molecular
terms.
The agr class of mutants
This class of mutants was reported by Bell and Maher
who first screened Arabidopsis seedlings for roots which
showed ‘abnormal’ gravitropism (Bell and Maher, 1990).
They then selected from that class a subset that showed
normal auxin sensitivity to exclude the much more
common aux1 alleles. Subsequent studies have shown
that the agr1, agr2 and agr3 are different alleles of the
same gene ( Utsuno et al., 1998) and, interestingly, the
phenotypes of each mutant differs. agr1 shows no gravitropic response but agr2 and agr3 produce an ‘incomplete’
gravitropic responses where the final angle achieved after
a period of differential growth differs from that of the
controls. Whether this is because the roots have a limited
capacity for differential growth or whether they have
changed their GSA is unknown and this needs
clarification.
agr1 has agravitropic roots but shows few other phenotypic changes, hypocotyl and inflorescence gravitropism
seems normal (Bell and Maher, 1990; Utsuno et al.,
1998). The AGR gene codes for a membrane-associated
protein with homology to a bacterial transporter ( Utsuno
et al., 1998; Chen et al., 1998). agr1 is allelic with the
ethylene-insensitive mutant eir1 (Luschnig et al., 1998),
with the polar auxin transport mutant Atpin2 (Müller
et al., 1998) and also the wav6–52 mutant. The phenotype
of these mutants can be interpreted as offering support
for model 2.2.1 in roots, indeed two of the above three
research groups reporting the characterization of the gene
( Utsuno et al., 1998; Müller et al., 1998) specifically
imply or state as much. However, Luschnig et al. outline
a role for this gene product which involves it acting to
relay an auxin asymmetry in the tip to the elongation
zone but does not ascribe a role to it in forming the
asymmetry (Luschnig et al., 1998). However, model 2.2.1
is challenged by the fact that swamping endogenous auxin
gradients does not impair the gravitropic response in the
manner expected, hence further thought needs to be given
to the information that the characterization of agr1
has given.
agr2 shows an ‘incomplete’ gravitropic response of its
roots (Bell and Maher, 1990; Utsuno et al., 1998). A
fuller physiological characterization is needed because the
explanation of the phenotype in terms of the molecular
lesion is incomplete.
agr3 has weakly gravitropic roots and shows an unusual
distribution of the root direction after re-orientation
(Maher and Bell, 1990; Sinclair et al., 1996; Utsuno et al.,
1998). As with agr2, a more complete physiological
analysis is needed and an explanation of the phenotype
in terms of the molecular lesion needs to be devised.
The rgr mutant
T-DNA insertional mutagenesis gave rise to an individual
which showed a reduced ability to reorient the root after
displacement and this was termed a ‘r2educed r2oot g2 ravitropism’ rgr mutant. The roots showed a characteriztic
curling growth habit on agar plates and had a reduced
number of lateral roots. The rgr roots were slightly shorter
than normal if grown in light but similar in length in the
dark. (Simmons et al., 1995; Mullen et al., 1998). The
roots were less sensitive to IAA, 2,4-D, NPA, and TIBA
than wild type, but had no change in sensitivity to 6-BA,
ABA or ethylene. Hypocotyl gravitropism and phototropism were apparently normal.
Mutants showing increased or decreased sensitivity to a
hormone
Exogenous auxin inhibits root elongation dramatically.
Hence it is relatively simple to screen for mutants with
an abnormal sensitivity to auxin by growing mutagenized
populations on agar containing either a natural or synthetic auxin and seeking individuals which show increased
or decreased sensitivity to auxin. In Arabidopsis such
auxin-insensitive mutants reported to date show abnormal
gravitropism ( Table 1). As can be deduced from Fig. 1,
the prediction that roots that are insensitive to auxin will
be agravitropic is a poor discriminator of the various
models—it is a prediction that could be made by most
of the many variants at every level of the hierarchy.
However, now that the molecular lesion of each mutant
phenotype is being characterized, the opportunity to
challenge or build specific or even mechanistic models is
increasing rapidly ( Estelle, 1996; Chen et al., 1999).
The aux class of mutants
This class of mutant was found as a result of screening
for Arabidopsis roots that elongated more than expected
in the presence of inhibitory concentrations of the auxinlike herbicide 2,4-D (Martindale and Maher, 1980). The
aux1 roots grow at a normal rate, but show no gravitropic
response. ( The care needed in interpreting the cause of
the abnormality is nowhere better illustrated than in the
case of aux1 where it was reported that statolith movement was changed in aux1 with the implication that
gravity sensing might be impaired (Olsen et al., 1984)
even though this now seems unlikely to be the cause of
the agravitropic root phenotype.) The AUX1 gene has
been cloned and has been shown to code for a permeaselike membrane-associated protein (Marchant et al., 1997).
The gene is expressed only in the root tip and especially
in the root epidermal cells. Bennett et al. interpreted this
evidence as indicating that gravitropism requires IAA to
be taken up into cells which transport the regulator back
to the elongation zone (Bennett et al., 1996). The IAA
permease would be part of the organ-level transduction
Models of gravitropism
chain. It was proposed that in aux1 an asymmetry in
auxin movement in the tip would not be transmitted to
the elongation-limiting cells. This explanation of root
gravitropism would be consistent with model 2.2.1.
However a serious inconsistency exists with this explanation. In aux1 roots grown on the synthetic auxin NAA,
gravitropism is restored ( Yamamoto and Yamamoto,
1998) even under conditions where the NAA is causing
considerable inhibition of elongation (hence is entering
the growth-limiting cells) and presumably swamping any
endogenous auxin gradients. This observation is related
to the most serious challenge to model 2.2.1 for root
gravitropism that was discussed previously—roots can
show a normal gravitropic response even when very high
concentrations of exogenous auxin are supplied (maize
roots for instance show a good gravitropic response at
10−4 M IAA, Katekar and Geissler, 1992). Thus although
aux1 has provided exciting new information about the
need for the AUX1 gene product for normal root gravitropism, none of the existing model variants can explain
the NAA restoration of gravitropism, hence the mutant
challenges rather than supports these models.
The axr class of mutants
Estelle and Somerville isolated a class of mutants in which
root elongation was resistant to inhibitory concentrations
of 2,4-D (Estelle and Somerville, 1987). The AXR and
AUX genes function in separate auxin-responsive pathways ( Timpte et al., 1995).
axr1: This is well characterized in physiological ( Estelle
and Somerville, 1987; Lincoln et al., 1990; Ishikawa and
Evans, 1997) and molecular terms (Leyser et al., 1993).
The mutation causes morphological changes in the roots,
hypocotyls and leaves. There is a reduction in hypocotyl
and inflorescence elongation but an enhancement in root
elongation. There have been no reports of abnormal
hypocotyl or inflorescence gravitropism, hence it would
appear to be normal but the rate of root gravitropism is
reduced. The AXR1 gene codes for a protein that has
homology with ubiquitin-activating enzyme E1, but it is
possible that the protein might be functionally distinct.
The mutant provides support for model 1. However, the
mutant does not offer a clear test of any of the more
specific models of gravitropism, partly because it is a
mutant that possesses considerable cross-resistance to
other growth substances and also because the molecular
characterization needed to test the specific models is
currently lacking.
axr2: An auxin-resistant root phenotype, resistant to
IAA-induced inhibition, axr2 is also insensitive to ethylene, ABA and 2,4-D ( Wilson et al., 1990). The root
growth is slightly reduced and the roots lack root hairs.
The growth rate changes following gravitropic stimulation
have been studied ( Evans et al., 1994). Seedlings grown
1335
on vertical agar plates have roots which initially grow at
no defined angle but later they begin to grow predominantly downwards. The hypocotyl elongation is inhibited
in the dark with no reported gravitropic abnormality
( Timpte et al., 1992). The inflorescence gravitropism
might be impaired because the organs often curve downwards although this response has not been fully characterized ( Wilson et al., 1990). As with axr1, this mutant
supports model 1 but because it is a mutant that possesses
considerable cross-resistance to other growth substances
and because the molecular characterization needed to test
the specific models is currently lacking, it has not yet
contributed much to our understanding of gravitropism.
axr3 was found after screening for roots that were
resistant to the ethylene precursor ACC. A similar phenotype was found screening for resistance to 2,4-D. The
roots also show a dramatic reduction to the growthinhibiting concentrations of the cytokinin 6-BA. The
phenotype is highly pleiotropic (Leyser et al., 1996).
Inflorescence growth, hypocotyl elongation in the dark
and root elongation are all reduced (but hypocotyl elongation in the light seems normal ). The leaves are curled,
the petioles are epinastic and anthocyanin levels are
elevated. The agr3 plants have few lateral branches.
Only the roots seem to have an abnormal gravitropic
response—the roots change direction but with no gravityinduced influence detected. The axr3 mutant is caused by
a change to the IAA17 gene which codes for a short-lived
nuclear protein that is a member of the AUX/IAA family
(Rouse et al., 1998). Model 2 would predict the existence
of this type of mutant.
axr4 was isolated from T-DNA tagged lines of
Arabidopsis using resistance of roots to 2,4-D as a screen
(Hobbie and Estelle, 1995). Unlike some other auxin
insensitive mutants, axr4 shows little cross resistance to
other hormones. The effects of the mutation are largely
confined to the root with only minor effects shown on
leaves. Model 2 would predict the existence of this type
of mutant.
Shoot gravitropism mutants
Although Arabidopsis hypocotyls show a strong gravitropic response when grown in darkness they maintain a
less upright habit when grown or exposed to light.
Although several interesting mutants were isolated by
screening for abnormal gravitropic behaviour of the hypocotyls after stimulation (Bullen et al., 1990), these have
been less well characterized than the inflorescence gravitropism mutants.
sgr inflorescence mutants
This class of mutant was isolated by seeking plants in
which the inflorescence could not show a gravitropic
response after displacement ( Fukaki et al., 1996a, b, c).
1336 Firn et al.
The several sgr mutants have been grouped into subgroups ( Tasaka et al., 1999) on the basis of in which
organs the lesion in gravitropism is expressed. None of
the sgr mutants show any abnormality of root gravitropism. There was no evidence of any changed sensitivity to
auxin in any of the sgr mutants.
sgr1 and sgr7: The inflorescence stems and the hypocotyls
of sgr1 and sgr7 show no gravitropic curvature after
stimulation. SGR1 has been shown to be allelic to SCR
(scarecrow) and SGR7 allelic to SHR (short-root) and
they are genes that influence the differentiation of the
root and the hypocotyl endodermis ( Fukaki et al., 1998).
The inflorescence stems and hypocotyls of sgr1 shows
phototropic curvature after stimulation ( Table 2). The
loss of the endodermis in sgr1 and sgr7 is interpreted as
supporting the proposal that statoliths, which are exclusively found in the starch sheath or endodermis in
Arabidopsis, are essential for gravitropic sensing in these
organs and that gravitropic sensing occurs in cells other
than the ones capable of showing a gravity-induced
change in elongation. It could be concluded that there is
a radial separation of the sites of gravity perception and
response in hypocotyls. These mutants, by providing
supporting evidence that the endodermis is the site of
gravity perception, challenge the idea that many cells in
hypocotyls could participate in the lateral movement of
auxin. Hence these mutant support the view that, if auxin
movement is part of the gravitropic response mechanism,
it must be localized.
sgr3, 5 and 6: The sgr3, sgr5 and sgr6 mutants show
abnormal gravitropism in the inflorescence stem, but the
response of the hypocotyl is apparently normal
( Yamauchi et al., 1997). sgr5 and sgr6 subjected to
continuous gravitropic stimulation appear to show a
normal latent period, but a reduced rate of curvature
development, giving less overshoot than wt. Limited
studies have failed to show any impairment of the phototropic response of sgr5 and sgr6 and what effects are
evident (a different shape of inflorescence stem after some
hours of phototropic exposure) could be a secondary
consequence of the impaired gravitropic response
(phototropism and gravitropism interact normally).
Interestingly, the angle at which the lateral branches are
maintained relative to the main axis is increased in sgr5
and sgr6 and it will be interesting to determine whether
that is a result of a change in the gravitropism of the
mutants. The molecular characterization of these mutants
is awaited with interest.
‘lazy’ phenotypes
A number of mutants of horticultural or agricultural
species have been found which have a very dramatic
trailing or ‘lazy’ habit. The ‘lazy’ mutations seems to
have no other deleterious effects and many commercially
important garden plants with this habit perform well. It
has been proposed that these phenotypes are mutants in
which the GSA (gravitropic setpoint angle) mechanism
has been impaired in some way, either the developmental
control of that GSA or the ability of the plant to use a
gravitropic response to align itself relative to the GSA
has been impaired ( Firn and Digby, 1997). Some of these
mutants have attracted the attention of physiologists.
Lazy maize
Jenkins and Gerhardt (Jenkins and Gerhardt, 1931)
described this mutant (la) which has a prostrate habit
and it was soon shown to be a simple recessive gene on
chromosome 4 ( Emerson et al., 1935) which actively grew
downwards ( Eyster, 1934). It was shown that the young
seedling growing in the dark beneath the soil grew
upwards as normal (van Overbeek, 1936). The molecular
characterization of this gene is known to be underway.
Lazy phenotypes of some other cereals are known (Jones
and Adair, 1938; Tagliani et al., 1986; Abe and Suge,
1993; Abe et al., 1994).
Lazy 2 tomato
This mutant shows normal root gravitropism and if grown
in darkness the hypocotyl develops normally. However if
grown in light, the angle at which elongation zone grows
gradually reduces until after some days the hypocotyl in
growing downwards (Roberts, 1987; Gaiser and Lomax,
1992). This sequence of events is similar to that found in
many trailing plants and it has been proposed that such
mutants have a lesion in some element that is controlling
the GSA of the plant (Firn and Digby, 1997).
Diageotropica tomato
The roots and shoots of dgt grow horizontally (GSA=
90°) with other effects being hyponastic leaves and a lack
of lateral roots (Zobel, 1973; Roberts and Gilbert, 1992;
Muday et al., 1995). Isolated hypocotyl sections from dgt
plants show an insensitivity to exogenous auxin and this,
rather than a defect in ethylene physiology (Zobel, 1973;
Jackson, 1979), was offered as an explanation of the
phenotype ( Kelly and Bradford, 1986). However, the fact
that dgt simply grows horizontally rather than vertically,
hence has a capacity to show a gravitropic response,
suggests that an auxin insensitivity is insufficient to
account for the observed phenotype.
Capsicum fruit stalks
Pendent and erect-fruited varieties of red pepper
(Capsicum annuum) were noted in the 19th century and
the inheritance was studied by a number of workers in
the early part of this century. Studies of the role gravitrop-
Models of gravitropism
ism in this phenotype being made some decades later
( Kaiser, 1935). Although not demonstrated experimentally yet, it would seem likely that the erect and pendent
varieties differ in that the flower stalks have different
GSAs. Kaiser very perceptively noted that the behaviour
of the pepper stalk mutants provided interesting and
challenging results with regards to the model of
Cholodny—he considered the fact that pepper stalks
could grow up or down depending on a single gene was
possibly more noteworthy than the fact that roots usually
grew down and shoots usually grew up.
Lazy 1 tomato shows no gravitropic response and lacks
sedimentable starch grains in its endodermis, hence its
insensitivity to gravity is ascribed to the lack of statoliths
(Roberts, 1984).
Lazy 1 is useful in providing supporting evidence for
the form and location of gravisensing in stems, a finding
confirmed by the sgr 1/7 Arabidopsis mutant. The fact
that auxin treatment restores the gravitropism of Lazy 1
and also allows the accumulation of starch supports this
conclusion. However, the IAA restoration of a sensing
function does rather complicate the role of auxin in
models of gravitropism because it implies that auxin may
be needed in some other capacity than as a regulator of
cell elongation.
So what do the lazy phenotypes tell us about the models of
gravitropism?
The behaviour of Lazy 2 tomato and of the many other
mutations which seem to act via a change in the developmental regulation of the GSA of the plant, are very hard
to explain with any of the models in Fig. 1. The fact that
the direction of gravity-induced organ movement can be
changed (reversibly in some organs) undermines one of
the fundamental concepts on which Cholodny based his
model. Cholodny produced a unifying model by arguing
there was a gravity-directed movement of a growth regulator in the direction of the gravity vector and that roots
bent down and shoot bent upwards because root growth
was inhibited, but shoot growth was enhanced by the
substance. However, the studies of the gravitropic
response of shoots with a GSA of >90<180° and of
lateral roots with a GSA of >0<90° show that shoots
and roots are capable of moving up or down following
gravitropic stimulation. Hence Cholodny’s original model
was based on a misconception about the gravitropic
behaviour of both roots and shoots. This misconception
has also influenced some mechanistic models of gravitropism. For instance, proposals that sedimenting organelles
activate some pump or valve which are asymmetrically
located in cells are unable to account for the ability of
the same organ to move up or down depending on the
gravitropic displacement. Thus the lazy mutants, the
earliest of all known gravitropic mutants, first identified
1337
before the Cholodny–Went model was widely accepted,
provide one of the greatest challenges to the concepts of
the model.
Conclusions
The first 30 years of this century took us from the broad
concept of chemical mediation of tropistic responses to
the quite specific model of Cholodny and Went. The last
70 years of the century has seen the Went and Thimann
version of the Cholodny–Went model acting as the focus
for thought and experiments, but it has also seen the
proliferation of other models of gravitropism, which
because they are based on the idea that auxin must play
a role in gravitropism, are sometimes erroneously
regarded as mere variants of the basic Cholodny–Went
scheme. Three key features of the Cholodny–Went model
of gravitropism are certainly not universally true—the
apex in not universally important, transorgan auxin movement is not an essential requirement for gravitropism
and roots and shoots do not always show opposite
directions of movement following gravitropism. The
Cholodny–Went model in any form certainly does not
provide a universal explanation of gravitropism and its
advocates have had to resort to weakening the definition
in order to allow the model to survive experimental
challenge. This has been a degradation not a development
of the model.
Studies of the phenotypes of many mutants with abnormal gravitropism initially only provided information
interpretable at the very broad or broad conceptual model
level (model 1 or model 2). This was convenient to
advocates of the Cholodny–Went framework because it
was at that level where the greatest consensus existed.
However, once the molecular characterization of the
mutants began to yield results, it was necessary to discuss
detailed or even mechanistic models once again. This has
presented two challenges. Firstly, there is less of a consensus around any model. Secondly, most of the existing
detailed models have been compromised at some stage
by experimental data. Instead of trying to use the new
information from the molecular characterization to devise
new, more appropriate detailed or even mechanistic
models there has been a tendency to try to accommodate
the inconsistencies of the existing models. The pieces of
the molecular mechanisms that would seem to be part
of the gravitropic response mechanism in roots have been
found, but it is questionable that any of the existing
detailed models which have been proposed to explain
root gravitropism in terms of auxin physiology actually
tell us how those pieces should be assembled. For instance,
the aux 1 and the agr1 mutants provide strong evidence
that auxin movement must be an important part of the
gravitropic response in roots yet these mutants have
sometimes been discussed as if they were providing strong
1338 Firn et al.
support for model 2.2.1 despite the fact that the apparently normal gravitropic behaviour of the aux 1 mutant
and wild-type roots in the presence of high concentrations
of exogenous auxin are inconsistent with the predictions
of model 2.2.1. Perhaps if the pieces of information
coming from the molecular and cellular studies of the
mutants can be assembled in some more imaginative way,
it will be possible to find a unity that has hitherto eluded
workers. Is it possible that the differences in root and
shoot gravitropism are due to the fact that similar mechanistic units are located and controlled differently within
the two very different organ? The phytotropins first
provided molecular evidence nearly 40 years ago that
shoot and root gravitropism must share some closely
related molecular target yet the very different anatomical
and physiological features of roots and shoots suggest
that the evidence of a unity at a molecular level might
need not imply a unity at a higher level of organization.
In summary, the discovery and molecular characterization of gravitropism mutants will provide us with knowledge as to the component parts of gravitropism, but
there will still be a need to assemble the parts into a
coherent mechanistic model. The building of that model
might best be done without following old, inappropriate
plans.
Acknowledgements
RDF and JD would like to thank the many undergraduate
students who have stimulated them to think more deeply about
gravitropism.
References
Abe K, Suge H. 1993. Role of gravitropic response in the dry
matter production of rice (Oryza sativa L.): an experiment
with a line having lazy gene. Journal of Plant Research
106, 337–343.
Abe K, Takahashi H, Suge H. 1994. Graviresponding sites in
shoots of normal and ‘lazy’ rice seedlings. Physiologia
Plantarum 92, 371–374.
Avery GS, Burkholder PR. 1936. Growth hormones in plants.
New York: McGraw-Hill.
Bandurski RS, Schulze A, Desrosiers M, Jensen P, Epel B,
Renecke D. 1990. Relationship between stimuli, IAA and
growth. In: Pharis RP, Rood SB, eds. Plant growth substances
1988. Berlin: Springer 341–352.
Bell PJ, Maher P. 1990. Mutants of Arabidopsis thaliana with
abnormal gravitropic responses. Molecular and General
Genetics 220, 289–293.
Bennett MJ, Marchant A, Green HG, May ST, Ward SP,
Millner PA, Walker AR, Schulz B, Feldmann KA. 1996.
Arabidopsis AUX1 gene: a permease-like regulator of root
gravitropism. Science 273, 948–950.
Brauner L. 1966. Versuche zur Analyse der geotropischen
Perzeption V Mitteilung. Uber den Einfluss des Schwerefeldes
auf die auxinempfindlichkelt von Helianthus-Hypokotylen.
Planta 69, 299–318.
Bullen BB, Best TR, Gregg MM, Barsel SE, Poff KL. 1990. A
direct screening procedure for gravitropism mutants in
Arabidopsis thaliana L. Plant Physiology 93, 525–531.
Cane AR, Wilkins MB. 1969. Independence of lateral and
differential longitudinal movement of indoly-acetic acid in
geotropically stimulated coleoptiles of Zea mays. Plant
Physiology 44, 1481–1487.
Chen R, Hilson P, Sedbrook J, Rosen E, Caspar T, Masson P.
1998. The Arabidopsis thaliana AGRAVITROPIC 1 gene
encodes a component of the polar-auxin-transport efflux
carrier. Proceeding of the National Academy of Sciences, USA
95, 15112–15117.
Chen R, Rosen E, Masson P. 1999. Gravitropism in higher
plants. Plant Physiology, 120, 343–350.
Ciesielski T. 1872. Untersuchungen über die Abwärtskrümmung
der Wurzel. Beitraege zur Biologie der Pflanzen 1, 1–30.
Copeland EB. 1900. Studies on the geotropism of stems.
Botanical Gazette 29, 185–196.
Cosgrove DJ. 1990. Rapid, bilateral changes in growth rate and
curvature during gravitropism of cucumber hypocotyls:
implications for mechanism of growth control. Plant, Cell
and Environment 13, 227–234.
Darwin C. 1880. The power of movements of plants. London:
John Murray.
Digby J, Firn RD. 1979. An analysis of the changes in growth
rate occurring during the initial stages of geocurvature in
shoots. Plant, Cell and Environment 2, 145–148.
Digby J, Firn RD. 1995. The gravitropic set-point angle GSA:
the identification of an important developmentally controlled
variable governing plant architecture. Plant, Cell and
Environment 18, 1434–1440.
Dolk HE. 1936. Growth and the geotropic substance. Recueil
des Travaux Botaniques Neerlandais 33, 509–685.
Emerson RA, Beadle GW, Fraser AC. 1935. A summary of
linkage analysis in maize. Cornell University Agricultural
Experimental Station Memoir 180.
Estelle M. 1996. Plant tropisms: the ins and outs of auxin.
Current Opinions in Plant Science 6, 1589–1591.
Estelle M, Somerville CR. 1987. Auxin-resistant mutants of
Arabidopsis thaliana with altered morphology. Molecular and
General Genetics 206, 200–206.
Evans ML. 1991. Gravitropism: interaction of sensitivity
modulation and effectors redistribution. Plant Physiology
95, 1–5.
Evans ML. 1992. What remains of the Cholodny–Went theory?
Plant, Cell and Environment 15, 767–768.
Evans ML, Ishikawa H, Estelle MA. 1994. Responses of
Arabidopsis roots to auxin studied with high temporal
resolution—comparison of wild-type and auxin-response
mutants. Planta 194, 215–222.
Evans ML, Ishikawa H. 1997. Cellular specificity of the
gravitropic motor response in roots. Planta 203, S115-S122.
Eyster WH. 1934. Genetics of Zea mays. Bibliographia Genetica
9, 187.
Firn RD. 1968. The effects of morphactin on some aspects of
plant growth. University of Adelaide. MAgSc dissertation.
Firn RD. 1986. Growth substance sensitivity: the need for
clearer ideas, precise terms and purposeful experiments.
Physiologia Plantarum 67, 267–272.
Firn RD. 1992. What remains of the Cholodny–Went theory?
Which one? Plant, Cell and Environment 15, 769–770.
Firn RD, Digby J. 1977. The role of the peripheral cell layers
in the geotropic curvature of sunflower hypocotyls: a new
model of shoot geotropism. Australian Journal of Plant
Physiology 4, 337–47.
Firn RD, Digby J. 1980. The establishment of tropic curvatures
in plants. Annual Review of Plant Physiology 31, 131–148.
Models of gravitropism
Firn RD, Digby J. 1997. Solving the puzzle of gravitropism—
has a lost piece been found? Planta 203, S159–S163.
Firn RD, Myers AB. 1987. Hormones and plant tropisms. The
degeneration of a model of hormonal control. In: Hoad GV,
Lenton JR, Jackson MB, Atkin RK, eds. Hormone action in
plant development. London: Butterworth, 251–264.
Firn RD, Tamimi S. 1985. Auxin transport and shoot tropisms:
the need for precise models. In: Bopp M, ed. Plant growth
substances 1985. Berlin: Springer, 236–240.
Fukaki H, Fujisawa H, Tasaka M. 1996a. How do plants bend
up? The initial step to elucidate the molecular mechanisms of
shoot gravitropism using Arabidopsis thaliana. Journal of
Plant Research 109, 129–137.
Fukaki H, Fujisawa H, Tasaka M. 1996b. Gravitropic response
of inflorescence stems in Arabidopsis thaliana. Plant
Physiology 100, 933–943.
Fukaki H, Fujisawa H, Tasaka M. 1996c. SGR1, SGR2 and
SGR 3: novl genetic loci involved in shoot gravitropism in
Arabidopsis thaliana. Plant Physiology 110, 945–555.
Fukaki H, Wysocka-Diller J, Kato T, Fujisawa H, Benfey PN,
Tasaka M. 1998. Genetic evidence that the endodermis is
essential for shoot gravitropism in Arabidopsis thaliana. The
Plant Journal 14, 425–430.
Gaiser JC, Lomax TL. 1992. The gravitropic tomato mutant
lazy-2 is altered in signal transduction. In: Karssen CM, van
Loon LC, Vreugdenhil D, eds. Progress in plant growth
regulation. Dordrecht: Kluwer, 928–937.
Gordon SA, Shen-Miller J. 1968. Auxin relations in phototropism of the coleoptile: a re-examination. In: Wightman F,
Setterfield G, eds. Biochemistry and physiology of plant growth
subtances. Ottawa: Runge Press, 1097–1108.
Hart JW. 1990. Plant tropism and other movements. London:
Unwin Hyman.
Hasegawa, K, Sakoda M, Bruinsma J. 1989. Revision of the
theory of phototropism in plants: a new interpretation of a
classical experiment. Planta 178, 540–544.
Hertel R. 1992. What remains of the Cholodny–Went theory?
All, but this is only a partial answer within a larger
mechanism of tropism. Plant, Cell and Environment 15,
771–772.
Hertel R, Leopold AC. 1963. Versuche zur Analyses des Auxintransports in der Koleoptile von Zea mays L. Planta
59, 535–562.
Hobbie L, Estelle M. 1995. The axr4 auxin-resistant mutants of
Arabidopsis thaliana define a gene important for root
gravitropism and lateral root initiation. The Plant Journal
7, 211–220.
Ishikawa H, Evans M. 1990. Gravity-induced changes in
intracellular potentials in elongating corical cells of mung
bean roots. Plant Cell Physiology 31, 457–462.
Ishikawa H, Evans M. 1993. The role of the distal elongation
zone in the response of maize roots to auxin and gravity.
Plant Physiology 102, 1203–1210.
Ishikawa H, Evans ML. 1997. Novel software for analysis of
root gravitropism: comparative response patterns of
Arabidopsis wild-type and axr1 seedling. Plant, Cell and
Environment 20, 919–928.
Iwami S, Masuda Y. 1974. Distribution of labelled auxin in
geotropically stimulated stems of cucumber and pea. Plant
and Cell Physiology 20, 973–987.
Jackson MB. 1979. Is the diagravitropica deficient in ethylene?
Plant Physiology 46, 347–351.
Jenkins MT, Gerhardt F. 1931. A gene influencing the
composition of the culm in maize. Iowa State College
Agricultural Experimental Station Research Bulletin 138,
121–151.
1339
Jones AM. 1992. What remains of the Cholodny–Went theory?
Assymetric redistribution of auxin need only occur over the
distance of one cell. Plant, Cell and Environment 15, 775–776.
Jones JW, Adair CR. 1938. A ‘lazy’ mutation in rice. Heredity
29, 315–318.
Juniper BE. 1976. Geotropism. Annual Review of Plant
Physiology 27, 385–406.
Kaiser S. 1935. The inheritence of geotropic response in
Capsicum fruits. Bulletin of the Torrey Botanical Club 62,
75–81.
Katekar GF, Geissler AE. 1992. On the role of the NPA receptor
in the root gravitropic response mechanism. In: Karssen CM,
van Loon LC, Vreugdenhil D, eds. Progress in plant growth
regulation. Dordrecht: Kluwer, 928–937.
Kelly MO, Bradford KJ. 1986. Insensitivity of the diagravitropica
tomato mutant to auxin. Plant Physiology 25, 23–27.
Konings H. 1995. Gravitropism of roots: an evaluation of
progress during the last three decades. Acta Botanica
Neerlandica 44, 195–223.
Leopold AC. 1992. What remains of the Cholodny–Went theory?
Valid but not universal. Plant, Cell and Environment 15,
777–778.
Leopold AC, Kriedermann PE. 1975. Plant growth and development, 2nd edn. New York: McGraw-Hill.
Leyser HMO, Lincoln CA, Timpte C, Lammer D, Turner J,
Estelle M. 1993. Arabidopsis auxin-resistance gene AXR1
encodes a protein related to ubiquitin-activating enzyme E1.
Nature 346, 161–164.
Leyser HMO, Pickett FB, Dharmasiri S, Estelle M. 1996.
Mutations in the AXR3 gene of Arabidopsis thaliana results
in altered auxin response including ectopic expression from
the SAUR-AC1 promoter. The Plant Journal 10, 403–413.
Li Y, Hagen G, Guilfoyle TJ. 1991. An auxin-responsive
promoter is differentially induced by auxin gradients during
tropisms. The Plant Cell 3, 1167–1175.
Lincoln C, Britton JH, Estelle M. 1990. Growth and development of axr1 mutants of Arabidopsis. The Plant Cell 2,
1071–1080.
Luschnig C, Gaxiola RA, Grisafi P, Fink GR. 1998. EIR1, a
root-specific protein involved in auxin transport, is required
for gravitropism in Arabidopsis thaliana. Genes and
Development 12, 2175–2187.
Macdonald IR, Hart JW. 1987. Tropisms as indicators of
hormone-mediated growth phenomena. In: Hoad GV, Lenton
JR, Jackson MB, Atkin RK, eds. Hormone action in plant
development. London: Butterworth, 231–250.
Maher P, Bell CJ. 1990. Abnormal responses to gravity and
auxin in mutants of Arabidopsis thaliana. Plant Science
66, 131–138.
Maher P, Martindale SJB. 1980. Mutants of Arabidopsis
thaliana with altered responses to auxin and gravity.
Biochemical Genetics 18, 1041–1053.
Marchant A, Kargul J, May ST, Muller P, Delbarre A, PerrotRechenmann C, Bennett MJ. 1997. Aux1 regulates root
gravitropism in Arabidopsis by facilitating auxin uptake
within root apical tissues. EMBO Journal 18, 2066–2073.
Martindale SJB, Maher P. 1980. Further studies of auxin
resistance in Arabidopsis thaliana. Arabidopsis Information
Service 17, 39–48.
Mertens R, Weiler EW. 1983. Kinetic studies on the redistribution of endogenous growth regulators in gravireacting plant
organs. Planta 158, 195–198.
Muday GK, Haworth P. 1994. Tomato root-growth, gravitropism and lateral development—correlation with auxin transport. Plant Physiology and Biochemistry 32, 193–203.
Muday GK, Lomax TL, Rayle DL. 1995. Characterization of
1340 Firn et al.
the growth and auxin physiology of roots of the tomato
mutant diageotropica. Planta 195, 548–553.
Mullen JL, Turk E, Johnson K, Wolverton C, Ishikawa H,
Simmons C, Soll D, Evans ML. 1998. Root-growth behavior
of the Arabidopsis mutant rgr1—roles of gravitropism and
circumnutation in the waving/coiling phenomenon. Plant
Physiology 118, 1139–1145.
Müller A, Guan C, Galweiler L, Tanzler P, Huijser P,
Marchant A, Parry G, Bennett M, Wisman E, Palme K. 1998.
AtPIN2 defines a locus of Arabidopsis for root gravitropism
control. EMBO Journal 17, 6903–6911.
Okada K, Shimura Y. 1992. Mutational analysis of root
gravitropism and phototropism in Arabidopis thaliana seedlings. Australian Journal of Plant Physiology 19, 439–448.
Olsen GM, Mirza JI, Mayer EP, Iversen T-H. 1984.
Ultrastructure and movements of cell organelles in the root
cap of agravitropic mutants and normal seedlings of
Arabidopsis thaliana. Physiologia Plantarum 60, 523–531.
Parsons AK, Firn RD, Digby J. 1986. The role of the coleoptile
apex in controlling organ elongation. I. The effect of
decapitation and apical incisions. Journal of Experimental
Botany 39, 491–495.
Pickard BG. 1985a. Roles of hormones, protons and calcium
in geotropism, In: Pharis RP, Reid DM, eds. Hormonal
regulation of development. III. Role of environmental factors.
Berlin: Springer, 233–247.
Pickard BG. 1985b. Early events in geotropism in seedling
shoots. Annual Review of Plant Physiology 36, 55–75.
Roberts JA. 1987. Mutants and gravitropism. In: Thomas H,
Grierson D, eds. Developmental mutants in higher plants.
Cambridge: Cambridge University Press, 135–153.
Roberts JA. 1984. Tropic responses of hypocotyls from normal
tomato plants and the gravitropic mutant Lazy 1. Plant, Cell
and Environment 7, 515–520.
Roberts JA, Gilbert I. 1992. Gravitropism research—will
mutants prevent us from going round the bend? In: Karssen
CM, vanLoon LC, Vreugdenhil D, eds. Progress in plant
growth regulation. Dordrecht: Kluwer, 913–920.
Roman G, Lubarsky B, Kieber JJ, Rothenberg M, Ecker JR.
1995. Genetic analysis of ethylene signal transduction in
Arabidopsis thaliana: five novel mutant loci intergrated into
a stress response pathway. Genetics 139, 1393–1409.
Rouse D, Mackay P, Stirnberg P, Estelle M, Leyser HMO.
1998. Changes in auxin response from mutations in an
AUX/IAA gene. Science 279, 1371–1373.
Sachs J. 1887. Lectures in plant physiology. HM Ward trans.
Oxford: Clarendon Press.
Salisbury FB, Rorabaugh P, White R, Gillespie L. 1986. A key
role for sensitivity to auxin in gravitropic stem bending. Plant
Physiology 80, S-26.
Sedbrook J, Chen R, Masson P. 1999. ARG1 Altered response
to gravitropism. encodes a novel DnaJ-like protein which
potentially interacts with the cytoskeleton. Proceedings of the
National Academy of Sciences, USA 96, 1140–1145.
Sievers A, Zieschang HE. 1992. What remains of the
Cholodny–Went theory? It does not fit the growth patterns
of cells during bending of a root. Plant, Cell and Environment
15, 789–790.
Simmons C, Migliaccio F, Masson P, Caspar T, Söll D. 1995.
A novel root gravitropism mutant of Arabidopsis thaliana
exhibiting altered auxin physiology. Physiologia Plantarum
93, 790–798.
Sinclair W, Oliver I, Maher P, Trewavas A. 1996. The role of
calmodulin in the gravitropic response of the Arabidopsis
thaliana agr-3 mutant. Planta 199, 343–351.
Tagliani L, Nissen S, Blake TK. 1986. Comparisons of growth,
exogenous auxin sensitivity and endogenous indolyl-3-acetic
acid content in roots of Hordeum vulgare L. and an
agravitropic mutant. Biochemical Genetics 24, 839–848.
Tamimi S, Firn RD. 1984. The basipetal auxin tranport system
and the control of cell elongation in hypocotyls. Journal of
Experimental Botany 36, 955–962.
Tasaka M, Kato T, Fukaki H. 1999. The endodermis and shoot
gravitropism. Trends in Plant Science 4, 103–107.
Timpte CS, Wilson AK, Estelle M. 1992. Effects of the axr2
mutation on cell shape in hypocotyls and inflorescence.
Planta 188, 271–278
Timpte CS, Lincoln C, Pickett FB, Turner J, Estelle M. 1995.
The AXR1 and AUX1 genes of Arabidopsis function in
separate auxin-response pathways. The Plant Journal 8,
561–569.
Trewavas A. 1992. What remains of the Cholodny–Went theory?
Plant Cell and Environment 15, 761–794.
Trewavas A. 1982. Growth substance sensitivity: a limiting
factor in plant development. Physiologia Plantarum 55, 60–72.
Utsuno K, Shikanai T, Yamada Y, Hashimoto T. 1998. AGR, an
agravitropic locus of Arabidopsis thaliana, encodes a novel
membrane-protein family member. Plant Cell Physiology 39,
1111–1118
van Overbeek J. 1936. ‘Lazy’, an a-geotropic form of maize.
Journal of Heredity 27, 93–96.
Went FW. 1956. Phototropism. In: Hollaender A, ed. Radiation
biology, Vol. 3. Visible and near-visible light. New York:
McGraw-Hill, 463–478.
Went FW, Thimann KV. 1937. Phytohormones. New York:
Macmillan.
Weyers JB, Patterson NW, A’Brook R, Peng Z-Y. 1987.
Quantitative analysis of the control of physiological phenomena by plant hormones. Physiologia Plantarum 95,
486–494.
Wilson AK, Pickett F, Turner JC, Estelle M. 1990. Dominant
mutation confers resistance to auxin, ethylene and abscisic
acid. Molecular and General Genetics 222, 377.
Wright M, Mousedale DMA, Osborne DJ. 1978. Evidence for a
gravity regulated level of endogenous auxin controlling cell
elongation and ethylene production during geotropic bending
of grass nodes. Biochemie und Biophysiologie der Pflanzen
172, 581–596.
Yamamoto M, Yamamoto KT. 1998. Differential effects of
1-naphthyleneacetic acid, indolyl-3-acetic acid and
2,4-dichlorophenoxyacetic acid on the gravitropic response
of roots in an auxin-resistant mutant of Arabidopsis, aux1.
Plant Cell Physiology 39, 660–664.
Yamauchi Y, Fukaki H, Fujisawa H, Tasaka T. 1997. Mutations
in the SGR4, SGR5 and SGR6 loci of Arabidopsis thaliana
alter shoot gravitropism. Plant Cell Physiology 38, 530–535.
Zobel RW. 1973. Some physiological characteriztics of ethylenerequiring tomato mutant diageotropica. Plant Physiology
52, 385–389.