Update on Cytokinins
Cytokinins. New Insights into a Classic Phytohormone
Georg Haberer and Joseph J. Kieber*
Department of Biology, University of North Carolina, Chapel Hill, North Carolina 27599–3280
Cytokinins were discovered in the search for factors
that promoted division of plant cells in culture.
Naturally occurring cytokinins are N6
-substituted
adenine derivatives that generally contain an isoprenoid
derivative side chain. These hormones influence
numerous aspects of plant development and physiology,
including seed germination, de-etiolation,
chloroplast differentiation, apical dominance, plantpathogen
interactions, flower and fruit development,
and leaf senescence. These processes are also influenced
by various other stimuli (e.g. light and other
phytohormones), and the physiological and developmental
outcomes reflect a highly integrated response
to these multiple stimuli. For example, the classical
reports of Skoog and Miller (1957) revealed that undifferentiated
callus cultures would form into roots
or shoots depending on the relative amount of cytokinins
and auxin in the medium; the ratio rather than
the absolute amount of these two hormones is critical.
A balanced ratio keeps the cells in an undifferentiated
state, while high cytokinin to auxin ratios promote
shoot and low ratios promote root development.
In the last decade, genetic and molecular analysis
of mutants has provided valuable insights into the
molecular mechanisms underlying the action of other
phytohormones. Cytokinin has lagged behind in
such studies, but several exciting recent reports show
that this is now changing. This review focuses on the
dramatic recent progress in understanding cytokinin
metabolism, perception, and signaling. The analysis
of all three aspects is required to understand the
different effects of cytokinin on plant physiology and
its role as a developmental signal.
BIOSYNTHESIS AND METABOLISM
Cytokinins occur as a bound form in the tRNA of
most organisms, including plants, but plants also
possess significant amounts of free cytokinins.
Isoprenoid-type cytokinins are the most abundant,
but several plant species contain adenine derivatives
with aromatic substituents. In addition, there are
synthetic cytokinins derived from diphenylurea
(DPU) that are structurally unrelated to the adeninetype
cytokinins.
The structure and conformation of the side chain
are critical to the activity of the respective cytokinin.
For example, one of the most abundant cytokinins in
higher plants, trans-zeatin, displays a high cytokinin
activity in bioassays, but the cis-isomer possesses a
significantly lower activity.
In general, a given plant tissue contains several
types of cytokinins and their modified forms. The
distribution of the various cytokinins differs significantly
between plant species. The phenotypic analysis
of biosynthetic mutants could reveal the roles of
specific cytokinin forms, but unfortunately neither
general nor specific mutants with reduced cytokinin
content have been isolated, and the overexpression of
biosynthetic genes generally affects the whole pool of
endogenous cytokinins. Furthermore, it is difficult to
quantify cytokinins in vivo, especially with respect to
their distribution within different plant tissues or
organs. Our knowledge about the in vivo role of
individual cytokinins and their modified forms is
therefore very limited.
BIOSYNTHESIS
The breakdown of tRNA was originally suggested
as a possible mechanism for cytokinin synthesis. The
released cis-zeatin could subsequently be converted
to active trans-zeatin by the enzyme cis-transisomerase
(Mok and Mok, 2001). However, the slow
turnover rate of tRNA is not sufficient to account for
the amount of cytokinins present in plants.
Enzymatic activity that converts AMP and dimethylallyl-diphosphate
(DMAPP) to the active
cytokinin isopentenyladenosine-5Ј-monophosphate
(iPMP) was firstly identified in Dictyostelium discoideum
(Taya et al., 1978). Subsequently, the tmr gene
(later designated as ipt) from Agrobacterium tumefaciens,
which results in root-like tumors when mutated,
was shown to encode an enzyme with similar
activity (Akiyoshi et al., 1984). ipt genes have also
been identified in several other bacteria, and IPT
activity was detected in crude extracts from a variety
of plant tissues, but the plant enzymes were not purified
and the corresponding genes were not cloned.
The completion of the Arabidopsis genomic sequence
enabled Takei et al. (2001a) and Kakimoto
(2001a) in independent studies to identify a total of
nine ipt-homologs, designated as AtIPT1 to AtIPT9,
by an in silico analysis. A phylogenetic analysis suggested
that AtIPT2 and AtIPT9 encode a putative
tRNA-ipt while the other seven AtIPTs formed a
1
This work was supported by the National Science Foundation
(grant no. MCB–9996354 to J.J.K.).
* Corresponding author; e-mail jkieber@unc.edu; fax 919–
962–1625.
www.plantphysiol.org/cgi/doi/10.1104/pp.010773.
354 Plant Physiology, February 2002, Vol. 128, pp. 354–362, www.plantphysiol.org © 2002 American Society of Plant Biologists
distinct clade more closely related to the bacterial
ipt/tmr gene. The expression of these seven genes in
Escherichia coli resulted in the secretion of the cytokinins
iP and zeatin, confirming that they encode cytokinin
biosynthetic genes (Takei et al., 2001a). Additionally,
calli overexpressing AtIPT4 under the
control of the Cauliflower mosaic virus (CaMV) 35S
promoter showed shoot regeneration even in the absence
of cytokinin, while CaMV 35S::AtIPT2 calli
were still dependent on cytokinin (Kakimoto, 2001a).
Surprisingly, unlike the bacterial IPT enzymes, purified
AtIPT4 utilized ATP and ADP preferentially
over AMP as a substrate (Kakimoto, 2001). The product
of the plant enzyme is likely to be isopentenyladenosine-5Ј-triphosphate
(iPTP) and isopentenyladenosine-5Ј-diphosphate
(iPDP), which can be
subsequently interconverted to zeatin (Fig. 1). Several
of the AtIPT genes display distinct, tissuespecific
patterns of expression, perhaps providing
new insights into the sites of cytokinin production
(Kakimoto, 2001b).
Another study in Arabidopsis indicates that an alternative
cytokinin biosynthesis pathway exists in
plants (Åstot et al., 2000). The authors compared the
biosynthetic rate of zeatin riboside-5Ј-monophosphate
(ZMP) and iPMP in wild type and transgenic plants
designed to inducibly overexpress the bacterial ipt
gene. iPMP is the direct product of the transfer of
DMAPP to AMP, and it can be converted to ZMP by
an endogenous hydroxylase activity. In vivo deuterium
labeling revealed a 66-fold higher biosynthetic
rate of ZMP than that of iPMP, the product of IPT. By
a feeding experiment with two tracers, which allowed
the simultaneous determination of iPMP-hydroxylase
activity and the de novo synthesis of ZMP, it was
shown that the major precursor for ZMP was not
cytoplasmic iPMP. The authors suggested the presence
of an iPMP-independent pathway, in which ZMP
Figure 1. Proposed biosynthetic and metabolic pathway for cytokinins. Left, The proposed biosynthesis of zeatin tri-/
diphosphate in Arabidopsis. Both ADP and ATP are likely substrates for the plant IPT enzyme, and these and their di- and
triphosphate derivatives are indicted together (e.g. ATP/ADP). The biosynthesis of cytokinins in bacteria (e.g. A. tumefaciens)
is compared next to it. Right, Several possible modifications and the degradation of zeatin. The diagram only depicts
reactions that are described in the text; cytokinin metabolism is more complex than the pathways shown (see Mok and Mok,
2001). See text for more details.
Cytokinins. New Insights into a Classic Phytohormone
Plant Physiol. Vol. 128, 2002 355
is directly synthesized by IPT from AMP and a yet
unidentified side chain precursor. This pathway could
be strongly inhibited by mevastatin, indicating a terpenoid
origin of the hypothetical precursor.
METABOLISM
Almost all cytokinins are present in plants as both
the free base and the corresponding nucleosides and
nucleotides. The interconversions among these forms
are likely to be performed by enzymes involved in
the general purine metabolism, but their biological
significance is unclear (Mok and Mok, 2001). The
adenine ring system can be glucosylated at the N3
-,
N7
-, and N9
-position. Glucosyl-conjugates at the
N7
- and N9
- but not at the N3
-position are usually
inactive in bioassays. It is assumed that the N7
- and
N9
- modifications irreversibly inactivate cytokinins,
but the precise in vivo function of these N-glucosylconjugates
is unknown (Mok and Mok, 2001).
O-Glucosyl-conjugates of the N6
-side chain are a
common modification in all plants. In addition,
O-xylosyl-conjugates have been reported in Phaseolus
spp. (Mok and Mok, 2001). The purified corresponding
glycosyltransferases of Phaseolus spp. exhibited a
high substrate specificity both to the cytokinins
zeatin and dihydrozeatin and to the glycosyl donor
(Dixon et al., 1989). The zeatin-O-glucosyltransferase
(ZOG1) strongly preferred UDP-Glc while the zeatinO-xylosyltransferase
(ZOX1) exclusively utilized
UDP-Xyl. The two genes, ZOG1 and ZOX1, have
been isolated from Phaseolus spp. and the binding site
of the UDP-Glc has been mapped by domain swap
experiments (Martin et al., 1999a, 1999b; Mok and
Mok, 2001). Recently the isolation of a cis-zeatin-Oglucosyltransferase
(cis-ZOG1) has been reported
that specifically glucosylates cis-zeatin but not transzeatin
or dihydrozeatin (Martin et al., 2001). The
authors suggested that the high specificity of cisZOG1
to the cis-isomer may indicate an important
role for cis-zeatin in plants.
O-Glycosylated cytokinins are resistant to the
cleavage of the N6
-side chain by cytokinin oxidases
(see below). Additionally, these forms can easily be
converted into active cytokinins by ␤-glucosidases
(Brzobohaty et al., 1993). Thus, it is believed that
O-glycosylated cytokinins are inactive, stable storage
forms that play an important role in balancing cytokinin
levels.
DEGRADATION
Free cytokinin bases and nucleosides with an unsaturated
N6
-side chain are irreversibly degraded by
cleavage of the side chain by cytokinin oxidases.
Cytokinin oxidases are found in many plant species,
and they represent a class of highly diverse proteins
(Jones and Schreiber, 1997). The genes of two highly
similar cytokinin oxidases have been isolated from
maize (Zea mays; Houba-He´rin et al., 1999; Morris et
al., 1999). Heterologous expression in Pichia pastoris
and Physcomitrella patens confirmed the predicted cytokinin
oxidase activity. Maize cytokinin oxidase and
its Arabidopsis homologs contain a signal sequence
for the secretory pathway, and they may be therefore
extracellular proteins, but no in vivo localization
studies have yet been reported.
Werner et al. (2001) have recently constructed
transgenic tobacco (Nicotiana tabacum) expressing
four Arabidopsis homologs of the maize cytokinin
oxidase from the CaMV 35S promoter. For the first
time, this allowed the phenotypic consequences of
reduced endogenous cytokinin levels to be determined.
Transgenic lines exhibited a higher cytokinin
oxidase activity and significantly reduced amounts
of iP and zeatin metabolites, including their glycosides
as compared with wild type. The transgenic
lines showed a dwarfed growth as a consequence of
a severely retarded shoot development. In contrast,
the growth of the root system was enhanced, indicating
that cytokinin may have opposing roles in shoot
and root development. These defects could be traced
back to alterations in the cell number and the rate of
cell division in the apical meristems. Shoot apical
meristems consisted of fewer cells with sizes comparable
to wild type. Leaves were formed from a significantly
decreased number of cells, which was
partly compensated for by an increased cell size.
Nevertheless, the transgenic leaves were about 15%
of the size of their wild-type counterparts. The reduced
endogenous cytokinin levels resulted in alterations
in cell proliferation in the apical meristems,
consistent with an in vivo role for cytokinins in the
regulation of cell division. In contrast, there was an
increased number and size of cells in the root apical
meristem, which contributed to a general enlargement
of the root.
TRANSPORT
Based on the occurrence of cytokinins in the xylem
sap and the identification of the root tip as a major
site of cytokinin biosynthesis, it is generally assumed
that cytokinins are transported in the xylem to exert
their effects on the aerial parts of plants. A purine
transporter, AtPUP1, has been isolated in Arabidopsis
by the functional complementation of a yeast
mutant deficient in adenine uptake (Gillissen et al.,
2000). Adenine uptake was competitively inhibited
by free cytokinin bases. Although no direct evidence
for cytokinin transport was presented, this competitive
inhibition suggests that AtPUP1 represents a
possible cytokinin/purine transporter involved in
the translocation of root-derived cytokinins to the
aerial plant parts.
Insight into the biological role of cytokinin transport
was provided by feeding roots of nitrogendepleted
maize with nitrate (Takei et al., 2001b). In
Haberer and Kieber
356 Plant Physiol. Vol. 128, 2002
response to the applied nitrate, cytokinin accumulated
first in the roots, subsequently in the xylem sap,
and finally in leaves. This suggests that in response to
nitrate, these plants made cytokinins in the root,
which was subsequently transported into the aerial
portion of the plant through the xylem. Thus, cytokinin
may represent a long-distance signal for nitrogen
availability from the root to the shoot, presumably
to coordinate shoot and root development.
Somewhat conflicting results with respect to the
biological significance of transported cytokinin were
obtained from reciprocal grafting experiments with
wild-type and cytokinin-overproducing tobacco
plants (Faiss et al., 1997). The phenotypic effects of
elevated cytokinin were restricted to the part of the
plant that was derived from the ipt overexpressing
mutant. Elevated level of cytokinins in the root led to
only a slight increase in cytokinin levels in the xylem
and had no phenotypic consequences in the scion.
Thus, it was concluded that cytokinin may act as
paracrine signal, at least with respect to apical dominance
and leaf senescence.
PERCEPTION AND SIGNAL TRANSDUCTION
The concept of cytokinin as a phytohormone implies
that the site of synthesis and the site of action
can spatially be separated. Thus, the search for receptors
for cytokinin is almost as old as its discovery.
Biochemical approaches led to the isolation of cytokinin
binding proteins, but none has been unequivocally
shown to be a receptor. The report of a
G-protein-coupled receptor, GCR1, as a candidate for
a cytokinin receptor has recently been retracted
(Kanyuka et al., 2001). In an activation-tagging mutagenesis,
a His kinase (CKI1) was identified by its
ability to confer cytokinin-independent callus growth
when overexpressed in Arabidopsis (Kakimoto,
1996). It was proposed that the elevated basal level of
CKI1 overcomes the cytokinin requirement for callus
growth and greening, and that CKI1 represents a
putative cytokinin receptor. Transient expression of
CKI1 in Arabidopsis protoplasts indeed showed activation
of the promoter of cytokinin primary response
gene (Hwang and Sheen, 2001). However, this
expression of ARR6 was constitutive, not responsive
to exogenous cytokinin. The constitutive induction of
ARR6 by CKI1 in this system might reflect cross talk
between CKI1 and the cytokinin signaling pathway
due to the overexpression of CKI1, and may not
reflect the situation in vivo. Although the phenotype
of the CKI1 overexpressing mutant implicates an important
role for CKI1 or a similar His kinase in the
regulation of cell proliferation and/or division, the
nature of the mutation (i.e. gain-of-function) does not
allow a definitive conclusion regarding the wild-type
function of the protein.
The involvement of a His kinase in cytokinin signaling
has recently been confirmed by two groups
who have reported the identification of a cytokinin
receptor in Arabidopsis. This gene, CRE1, encodes a
His kinase that shows weak similarity to the amino
acid sequence of CKI1 (Inoue et al., 2001; Suzuki et
al., 2001a). A screen of 19,000 ethyl methanesulfonate-mutagenized
Arabidopsis plants identified a
single mutant line that displayed reduced sensitivity
to cytokinin in tissue culture, including shoot formation
and proliferation of green callus (Inoue et al.,
2001). Additionally, cre1 mutant seedlings displayed
resistance to cytokinin in a root elongation assay. An
independently isolated T-DNA-tagged allele of cre1
showed a similar phenotype. Both mutations appear
to be partial loss-of-function alleles by molecular
analysis (see also below), but are semidominant in
the root elongation assay. This semidominance of a
hypomorphic allele implies haploinsufficiency, indicating
a dose-dependent action of the receptor in this
response. CRE1 encodes an unusual type of hybrid
kinase with two C-terminal response regulator domains.
In addition, CRE1 contains a His kinase domain
and two transmembrane regions flanking a
novel presumed extracellular domain. All identified
mutations affect the N-terminal part of the receptor,
and the functional relevance of the two response
regulator domains is unknown.
The phenotype of cre1 mutants and the homology
to His receptor kinases are consistent with a role in
cytokinin signaling, but compelling evidence that cytokinin
can indeed act as a ligand for CRE1 was
provided by a series of elegant experiments using
yeast and E. coli (Inoue et al., 2001; Suzuki et al.,
2001a). The experiments were based on the idea that
endogenous responses can be influenced by expression
of CRE1 in a suitable genetic background of a
heterologous system, and that these responses can be
altered in a cytokinin-dependent manner. In yeast,
the sensor His kinase SLN1, which is involved in
osmosensing, inhibits the activity of SSK2 (a
mitogen-activated protein kinase kinase kinase) by a
phosphorelay consisting of SLN1 itself, the His transfer
protein YPD1, and the response regulator SSK1
(Fig. 2A; Posas et al., 1998). In wild-type yeast, unphosphorylated
SSK1 enables activation of SSK2 by
autophosphorylation. Under normal osmotic conditions,
this activation is inhibited by phosphorylation
of SSK1 by SLN1 and YPD. Deletion of SLN1 is lethal
due to the overactivation of SSK2 by unphosphorylated
SSK1. Viability can be restored by the expression
of the phosphatase PTP2 that inhibits SSK2.
Thus, a yeast strain deleted in SLN1 and carrying the
PTP2- gene under the control of a Gal-inducible promoter
is conditionally lethal, allowing growth only in
the presence of Gal. In the absence of Gal, the expression
of CRE1 suppressed the conditional lethality,
but only when cytokinins were present in the medium.
The cytokinin-dependent rescue by CRE1 required
YPD1, indicating that CRE1 acts through the
endogenous yeast phosphotransfer mechanism. FurCytokinins.
New Insights into a Classic Phytohormone
Plant Physiol. Vol. 128, 2002 357
thermore, only cytokinins that exhibit high activity in
bioassays, such as trans-zeatin, or benzyladenine,
could suppress lethality on Glc-containing media; the
weakly active cis-zeatin was ineffective (Inoue et al.,
2001). Interestingly, thidiazuron, a N,NЈ-diphenylurea
(DPU)-type cytokinin, also activates CRE1 in this
assay. DPU-type cytokinins strongly inhibit cytokinin
oxidase activity in vitro, and this has been postulated
as their mode of action in vivo (Jones and
Schreiber, 1997). However, the activity of thidiazuron
in the yeast assay indicates that DPU-type cytokinins
likely also exert their in vivo effects by direct stimulation
of the receptor. This is quite remarkable given
the profound structural differences between DPUand
zeatin-type cytokinins. Recent results from heterologous
expression of CRE1 in yeast have confirmed
that cytokinin does bind directly to CRE1
with high affinity (Yamada et al., 2001).
What is the biological role of CRE1? Surprisingly,
cre1 has been shown to be allelic to the wooden leg
(wol) mutation, which causes defects in the development
of vascular tissue (Ma¨ho¨nen et al., 2000). wol
seedlings have a reduced number of cells in the vascular
cylinder of the root and the lower part of the
hypocotyl, and these consist almost entirely of xylem
cells (Scheres et al., 1995). Thus, phloem and metaxylem
are absent in these parts of wol. The defects in wol
can be traced back to the late torpedo stage of the
embryo, which has a reduced number of vascular
initials due to the lack of specific asymmetric cell
divisions (Scheres et al., 1995; Ma¨ho¨nen et al., 2000).
The absence of phloem and metaxylem cells is presumably
caused by reduced cell division rather than
aberrant cell specification because the cre1 mutant
phenotype can be suppressed by the fass mutation,
which has supernumerary cell layers (Scheres et al.,
1995). It was proposed that the determination of xylem
cells precedes the specification of phloem cells
and that the reduced number of initials are completely
used up as xylem precursors in wol, leaving
no initials left to specify phloem (Scheres et al., 1995).
This model of the wol phenotype nicely brings together
the role of CRE1/WOL as a cytokinin receptor
and the stimulatory effects of cytokinin on cell division.
However, CRE1/WOL may also have additional
roles in the differentiation of vascular cells. wol
has a substantially stronger phenotype than the other
cre1 alleles (Ma¨ho¨nen et al., 2000; Inoue et al., 2001),
although it is unknown if it is a null allele. Interestingly,
the wol mutation is a missense mutation affecting
the extracellular domain of the protein, which is
the putative ligand-binding site (Ma¨ho¨nen et al.,
2000). In vitro binding assays have recently shown
that the wol mutation disrupts binding of cytokinin to
CRE1 (Yamada et al., 2001).
CRE1/WOL is expressed primarily in the root, and
in the embryo it is restricted to the provascular tissue
from the root to the shoulder region of the cotyledons
(Ma¨ho¨nen et al., 2000). This expression pattern may
explain the rather unexpected absence of a phenoFigure
2. A, Evidence that CRE1 is a cytokinin receptor. The left-most pathway depicts the osmosensing pathway in
wild-type yeast: the His kinase SLN1, which suppresses the activity of SSK2 (a mitogen-activated protein kinase kinase
kinase) activity via a phosphorelay consisting of YPD1 (an Hpt) and SSK1 (a response regulator). A deletion mutant of SLN1
is lethal due to overactivation of SSK2. CRE1 can suppress the growth defect in an SLN1 deletion only in the presence of
cytokinins (two right pathways). B, Model of cytokinin signaling in Arabidopsis. Cytokinin binds to the N-terminal domain
of CRE1 (and likely other similar sensor kinase) and activates its His kinase activity. CRE1 phosphorylate the AHPs, which
in turn transfer the phosphate to the receiver domain of ARR1 (and presumably to other type-B ARRs), thus activating their
output (transcriptional activator) domain. Type-A ARRs (and possibly other primary target genes) are transcriptionally
induced by the activated type-B ARRs. The type-A ARRs also interact with the AHPs, and are also likely phosphorylated. The
activated type-A ARRs, perhaps in parallel and/or in combination with the activated type-B proteins, interact with various
effectors to alter cellular function, including the a feedback inhibition of their own expression. The curved arrows indicate
phosphotransfer. CK, Cytokinins. See text for additional details.
Haberer and Kieber
358 Plant Physiol. Vol. 128, 2002
type in the aerial portions of wol mutant seedlings
and adults. In contrast to the lack of a shoot phenotype
in intact plants, cre1 mutants show a strong
reduction in the regeneration of shoots from tissue
culture cells in response to cytokinin. It may be that
the formation of a shoot in tissue culture requires the
elaboration of a proper vasculature. In this context, it
is interesting that regenerated roots from wol callus
exhibit identical phenotypic aberrations like roots of
wol seedlings (Scheres et al., 1995). Alternatively,
CRE1 may be redundant with other genes in the
shoots of intact plants, but not with respect to shoot
formation in vitro. Indeed, there are two closely related
His kinases (AHK2 and AHK3) in Arabidopsis
that are mainly expressed in the aerial parts of wildtype
plants (Ueguchi et al., 2001). Both His kinases
are highly homologous to the entire CRE1 sequence,
including the putative ligand binding domain, and
also share with CRE1 the presence of two response
regulator domains in their C terminus. Furthermore,
in a mesophyll protoplast transient expression system,
both His kinases may regulate the promoter of a
primary cytokinin response gene in a cytokinindependent
manner (Hwang and Sheen, 2001; see also
below). These results strongly suggest that AHK2
and AHK3 indeed act as cytokinin receptors in
Arabidopsis.
The involvement of a hybrid His kinase in cytokinin
signaling suggests that additional homologs of
bacterial phosphorelay elements may be involved in
the cytokinin signal transduction pathway. Genes
encoding homologs of these other components,
phospho-His transfer proteins and response regulators,
have been identified in Arabidopsis. The possible
role of several of these genes in cytokinin signaling
has been addressed by numerous groups.
Recently, for example, the induction of the promoter
of a cytokinin primary response gene, the type-A
ARR6 (see below), was monitored by a reporter in
Arabidopsis mesophyll protoplasts that transiently
expressed wild-type and mutated forms of these
genes (Hwang and Sheen, 2001). The results, along
with numerous findings of other groups, provide
important and valuable insights into the putative
hardwiring of cytokinin signaling components. The
data for each of the three types of components acting
downstream of the receptors will first be discussed
separately, then we will integrate the information
about these components into a working hypothesis of
cytokinin signal transduction.
There are five Arabidopsis genes encoding putative
His phosphotransfer proteins (AHPs), which function
as a bridge in the primary phosphotransfer between
the Asp residues present in the receiver domain
of the hybrid sensor kinase and the receiver
domain of the response regulator (Suzuki et al.,
2000). Suzuki et al. (2001a) established a heterologous
system in E. coli analogous to the above-described
yeast system to verify that CRE1 could act as a cytokinin
receptor, and to implicate the AHPs as downstream
targets of CRE1. The His phosphotransfer
protein in E. coli used in this analysis was YojN, and
the cognate response regulator was RcsB. The activation
of RcsB by CRE1 could be quantified by a RcsBresponsive
promoter fused to the reporter LacZ. Coexpression
of four different AHPs differentially
inhibited the reporter expression. It was proposed
that the coexpressed AHPs competed with the endogenous
YojN for phosphotransfer by CRE1, consistent
with an interaction between that CRE1 and
these AHPs. The different effectiveness of specific
AHPs might reflect specificity of the proposed CRE1AHP
interactions. However, the individual expression
levels of the four AHPs in this system were not
determined, and thus quantitative measurements
should be interpreted carefully. Nuclear localization
studies also support the hypothesis that at least some
of the AHPs are involved in cytokinin signaling
(Hwang and Sheen, 2001). It was shown that AHP1
and AHP2, but not AHP5, transiently translocate
from the cytoplasm to the nucleus of Arabidopsis
protoplasts in response to exogenously applied transzeatin.
Finally, several groups have reported yeast
two-hybrid interactions between His kinases and
AHPs. Taken together, these reports indicate that at
least a subset of AHPs may be involved in cytokinin
signaling. These AHPs likely transmit the signal from
its site of perception, probably the plasma membrane,
to the effectors in the nucleus. Numerous results
indicate that these effectors consist, at least in
part, of the third class of components of phosphorelays,
the response regulators.
The response regulators in Arabidopsis (ARRs)
form a gene family that includes the type-A and the
type-B ARRs (D’Agostino and Kieber, 1999; Imamura
et al., 1999). The type-A ARRs are composed solely of
a receiver domain; the type-B ARRs have an output
domain fused to the receiver. This C-terminal output
domain has been demonstrated to activate transcription
in yeast and in planta and to bind to DNA in a
sequence-specific manner. The type-B ARRs have
been shown to be nuclear-localized proteins (Lohrmann
et al., 1999; Sakai et al., 2000; Hwang and
Sheen, 2001). The steady-state levels of mRNA of
most of the type-A, but not the type-B ARRs are
rapidly and specifically elevated in response to cytokinin
(Brandstatter and Kieber, 1998; Taniguchi et al.,
1998; Kiba et al., 1999; D’Agostino et al., 2000). Homologous
response regulators in maize exhibit very
similar induction characteristics indicating that the
function of type-A ARRs is highly conserved in
plants (Sakakibara et al., 1998). The induction by
cytokinin of several of the type-A ARRs is due, at
least in part, to the transcriptional activation of these
genes and is not blocked by cycloheximide, which
indicates that these are primary response genes
(Brandstatter and Kieber, 1998; D’Agostino et al.,
2000). The similarity to bacterial signaling elements
Cytokinins. New Insights into a Classic Phytohormone
Plant Physiol. Vol. 128, 2002 359
predicted to act downstream of sensor His kinases,
combined with the rapid induction by cytokinin, suggest
that these type-A ARRs may play a role in cytokinin
signaling. Consistent with this, the ARR5 gene
has been found to be expressed primarily in the shoot
and root apical meristems, likely sites of cytokinin
action (D’Agostino et al., 2000). Recently, studies in
protoplasts have shown that the overexpression of
type-A ARRs represses their own expression, suggesting
that type-A ARRs may provide a negative
feedback regulation of cytokinin signaling (Hwang
and Sheen, 2001). Interestingly, mutant forms of
these genes lacking a conserved Asp, the predicted
phosphoaccepting site, were equally effective in their
feedback inhibition. Thus, wild-type and mutant proteins
may interact with cytokinin signaling elements,
such as the AHPs, making these components inaccessible
to the pathway. Type-A ARRs may also interact
with additional partners and could have dual functions.
Their activation by cytokinin may negatively
feedback into cytokinin signaling, but would also
trigger yet unknown, secondary responses to cytokinin.
This model may also explain why there are 10
type-A ARRs in Arabidopsis.
In two recent publications, the transcriptional control
of the type-A ARRs was shown to be mediated, at
least partly, by type-B ARRs (Sakai et al., 2001;
Hwang and Sheen, 2001). An arr1 (a type-B ARR)
knockout mutant was found to be partially resistant
to cytokinin in shoot regeneration and root elongation
assays, and an ARR1 overexpressing transgenic
line displayed increased sensitivity to cytokinin in
these assays. These results suggest that this type-B
gene plays a role in cytokinin responsiveness. Examination
of the expression of the type-A genes in the
arr1 mutants revealed that the loss-of-function allele
exhibited reduced expression of several type-A ARR
transcripts and that overexpression of ARR1 resulted
in an enhanced induction of the type-A genes, even
in the presence of cycloheximide. These results, coupled
with the presence of multiple ARR1 binding
sites in the promoter of several type-A ARRs suggest
that ARR1 directly regulates the transcription of
these primary targets of cytokinin signaling (Sakai et
al., 2001). The activity of the ARR1 output domain is
inhibited by the N-terminal response regulator domain
(Sakai et al., 2000). Thus, it was proposed that
the inhibition of the transcriptional activator domain
by the response regulator domain was modulated by
cytokinin. Similar results were independently obtained
by transient expression studies in Arabidopsis
protoplasts (Hwang and Sheen, 2001). Overexpression
of three type-B ARRs, ARR1, ARR2, and ARR10,
highly increased both the basal level and the induction
of the promoter of a type-A ARR, ARR6. ARR2
exhibited the strongest induction and was consequently
analyzed in greater detail. Expression of the
DNA binding domain of ARR2 without the receiver
and transactivator domain abolished the cytokinin
responsiveness of the ARR6 promoter, presumably
acting as a dominant negative form. The overexpression
of ARR2 in planta promoted rapid cell proliferation
and shoot and leaf formation in tissue cultures
even in the absence of exogenous cytokinin. The
transgenic plants also showed delayed leaf senescence
as expected for cytokinin overaction.
These results implicate at least several of the type-B
ARRs as positive regulators of cytokinin responses.
Surprisingly, mutated proteins in the transient assay
in which the Asp, the probable site of phosphorylation,
was altered to Asn, were equally effective in the
induction of the ARR6::reporter by cytokinin (Hwang
and Sheen, 2001). This raises the question of whether
cytokinin signaling acts as a classical His-to-Asp
phosphorelay, at least with regard to the induction of
the type-A ARR genes. The authors propose that an
endogenous repressor may be liberated by the overexpression
of the type-B ARRs. Several indirect arguments
support the phosphorelay hypothesis. The
conservation of the presumed phosphorylation sites
within all 10 type-A and about 12 type-B ARRs in
Arabidopsis as well as the conservation of these sites
between maize, rice (Oryza sativa), and Arabidopsis
homologs strongly suggests the functional relevance
of these sites. At least at the level of the receptor, the
phosphorelay hypothesis is also experimentally verified.
The cytokinin-dependent complementation in
yeast and E. coli requires an interaction of CRE1 with
endogenous His transfer proteins similar to their interaction
with their cognate His kinases. In addition,
mutant forms of CRE1 in which these sites have been
altered are unable to induce the ARR6::reporter in
protoplasts by application of cytokinin. Thus, it is
highly likely that CRE1 acts both in the transient
expression system and in the cytokinin-dependent
complementation in yeast and E. coli as a His kinase.
Biochemical studies and loss-of-function mutants of
the AHPs and ARRs should enable us to clarify the
significance of their presumed phosphorylation sites.
These results are consistent with the following
model of cytokinin action. Binding of cytokinin to
CRE1 (and possibly other related His kinases) initiates
a phosphorelay that culminates, via an AHP
intermediate, in the phosphorylation and activation
of a set of the type-B ARR(s). These then activate the
transcription of the type-A genes, which in turn negatively
feedback into the pathway. The type-A ARRs,
perhaps together and/or in parallel with the activated
type-B ARRs, interact with various effectors to
mediate the changes in cell function appropriate to
cytokinin (Fig. 2B). The near future should see significant
refinement of this model as the tools are now
in hand to analyze the interactions among these
elements.
CONCLUSION
The field of cytokinin research has seen dramatic
advancements during the last 2 years. Several of the
Haberer and Kieber
360 Plant Physiol. Vol. 128, 2002
key elements of cytokinin action and metabolism
have been identified, but many questions need to be
addressed. The in vivo role of most of the members of
the phosphorelay in Arabidopsis remains unknown,
including their putative function in cytokinin signaling.
The function and specificity of the interactions
among the various components in vivo represents
one of the key questions. The analysis of expression
patterns and the identification of mutants in these
genes will enable us to understand the numerous
effects of cytokinin on plant physiology and development
at the molecular level. Furthermore, these
genes can now be used to identify additional regulators
and downstream targets, which should help unveil
the complex network of cytokinin metabolism
and signaling.
LITERATURE CITED
Akiyoshi DE, Klee H, Amasino RM, Nester EW, Gordon
PM (1984) T-DNA of Agrobacterium tumefaciens encodes
an enzyme of cytokinin biosynthesis. Proc Natl Acad Sci
USA 81: 5994–5998
Åstot C, Dolezal K, Nordstro¨m A, Wang Q, Kunkel T,
Moritz T, Chua N-H, Sandberg G (2000) An alternative
cytokinin biosynthesis pathway. Proc Natl Acad Sci USA
97: 14778–14783
Brandstatter I, Kieber JJ (1998) Two genes with similarity
to bacterial response regulators are rapidly and specifically
induced by cytokinin in Arabidopsis. Plant Cell 10:
1009–1020
Brzobohaty B, Moore I, Kristoffersen P, Bako L, Campos
N, Schell J, Palme K (1993) Release of active cytokinin by
a ␤-glucosidase localized to the maize root meristem.
Science 262: 1051–1054
D’Agostino IB, Deruere J, Kieber JJ (2000) Characterization
of the response of the Arabidopsis response regulator
gene family to cytokinin. Plant Physiol 124: 1706–1717
D’Agostino IB, Kieber JJ (1999) Phosphorelay signal transduction:
the emerging family of plant response regulators.
Trends Biochem Sci 24: 452–456
Dixon SC, Martin RC, Mok MC, Shaw G, Mok DWS
(1989) Zeatin glycosylation enzymes in Phaseolus: isolation
of O-glucosyltransferase from P. lunatus and comparison
to O-xylosyltransferase from P. vulgaris. Plant
Physiol 90: 1316–1321
Faiss M, Zalubilova J, Strnad M, Schmulling T (1997)
Conditional transgenic expression of the ipt gene indicates
a function for cytokinins in paracrine signaling in
whole tobacco plants. Plant J 12: 401–415
Gillissen B, Burkle L, Andre B, Kuhn C, Rentsch D,
Brandl B, Frommer WB (2000) A new family of highaffinity
transporters for adenine, cytosine, and purine
derivatives in Arabidopsis. Plant Cell 12: 291–300
Houba-He´rin N, Pethe C, d’Alayer J, Laloue M (1999)
Cytokinin oxidase from Zea mays: purification, cDNA
cloning and expression in moss protoplasts. Plant J 17:
615–626
Hwang I, Sheen J (2001) Two-component circuitry in Arabidopsis
cytokinin signal transduction. Nature 413:
383–389
Imamura A, Hanaki N, Nakamura A, Suzuki T, Taniguchi
M, Kiba T, Ueguchi C, Sugiyama T, Mizuno T (1999)
Compilation and characterization of Arabidopsis thaliana
response regulators implicated in His-Asp phosphorelay
signal transduction. Plant Cell Physiol 40: 733–742
Inoue T, Higuchi M, Hashimoto Y, Seki M, Kobayashi M,
Kato T, Tabata S, Shinozaki K, Kakimoto T (2001)
Identification of CRE1 as a cytokinin receptor from Arabidopsis.
Nature 409: 1060–1063
Jones RJ, Schreiber BMN (1997) Role and function of
cytokinin oxidases in plants. Plant Growth Regul 23:
122–134
Kakimoto T (1996) CKI1, a histidine kinase homolog implicated
in cytokinin signal transduction. Science 274:
982–985
Kakimoto T (2001a) Identification of plant cytokinin biosynthetic
enzymes as dimethylallyl diphosphate:ATP/
ADP isopentenyltransferases. Plant Cell Physiol 42:
677–685
Kakimoto T (2001b) Biosynthesis and perception of cytokinins
(abstract). In 17th International Conference on
Plant Growth Substances Brno, Czech Republic, July 1–6,
2001, p 56
Kanyuka K, Couch D, Hooley R (2001) A higher plant
seven-transmembrane receptor that influences sensitivity
to cytokinins. Curr Biol 11: 535
Kiba T, Taniguchi M, Imamura A, Ueguchi C, Mizuno T,
Sugiyama T (1999) Differential expression of genes for
response regulators in response to cytokinins and nitrate
in Arabidopsis thaliana. Plant Cell Physiol 40: 767–771
Lohrmann J, Buchholz G, Keitel C, Sweere C, Kircher S,
Ba¨urle I, Kudla J, Harter K (1999) Differentiallyexpressed
and nuclear-localized response regulator-like
proteins from Arabidopsis thaliana with transcription factor
properties. J Plant Biol 1: 495–506
Ma¨ho¨nen AP, Bonke M, Kaupinnen L, Riikonen M, Benfey
PN, Helariutta Y (2000) A novel two-component
hybrid molecule regulates vascular morphogenesis of the
Arabidopsis root. Genes Dev 14: 2938–2943
Martin RC, Mok MC, Habben JE, Mok DWS (2001) A
maize cytokinin gene encoding an O-glucosyltransferase
specific to cis-zeatin. Proc Natl Acad Sci USA 98:
5922–5926
Martin RC, Mok C, Mok DWS (1999a) A gene encoding
the cytokinin enzyme zeatin O-xylosyltransferase of
Phaseolus vulgaris. Plant Physiol 120: 553–557
Martin RC, Mok C, Mok DWS (1999b) Isolation of a
cytokinin gene, ZOG1, encoding zeatin O-glucosyltransferase
of P. lunatis. Proc Natl Acad Sci USA 96:
284–289
Mok DW, Mok MC (2001) Cytokinin metabolism and action.
Annu Rev Plant Physiol Plant Mol Biol 52: 89–118
Morris RO, Bilyeu KD, Laskey JG, Cheikh NN (1999)
Isolation of a gene encoding a glycosylated cytokinin
oxidase from maize. Biochem Biophys Res Commun 255:
328–333
Cytokinins. New Insights into a Classic Phytohormone
Plant Physiol. Vol. 128, 2002 361
Posas F, Takekawa M, Saito H (1998) Signal transduction
by MAP kinase cascades in budding yeast. Curr Opin
Microbiol 1: 175–182
Sakai H, Aoyama T, Oka A (2000) Arabidopsis ARR1 and
ARR2 response regulators operate as transcriptional activators.
Plant J 24: 703–711
Sakai H, Honma T, Aoyama T, Sato S, Kato T, Tabata S,
Oka A (2001) ARR1, a transcription factor for genes
immediately responsive to cytokinins. Science 294:
1519–1521
Sakakibara H, Suzuki M, Takei K, Deji A, Taniguchi M,
Sugiyama T (1998) A response-regulator homologue
possibly involved in nitrogen signal transduction mediated
by cytokinin in maize. Plant J 14: 337–344
Scheres B, DiLaurenzio L, Willemsen V, Hauser M-T,
Janmaat K, Weisbeek P, Benfey PN (1995) Mutations
affecting the radial organization of the Arabidopsis root
display specific defects throughout the embryonic axis.
Development 121: 53–62
Skoog F, Miller CO (1957) Chemical regulation of growth
and organ formation in plant tissues cultured in vitro.
Symp Soc Exp Biol 11: 118–131
Suzuki T, Miwa K, Ishikawa K, Yamada H, Aiba H,
Mizuno T (2001a) The Arabidopsis sensor kinase, AHK4,
can respond to cytokinin. Plant Cell Physiol 42: 107–113
Suzuki T, Sakurai K, Imamura A, Nakamura A, Ueguchi
C, Mizuno T (2000) Compilation and characterization of
histidine-containing phosphotransmitters implicated in
His-to-Asp phosphorelay in plants: AHP signal transducers
of Arabidopsis thaliana. Biosci Biotechnol Biochem
64: 2486–2489
Suzuki T, Sakurai K, Ueguchi C, Mizuno T (2001b) Two
types of putative nuclear factors that physically interact
histidine-containing phosphotransfer (Hpt) domains,
signaling mediators in His-to-Asp phosphorelay, in Arabidopsis
thaliana. Plant Cell Physiol 42: 37–45
Takei K, Sakakibara H, Sugiyama T (2001a) Identification
of genes encoding adenylate isopentenyltransferase, a
cytokinin biosynthesis enzyme, in Arabidopsis thaliana.
J Biol Chem 276: 26405–26410
Takei K, Sakakibara H, Taniguchi M, Sugiyama T (2001b)
Nitrogen-dependent accumulation of cytokinins in root
and the translocation to leaf: implication of cytokinin
species that induces gene expression of maize response
regulator. Plant Cell Physiol 42: 85–93
Taniguchi M, Kiba T, Sakakibara H, Ueguchi C, Mizuno
T, Sugiyama T (1998) Expression of Arabidopsis response
regulator homologs is induced by cytokinins and nitrate.
FEBS Lett 429: 259–262
Taya Y, Tanaka Y, Nishimura S (1978) 5Ј-AMP is a direct
precursor of cytokinin in Dictyostelium discoidum. Nature
271: 545–547
Ueguchi C, Koizumi H, Suzuki T, Mizuno T (2001) Novel
family of sensor histidine kinase genes in Arabidopsis
thaliana. Plant Cell Physiol 42: 231–235
Yamada H, Suzuki T, Terada K, Takei K, Ishikawa K,
Miwa K, Mizuno T (2001). The Arabidopsis AHK4 histidine
kinase is a cytokinin-binding receptor that transduces
cytokinin signals across the membrane. Plant Cell
Physiol 42: 1017–1023
Werner T, Motyka V, Strnad M, Schmu¨lling T (2001)
Regulation of plant growth by cytokinin. Proc Natl Acad
Sci USA 98: 10487–10492
Haberer and Kieber
362 Plant Physiol. Vol. 128, 2002