Update on Development
Multiple Signaling Pathways Control Tuber
Induction in Potato
Stephen D. Jackson
Horticulture Research International, Wellesbourne, Warwick CV35 9EF, United Kingdom
Potatoes (tubers of Solanum tuberosum) are grown and
eaten in more countries than any other crop, and in the
global economy they are the fourth most important crop
after the three cereals maize, rice, and wheat. Therefore,
research into potato tuber initiation and development,
which enables our understanding and possible manipula-
tion of these processes, is of great relevance. In addition to
improving the yield and quality of potato harvests and
increasing resistance to pathogen infection, research is also
directed at improving the nutritional content of the tuber,
and "pharming" which is removing the starch in the potato
tuber and instead producing organic compounds such as
proteins that are too expensive or cannot be produced in
bacterial or yeast culture systems.
Research on potatoes has many advantages in that they
are easily transformable and therefore amenable to genetic
manipulation, and can be propagated rapidly both in tissue
culture and through cuttings. Also, microtubers can be
induced to form in tissue culture and are used in experi-
mental systems in some laboratories. Other laboratories
have used stem cuttings as small models of the whole
plant. Last but not least, the potato is very closely related to
the tomato, for which there is a good genetic map. The
main drawback to the use of potatoes in research is the fact
that most potato species are polyploid, which means that
classical genetic experiments cannot be performed.
What is a potato tuber? It is not formed from a root, as is
often supposed, but from an underground stem called a
stolon. In conditions that are noninductive for tuberization,
e.g. LD, the stolons often grow upward and emerge out of
the soil to form a new shoot (Fig. 1). In tuber-inducing
conditions, e.g. SD, however, the stolons grow under-
ground until the tip of the stolon swells to form the tuber.
The swelling is caused when the stolon ceases to elongate
and cells in the pith and cortex enlarge and divide trans-
versely. Later, cells in the perimedullary region enlarge
and divide in random orientations to form the bulk of the
tissue of the mature tuber. If the plant is put back into
noninducing conditions after a tuber has been formed, the
plant loses its induced state, and after a lag of up to 2 weeks
stolon growth may resume from the tuber. Stolon forma-
tion occurs in both tuber-inducing and noninducing con-
ditions; however, the angle and amount of stolon growth
has been correlated with the strength of the inductive
signal. Very strong induction results in "sessile" tuber
formation with no prior stolon growth (Fig. 2; Van den
Berg et al., 1996).
Tubers can actually form on other parts of the plant
above ground, normally from axillary nodes on the stem,
and in specific circumstances they can even form from
flowers (Ewing and Struik, 1992). These aerial tubers are
usually formed only on injured or diseased plants, where
translocation of assimilates below ground has been pre-
vented, or in plants grown in very strong inducing
conditions.
This Update cannot possibly summarize all of the knowl-
edge available about potato tuberization, much of which
can be found elsewhere (Li, 1985; Ewing and Struik, 1992,
and refs. therein). Therefore, it will principally focus on the
role of the environment and possible hormonal signals
involved in the induction of tuberization rather than on the
postinduction processes such as starch and storage protein
accumulation that occur during tuber formation.
ENVIRONMENTAL AND HORMONAL FACTORS
AFFECTING TUBERIZATION
There are many factors that affect tuber formation--even
the bacteria living in the root zone are reported to have an
influence--but nitrogen levels, temperature, and light have
the greatest effect. Although the overall effects of various
environmental factors are generally consistent, the geno-
type and physiological age and state of the plant (e.g.
whether still attached to the mother tuber or derived from
cuttings) can cause considerable variations in the degree to
which a plant responds to a particular environmental stim-
ulus. Analyses have been performed on a population seg-
regating for the ability to tuberize under a specific set of
conditions to try to identify quantitative trait loci that affect
tuberization (Van den Berg et al., 1996).
High Nitrogen Levels Inhibit Tuberization
Elegant experiments by Krauss and co-workers (for re-
view, see Krauss, 1985) demonstrated that tuberization
could readily be manipulated by altering the supply of
* Corresponding author; e-mail stephen.jackson@hri.ac.uk; fax
44­1789­470552.
Abbreviations: JA, jasmonic acid; LD, long day(s); PHYA and
PHYB, phytochrome A and B; SD, short day(s).
Plant Physiology, January 1999, Vol. 119, pp. 1­8, www.plantphysiol.org  1999 American Society of Plant Physiologists
1
nitrogen (either ammonium or nitrate ions) to the plant.
Whereas previous studies had looked at the effect of high
doses of nitrogen fertilizer on field-grown plants, Krauss
and co-workers grew the plants in liquid media (hydro-
ponic culture) in which the level of nitrogen supplied to
the plant could be precisely controlled. They found that a
continuous supply of between 1 and 3 mm nitrogen com-
pletely inhibited or severely delayed tuberization in oth-
erwise inducing conditions. However, interrupting the
nitrogen supply by putting the plants temporarily in
nitrogen-free media for 4 to 6 d allowed tuberization to
occur. If the plants are put into "excessive" nitrogen sup-
ply after they have started tuberizing, then tuber forma-
tion will cease and stolon growth may be resumed.
Removal of the nitrogen from the media will then cause
initiation of a second tuber from this stolon (secondary
growth). Repeated cycles of high nitrogen/nitrogen with-
drawal can result in the formation of "chain tubers," dem-
onstrating that nitrogen levels play an important role in the
control of tuber formation.
It is interesting that high nitrogen supply to the leaf did
not prevent tuberization, even though the nitrogen content
of the plants was comparable to those receiving high nitro-
gen through the roots. Furthermore, reducing nitrogen lev-
els in normally noninducing conditions such as LD or high
temperatures (see below) will not result in tuberization,
indicating that nitrogen is probably not involved in the
induction of tuberization but that it is able to repress tuber
formation once induction has taken place.
It is not yet known how nitrogen levels cause the inhi-
bition of tuberization, although there are reports that ni-
trogen withdrawal affects phytohormone levels, causing a
reduction in GA levels and an increase in ABA levels
(Krauss, 1985). An alternative hypothesis is that the ratio of
carbohydrate to nitrogen is important. High levels of car-
bohydrates in the form of sugars and starch favor the
formation of storage organs, i.e. tuberization, whereas high
nitrogen levels are known to promote shoot and root
growth that would utilize much of the available carbohy-
drate and thereby reduce the amount available for tuber
formation. Observations consistent with this hypothesis
have been made in in vitro tuberization experiments in
which the inhibitory effect of increased nitrogen levels on
tuberization were observed only at 2% Suc but not at
higher concentrations (Koda and Okazawa, 1983), at which
the high carbohydrate levels may be masking the effects of
altering the nitrogen levels. The high Suc concentrations
(up to 8%, w/v) often used to obtain uniform in vitro
tuberization, along with the possible addition of other
growth regulators, favor tuberization so much that the
interpretation of results of in vitro experiments in soil-
grown plants should be made with caution.
High Temperatures Inhibit Tuberization
High temperatures are inhibitory for tuberization in both
short and long photoperiods, although the inhibitory effect
is much greater in long photoperiods. High temperatures
Figure 1. Solanum demissum plants grown in noninducing long-day
(left) or inducing short-day (right) conditions.
Figure 2. The tuberization response of cuttings that have been induced to differing degrees. From left to right, noninduced
(no stolon or tuber) to strongly induced (sessile tuber). Photo courtesy of E. Ewing (Cornell University, Ithaca, NY).
2 Jackson Plant Physiol. Vol. 119, 1999
affect the partitioning of assimilates by decreasing the
amount going to the tubers and increasing the amounts to
other parts of the plant; similar effects are also observed in
long photoperiods. It was established, by varying the tem-
perature of the soil or the air, and thus treating different
parts of the plant with different temperature regimes (high,
30°C­35°C; low, 17°C­27°C), that high temperatures given
to the shoots had the greatest inhibitory effect on the
induction to tuberize (as determined by tuberization of
cuttings taken from the plants after the treatment). High
soil temperature did not affect the production of the induc-
ing signal but prevented stolons from developing into tu-
bers (Ewing and Struik, 1992). At high soil temperatures
stolons grow upward, and once they reach the soil surface
and the cooler air, tuberization can occur. Hot weather can
cause secondary growth of a tuber, i.e. resumption of sto-
lon growth from the tuber, in a process known as heat
sprouting. If the temperature becomes cooler after heat
sprouting, then a new tuber will start to form at the stolon
tip, forming a "chain tuber" in a manner similar to that
obtained by cycles of alternating high/low nitrogen levels.
There is some evidence that the inhibitory effect of high
temperatures is mediated through increased GA levels.
Treating plants or cuttings with chloroethyltrimethylam-
monium chloride, an inhibitor of GA biosynthesis, over-
came the inhibition of tuberization caused by high temper-
atures. This did not occur, however, if the plants had been
disbudded, indicating that the increase in GA biosynthesis
in response to high temperatures occurs in the buds and
that this is the site of action of the chloroethyltrimethylam-
monium chloride. This is supported by measurements of
GA activity, which indicated that higher temperatures
caused higher levels of activity in the buds but not in the
leaves and that this was associated with increased inhibi-
tion of tuberization (Menzel, 1983).
High Light and High Suc Promote Tuberization
Low light levels delay tuberization, and this has been
shown with field-grown plants as well as with plants
grown in controlled environments. The effects of low light
intensities on growth resemble the effects of high temper-
atures, and the promotive effects of high levels of irradi-
ance can ameliorate the inhibitory effects of high temper-
ature. Menzel (1985) suggested that the effects of both
temperature and irradiance may be mediated through the
same control process, possibly involving GAs. Although
low light intensities have been shown to increase the acidic
GA levels in potato leaves (Woolley and Wareing, 1972),
little evidence has been presented to refute the argument
that the effect of low light levels on tuberization is due to
reduced Suc levels as a direct consequence of lower pho-
tosynthetic rates.
As mentioned before, in vitro tuberization is highly de-
pendent on Suc concentration (Xu et al., 1998), and Suc is
known to induce several genes that are also induced in the
potato tuber, e.g. patatin, proteinase inhibitor II, and ADP-
Glc pyrophosphorylase. Xu et al. (1998) reported much
higher levels of GA1 (but not ABA) in the tips of stolons
growing in media with 1% Suc, as opposed to 8% Suc, and
suggested that Suc can modulate endogenous GA levels in
the stolon tip. Perata et al. (1997) showed that Glc can
repress both GA signaling and GA biosynthesis in barley
embryos. It has also been reported that the reverse is true,
i.e. that GAs can affect carbohydrate metabolism. GA3
reduces the activity of ADP-Glc pyrophosphorylase and
thus starch synthesis but increases the activity of UDPG
pyrophosphorylase, an enzyme involved in the production
of nucleotide sugars that can be used in cell wall synthesis
(Mares et al., 1981).
Evidence supporting the role of Suc as an inducing signal
includes the fact that an increase in leaf starch accumula-
tion (and presumably, therefore, of Suc export from the
leaf) can be detected after as few as 2 d in inducing condi-
tions, and recent results show that Suc can repress
phytochrome-mediated responses (Dijkwel et al., 1997).
Increasing the level of Suc in the stolons by antisensing the
ADP-Glc pyrophosphorylase and thus preventing starch
formation in the tubers led to an increased number of
tubers being initiated, even though they did not grow very
large (Muller-Rober et al., 1992).
SD Promote Tuberization
The potato is a short-day plant, although the critical
night length for tuberization and the strength of the pho-
toperiodic response varies with different genotypes (Sny-
der and Ewing, 1989). Potato species such as S. demissum
and S. tuberosum ssp. andigena are qualitative short-day
plants that require daylengths of 12 h or less to tuberize,
and because of their strict photoperiodic response, they are
often used in experiments on photoperiodic effects on tu-
berization. With photoperiodic responses it is actually the
length of the dark period rather than the light period that
is important, i.e. a SD has a long night and vice versa. This
is illustrated by the fact that interrupting an inducing long
night with a light treatment (night break) will prevent
tuberization, whereas a dark treatment in the middle of a
long light period will have no effect (Fig. 3A). SD promote
higher rates of photosynthesis per unit leaf dry weight and
Figure 3. A, Tuberization response of S. tuberosum ssp. andigena to
different photoperiodic treatments. White boxes, Lights on; black
boxes, lights off. B, Tuberization response to night breaks of red (R)
and far-red (FR) light.
Control of Tuber Induction in Potato 3
more starch accumulation in the leaf during the day. As-
similate export from leaves was also found to be higher in
SD than in LD (Lorenzen and Ewing, 1992). These effects
were observable fairly soon (2 d) after the short-day treat-
ment started and preceded tuber initiation, which is usu-
ally observed after 1 to 2 weeks.
The principal site of perception of the photoperiodic
signal is in the leaves. Photoperiodic responses can be
readily observed in single leaf cuttings (Ewing and Ware-
ing, 1978), where it was shown that increasing the number
of SD shifted growth from aboveground meristems to those
below ground, where stolons and tubers were formed. The
effect of photoperiod on tuberization appears to be medi-
ated at least in part by GA application, which prevents or
delays tuberization in inducing SD, whereas inhibiting GA
biosynthesis using inhibitors such as ancymidol will allow
tuberization to proceed in normally noninducing LD (Jack-
son and Prat, 1996).
Reports of the effects of photoperiod on tuberization in
vitro are inconsistent, with some studies finding that LD
rather than SD are more favorable for tuberization. Apart
from the high levels of Suc, the growth regulators added to
the media (some of which are inhibitors of GA biosynthe-
sis) may be complicating the picture. In addition, some
studies were performed with leafless stem sections or even
stolons, which, without any expanded leaves, cannot be
expected to exhibit a strong photoperiodic response.
PHYB Is Involved in the Photoperiodic Response
Interrupting a long dark period with a night break of red
light inhibits tuberization and this inhibition can be par-
tially reversed by a subsequent treatment with far-red light
(Fig. 3B; Batutis and Ewing, 1982). This photoreversibility
is a defining characteristic of responses mediated by phy-
tochrome. There are at least five different types of phyto-
chrome identified in tomato, and because potato and to-
mato are so closely related, it can safely be assumed that
equivalent types are present in potato. Using an antisense
approach in the short-day S. tuberosum ssp. andigena to
obtain plants with reduced phytochrome levels enables the
roles of different phytochromes in the photoperiodic con-
trol of tuberization to be studied (because S. tuberosum ssp.
andigena is tetraploid, it is not possible to screen for mu-
tants in which phytochrome genes are knocked out, as has
been done with tomatoes and Arabidopsis). To date, only
the role of PHYB has been reported (Jackson et al., 1996).
Reduced levels of PHYB in transgenic antisense S. tubero-
sum ssp. andigena plants enables them to tuberize in both
SD and LD, whereas wild-type plants form only stolons
and do not tuberize in LD (Fig. 4). Tubers form on the
antisense plants with little or no stolon formation, even in
continuous light, reflecting a strongly induced state of
these plants to tuberize (Ewing and Wareing, 1978; Ewing
and Struik, 1992). Thus, the antisense plants have lost the
inhibitory effect on tuberization caused by LD; in other
words, PHYB appears to play a role in inhibiting tuberiza-
tion in LD.
PHYB also appears to be involved in the photoperiodic
control of flowering, especially in short-day plants. The
ma3
R
mutant of Sorghum bicolor is now known to be a phyB
mutant, and this mutant has a reduced sensitivity to pho-
toperiod in its flowering response. Even in the long-day
plant Arabidopsis, flowering is earlier in the phyB mutant
than in wild type. PHYA has also been shown to be in-
volved in the photoperiodic control of flowering in Arabi-
dopsis, and it is very likely that it will be involved in other
responses controlled by photoperiod, such as tuberization
in potato.
Apart from the influence of PHYB, there are other simi-
larities between the flowering and tuberization processes.
Like tuberization, flowering is also affected by nitrogen
levels, temperature, and light levels; indeed, they may even
share the same transmissible signals (see below).
TRANSMISSIBLE SIGNALS ARE INVOLVED IN THE
CONTROL OF TUBERIZATION
Photoperiodic perception occurs in the leaf. Some sort of
signal must therefore be produced in response to the pho-
toperiodic stimulus that is then transmitted from the leaves
of the plant to the underground stolons, where tuber for-
mation occurs. Such a signal can be transmitted across a
graft union, as was demonstrated in experiments by Greg-
ory (1956) and Chapman (1958). In these experiments
leaves from potato plants that were induced to tuberize
caused noninduced stocks onto which they were grafted to
tuberize, even though after grafting they were maintained
in noninducing conditions. Furthermore, the signal pro-
duced in leaves of tobacco plants that have been induced to
flower is similar to or the same as the signal that induces
tuberization in potato plants. Grafting leaves from tobacco
plants induced to flower onto potato plants maintained in
noninducing conditions led to tuberization of the potato
plants, whereas grafted leaves from noninduced tobacco
plants did not cause tuberization (Table I). Thus, the pro-
cesses leading to the production of this signal in response
to an inducing photoperiod appear to be similar in potato
and tobacco for tuberization and flowering, respectively.
Similar results have been obtained with leaves of induced
sunflowers, which were able to cause tuberization of
Figure 4. Tuberization response of two wild-type (left) and two
antisense PHYB (right) potato plants grown in LD. Tuber formation
occurred only on the antisense PHYB plants. Notice the lack of
stolon formation.
4 Jackson Plant Physiol. Vol. 119, 1999
Jerusalem artichoke stocks, and therefore the phenomenon
does not seem to be restricted to tobacco and potato plants.
The nature of this transmissible signal is not known but
it is thought to be hormonal for a number of reasons. It
moves through the phloem both acropetally and basipe-
tally but is prevented from doing so by "heat girdling" of
the stem, which results in tubers forming from axillary
buds above the site of girdling but not below on the sto-
lons. Such a signal may have more than one component,
e.g. an inducing substance that increases under inductive
conditions and/or an inhibitory substance that decreases
under inductive conditions. As mentioned above, PHYB is
involved in the inhibition of tuberization in LD rather than
the induction of tuberization in SD, since removal of PHYB
results in tuberization in both LD and SD. Tuberization of
the antisense PHYB plants in LD could be caused by a
reduction in the levels of an inhibitor or by the production
of an inducing substance in normally noninducing LD.
That PHYB is involved in the production of a transmissible
signal(s) has been shown by grafting experiments in which
a wild-type S. tuberosum ssp. andigena plant could be in-
duced to tuberize in LD by grafting on a shoot from an
antisense PHYB plant but not by a graft from another
wild-type plant (Fig. 5; Jackson et al., 1998). Tuberization of
such graftings does not occur, however, if some leaves are
left on the wild-type stock plant. Furthermore, with the
reciprocal grafting of a wild-type shoot grafted onto an
antisense PHYB plant, tuberization of the antisense plant
that would normally occur in LD is inhibited by the wild-
type graft. These results indicate that an inhibitor of tuber-
ization exists in the leaves of wild-type potato plants in LD
and that the lower levels of PHYB in the antisense plants has
led to reduced levels of this inhibitor, thus allowing tuber-
ization to occur in LD. PHYB thus appears to be involved in
the production of the inhibitor in noninducing LD.
GAs INHIBIT TUBERIZATION AND PLAY A ROLE IN
THE CONTROL BY PHOTOPERIOD
As already mentioned, nitrogen levels, temperature, and
light intensity are all thought to have an effect on GA
levels. Photoperiod also has an effect; in many species
levels of GAs are higher in LD than in SD. It has been
shown, for example, that levels of GA-like activity decrease
in leaves of S. tuberosum ssp. andigena plants upon transfer
from LD or night-break conditions to SD (Machackova et
al., 1998). Although other studies of GA12-aldehyde metab-
olism in S. tuberosum ssp. andigena plants grown in SD and
LD did not find any difference between the photoperiods
(Van den Berg et al., 1995), the step controlled by photo-
period could lie before GA12­aldehyde in the biosynthetic
pathway. Certain steps in the GA biosynthetic pathway are
known to be affected by photoperiod, as has been shown in
spinach and pea. In spinach bolting is prevented in SD by
a lower activity of GA 20-oxidase, which results in less
GA20 and GA1. In pea senescence is prevented in SD by an
increased production of GA53 from GA12-aldehyde.
GAs inhibit tuberization and appear to play a role in the
photoperiodic control of tuberization by preventing tuber-
ization in LD. A dwarf mutant of S. tuberosum ssp. andigena
that is able to tuberize in LD as well as in SD has been
shown to have a partial block in its GA biosynthetic path-
way (Van den Berg et al., 1995). Furthermore, wild-type S.
tuberosum ssp. andigena plants treated with ancymidol, an
inhibitor of GA biosynthesis, will tuberize in LD (Jackson
and Prat, 1996). This ancymidol treatment of wild-type
plants resulted in sessile tuber formation, with little or no
stolon formation, in a manner very similar to the formation
of tubers on the antisense PHYB plants. These results indi-
Figure 5. Graftings of wild-type and antisense PHYB plants main-
tained in LD. A wild-type scion grafted onto a wild-type stock (left)
did not tuberize, whereas an antisense PHYB scion could induce a
wild-type stock to tuberize in the long-day conditions (right).
Table I. Results of grafting scions from different photoperiodic
tobacco species onto potato stocks
The grafted plants were kept in LD or SD. Mammoth is a short-day
species, Xanthi is a day-neutral species, and Sylvestris a long-day
species. (Summarized from Ewing, 1995.) , Tuberization occurred;
, tuberization did not occur.
Tobacco
Scion
SD LD
Tobacco
scion
flowering
Potato
stock
tuberizing
Tobacco
scion
flowering
Potato
stock
tuberizing
Mammoth
Xanthi
Sylvestris
Control of Tuber Induction in Potato 5
cate that a decrease in GA levels may be involved in the
photoperiodic induction of tuberization in potato plants
and that the reduced levels of PHYB in the antisense plants
may lead to reduced levels of, or sensitivity to, GA, there-
fore enabling them to tuberize in LD.
Whereas the observations mentioned above may appear
to contradict reports of increased GA levels or sensitivity in
phyB mutants such as the Brassica ein, Sorghum ma3
R
, Ara-
bidopsis phyB, and cucumber lh mutants, there is a range of
different biologically active GAs affecting different re-
sponses. It is known that the 3 -hydroxylated GA1 is the
most active GA with respect to stem elongation, and there
is evidence from Lolium that non-3 -hydroxylated GAs are
more active in promoting flowering than 3 -hydroxylated
GAs (Evans et al., 1994). Reduced levels of PHYB may not,
therefore, be causing a general reduction in the levels of all
GAs, but only in specific ones, which would alter the
relative levels of different GA species. By affecting the
expression or activity of one or more enzymes involved in
the GA biosynthetic pathway, PHYB could, for example,
change the ratio of 3 -hydroxylated to non-3 -hy-
droxylated GAs and thus change the development of the
plant away from stem elongation and vegetative growth
and toward flowering and reproductive growth. PHYA has
also been shown to affect GA levels; overexpressing PHYA
in tobacco results in reduced GA levels and a dwarfed
phenotype. Thus, it may be the case that both PHYA and
PHYB affect photoperiodic responses such as tuberization
and flowering by modifying GA metabolism/response.
Studies of the effect of GAs on in vitro tuberization have
shown that concentrations of GA1 vary throughout the
stolon, with the highest concentration located in the stolon
tip. The stolon tip is also where the greatest differences in
GA1 concentrations were observed between inducing (8%
Suc) and noninducing (1% Suc) conditions (Xu et al., 1998).
It was also shown that, by putting the cuttings in alternat-
ing low and high GA-containing media, chain tubers could
be induced to form in vitro. GAs are known to promote cell
elongation in meristematic tissue, and GA3 has been shown
to cause microtubules and microfibrils to become orien-
tated transversely to the cell axis, resulting in longitudinal
cell expansion (Shiboaka, 1993) and thus stolon elongation
(Fujino et al., 1995). Reducing the levels of GA will result in
the microtubules and microfibrils becoming orientated lon-
gitudinally (as has been shown to occur during treatment
with uniconazol, a GA biosynthesis inhibitor; Shiboaka
[1993]), thus allowing lateral cell expansion and division.
NO CLEAR ROLE HAS BEEN DEFINED FOR ANY
OTHER PLANT HORMONE
To date there is little evidence that shows a role for any
other hormone in the control of tuber induction. Numerous
measurements have been made on auxin and cytokinin
levels, but the results have been inconsistent. ABA levels
have been shown to be affected by photoperiod with up to
4-fold higher levels being measured in S. tuberosum ssp.
andigena plants in inducing SD conditions as opposed to
noninducing long-day or short-day plus night-break con-
ditions (Machackova et al., 1998). ABA levels in shoots and
roots of potato have also been shown to be affected by
nitrogen levels (Krauss, 1985). However, as an ABA-
deficient mutant of potato, Droopy is able to tuberize (Quar-
rie, 1982), and it is clear that ABA is not an essential
component of the tuberization stimulus. The promotive
effects of ABA on tuberization, both in soil-grown plants
and in vitro, appear to be due to the antagonistic effects of
ABA and GA (Xu et al., 1998). Such antagonism could be at
the level of cortical microtubules, where ABA was shown
to promote longitudinal arrays of microtubules and was
able to reverse the effect of GA3 on microtubule orientation
(Shiboaka, 1993).
Ethanol extraction of induced potato leaves led to the
identification and isolation of an acidic substance that
showed tuber-inducing activity in vitro. This substance,
called tuberonic acid, was found to be structurally related
to JA, which also showed similar levels of tuber-inducing
activity in vitro (Koda et al., 1991). JA itself, when repeat-
edly sprayed on noninduced S. tuberosum ssp. andigena
plants, and taken up and transported throughout the plant
in sufficient quantities to induce a systemic wound re-
sponse (an established role of JA in plants), did not result
in tuberization (Jackson and Willmitzer, 1994). No differ-
ences in the endogenous levels of JA were observed in
leaflets of photoperiodic S. demissum plants grown in SD
and LD. Furthermore, application of salicylhydroxamic
acid, an inhibitor of one step in the JA biosynthetic path-
way, did not prevent tuberization in short-day conditions
(Helder et al., 1993). These results indicate that differences
in the levels of JA itself do not control tuberization. This
does not exclude the possibility that tuberonic acid or other
JA-related compounds may be able to cause tuberization in
noninductive conditions, but as yet there are no reports of
these compounds having been tested on soil-grown plants.
JA may promote tuberization in vitro by disrupting cortical
microtubules of the cells and thus allowing their lateral
expansion (Matsuki et al., 1992). Consequently, JA may act
in a manner similar to that proposed for ABA (see above)
and exert its effect principally by antagonizing the effect of
GA on microtubule orientation.
While the debate continues about whether JA or related
compounds are involved in inducing tuberization in soil-
grown plants (Koda, 1997), another compound, called coro-
natine, with at least 1000-fold greater in vitro tuber-
inducing activity than JA, has been discovered (Koda et al.,
1996). Coronatine is a phytotoxin isolated from Pseudomo-
nas syringae and, in addition to its tuber-inducing activity,
has been shown to induce cell expansion in tissue from
potato tubers (at a concentration of 1:100 of that required
by JA to produce the same effect). The ability of coronatine
to induce tuberization of soil-grown plants maintained in
noninducing conditions remains to be tested.
WHAT IS THE SIGNAL TRANSDUCTION PATHWAY?
The signal transduction pathway(s) is only beginning to
be elucidated (Fig. 6), and there is good evidence that
shows the involvement of phytochrome in the response to
photoperiod. PHYB is known to affect GA levels and/or
response, and this is probably the mechanism by which
6 Jackson Plant Physiol. Vol. 119, 1999
tuberization is affected in the PHYB-deficient antisense
plants. Photoperiod, however, is also known to affect the
production and export of Suc, another signaling molecule.
Whether this effect of photoperiod is mediated through
phytochrome is not yet known, although there is known to
be a close interaction between Suc and light signaling
pathways, with Suc being able to repress phytochrome-
mediated responses.
There is also evidence that indicates the involvement of
Ca2
/calmodulin at some stage downstream in the induc-
tion pathway. Studies with single leaf cuttings have shown
that including Ca2
chelators together with a Ca2
iono-
phore in the liquid medium prevented tuberization, but
tuberization occurred when the cuttings were transferred
to Ca2
-containing medium. Calmodulin antagonists were
also found to inhibit tuberization of the cuttings (Balamini
et al., 1986). Transgenic plants overexpressing a potato
calmodulin isoform, PCM1, were found to be inhibited in
their tuberization response (Poovaiah et al., 1996). These
plants exhibit a phenotype reminiscent of GA-treated
plants. Such results suggest that Ca2
and calmodulin are
somehow involved in the tuberization process, which may
not be surprising since they have been shown to be in-
volved in at least one phytochrome signal transduction
pathway. Furthermore, a Ca2
-dependent protein kinase
has been isolated from potatoes and shown to increase in
activity at the onset of in vitro tuberization, implying that
more than one Ca2
-signaling pathway may be involved in
the induction of tuberization.
The identification of a tuber-inducing signal remains an
elusive goal. Although the majority of opinions favor a
mechanism whereby the relative levels of two or more
factors (inducing and inhibitory) determine whether tuber-
ization occurs, so far the only strong candidate for the
inhibitor that has been identified is GA. All environmental
conditions and other hormones that have an effect on tu-
berization appear to affect GA levels or to antagonize the
effects of GA, e.g. on microtubule orientation.
In addition to controlling microtubule orientation, GA
levels may control carbohydrate metabolism and direct Suc
utilization toward storage (tuber formation at low GA) or
cell wall synthesis (continued stolon growth at high GA).
At the same time Suc may exert a positive influence
(thereby being classified as a tuber-inducing signal) by
regulating endogenous GA levels and responses in the
stolon tip. Conditions such as high light or short photope-
riods, which lead to high Suc levels, would thus also cause
a reduction in GA levels and promote tuberization. A high
level of photoassimilate was once thought to be an induc-
ing signal for the formation of tubers, but this was later
modified to incorporate the effect of nitrogen, and the ratio
of carbohydrate to nitrogen was proposed to be the impor-
tant factor. This may eventually turn out to be the case if
indeed Suc reduces endogenous GA levels, whereas nitro-
gen (along with other noninducing conditions such as high
temperatures) increases them.
Received August 18, 1998; accepted September 1, 1998.
LITERATURE CITED
Balamini V, Veluthambi K, Poovaiah BW (1986) Effect of calcium
on tuberization in potato (Solanum tuberosum L.). Plant Physiol
80: 856­858
Batutis EJ, Ewing E (1982) Far-red reversal of red light effect
during long night induction of potato (Solanum tuberosum L.)
tuberization. Plant Physiol 69: 672­674
Chapman HW (1958) Tuberization in the potato plant. Physiol
Plant 11: 215­224
Dijkwel PP, Huijser C, Weisbeek PJ, Chua N-H, Smeekens SCM
(1997) Sucrose control of phytochrome A signaling in Arabidop-
sis. Plant Cell 9: 583­595
Evans LT, King RW, Mander LN, Pharis RP (1994) The relative
significance for stem elongation and flowering in Lolium temu-
lentum of 3 -hydroxylation of gibberellins. Planta 192: 130­136
Ewing EE (1995) The role of hormones in potato (Solanum tubero-
sum L.) tuberization. In PJ Davies, ed, Plant Hormones and Their
Role in Plant Growth and Development. Martinus Nijhoff, Dor-
drecht, The Netherlands, pp 698­724
Ewing EE, Struik PC (1992) Tuber formation in potato: induction,
initiation and growth. Hortic Rev 14: 89­197
Ewing EE, Wareing PF (1978) Shoot, stolon and tuber formation on
potato (Solanum tuberosum L.) cuttings in response to photope-
riod. Plant Physiol 61: 348­353
Fujino K, Koda Y, Kikuta Y (1995) Reorientation of cortical mi-
crotubules in the sub-apical region during tuberization in single
note stem segments of potato in culture. Plant Cell Physiol 36:
891­895
Gregory LE (1956) Some factors for tuberization in the potato. Ann
Bot 41: 281­288
Helder H, Miersch O, Vreugdenhil D, Sembdner G (1993) Oc-
currence of hydroxylated jasmonic acids in leaflets of Solanum
demissum plants grown under long- and short-day conditions.
Physiol Plant 88: 647­653
Jackson S, Heyer A, Dietze J, Prat S (1996) Phytochrome B medi-
ates the photoperiodic control of tuber formation in potato. Plant
J 9: 159­166
Jackson S, James P, Prat S, Thomas B (1998) Phytochrome B
affects the levels of a graft-transmissible signal involved in
tuberization. Plant Physiol 117: 29­32
Jackson S, Prat S (1996) Control of tuberization in potato by
gibberellins and phytochrome B. Physiol Plant 98: 407­412
Jackson S, Willmitzer L (1994) Jasmonic acid spraying does not
induce tuberization in short-day requiring potato species kept in
non-inducing conditions. Planta 194: 155­159
Koda Y (1997) Possible involvement of jasmonates in various
morphogenic events. Physiol Plant 100: 639­646
Koda Y, Kikuta Y, Tazaki H, Tsujino Y, Sakamura S, Yoshihara
T (1991) Potato tuber-inducing activities of jasmonic acid and
related compounds. Phytochemistry 30: 1435­1438
Koda Y, Okazawa Y (1983) Influences of environmental, hormonal
and nutritional factors on potato tuberization in vitro. Jpn J Crop
Sci 52: 582­591
Figure 6. Environmental factors and signaling molecules affecting
the induction of tuberization. The transduction pathway still has to
be defined.
Control of Tuber Induction in Potato 7
Koda Y, Takahashi K, Kikuta Y, Greulich F, Toshima H, Ichihara
A (1996) Similarities of the biological activities of coronatine and
coronafacic acid to those of jasmonic acid. Phytochemistry 41:
93­96
Krauss A (1985) Interaction of nitrogen nutrition, phytohormones
and tuberization. In PH Li, ed, Potato Physiology. Academic
Press, London, pp 209­231
Li PH (1985) Potato Physiology. Academic Press, London
Lorenzen J, Ewing E (1992) Starch accumulation in leaves of
potato (Solanum tuberosum L.) during the first 18 days of photo-
period treatment. Ann Bot 69: 481­485
Machackova I, Konstantinova TN, Seergeva LI, Lozhnikova VN,
Golyanovskaya SA, Dudko ND, Eder J, Aksenova NP (1998)
Photoperiodic control of growth, development and phytohor-
mone balance in Solanum tuberosum. Physiol Plant 102: 272­278
Mares DJ, Marschner H, Krauss A (1981) Effect of gibberellic acid
on growth and carbohydrate metabolism of developing tubers of
potato (Solanum tuberosum). Physiol Plant 52: 267­274
Matsuki T, Tazaki H, Fujimori T, Hogetsu T (1992) The influences
of jasmonic acid methyl ester on microtubules in potato cells and
formation of potato tubers. Biosci Biotechnol Biochem 56: 1329­
1333
Menzel CM (1983) Tuberization in potato at high temperatures:
gibberellin content and transport from buds. Ann Bot 52:
697­702
Menzel CM (1985) Tuberization in potato at high temperatures:
interaction between temperature and irradiance. Ann Bot 55:
35­39
Muller-Rober B, Sonnewald U, Willmitzer L (1992) Inhibition of
the ADP-glucose pyrophosphorylase in transgenic potatoes
leads to sugar-storing tubers and influences tuber formation and
expression of tuber storage protein genes. EMBO J 11: 1229­1238
Perata P, Matsukura C, Vernieri P, Yamaguchi J (1997) Sugar
repression of a gibberellin-dependent signaling pathway in bar-
ley embryos. Plant Cell 9: 2197­2208
Poovaiah BW, Takezawa D, An G, Han T-J (1996) Regulated
expression of a calmodulin isoform alters growth and develop-
ment in potato. J Plant Physiol 149: 553­558
Quarrie SA (1982) Droopy: a wilty mutant of potato deficient in
abscisic acid. Plant Cell Environ 5: 23­26
Shibaoka H (1993) Regulation by gibberellins of the orientation of
cortical microtubules in plant cells. Aust J Plant Physiol 20:
461­470
Snyder, Ewing EE (1989) Interactive effects of temperature, pho-
toperiod and cultivar on tuberization of potato cuttings. Hortic
Sci 24: 336­338
Van den Berg JH, Ewing E, Plaisted RL, McMurry S, Bonierbale
MW (1996) QTL analysis of potato tuberization. Theor Appl
Genet 93: 307­316
Van den Berg JH, Simko I, Davies PJ, Ewing EE, Halinska A
(1995) Morphology and (14
C)gibberellin A12 aldehyde metabo-
lism in wild-type and dwarf Solanum tuberosum ssp. andigena
grown under long and short photoperiods. J Plant Physiol 146:
467­473
Woolley DJ, Wareing PF (1972) Environmental effects on endog-
enous cytokinins and gibberellin levels in Solanum tuberosum.
New Phytol 71: 1015­1025
Xu X, van Lammeren AA, Vermeer E, Vreugdenhil D (1998) The
role of gibberellin, abscisic acid, and sucrose in the regulation of
potato tuber formation in vitro. Plant Physiol 117: 575­584
8 Jackson Plant Physiol. Vol. 119, 1999