Tansley Review No. 106

Printed from the CJO service for personal use only by...

New Phytol. (1999), 143, 427–455

Tansley Review No. 106

Cyclic nucleotides in higher plants: the

enduring paradox

RUSSELL P. NEWTON "*, LUC ROEF # , ERWIN WITTERS #

HARRY VAN ONCKELEN #

"Biochemistry Group, School of Biological Sciences, University of Wales Swansea,

Singleton Park, Swansea SA2 8PP, UK

#Laboratorium voor Plantenbiochemie en -fysiologie, Department of Biology,

Universiteit Antwerpen (UIA), Universiteitsplein 1, B-2610, Antwerp, Belgium

Received 23 October 1998; accepted 17 May 1999

Summary

I. 427

II. c

431

III.

c 432

IV. 435

V. 437

For three decades, hypotheses relating to the occurrence and function of cyclic nucleotides in higher plants have

been highly controversial. Although cyclic nucleotides had been shown to have key regulatory roles in animals and

bacteria, investigations with higher plants in the 1970s and early 1980s were criticized on the basis of (i) a lack of

specificity of effects apparently elicited by cyclic nucleotides, (ii) the equivocal identification of putative

endogenous cyclic nucleotides and (iii) ambiguity in the identification of enzymes connected with cyclic

nucleotide. More recent evidence based on more rigorous identification procedures has demonstrated conclusively

the presence of cyclic nucleotides, nucleotidyl cyclases and cyclic nucleotide phosphodiesterases in higher plants,

and has identified plant processes subject to regulation by cyclic nucleotides. Here we review the history of the

debate, the recent evidence establishing the presence of these compounds and their role ; future research objectives

are discussed.

I.

The hypothesis that adenosine 3«,5«-cyclic mono-

phosphate (cAMP) performs a regulatory and}or

signal transduction role in higher plants has been

variously described over the past three decades as

‘non-existent’, ‘unequivocally established’, ‘un-

*Author for correspondence (fax 44 1792 295447; e-mail r.p.newton!swansea.ac.uk).

Abbreviations: ATF, activating transcription factor; cAMP, adenosine 3«,5«-cyclic monophosphate; CBP, CREB-binding protein;

cdTMP, 2«-deoxythymidine 3«,5«-cyclic monophosphate; cCMP, cytidine 3«,5«-cyclic monophosphate; cGMP, guanosine 3«,5«-cyclic

monophosphate; CID, collision-induced dissociation; cIMP, inosine 3«,5«-cyclic monophosphate; cPKA and cPKB, cAMP-dependent

protein kinases; cUMP, uridine 3«,5«-cyclic monophosphate; CRE, cAMP-response element; CREB, cAMP-response-element-

binding protein; ESI, electrospray ionization; FAB, fast atom bombardment; G-protein, GTP-binding protein; Ins(1,4,5)P$, inositol

1,4,5-trisphosphate; MIKE, mass-analysed kinetic energy; MSMS, tandem MS; PAL, phenylalanine ammonia-lyase; VBP1, Vicia

faba DNA-binding protein.

VI. c- , ,

c- -

- 439

VII. c 442

VIII. c 445

IX. 447

References 449

likely’, ‘probable’ and ‘controversial ’. Few with

active interests in the relevant areas of plant

biochemistry and physiology have retained a neutral

or indifferent viewpoint on the concept. A recent

critique (Trewavas, 1997) described believers, of

which company we have long been members, ‘who

reported meaningful micromolar concentrations,


428 R. P. Newton et al.

Stimulatory hormone/neurotransmitter

Cellmembrane

RsGs Adenylyl

cyclase

Gi

Inhibitory hormone/neurotransmitter

Ionchannel

M+(+)ATP cAMP

PDE

AMP

R – R

C – CA-PK

R – RProtein

Protein – PO4

C C

Cytoplasm

NucleusCCREB

CREMCREB – PO4CREM – PO4

CREActivation ofcAMP-induced genes

Ri

Fig. 1. Role of cyclic AMP in mammals. Diagram of mechanisms of action of cAMP in the mammalian cell.

Molecular conversions are represented by solid arrows, regulatory effects by dotted arrows. Rs/i

, receptor; Gs/i

,

G-protein; PDE, phosphodiesterase; A-PK, cAMP-dependent protein kinase; R-R and C-C, regulatory and

catalytic subunits ; CREB and CREM, cAMP-responsive elements.

albeit with weaker technology’ being ‘few in num-

ber’ and making ‘occasional forays to semi-respect-

able journals ’. The objective of the present review is

to survey and discuss the evidence and to attempt to

convert a majority of the sceptics to acceptance of the

existence in higher plants of cyclic nucleotides and

enzymes connected with them, and the likelihood

that cyclic nucleotides function in signal transduction

and regulation in plants.

cAMP has been established as a signalling mol-

ecule in both eukaryotes and prokaryotes, including

lower plants, thus posing the question ‘why should

higher plants be different?’. Before attempting to

answer this we must first consider the mechanisms of

action of cAMP in these other organisms. From the

initial discovery of cAMP by the Nobel Prize winner

Earl Sutherland (Rall et al., 1957), and the sub-

sequent demonstration of its role in mediating the

action of mammalian hormones on carbohydrate

metabolism in the liver, the secondary-messenger

concept developed. According to this concept,

mammalian hormones and neurotransmitters, acting

as primary messengers, remain outside the cell and

transmit their signal to the interior via receptors and

a membrane-sited enzyme system, which releases

the secondary messenger inside the cell. The

secondary-messenger concept has been extended

beyond cyclic nucleotides to include the action of

inositol phosphates and Ca#+. The original secon-

dary-messenger concept involving receptors directly

linked to a membrane-bound adenylyl cyclase has

been developed into the present model for the

mechanism of action of cAMP in eukaryotic signal

transduction depicted in Fig. 1.

Two sets of receptor units are associated with a

single adenylyl cyclase catalytic subunit in the

membrane. Both sets are of the ‘seven-pass’ struc-

ture, comprising a series of three-and-a-half loops

across the membrane: one set is stimulatory and is

designated Rs; the other is inhibitory and is desig-

nated Ri. The binding of a ligand to a specific

receptor induces a conformational change in the

receptor, enabling it to interact with a GTP-binding

protein (G-protein) and influence its activity. Two

forms of G-protein are also present: Gs, which

stimulates adenylyl cyclase, and Gi, which inhibits

that enzyme. Both undergo a cycle in which they

exist as heterotrimers, to which either GDP or no

guanosine nucleotide is bound. These heterotrimers

dissociate on binding GTP and the free Gsα subunit

undergoes a conformational change enabling it to

interact with and stimulate the catalytic unit of


Cyclic nucleotides in higher plants 429

Light

GTP Rhodopsintransducin

PDEGMPcGMP

lon channelsG-PK

Nitricoxide

Guanylin

Natriureticpeptides

Guanylylcyclase

Fig. 2. Role of cGMP in mammals. Diagram of mechanisms of action of cGMP in the mammalian cell. PDE,

phosphodiesterase; G-PK, cGMP-dependent protein kinase.

adenylyl cyclase. The dissociation of the Gshetero-

trimer is transient; after the hydrolysis of GTP to

GDP, reassociation takes place and a further dis-

sociation occurs only after the GDP has been

replaced by GTP. The dissociated Giα subunit

exerts an inhibitory effect on adenylyl cyclase

(Taussig et al., 1993; Taussig & Gilman, 1995);

however, some types of adenylyl cyclase are not

sensitive to Giα. Adenylyl cyclase I is inhibited by

both Giα and Gβγ ; others (types V and VI) are

stimulated by the βγ subunit of Gi

(Taussig &

Gilman, 1995).

After stimulation, adenylyl cyclase, already func-

tioning at a basal level in the absence of agonist,

catalyses the conversion of ATP to cAMP, which is

then released into the interior of the cell. The cAMP

signal is switched off by another set of enzymes, the

phosphodiesterases, which hydrolyse cAMP to

AMP. Compared with the enormous research effort

concentrated on the adenylyl cyclase system, little

interest was shown initially in these enzymes.

However, more recently they also have been found to

be subject to tight regulation; a number of phos-

phodiesterase ‘families’, varying in substrate speci-

ficity and effector sensitivity, have been investigated

(Beavo, 1990; Conti et al., 1995).

On release into the cytosol, cAMP elicits a

response in two main ways. The first established

mechanism is via the stimulation of two isoforms of

cAMP-dependent protein kinase. Binding of cAMP

to this kinase, which is composed of two types of

subunit, causes the kinase to dissociate into a

regulatory dimer, to which four molecules of cAMP

are bound, and two catalytic monomers, which are

then capable of phosphorylating a wide range of

protein substrates. Phosphorylation alters the ac-

tivity of the substrate as a result of a change in

surface charge and the subsequent change in con-

formation. An immediate cellular response is the

result. However, phosphorylation by cAMP-depen-

dent protein kinases does not exclusively target

cytoplasmatic proteins. cAMP also exerts a second,

intranuclear, effect. The catalytic subunits migrate

to the nucleus, where they regulate the gene

expression of cAMP-regulated genes through a set of

transcription factors, called cAMP-response element

(CRE)-binding (CREB) proteins (Montminy et al.,

1986). In this scheme, de novo synthesis of proteins

provokes the cellular response.

Although much interest was initially focused on

kinase-effected phosphorylation, it later became

apparent that a series of phosphatases catalysing the

dephosphorylation of the kinase substrates were also

subject to regulation and were an integral part of the

control mechanism (Cohen, 1989; Mumby & Walter,

1993; Shenolikar, 1994).

In mammalian tissues, with the exception of the

enucleate red blood cell, cAMP has been established

as ubiquitous, and mediates the action of a wide

range of hormones and neurotransmitters. Also in

mammals a second cyclic nucleotide, guanosine 3«,5«-cyclic monophosphate (cGMP), has been shown to

have a more restricted role (Fig. 2), for example:

altering the permeability of the cell membrane in the

retinal rods to Na+ ions in response to activation of

the visual pigment by light, via a G-protein in-

teraction with cGMP phosphodiesterase (Fesenko

et al., 1985; Koch & Kaupp, 1985; Stryer, 1986);

regulating the movement of Na+ ions and water

across membranes in response to guanylin and

natriuretic peptides (Hofmann et al., 1992); and

mediating the response to nitric oxide in smooth

muscle (Moncada et al., 1992). Although significant

differences occur in the physiological effects of the



two nucleotides, enzymes for cGMP analogous to

those for cAMP are present, namely guanylyl

cyclase, cGMP-dependent protein kinase and cGMP

phosphodiesterases.

The action of cAMP in the mammalian system is

not mediated solely by phosphorylation}dephos-

phorylation phenomena. In the process of olfaction,

cAMP mediates the cellular response through the

activation of ion channels in the plasma membrane in

a way that is very similar to the action of cGMP, i.e.

through a direct interaction with an ion channel. Air-

or water-borne chemicals are recognized at specific

receptors in the plasma membrane of olfactory cilia

that are connected to adenylyl cyclase through

specific Golf

-proteins. cAMP is released into the

cytoplasm, the subsequent activation of cyclic-

nucleotide-gated channels results in the depolari-

zation of the cell membrane, bringing about the

propagation of an electrical signal through the

olfactory nerve. Ca#+ that enters the cells through the

opened channels modulates the response by in-

creasing the activity of Ca#+}calmodulin-activated

phosphodiesterases (Dhallan et al., 1990; Zufall et

al., 1994; Finn et al., 1996).

In most non-mammalian eukaryotic organisms

other than higher plants, cAMP signal transduction

has been shown to exist that is very similar to that in

the mammalian model system. The basic com-

ponents, adenylyl cyclase, phosphodiesterase and

cAMP-dependent protein kinase, are almost ubiqui-

tous but are very diverse. In contrast to that in

Schizosaccharomyces pombe, the Saccharomyces cere-

visiae adenylyl cyclase, for instance, is different from

the prototypical mammalian adenylyl cyclase in that

it is a peripheral membrane protein (Masson et al.,

1984; Kataoka et al., 1985) that is regulated not by

heterotrimeric G-proteins but by the monomeric

GTP-binding Ras proteins (Toda et al., 1987). The

slime mould Dictyostelium discoideum, in which

cAMP acts as both a primary and a secondary

messenger, has two kinds of adenylyl cyclase. One is

very similar to the mammalian enzyme with 12

membrane-spanning helices and is connected to a G-

protein-coupled cAMP receptor (‘cAR’). The other

contains a single transmembrane span and is ex-

pressed only during germination (Parent & De-

vreotes, 1995). The Dictyostelium discoideum protein

kinase A is also different from mammalian cAMP-

dependent protein kinases. It exists as a dimer and

probably contains only one functional cAMP-bind-

ing site (Mutzel et al., 1987). Despite these marked

differences in building blocks, the overall scheme of

eukaryotic signal transduction is well conserved.

For decades, cAMP signal transduction has been

presented as having two prototypical modes of

action: the eukaryotic system described above, which

acts mainly through protein phosphorylation, and

the prokaryotic scheme, for which the model system

is that of Escherichia coli. The model of prokaryotic

cAMP signalling is catabolite repression, a process

that ensures adequate utilization of carbohydrate

resources. cAMP content in Escherichia coli is highly

regulated by the presence of glucose in the growth

medium (Rickenberg, 1974). In its absence, [cAMP]

is high and drives gene expression of the lac operon,

enabling the bacterium to use lactose as an energy

source. cAMP acts through direct binding to a

transcription factor known as catabolite activator

protein (‘CAP’) or cAMP receptor protein (‘CRP’).

After binding to cAMP, CRP changes its con-

formation and binds to the promoter region of lac,

enabling RNA polymerase to start transcription. In

the presence of glucose, the intracellular [cAMP]

decreases, cAMP-CRP complexes are no longer

formed and transcription is stopped. Although this is

probably the most important mode of action of

cAMP in prokaryotes, there are indications that

‘eukaryotic-type’ signal transduction chains also

exist in prokaryotes.

Many bacteria possess serine}threonine kinases

resembling eukaryotic protein kinases. The eubac-

teria Myxococcus xanthus (Mun4 oz-Dorado et al.,

1993), Yersinia pseudotuberculosis (Galyov et al.,

1993), Streptomyces coelicolor (Urabe & Ogawara,

1995) and Thermomonospora curvata (Janda et al.,

1996) possess protein kinases with a eukaryotic

character. Myxococcus xanthus has at least eight

genes that have homology with eukaryotic kinases.

Sequencing of the genome of the cyanobacterium

Synechocystis has revealed the presence of a regu-

latory subunit of a cAMP-dependent protein kinase

(Kaneko et al., 1996). Some eubacteria also have

GTP-binding proteins reminiscent of heterotrimeric

G-proteins. The hydrolysis of cGMP in Halo-

bacterium halobium is enhanced by the addition of

GTP, guanosine 5«-[γ-thio]triphosphate (‘GTPγS’)

and AlF$, which are activators of G-proteins. Besides

having an effect on behavioural changes, light also

influences the endogenous [cGMP] in Halobac-

terium : the process is possibly analogous to visual

perception in mammals, because a G-protein-

regulated cGMP-phosphodiesterase is present. On

the basis of these enzymatic and additional immuno-

logical data, Schimz et al. (1989) postulate the

existence of a Gα subunit in this bacterium.

A number of systems exist in the lower plants for

which the physiological role of cAMP and its

metabolism are reasonably well understood. cAMP

is an important signalling molecule during the sexual

interaction between mt+ and mt- gametes of the

diflagellate green alga Chlamydomonas. Both in

Chlamydomonas reinhardtii (Saito et al., 1993; Zhang

& Snell, 1994) and in Chlamydomonas eugametos

(Kooijman et al., 1990) the intracellular [cAMP]

increases after the agglutination of compatible

mating types. The elevated [cAMP] stimulates a

sequence of mating responses, such as the excretion

of serine proteases, cell wall breakdown and actin



polymerization into a mating structure, that eventu-

ally lead to cell fusion. These responses can also be

evoked in cells from one single mating type by the

addition of dibutyryl-cAMP. A number of com-

pounds known to inhibit cAMP accumulation also

inhibit the mating response. At the same time,

cAMP is assumed to govern the motility of flagella of

vegetative cells. The regulation of both non-related

phenomena is believed to occur at the level of

adenylyl cyclase activity (Zhang & Snell, 1993).

Chlamydomonas reinhardtii possesses two adenylyl

cyclase activities with distinctive properties. One is

expressed only in gametes and is strongly regulated

during the sexual interaction. In a fast response

(approx. 15 s) this adenylyl cyclase activity rises 2–3-

fold as soon as compatible gametes are mixed.

Activity does not seem to be regulated by G-

proteins. It is dependent on regulation by phos-

phorylation}dephosphorylation through an antag-

onistic action of a constitutive inhibitor kinase and a

facultative activator kinase. Ca#+ might be important

in the process. The vegetative adenylyl cyclase is not

regulated by G-proteins either, but differs from the

gametic adenylyl cyclase in that it shows a 3–5-fold

lower activity. It is not stimulated by Mn#+, it is not

inhibited by Ca#+ and ATP, and it is not sensitive to

the addition of staurosporine. It is still not known

whether there are two different gene products or one

adenylyl cyclase that is differently regulated at

different stadia during differentiation.

One of the best studied cAMP signal transduction

systems of plant origin is that of Euglena gracilis.

Both cAMP and cGMP are crucial in the regulation

of cell division by day-night rhythm in this flagellate

alga. All indispensable components of cAMP-

metabolism have been found in this organism

(Edmunds, 1994). Changes in cAMP were found to

occur during the cell cycle and cAMP is believed to

be the link between the internal clock and the cell

cycle, in which it permits transition through the

G"}S and G

#}M boundaries. Experiments in which

intracellular cAMP concentrations were manipulated

through the application of cAMP, isobutylmethyl-

xanthine or forskolin at very low doses (1–5 nM)

have helped in elucidating the mode of action of

cAMP in this phenomenon (Edmunds, 1994).

In the light of the crucial roles of cyclic nucleotides

in other organisms described above, why should

higher plants not possess analogous functions for

these compounds? In comparison with mammals

and many other animals, an immediate difference is

the absence of neurotransmitters ; the presence of

plant cell walls might seem to compromise the

adenylyl cyclase model depicted in Fig. 1. However,

the cell wall probably does not constitute a problem

because it does not interfere in phytohormone action,

and reports exist describing putative G-protein-

coupled receptors with the seven-transmembrane-

region signature (Josefsson & Rask, 1997; Plakidou-

Dymock et al., 1997). Both reports describe a

putative seven-transmembrane-region G-protein-

coupled receptor bearing greatest homology with the

Dictyostelium discoideum cAMP receptor; in the

former case, a role in cytokinin signal transduction is

postulated.

Thus, at first glance, a comparison of cAMP

systems in higher plants with those of lower

organisms raises few intrinsic problems, particularly

given the established role of cAMP in bacteria and

algae together with the possible endosymbiotic origin

of plant cell organelles ; this is also true of possible

analogies relating to cGMP action in animals and

higher plants. In summary, we can conclude that the

existence and established functions of cAMP in

mammals and other organisms do not allow the

simple prediction that it is certain to have analogous

functions in higher plants. What, therefore, is the

evidence that it does, and}or that other cyclic

nucleotides are involved in signal transduction in

higher plants?

II. c

The early reports on the existence of cAMP in plants

were criticized on the basis that they were either

presumptive deductions from the observed physio-

logical effects of endogenously supplied cAMP or

cAMP analogues, or conclusions based solely on

insufficiently rigorous chromatographic identifi-

cation. As an example of the former, Salomon &

Mascarenhas (1971) reported that cAMP delayed

petiole abscission in Coleus, but did not demonstrate

that this was a process in vivo, merely that the cyclic

nucleotide could replicate the action of auxin. In the

latter category, Pollard (1970) obtained a radio-

labelled product from the incubation of [8-"%C]

adenine with germinating barley seeds that chro-

matographed together with cAMP in ten chromato-

graphic systems. The latter report was criticized on

the basis that the chromatographic systems would

not resolve the putative secondary messenger cAMP

from the RNA catabolism intermediate, adenosine

2«,3«-cyclic monophosphate. To overcome such

criticism the putative radiolabelled cAMP was

hydrolysed to AMP by cAMP phosphodiesterase,

which was then determined enzymatically (Nara-

yanan et al., 1970); however this expedient was

criticized in that the phosphodiesterase was not of

demonstrated absolute specificity for cAMP.

Concomitantly with these reports of cAMP,

reports of the existence of cAMP-based signalling

enzymes were made; these reports were also criti-

cized, those of phosphodiesterase activity on the

basis of substrate specificity (see section V) and those

of adenylyl cyclase on the basis of product identi-

fication (see section IV). In an attempt to identify

cAMP conclusively as an endogenous component of



plant cells, a sequential chromatographic and electro-

phoretic procedure for the extraction and isolation of

cAMP was developed (Brown & Newton, 1973). The

identity of the putative cAMP was confirmed by co-

chromatography with an authentic sample in five

paper and three thin-layer chromatography systems

and by high-voltage electrophoresis in three different

buffers. Collectively, these steps were capable of

separating cAMP from all then known naturally

occurring adenine nucleotides, including 2«,3«-cyclic

AMP. Nevertheless, some authors still claimed that

the evidence was equivocal (Keates, 1973; Amrhein,

1974a; Lin, 1974), considering that hitherto un-

identified adenine compounds, with identical chro-

matographic properties to those of cAMP in these

systems, existed in higher plants. During and after

this initial phase of investigation in this area, a

considerable number of reports quantifying cAMP

in various plant species were made, with concen-

trations of a similar order ranging from 2.1–3.5 pmol

cAMP g−" wet wt in Zea (Tarantowicz-Marek &

Kleczkowski, 1978) to 220–280 pmol g−" wet wt in

Lactuca (Kessler & Levenstein, 1974): a compre-

hensive listing of concentrations then reported is

reviewed in Newton & Brown (1986). Nevertheless,

some authors reported that concentrations of cAMP

were close to and below the sensitivity of their

methods (Niles & Mount, 1973; Amrhein, 1974a;

Bressan et al., 1976) and as a consequence several

reviews at the time concluded that cAMP was not

present in plants (Keates, 1973; Lin, 1974; Amrhein,

1974a, 1977); others suggested that any cAMP

present was a result of bacterial infection (Bonnafous

et al., 1975). The contamination concept was refuted

by Ashton & Polya (1977), who calculated that less

than 0.1% was contributed by bacteria and demon-

strated the presence of cAMP in axenic cell cultures

of rye grass (Ashton & Polya, 1978), supporting

earlier reports of cAMP in axenic cultures of soybean

callus tissue (Brewin & Northcote, 1973) and of

tobacco cells (Lundeen et al., 1973).

Although a range of physiological processes and

enzymatic reactions in plants were suggested to be

responsive to cAMP (Brown & Newton, 1981;

Newton & Brown, 1986), for credibility to be

sustainable it was essential to demonstrate un-

equivocally the identity of the putative cAMP

obtained in tissue and cell extracts and as the product

of incubations with adenylyl cyclase. This was

successfully accomplished by the use of physical

techniques including MS and NMR spectroscopy,

as detailed in section III; the inclusion of suitable

controls has also been used to demonstrate that the

identified cAMP is not an artefact. As a consequence,

although there were a few reports after the pro-

duction of MS evidence of the identity of cAMP in

plant extracts that claimed the absence of cAMP

from plant cells (Spiteri et al., 1989), several reviews

have appeared that show a shift in the balance of

opinion. Although most reviews in the initial phase

expressed the opinion that cAMP did not, or was

unlikely to, function in higher plants (Keates, 1973;

Lin, 1974; Amrhein, 1974a, 1977), these have been

superseded by commentaries suggesting potential

functions (Brown & Newton, 1981; Francko, 1983;

Newton & Brown, 1986; Assmann, 1995; Bolwell,

1995; Trewavas, 1997). As will be detailed in the

sections below, conclusive evidence of the existence

of cAMP, adenylyl cyclase, phosphodiesterase and

cAMP-binding proteins is now available and sys-

tematic studies of the function of the cyclic nucleo-

tide are appearing, for example in studies of its role

in the cell cycle, in stress response systems and in the

regulation of ion channels.

III.

c

In the 1970s, studies on the occurrence and effect of

cAMP in higher plants were a mere extrapolation of

the investigation going on in animal and fungal

systems; in retrospect it seems naive to assume that

plant 3«,5«-cAMP-mediated signal transduction is

virtually identical to that in other kingdoms. Early

papers (Pollard, 1970; Oota, 1972; Keates, 1973;

Truelsen et al., 1974) were based on observations of

various physiological and metabolic responses after

the exogenous application of cAMP, cAMP ana-

logues and phosphodiesterase inhibitors known at

that time. This shotgun-like approach was per-

formed without any knowledge of the underlying

metabolism in plants. A major point of interest in

this respect is the discrepancy between the en-

dogenous concentrations present in plants and in

animals. Animal and microbial cells contain cAMP

concentrations in the nanomolar to micromolar

range, whereas in plants the [cAMP] is much lower.

This suggests that either the global metabolic activity

of adenylyl cyclase and phosphodiesterase is low or

that it is subject to a strict temporal and spatial

regulation (Brown & Newton, 1992). However, the

term ‘low concentration’ might be misleading;

although the cAMP concentration per unit weight is

much higher in animals than in plants, the picture

might be distorted by differences in cell structure

(such as the presence of vacuoles and a cell wall).

The ratio of cAMP to ATP, for example, is similar

in animals and plants: between 1:100 and 1:10000.

This low [cAMP] definitely constitutes a major

problem in cAMP research in plants. The initial

application of the same extraction, purification and

detection techniques as those described for animal

tissues without considering the problems peculiar to

plant matrices led to diverse results, which were

often interpreted as contradictory. Without modi-

fication of the separation techniques (if indeed any

separation was included, because the original work

of Cailla et al. (1973) prescribed no requirement for



the purification of cAMP by radioimmunoassay at

the femtomole level), ubiquitous interfering plant

metabolites co-eluted and compromised the data.

Furthermore, because the limit of detection had to

be made very low, extra precautions had to be taken

to prevent bacterial contamination (Bonnafous et al.,

1975) and physicochemical cyclization of ATP to

cAMP (Cook et al., 1957) catalysed by the presence

of bivalent ions during extraction at basic pH

(Brooker et al., 1979). In the late 1970s and early

1980s, the experimental set-up for cAMP analysis

was adapted to a great extent to cope with plant

matrices. Taking care to minimize the artefactual

origin of cAMP, various research groups produced

more reproducible and coherent data (Brown et al.,

1977; Katsumata et al., 1978; Hilton & Nesius,

1978; Tu, 1979). In the 1980s, chromatographic

techniques in plant cAMP analysis such as ad-

sorption, ion-exchange, paper and thin-layer chro-

matographies were largely reduced to preparative

steps and replaced by the far superior HPLC, which

is able to separate the 3«,5«-cyclic nucleotides and

their 2«,3« isomers (Van Onckelen et al., 1982;

Brown, 1983).

Although widely applied and among the most

sensitive methods over the past decades, immuno-

sorbent assays for cAMP, although having a fem-

tomole dynamic range, lack accuracy. For most of

these assays cAMP needs to be derivatized; in-

terference by unknown contaminants can be ruled

out only if extensive controls are built in into the

assay. Moreover, the presence of the cyclic nucleotide

derivatives 2«-deoxyadenosine-3«,5«-cyclic mono-

phosphate, 2«-deoxyguanosine-3«,5«-cyclic mono-

phosphate, 2«-O-glutamyl-3«,5«-cyclic AMP, 2«-O-

aspartyl-3«,5«-cyclic AMP, 2«-O-glutamyl-3«,5«-cyc-

lic GMP, 5«-phosphoadenosine-2«,3«-cyclic mono-

phosphate and 2«-phosphoadenosine-3«,5«-cyclic

pyrophosphate was demonstrated in Porphyra um-

bilicalis used as a plant model system (Newton et al.,

1995); such derivatives of cyclic nucleotides have

been shown to interfere in radioimmunoassays of

cyclic nucleotides (Newton et al., 1994), illustrating

the inadequacy of the method as a sole detection

technique. However, the use of anti-cAMP anti-

bodies in an immunopurification step results in a

very powerful sample clean-up because most in-

terfering compounds in further quantitative steps

are thereby disposed of. Polyclonal chicken egg yolk

anti-cAMP antibodies in combination with UV-

PDA (photo diode array) HPLC has proved to be a

very powerful analytical method for cAMP quanti-

fication in higher-plant matrices such as plasma

membranes, chloroplasts and protoplasts (Roef et

al., 1996; Witters et al., 1996).

Despite the significant technical improvements,

UV absorption spectra or fluorescence spectra

together with chromatographic retention times were

still considered ambiguous identification criteria. It

was mass spectrometric analysis that unequivocally

established cAMP as being endogenous to plant

tissues (Newton et al., 1980; Johnson et al., 1981;

Janistyn, 1983). These first observations were ob-

tained by electron-impact GC-MS of volatile tri-

methylsilyl derivatives of cAMP, as reported by

Lawson et al. (1971) with the chemically synthesized

compound. Although electron-impact GC-MS is a

sensitive method, its major pitfall is the requirement

for the non-homogenous silylation of cAMP extracts.

The advent of the soft-ionization mass spectrometric

techniques of fast atom bombardment (FAB) and

electrospray ionization (ESI) was a big step forward

in nucleotide research, because it removed this

requirement for derivatization. For an overview of

the basic principles of MS and its application to

biomolecular research, the reader is recommended to

consult Newton & Walton (1996) and Caprioli et al.

(1997).

The potency of static FAB-MS in plant nucleotide

research was first demonstrated by Newton and co-

workers. FAB ionization readily provided molecular

mass information of non-volatile polar compounds

including nucleotides. However, because it is a soft

ionization technique, the major drawback of the

spectra obtained is the absence of diagnostic frag-

ments. FAB mass spectra do not permit detailed

structural analysis ; isomeric compounds, for ex-

ample, cannot be differentiated because they produce

very similar mass spectra. To overcome this prob-

lem, collision-induced dissociation (CID) of the

protonated molecule selected from the FAB mass

spectrum provides a mass-analysed kinetic energy

(MIKE) spectrum that can be used to generate

structural information, including the differentiation

of cyclic nucleotide isomers (Kingston et al., 1984,

1985; Newton et al., 1984b, 1986, 1989). This

tandem MS permits the identification of diagnostic

fragments in the MIKE spectrum from the pro-

tonated molecule and has thereby allowed the

unambiguous identification of 3«,5«-cAMP (Fig. 3)

and 3«,5«-cGMP (section VIII). Furthermore, FAB-

CID}MIKE analysis of partly purified extracts

from meristematic and non-meristematic tissue

from Pisum sativum has demonstrated the natural

occurrence of inosine 3«,5«-cyclic monophosphate

(cIMP), uridine 3«,5«-cyclic monophosphate

(cUMP), cytidine 3«,5«-cyclic monophosphate

(cCMP) and 2«-deoxythymidine 3«,5«-cyclic mono-

phosphate (cdTMP) in addition to cAMP and

cGMP (Newton et al., 1991).

ESI-MS has become a very popular detection

technique for the analysis of polar biomolecules and

is replacing many of the FAB-MS applications by

virtue of its greater sensitivity. As with FAB

ionization, ESI produces mainly the quasi-molecular

ion; to acquire information on the molecular struc-

ture, tandem MS (MSMS) needs to be performed

(Fig. 3). When combined with separation techniques



100

50

0

Rel

ativ

e in

ten

sity

(%

)

277

299

321

330

352

369

391

413 422

450400350300

(a)

m/z

100

50

0

Rel

ativ

e in

ten

sity

(%

)

(b)(c)

136

119 184178

313

0 2000 4000 6000 8000 0.400 0.675

Electric sector potential

100

0

m/z 136

164

178202

Potential (V)

(d)

m/z313 N N

NN

OO

OO

HO P

OH

N N

NN

N N

N N

O = C – HN N

N N

HOHC = CH

S1

S2

m/z 119

m/z 136

m/z 164

m/z 178

cAMP([MH]+)m/z 330

+NH3+NH3

+NH3

+NH3

CH2

Fig. 3. For legend see opposite.

such as capillary zone electrophoresis or capillary

HPLC, ESI-MS becomes a very powerful analytical

technique. The sensitivity is reached primarily by

virtue of the ESI process, which behaves as a

concentration-sensitive phenomenon (Hopfgartner

et al., 1993). Reducing the dimensions of the LC set-

up enhances the sensitivity exponentially (Chervet

et al., 1996; Vanhoutte et al., 1997; Witters et al.,

1997a) and the introduction of a capillary column

switching method yields detection limits as low as

25 fmol (Witters et al., 1997b, 1998). As will be

discussed in section VII, the use of this sensitive

LC-ESI-MSMS set-up has enabled a cell cycle-

regulated cAMP accumulation to be demonstrated

in a Nicotiana tabacum BY2 cell culture (Ehsan et al.,

1998).



100

0

%

(e)

100

0

%

100

0

%

100

0

%

20 100 180 260 340

329.9311.8203.9176.8

136.0

97.1

99.1

136.0

194.8 329.9

327.8

327.8

134.0

79.0

134.0

m/z

Fig. 3. Mass spectrometric identification of cyclic nucleotides. (a) Positive-ion FAB mass spectrum of cAMP,

showing a protonated molecule at m}z 330. (b) CID}MIKE spectrum of m}z 330 from cAMP including

diagnostic fragments at m}z 136 (protonated base), m}z 164 (protonated base 28) and m}z 178 (protonated

base 42). (c) Partial CID}MIKE spectra of m}z 330 from cAMP (solid line) and 2«,3«-cyclic AMP (dotted

line), demonstrating differentiation between them. (d) Fragmentation of cAMP in a FAB-MS-CID}MIKE

spectrum scan showing the origin of m}z 164 and m}z 178 peaks by S#

and S"

cleavage: 2«,3«-cyclic AMP is

unable to produce m}z 178 because of the substitution at the 2«-O position. (e) Product ion spectra of 2«,3«-cyclic AMP (first and third panels) and cAMP (second and fourth panels) in both negative (top two panels) and

positive (bottom two panels) ionization modes. Both sets of spectra contain the base fragment as the major peak,

at m}z 134 and 136 respectively. The absence of the PPiion at m}z 79 for 2«,3«-cyclic AMP and its presence

in the cAMP spectrum differentiates between the isomers in negative mode: in positive ionization mode the

presence of peaks at m}z 177 and 312 in the cAMP spectrum and at m}z 195 in the 2«,3«-cyclic AMP spectrum,

together with the relative heights of the peaks at m}z 97 and 99, is a further means of differentiation between

the cyclic nucleotide isomers by ES-MSMS.

IV.

Adenylyl cyclase activity has been demonstrated in

higher-plant material by the use of both histo-

chemical and biochemical procedures. Histochemi-

cal methods are based predominantly on the standard

Wachstein-Meisel lead phosphate precipitation tech-

nique (Wachstein & Meisel, 1957), with ATP as a

substrate for adenylyl cyclase: an electron-dense

precipitate of lead or cerium ions in the presence of

PPiis released on the formation of cAMP from ATP

and is detected by electron microscopy. Because cells

harbour tremendous quantities of ATP-hydrolysing

enzymes that subsequently release Pi, this procedure

produced lots of false positive results that could not

be overcome by the use of compounds to inhibit

these contaminating activities. The specificity of the

reaction was considerably improved by the replace-

ment of ATP with a specific substrate analogue for

adenylyl cyclase such as adenosine 5«-[β,γ-imido]

triphosphate (Yount et al., 1971) or adenosine 5«-[α,β-methylene]triphosphate (Mayer et al., 1985),

which are much less sensitive to ATPase and other

phosphatase activities. (For a recent review on

histochemical methodology the reader is referred to

Richards & Richards (1998).) By such methods,

early indications of the presence of adenylyl cyclase

activity were found in plasma membrane, in en-

doplasmic reticulum and nuclear membranes in Zea

mays root tips (Al-azzawi & Hall, 1976), on internal

membranes of cytoplasmatic vacuoles in Pisum

sativum (Hilton & Nesius, 1978), on the external side

of the host plasma membrane and membranes

surrounding the endophyte in root nodules of Alnus

glutinosa (Gardner et al., 1979), and on the external

side of the plasma membrane of Pisum sativum

(Nougare' de et al., 1984). Physiological roles have

been proposed for plant adenylyl cyclase: Rougier et

al. (1988) postulated that adenylyl cyclase activity is

a determining factor in the compatibility of pollen



tube formation in Populus spp., whereas Curvetto &

Delmastro (1990) located an adenylyl cyclase activity

in Vicia faba guard cells that was selectively

stimulated by IAA, Ca#+, caffeine, GTP, forskolin

and, to a smaller extent, ABA. According to these

authors cAMP is involved in the IAA signal

transduction chain during stomatal movement

through a G-protein-mediated mechanism. An

adenylyl cyclase activity was also found on the

external side of the plasma membrane and on

thylakoid membranes in primary leaves of Phaseolus

vulgaris, a finding corroborated by the immuno-

localization of cAMP in the chloroplast and cell wall

(Gadeyne, 1992).

A histochemical approach is invaluable for pin-

pointing the exact location of adenylyl cyclase

activity and is thus extremely helpful in providing

clues to a physiological role. However, the methods

available are not definitive. In spite of the fact that

the substrate now used is insensitive to phosphatase

activity, little is known of the impact of apyrases and

diphosphohydrolases ; some caution has therefore to

be observed in the interpretation of these data. Data

obtained by a biochemical approach, in which the

actual formation of radiolabelled cAMP from radio-

labelled precursor (ATP or adenosine 5«-[β,γ-imido]

triphosphate) is measured, are far more reliable in

this respect. Nevertheless, the first such reports were

heavily criticized for inadequate identification of the

newly formed compound, but more rigorous separa-

tory procedures provided more credible evidence.

For example Carricarte et al. (1988) described a

soluble adenylyl cyclase in roots of Medicago sativa

with an estimated molecular mass of 84 kDa. The

enzyme was active in the presence of Ca#+ and Mg#+.

A DEAE-purified extract produced 204 pmol cAMP

min−" mg−" protein in the presence of Ca#+, an

activity that was stimulated 15-fold by bovine

calmodulin. The effect of the addition of Spinacea

oleracea calmodulin was less marked but was never-

theless significant (approx. 7-fold stimulation). The

stimulation was abolished by the addition of EGTA

and chlorpromazine, an inhibitor of calmodulin

function. GTP, guanosine 5«-[β,γ-imido]triphos-

phate, forskolin, fluoride and cholera toxin were

ineffective, indicating that the enzyme was not

dependent on G-protein function.

By contrast, Lusini et al. (1991) described a

sedimentable adenylyl cyclase in the roots of Ricinis

communis. An enzyme activity of approx. 20 pmol

min−" mg−" protein was measured in the presence of

3 mM MnCl#. In contrast with the alfalfa enzyme,

MgCl#

did not stimulate this adenylyl cyclase, but

NaF and GTP did; thus this enzyme might well be

G-protein regulated. Although this radiometric

evidence was still not universally accepted, the

advent of highly reliable MS techniques (as de-

scribed in section III) capable of identifying cAMP

unambiguously as the reaction product, has now

produced a number of reports proving unequivocally

the presence of adenylyl cyclase activity in higher

plants.

A sedimentable adenylyl cyclase activity was

identified in Pisum sativum (Pacini et al., 1993) by

using mass spectrometric techniques for the first

time, producing unambiguous identification of the

reaction product. This enzyme utilizes Mg#+-ATP as

a substrate and is stimulated by GTP at 100 nM.

Higher concentrations of GTP (110 µM) inhibit the

activity, probably owing to competition with ATP.

The application of zeatin, GA$, IAA and Ca#+}

calmodulin yielded surprising results. No simple

dose-dependent response was observed; all four

effectors stimulated activity by up to 50% at an

optimum concentration, but higher concentrations

were less effective and with zeatin and GA$

they

even became inhibitory. A plausible explanation is

the involvement of an additional regulatory com-

ponent that can become saturated by these effectors

or becomes limiting by another, effector-indepen-

dent, mechanism.

Medicago sativa cell cultures exposed to the

elicitor of the phytopathogenic fungus Verticillium

albo-atrum respond with an increased adenylyl

cyclase activity (Cooke et al., 1994). Again, cAMP

formation was confirmed unequivocally by mass

spectrometric analysis. Adenylyl cyclase activity was

dependent on Mg#+ and was stimulated by Ca#+.

Basal activity was very low (maximum 400 fmol

min−" mg−" protein) but increased by 300% within a

time span of 4 min on application of the elicitor. The

transient rise in adenylyl cyclase activity was ac-

companied by an increase in intracellular [cAMP]

and was followed by a transient increase in phos-

phodiesterase activity (with a maximum at 100 min).

A role in the defence mechanism of higher plants on

attack by pathogens is proposed (see section VII).

More recently a MgCl#-stimulated adenylyl cy-

clase activity has been demonstrated by ESI-MS in

plasma membrane preparations from apical hooks

from Phaseolus vulgaris (Roef et al., 1996; Roef,

1997), confirming the histochemical and immuno-

chemical data of Gadeyne (1992). This presence of

adenylyl cyclase activity associated with plant plasma

membrane preparations seems compatible with a

mammalian-type secondary-messenger system.

The first paper reporting a plant gene sequence

showing high homology with that of mammalian

adenylyl cyclase (Ichikawa et al., 1997) and detailing

aspects of its regulation has now unfortunately been

withdrawn (Ichikawa et al., 1998) as the data cannot

be reproduced, the circumstances being discussed at

length in Balter (1999). Although we feel obliged to

indicate our knowledge of these three reports for

completeness, we do not consider it appropriate to

comment further at this stage, other than to state that

an acceptable demonstration of such a plant gene

sequence would have a very significant impact.



V.

Even before the first reports of cyclic-nucleotide-

containing extracts from higher plants, the enzyme

phosphodiesterase, which is capable of hydrolysing

cAMP to AMP, was reported in pea seedlings (M.

Liberman & A. T. Kunishi, unpublished); this was

quickly followed by demonstrations of activity in

such diverse sources as tobacco (Wood et al., 1972),

barley seeds (Vandepeute et al., 1972), carrot leaves

(Venere, 1972), potato (Shimoyama et al., 1972;

Ashton & Polya, 1975) and Jerusalem artichoke

tubers (Giannattasio et al., 1974b). The occurrence

of phosphodiesterase activity in these plants was

interpreted by several of these authors as indicating

that the substrate, cAMP, must be an endogenous

component of the tissue and that it would possess

functions analogous to those of cAMP in other

organisms. However, a conflicting view was ex-

pressed by Lin & Varner (1972), who reported that

unlike its mammalian counterpart the phospho-

diesterase from pea seedlings had an acidic pH

optimum, was insensitive to methylxanthines,

yielded 3«-AMP rather than 5«-AMP as the major

hydrolytic product, and, most significantly, had

substantially greater activity with the RNA break-

down intermediate 2«,3«-cyclic AMP as substrate

than with the putative secondary-messenger isomer,

3«,5«-cAMP. Because at that time the known mam-

malian phosphodiesterases functioning in the cAMP

secondary-messenger cascade produced only the 5«-mononucleotide product and would not hydrolyse

2«,3«-cyclic AMP to any significant extent, Lin &

Varner (1972) concluded that the pea phosphodi-

esterase was functioning not in a plant signal

transduction system but as part of a catabolic

sequence of RNA. This view had an immediate

negative impact on theories relating to a regulatory

role for cAMP in plants, which were dealt a further

blow by a survey of phosphodiesterases from a range

of plant species and tissues that concluded, on the

basis of pH optima and substrate specificity, that

3«,5«-cAMP was not their natural substrate (Am-

rhein, 1974b). This view was endorsed by evidence

that the phosphodiesterase preparations from barley

seeds (Vandepeute et al., 1972), carrot leaves (Niles

& Mount, 1974) and tobacco (Brennicke & Frey,

1977) had at least an equal activity with a 2«,3«-cyclic

AMP substrate as with the 3«,5« isomer.

However, further examination of more purified

plant phosphodiesterases indicated that more than

one form is present. French Dwarf bean seedlings

were found to contain a phosphodiesterase that,

when partly purified, possessed properties more

similar than the plant phosphodiesterases discussed

above to those of the mammalian phosphodiesterases

(Brown et al., 1975, 1977). It was active towards

several 3«,5«-cyclic nucleotides as substrate but

inactive with 2«,3«-cyclic nucleotides, produced 5«-

mononucleotides as the major product, and the Km,

pH optimum and sensitivity to methylxanthines

were also more like those of the mammalian enzyme

than those reported by Lin & Varner (1972) and

Amrhein (1974b). Interestingly, the enzyme was

stimulated by an endogenous protein with which it

was able to form a complex; this protein was only

partly purified and was not characterized further; it

had been obtained before the demonstration of

calmodulin in higher plants, but it was also found to

stimulate bovine brain calmodulin-sensitive phos-

phodiesterase, suggesting a further parallel between

the French Dwarf bean and mammalian phosphodi-

esterases.

Examinations of the subcellular distribution of

phosphodiesterase activity have confirmed the exist-

ence of more than one phosphodiesterase type in

plant cells. In spinach, three forms of phosphodi-

esterase were observed: one, designated Ic, had its

major subcellular site in the chloroplast, and a

second, predominantly outside the chloroplast, had

its major yield in the microsomal fraction designated

Im. Type I

mphosphodiesterase conformed to the

profile described by Lin & Varner (1972) and

Amrhein (1974b) in other plant species, having an

acidic pH optimum of 4.9, relative insensitivity to

methylxanthine inhibitors, and greater activity with

2«,3«- rather than 3«,5«-cyclic nucleotide substrates

(Brown et al., 1979b). In contrast, the Ic

phospho-

diesterase had highest activity with 3«,5«-cGMP and

3«,5«-cAMP and little activity with their 2«,3«isomers, had a less acidic pH optimum of 6.1, was

sensitive to inhibition by methylxanthines, and

liberated 5«-mononucleotides as the main product; it

also displayed sensitivity to endogenous protein

effectors and was activated by Ca#+ (Brown et al.,

1979b). Further examination of Icphosphodiesterase

revealed that it occurred in multienzyme complexes

of molecular mass 187 and 370 kDa in association

with acid phosphatase, ribonuclease, nucleotidase

and ATPase (Brown et al., 1980a).

A multiplicity of phosphodiesterases has also been

reported in other species. In potato (Ashton & Polya,

1975) three phosphodiesterases are present: one with

greatest activity with 3«,5«-cyclic nucleotide sub-

strates, one with greatest activity with 2«,3«-cyclic

nucleotides, and one with NAD pyrophosphate as

the major activity. Similarly, in Portuluca (Endress,

1979) three phosphodiesterases are present: two

show Michaelis-Menten kinetics, one having greater

activity with 3«,5«-substrates and one with 2«,3«-substrates, and the third is arguably the most

interesting because it shows positive cooperativity

and is sensitive to allosteric regulation by nucleo-

tides, with for example the presence of cGMP

stimulating a high activity towards 3«,5«-cAMP and

3«,5«-cGMP as substrates, a kinetic process anal-

ogous to that in one of the established mammalian

phosphodiesterase families (Manganiello et al.,



1990). In carrot cell cultures two phosphodiesterases

with distinct kinetic parameters have been reported:

constitutive phosphodiesterase activity did not de-

pend on either Ca#+ or calmodulin, but a calmodulin-

dependent isoform could be induced by increased

[Ca#+] (Kurosaki & Kaburaki, 1995). Kinetic analy-

ses suggested that the constitutive phosphodiesterase

has a role in the maintenance of the resting state of

the carrot cells by keeping cellular [cAMP] and

[Ca#+] very low, whereas the calmodulin-sensitive

phosphodiesterase induced in the excited cells hy-

drolyses cAMP rapidly under conditions of high

[cAMP] and [Ca#+] as one of the response-decay

mechanisms.

The presence of several forms of phosphodi-

esterase offers a ready explanation of the apparent

incompatibility of the data obtained and conclusions

drawn by different groups in the earlier reports of

plant phosphodiesterases, with different extraction

and purification protocols selecting for one or other

of the phosphodiesterase types. In addition, the

observation that in at least one plant species the

phosphodiesterase is present in a complex also

containing nucleotidase suggests that identification

of one or other mononucleotide isomer as the major

product of phosphodiesterase activity might not be

as clearcut as it seems at first. For example a 3«,5«-cAMP phosphodiesterase activity from Phaseolus

vulgaris seedlings 7 d old had an acidic pH optimum,

was strongly stimulated by Mn#+, Mg#+ and Ca#+ and

imidazole, was inhibited by NaF, PPiand Fe$+ and

was insensitive to butylmethylxanthine; purification

away from a contaminating monoesterase activity

revealed that the protein hydrolysed the 3«-esterlinkage exclusively (Dupon et al., 1987).

The existence of multiple forms complicates the

interpretation of phosphodiesterase function; this is

compounded by the complex kinetics of the in-

dividual enzymes. Several of the phosphodiesterases

examined had activity not only with cAMP but also

with cGMP. The greater activity with 3«,5«-cAMP as

substrate than with the 2«,3 isomer can be considered

indicative of the function of the enzyme’s being

hydrolysis of the putative signalling molecule. How-

ever, greater activity with cGMP might suggest that

the latter is in fact the natural substrate, for example

with a Phaseolus chloroplast phosphodiesterase,

which has a low Km

for cGMP of 77 µM and is more

than 3-fold more active with cGMP as substrate than

with cAMP (Newton et al., 1984a).

The activity of at least some plant phosphodi-

esterases is not confined to purine cyclic nucleotide

substrates : a lettuce phosphodiesterase with a mol-

ecular mass of 62 kDa (and thus a smaller entity than

the enzymes from spinach and French Dwarf bean)

showed significant similarity to the multifunctional

phosphodiesterase isolated initially from pig liver

(Helfman et al., 1981). This lettuce phosphodi-

esterase differs from other plant 3«,5«-cyclic nucleo-

tide phosphodiesterases in that it exhibits com-

parable activity with both pyrimidine and purine

cyclic nucleotide substrates, hydrolysing cytidine

3«,5«-cyclic monophosphate (cCMP) at a similar rate

to cAMP and cGMP, with Km

values of 1.1, 0.71 and

0.64 mM and Vmax

}Km

values of 5.1¬10$, 3.7¬10$

and 3.4¬10$ l min−" mg−" protein for cAMP, cGMP

and cCMP, respectively (Chiatante et al., 1986). A

unique feature of this enzyme among plant phospho-

diesterases is that it is stimulated to the greatest

extent by Fe$+ ions, a feature previously observed

only in the mammalian multifunctional (Kuo et al.,

1978) and cCMP-specific (Newton et al., 1990)

phosphodiesterases. This lettuce enzyme was able to

hydrolyse both 3«,5«- and 2«,3«-cyclic nucleotide

substrates. With the 3«,5« substrates both 5«- and 3«-mononucleotide products were released and the 5«isomer was the major form. With the 2«,3«-cyclic

nucleotide substrates the point of cleavage was

affected by the nature of the base: 3«-CMP was the

major product from 2«,3«-cyclic CMP, 2«-GMP was

the sole product of 2«,3«-cyclic GMP hydrolysis, and

equimolar proportions of 2«- and 3«-AMP were

liberated from 2«,3«-cyclic AMP (Chiatante et al.,

1987). Kinetic analysis of this enzyme revealed a

complex picture in which the presence of one cyclic

nucleotide affects the hydrolysis of another: with a

3«,5«-cAMP substrate other 3«,5«- and 2«,3«-cyclic

nucleotides exhibit mixed-type inhibition, with the

Ki

values of for example cGMP and 2«,3«-cyclic

AMP with cAMP as substrate being two orders of

magnitude lower than the Km

values of the former

two compounds when they are sole substrates. This

suggests that there is more than one binding site for

each cyclic nucleotide, although no cooperative effect

seems to exist for a single cyclic nucleotide as

substrate. In contrast to the mixed inhibition above,

the hydrolysis of cGMP was stimulated by the

presence of cAMP and cCMP. In this instance at

least it seems that a major factor regulating the

hydrolysis of one cyclic nucleotide is the presence of

others. With a highly purified preparation, from

which endogenous nucleotidase activity had been

removed, this enzyme was found to have greater

activity with 3«,5«-cAMP than 2«,3«-cyclic AMP and

produced 5«-AMP as the major product from the

former. With cGMP and cCMP the 2«,3« isomers

were the preferred substrates, but kinetic data

confirmed that there were distinct catalytic sites for

the 2«,3«- and 3«,5«-cyclic nucleotides (Chiatante et

al., 1988).

Although the phosphodiesterase activity in mam-

mals seems tightly regulated and integrated to

maintain a consistent turnover of cyclic nucleotides,

this pattern alters under various conditions, for

example a phosphodiesterase rebound activation

after stimulation of nucleotidyl cyclase, during the

cell cycle and, for cGMP phosphodiesterase, in the

visual cycle. In plants a phosphodiesterase rebound



is observed after the stimulation of adenylyl cyclase

by an elicitor as part of the plant cell’s defence

against pathogens, as discussed in section VII. In the

context of the suggested involvement of phospho-

diesterases in the regulation of the cell cycle (Levi

et al., 1981) the presence of two major forms of

phosphodiesterase in the meristems of peas is of

relevance. One form hydrolysed cAMP, cGMP and

cCMP; the second was unique in that it had a

preference for cCMP over cAMP but was devoid of

activity with cGMP. It was stimulated by Fe$+ but

not by calmodulin and was inhibited by methyl-

xanthines; most interestingly in respect of cell cycle

regulation, the enzyme was inhibited by the cyto-

kinins kinetin and kinetin riboside, which were also

demonstrated to inhibit the growth of the pea roots

(Chiatante et al., 1990).

The other apparent effector of plant phospho-

diesterase activity identified so far is light. As

described in section I, in mammals a light-sensitive

cGMP phosphodiesterase is an integral component

of the visual cycle (Gillespie, 1990). In plants a light-

dependent response of this enzyme was first shown

in the spinach chloroplast : the chloroplast phospho-

diesterase had similar Vmax

values in light- and dark-

grown seedlings, but in the dark-grown plants the

Km

was 27 µM in comparison with 870 µM in the

light, suggesting greater activity in the dark; How-

ever, in the light the enzyme was sensitized to

Ca#+}calmodulin: the enzyme in the light-exposed

chloroplast was stimulated 3-fold by calmodulin but

the enzyme from the dark-grown plant showed no

response (Brown et al., 1989). A non-transient

activation of cAMP phosphodiesterase activity by

red light in etiolated corn sprouts has also been

reported (Kasumov et al., 1991) and light is again

involved in the calmodulin response in this species:

one phosphodiesterase from the dark-grown seed-

lings has a calmodulin sensitivity that is dependent

on [Mg#+] and the season (Fedenko et al., 1992). In

spring it is activated by calmodulin irrespective of

[Mg#+] ; in autumn at high [Mg#+] it is inhibited by

calmodulin but activated at low [Mg#+], with the

high-[Mg#+] autumn inhibition being converted to

activation if the corn seedlings are pre-illuminated

with phytochrome-absorbed red light. In maize

sprouts the effect of the GTPase-resistant GTP

analogue guanosine 5«-[β,γ-imido]triphosphate in

inhibiting only dark-grown seedlings exposed to red

light led the authors to propose a role for G-proteins

and phosphodiesterase in light signal transduction in

this plant (Fedenko & Kasumov, 1993).

VI. c- ,

, c- -

-

To identify conclusively any physiological role(s) for

cAMP in higher plants it is necessary to establish

cellular targets for cAMP action. It has long been

established that cAMP action in eukaryotes is

predominantly mediated by the phosphorylation of

target proteins via cAMP-dependent protein kinase;

it was only comparatively recently that cAMP was

shown to exert some of its effects through a direct

interaction with ion channels (Zufall et al., 1994).

Consequently, the search for cAMP targets in plants

has concentrated primarily on the quest for cAMP-

dependent protein kinases. At present there have

been no reports of the purification of a plant cAMP-

dependent protein kinase to homogeneity; most

indications of the existence of cAMP-dependent

protein kinase result from experiments in which the

phosphorylation of specific substrates is regulated by

cAMP.

Three cyclic-nucleotide-responsive protein kin-

ases have been reported in Lemna paucicostata (Kato

et al., 1983). Each could phosphorylate histone: one

was stimulated by 10 µM cAMP, cGMP and cIMP,

a second was inhibited by these nucleotides, and the

third was cAMP-independent but sensitive to cGMP

and cIMP. The protein extract showing cAMP-

stimulated protein kinase activity was also found to

contain a cAMP-binding protein, but a possible

interaction between the two proteins was not re-

ported.

cAMP-dependent protein kinase was also shown

in Zea mays seedlings (Janistyn, 1988). This electro-

phoretically purified protein of molecular mass 36

kDa had a strong dependence on MnSO%: replace-

ment of MnSO%

by NiSO%, CoSO

%or FeSO

%

abolished the cAMP dependence of the kinase

activity. In contrast to the data from Lemna (Kato et

al., 1983) other cyclic nucleotides (cGMP, cIMP,

cCMP and cUMP) did not exhibit any stimulating

effect on the protein kinase activity. Janistyn (1989)

has also described the cAMP-dependent phosphory-

lation of protein components present in dialysed

coconut milk. At 1–10 µM, cAMP enhances the

phosphorylation of two endogenous proteins of

molecular masses 60 and 70 kDa, this phosphory-

lation being inhibited by a protein component

(molecular mass 9 kDa) present in non-dialysed

coconut milk. The apparent complexity of the

coconut milk system is illustrated by the fact that

cAMP also inhibited the phosphorylation of other

endogenous proteins with molecular masses between

27 and 30 kDa, implying that more than one cAMP-

responsive kinase was present, that the specificity of

a single kinase was altered by cAMP, or that cAMP

in some way directly or indirectly stimulates protein

phosphatase activity.

More recently, Komatsu & Hirano (1993) identi-

fied a cAMP-stimulated protein kinase activity in

rice leaves and roots. At 10 nM, cAMP enhanced the

phosphorylation of histone II-A, and the phosphory-

lation of three endogenous proteins (molecular

masses 40, 50 and 55 kDa) of rice seedlings 11 d old



was specifically stimulated by the addition of cAMP.

cGMP, phorbol ester or Ca#+ did not induce the

same effect. A similar cAMP-responsive phosphory-

lating activity was present in the rice embryo and

roots.

In each of above cases the kinase substrate was

either a histone or an endogenous, unidentified,

protein. The evidence relating to a partly purified

protein from Petunia petals (Polya et al., 1991) is

perhaps more convincing on the basis of substrate

information. A basic protein fraction (molecular

mass 30 kDa) phosphorylated both histone III-S

and Kemptide, a specific substrate for cAMP-

dependent protein kinase. Its activity was inhibited

by both the Walsh-Krebs inhibitor peptide and the

regulatory subunit of cAMP-dependent protein

kinase from beef heart (50% inhibition at 0.4 µg}ml),

with the latter inhibition being completely abolished

by the addition of 3 µM cAMP. The apparent Km

value of the Petunia Kemptide kinase for Kemptide

was 24 µM as opposed to 3.6 µM for its mammalian

counterpart in beef. The authors did not find a

regulatory subunit and therefore concluded that the

Petunia Kemptide kinase was regulated differently

from the animal protein kinase. Although it is

conceivable that regulatory and catalytic subunits

have separated during the course of the purification,

it is also possible, given the apparent broad specificity

of for example the Lemna cyclic-nucleotide-respon-

sive protein kinase, that the plant kinase is more

similar to the mammalian cGMP-dependent kinase,

which does not dissociate into separate subunits.

However, their finding that a mammalian regulatory

subunit inhibits activity, an effect that can be

alleviated by the addition of cAMP, possibly activa-

ting dissociation, would strongly suggest a model

similar to the animal cAMP-dependent protein

kinase.

The application of modern molecular biology

techniques has provided further evidence of cyclic-

nucleotide-responsive protein kinases in plants. A

number of cAMP-dependent protein kinases from

various organisms have been cloned and charac-

terized and show a well-conserved primary structure.

Recently, molecular biological evidence was put

forward that indicated the presence in higher plants

of protein kinases with high homology with cyclic-

nucleotide-dependent protein kinases from other

organisms. Although the indications discussed below

are based solely on sequence similarity and still need

to be consolidated by biochemical evidence, they

nevertheless constitute a significant step forward.

Lawton et al. (1989) isolated some candidate

serine}threonine protein kinase genes from Phaseolus

vulgaris and Oryza sativa cDNA libraries. PVPK-1

(bean) and G11A (rice) have high homology with the

catalytic subunit of both protein kinase A and protein

kinase C. The open reading frames of these cDNA

species contain all except one signature of serine}

threonine protein kinase; all invariant amino acids

necessary for ATP binding and phosphotransfer are

present. Thr-197, which is autophosphorylated in

cAMP-dependent protein kinase, was replaced by a

serine residue, which can serve the same purpose.

The presence of the sequence DLKPEN (single-

letter amino acid codes) (DLKLDN in protein

kinase C) and a GTHEYLAPE sequence (GTPEY-

LAPE in cyclic-nucleotide-dependent protein kin-

ases and GTPDYIAPE in protein kinase C) points

to a greater resemblance to the cyclic nucleotide

kinases than to protein kinase C. The plant sequence

carries an additional strand of 79–81 nucleotides

between these sequences. This insertion was also

found in Saccharomyces cerevisiae without apparent

loss of catalytic activity. The amino termini of the

highly conserved region comprising these sequences

in both cDNA species carry 238 additional amino

acid residues. These regions differ significantly

between both enzymes and probably contain regu-

latory sequences. This region contained no similarity

to other enzymes. Highly homologous protein

kinases were found in Zea mays, 90.7 (Biermann et

al., 1990); Pisum sativum PsPK1 (Lin et al., 1991)

and Arabidopsis thaliana Atpk7 (Hayashida et al.,

1992), Atpk64 (Mizoguchi et al., 1992) and Atpk5

(Hayashida et al., 1993). Their DNA sequences

contain one open reading frame carrying all con-

served residues occurring in serine}threonine kinases

except that corresponding to Thr-197; there, as in

Phaseolus vulgaris and Oryza sativa, they carry a

serine residue. All show the amino-terminal can-

didate regulatory domain comprising 108–238 amino

acid residues, rich in threonine and serine residues as

potential phosphorylation sites. They show the

highest degree of homology with cAMP-dependent

protein kinase and are categorized as cAMP-de-

pendent protein kinases solely on this basis (Hunter,

1991).

Although there is thus an increasing literature

indicating cyclic-nucleotide-responsive protein kin-

ases in plants, it is the identification of the protein

substrates that will provide the best clues as to which

physiological processes are governed by cAMP. The

phosphorylation of a substantial number of plant

proteins is influenced by cAMP. For example, Li

et al. (1994) have recently described an outward-

rectifying K+ channel in Vicia faba that is regulated

by cAMP. This regulation is dependent on a

phosphorylation event; additional evidence for a

cAMP-dependent protein kinase is that the event

can be mimicked by addition of a mammalian protein

kinase A. In fact, a number of plant proteins are

substrates for mammalian protein kinase A. Phyto-

chrome (Wong et al., 1986), phosphoenolpyruvate

carboxylase (Terada et al., 1990) and sucrose

phosphate synthase (Huber & Huber, 1991) are

examples of proteins that are influenced in their

function by such phosphorylation. Though they are



not in themselves sufficient to postulate the existence

of the same mechanism in plants, these data provide

a valuable indication of which processes are sus-

ceptible to regulation through a cAMP-dependent

phosphorylation.

As cited in section 1, cAMP regulates a con-

siderable number of mammalian physiological pro-

cesses by interfering with gene expression. Induction

is generally fast and independent of intermediate

protein synthesis de novo. In a search for cis-acting

sequences governing the sensitivity of somatostatin

gene expression to cAMP, a short palindromic

sequence motif (5«-TGACGTCA-3«) was found that

was highly conserved in the promoter region of

many cAMP-induced genes (Montminy et al., 1986).

This CRE is a classic enhancer sequence in that it

stimulates transcription irrespectively of its orien-

tation and distance from the transcription starting

point ; when placed upstream of a normally non-

induced gene, cAMP inducibility is introduced.

Further analysis of cAMP-inducible promoters

revealed the presence of a smaller functional motif

(5«-CGTCA-3«) that is present in two perfect copies

on one or both of the DNA strings (Fink et al.,

1988). The potential of CREs to act synergistically

when placed in tandem indicates that the number of

CGTCA motifs in the promoter region is an

important factor for the degree of transcriptional

inducibility by cAMP.

In animal cells, CREs are recognized by members

of the DNA-binding protein family called CREB}ATF (CREB}activating transcription factor). Pro-

teins belonging to this family consist of two domains.

A carboxy-terminal domain with a basic leucine

zipper (‘bZIP’) motif is involved in the sequence-

specific binding of DNA and dimerization, whereas

the activation domain possesses regions interacting

with other components of the transcription ma-

chinery and other signal transduction chains.

CREB}ATF proteins bind as dimers to conserved

CREs. The formation of heterodimers between

CREB (Yamamoto et al., 1988; Gonzalez & Mont-

miny, 1989), ATF-1 (Hurst et al., 1991) and CRE-

modulator protein (‘CREM’) (Foulkes et al., 1991)

yields a number of different transcription regulators

affecting transcription in a positive or a negative

manner. CREB}ATF proteins also dimerize with

members of the AP-1 transcription factor family

such as Jun and Fos (Hai et al., 1989). Eventually, an

abundant number of new tissue-specific and func-

tionally different homodimer and heterodimer com-

plexes can be formed, leading to extremely accurate

cell-specific regulation.

Even in mammals the exact mode of action of

CREB is still not completely known. An important

prerequisite for its action is a cAMP-dependent

phosphorylation of CREB at Ser-133 (Gonzalez &

Montminy, 1989). This phosphorylation is necessary

for interaction with a CREB-binding protein (CBP)

(Kwok et al., 1994) and this CREB-CBP complex

then activates transcription, although an an ad-

ditional cAMP-dependent phosphorylation of CBP

might be necessary for the interaction with CREB

(Parker et al., 1996).

The first indications of the presence of CREB in

higher plants come from a study by Inamdar et al.

(1991). Radioactively labelled oligonucleotides bear-

ing the CRE motif were recognized by DNA-binding

proteins in pea, soybean, cauliflower and wheat.

Methylation of the oligonucleotide in the CRE motif

resulted in a marked loss of binding capacity that

was more pronounced than that in animal systems

(Iguchi-Ariga&Schaffner, 1989). For this reason this

protein was named MIB-1 (methylation-inhibited

binding protein 1). In 1992 the same group isolated

a cDNA clone from Vicia faba showing extreme

resemblance to the animal CREB protein, both in

sequence and in biochemical properties (Ehrlich et

al., 1992). As in the animal protein, the amino acid

sequence derived from the VBP1 protein contains a

basic domain located next to a leucine zipper motif.

The protein shows greatest homology with a tobacco

DNA-binding protein, TGA1a, that was originally

isolated as a CREB homologue (Katagiri et al.,

1989), but in contrast with TGA1a the preferred

DNA-binding motif of VBP1 was a perfect -TG-

ACGTCA- palindrome. Methylation of the CRE

motif destroyed the binding capacity of VBP1 almost

completely. Although VBP1 lacks the PKA-phos-

phorylation site characteristic of CREB (R-R-X-

S}T (Gonzalez et al., 1989)), it still is a good

candidate for a modulator of cAMP-mediated gene

expression in higher plants. It contains three de-

generate phosphorylation sites for a cAMP de-

pendent protein kinase (R-X-S}T (Kennelly &

Krebs, 1991)). Recently, two cDNA fragments were

amplified by PCR in Cichorium intybus having

extensivehomologywithTGA1andVBP1 (Messiaen

& Van Cutsem, 1996).

Plant genes have been shown to carry CRE motifs

in their promoter region: the motif in a soybean

proline-rich protein has a perfect palindrome (Hong

et al., 1987); the sweet-potato α-amylase gene carries

a tandem repeat CGTCA motif in its promoter

region, and an analysis of its binding factor indicates

that it actually prefers the perfect palindrome

(Ishiguro et al., 1993). Unfortunately, at present, a

role for cAMP in their regulation is merely specu-

lative because data on the induction of these genes by

cAMP are lacking.

As stated earlier, cAMP does not act solely

through the phosphorylation of proteins. Olfactory

perception, for example, is governed by a direct

binding of cAMP to cyclic-nucleotide-gated chan-

nels. A search for cAMP-binding activity in plants

readily yielded cAMP-specific binding activities

without accompanying protein kinase activity. A

Helianthus binding protein was very specific for



cAMP and 8-bromo-cAMP. 5«-AMP and other 3«,5«-cyclic nucleotides did not bind to this protein

(Giannattasio et al., 1974a). High-affinity binding

activities were also found in Phaseolus and Hordeum

(Brown et al., 1979a; Smith et al., 1979). The partly

purified binding protein from Hordeum (Brown et

al., 1980b) was highly specific for cAMP; it had a

molecular mass of 170 kDa, a pH optimum of 6.5

and a Kd

of 8 nM; this last value is very similar to

that for mammalian cAMP-binding proteins (2–3

nM). The cAMP-binding activity was found in a

protein fraction that also contained glucose-6-phos-

phatase, ATPase, 5«-nucleotidase and fructose-1,6-

diphosphatase activity; cAMP binding had no effect

on glucose phosphatase or fructose diphosphatase

activities, but acted as a weak competitive inhibitor

of ATPase and a mixed inhibitor of nucleotidase.

Previously, 5«-nucleotidases had been isolated from

Triticum vulgare and Solanum tuberosum that were

competitively inhibited by cAMP with micromolar

Kivalues (Polya & Ashton, 1973; Polya, 1975).

The same group have also described phosphatase

activities in Beta vulgaris and Solanum tuberosum

that were inhibited by cyclic nucleotides (Polya &

Hunziker, 1987; Polya & Wettenhall, 1992). Both

enzymes, with molecular masses of approx. 30 kDa,

were inhibited by cGMP and cAMP. The Kdvalues

of the Beta enzyme for cGMP and cAMP were 0.4

and 3.3 µM respectively, and those of the Solanum

enzyme were 0.8 and 2.1 µM respectively. Both

enzymes catalyse the dephosphorylation of nucleo-

side monophosphates and O-phosphotyrosine, but

not O-phosphoserine and O-phosphothreonine; they

are thought to be involved in the dephosphorylation

of tyrosine-phosphorylated proteins.

The use of techniques of molecular biology has

also indicated binding sites for cAMP relating to ion

channels, for example two channels in Arabidopsis

thaliana (KAT1 and AKT1) belonging to the Shaker

superfamily of K+-channels. Within this family,

KAT1 (Anderson et al., 1992) and AKT1 (Sentenac

et al., 1992) show greatest similarity to the eag

channel in Drosophila (Warmke et al., 1991), which

itself bears a strong sequence similarity to cyclic-

nucleotide-modulated channels (Guy et al., 1991).

The eag channel is regulated by intracellular cAMP

(Bru$ ggemann et al., 1993). Apart from a hydro-

phobic domain consisting of six putative trans-

membrane stretches also found in the animal K+

channels, both plant genes possess a putative cyclic-

nucleotide-binding domain. Although not yet

proved, a direct interaction of cGMP with this

binding domain might explain the sensitivity of the

channels to cGMP (Hoshi, 1995; Gaymard et al.,

1996). Although KAT1 activity is thus influenced by

cGMP (see section VIII), cAMP and protein kinase

A did not produce the same effect (Hoshi, 1995);

however, the data from a K+ channel in carrot

(Kurosaki, 1997) do suggest an involvement of

cAMP. Treatment of carrot cells with dibutyryl-

cAMP or the adenylyl cyclase activator forskolin

produced an appreciable, but transient, decrease in

extracellular [K+] which could be inhibited by K+-

channel blockers. An increase in intracellular [Ca#+]

in response to cAMP was also inhibited by K+-

channel blockers but was stimulated by a K+ current

evoked by an ionophore.

These data have led to the proposal that the K+

channel has a role in cross-talk between cAMP and

Ca#+, with the gating of some of the inward K+

channels located at the plasma membrane controlled

by intracellular [cAMP], and the increased K+

current in response to elevated [cAMP] eliciting a

Ca#+ influx into the cells possibly by the activation of

voltage-dependent Ca#+ channels (Kurosaki, 1997).

VII. C

A large number of physiological processes in plants

are potentially sensitive to alterations in [cAMP], as

suggested by reports of the effects of cell-permeating

cAMP analogues and of alterations in [cAMP] and

cAMP-related enzyme activities during physiologi-

cal events (Brown & Newton, 1981; Newton &

Brown, 1986; Assmann, 1995; Bolwell, 1995).

However, it is questionable whether many of these

observations have any relevance in vivo. Reliable

indications of cAMP functions in several plant

processes are beginning to emerge, for example a role

in ion transport (section VI). It can also be argued

that a role in the chloroplast is likely, given the

apparent presence of the complete cAMP machinery

in this organelle. In this context it is interesting that

in the experiments demonstrating that cGMP has a

pivotal role in phytochrome phototransduction (as

discussed in section VIII) the effect of the micro-

injection of cAMP and Ca#+ into phytochrome-

deficient mutants was greater than that of cGMP and

Ca#+, and that the effects of cAMP derivatives

analogous to Rp-cGMPS (Rp-Guanosine 3«,5« cyclic

monophosphothioate) and GTPγS (Guanosine 5«-O-

3-thiotriphosphate), which are instrumental to the

demonstration of the involvement of cGMP, were

not examined (Bowler et al., 1994a).

Other systems have emerged as being very prom-

ising with regard to a physiological role for cAMP.

Ehsan et al. (1998) report on fluctuations in [cAMP]

that are tightly connected to cell cycle progression in

tobacco BY-2 cells. Peaks of [cAMP] were observed

during S-phase and to a smaller extent in G"(Fig. 4).

The addition of indomethacin, a drug that inhibits

adenylyl cyclase activity (Wang et al., 1978), resulted

in the loss of the peak in S-phase, an event that was

accompanied by a marked decrease in mitotic

activity. The authors postulate the existence of an

adenylyl cylase activated by prostaglandin or prosta-

glandin-like compound (e.g. jasmonic acid) that is

highly regulated during the cell cycle.



120

80

40

0

30

20

10

0

H4

mR

NA

an

d m

ito

tic

ind

ex (

%)

[cA

MP

] (p

mo

l g–1

f.w

t)

0 2 6 8 10 124Time after aphidicolin release (h)

G2 G1MS

Fig. 4. Fluctuation of cAMP levels during the cell cycle of

aphidicolin-synchronized tobacco BY-2 cells. (Modified

from Ehsan et al. (1998).) Circles, H4 mRNA; squares,

mitotic index; bars, [cAMP].

The significance of these cAMP fluctuations

becomes apparent on comparison with cell cycle

features in other organisms. cAMP is prominent in

cell cycle control in animal and fungal systems; its

concentration fluctuates during cell cycle progression

(Negishi et al., 1982) and, depending on the cell

type, it exhibits stimulatory or inhibitory effects on

cell proliferation (Dumont et al., 1989; Magnaldo

et al., 1989; Roger et al., 1995). A transient rise in

cAMP before the onset of S-phase is part of a series

of events leading to DNA synthesis (Boynton &

Whitfield, 1983). The expression of key regulators of

the cell cycle such as cyclin A, cyclin D and cyclin E

is affected by cAMP (Desdouets et al., 1995; Barlat

et al., 1995; Ward et al., 1996; L’Allemain et al.,

1997). Forskolin and 8-Br-cAMP inhibit cyclin A-

and cyclin E-dependent histone H1 kinase activity in

an astrocytic cell line (Gagelin et al., 1994). With

cyclin D"a direct phosphorylation in the cyclin box

by a cAMP-dependent protein kinase is thought to

regulate its activity (Sewing & Mu$ ller, 1994). The

Saccharomyces cerevisiae cell cycle is also highly

regulated by the Ras}cAMP signal transduction

pathway (Baroni et al., 1994; Tokiwa et al., 1994).

The same seems to apply to organisms close to

higher plants. Sakuanrungsirikul et al. (1996) iso-

lated three Chlamydomonas reinhardtii mutants

showing cell cycle arrest in G"when grown at a non-

permissive temperature. This cell cycle arrest could

be attributed to a decreased adenylyl cyclase activity

and cAMP content of the cells. Blocked cells could

be rescued by the addition of cAMP, indicating a

function of the cyclic nucleotide in cell division.

cAMP was also shown to be a key regulator of the

circadian-rhythm-driven cell cycle of the unicellular

alga Euglena gracilis (Carre! & Edmunds, 1992, 1993;

Edmunds, 1994). It is believed to form the link

between the internal clock and the cell cycle by

negotiating the transition through the G"}S and

G#}M boundaries. A study of adenylyl cyclase and

phosphodiesterase revealed that both activities

showed oscillating changes occurring in counter-

phase (Tong et al., 1991). The changing adenylyl

cyclase activity is probably due to the regulation of a

constant enzyme pool by an oscillating modulator,

rather than through a change in the amount of

enzyme. At all times of the day the adenylyl cyclase

could be stimulated to the same maximal activity

with forskolin. The addition of forskolin also

resulted in a dampening of the amplitude of the

cAMP oscillations and a loss of rhythmicity in cell

division. Experiments with isobutylmethylxanthine

showed that phosphodiesterase activity was inhibited

in a manner dependent on the time of application.

However, isobutylmethylxanthine is an inhibitor

that influences the activity of various phosphodi-

esterases, each to a different extent. The difference in

efficiency of isobutylmethylxanthine addition was

therefore a result of the presence of heterogeneous

types of phosphodiesterase during the cell cycle. One

specific phosphodiesterase (or a small subset) might

be responsible for the oscillations in cAMP content.

In a search for the clock directing cAMP content,

Tong & Edmunds (1993) studied the role of Ca#+,

calmodulin, inositol 1,4,5-trisphosphate (Ins(1,4,5)

P$) and cGMP in the regulation of adenylyl cyclase

and phosphodiesterase. The cGMP content showed

oscillations that preceded the oscillations in cAMP

by 2 h. The cGMP analogue 8-Br-cGMP and the

guanylyl cyclase inhibitor LY83583 (6-aniloquino-

line-5,8-quinone) also had an effect on the activities

of adenylyl cyclase and phosphodiesterase. cGMP is

therefore a good candidate for a mediator of cAMP

metabolism.

The effect of cAMP on the cell cycle is dependent

on the time of addition. The cell cycle was delayed

by the addition of cAMP during the subjective day

and enhanced during the night. This can be

explained by the presence of different receptors for

cAMP that selectively regulate one or the other of

two regulatory pathways. Two types of cAMP-

dependent protein kinase have been revealed in

Euglena (cPKA and cPKB) with differing affinities

for cAMP and some analogues (Carre! & Edmunds,

1992). The use of specific cAMP analogues activating

the two kinases in an uneven fashion revealed that

the two kinases had different effects on the cell cycle:

cPKA delayed the cell cycle; cPKB promoted it

(Carre! & Edmunds, 1993). According to Edmunds

(1994), an increase in [cAMP] inhibits DNA syn-

thesis, keeping the cell in G#, thereby resulting in the

inhibition of cell division during the day. As soon as

[cAMP] decreases this block is abolished. The G#}M

transition, or maybe mitosis itself, is stimulated by a

second peak of [cAMP]. In this way, mitosis is put

into phase with night. The delaying effect of cAMP

during the subjective day is regulated by the

activation of cPKA; the stimulation of mitosis during



the night is regulated by cPKB. The activation of

one of the two results in the phosphorylation of a

specific set of target proteins. Possibly the expression

of cPKA and cPKB differs during the cell cycle, with

cPKA being expressed during the day and cPKB

during the night, but it is equally possible that the

expression of the protein targets varies in such a way

that cPKA activation can exert its effect only during

the day and cPKB only during the night. The search

for targets for cAMP-dependent phosphorylation is

under way (Edmunds, 1994).

There is a strong argument that it is very unlikely

that a compound with such an impact on a function

of the cell as fundamental as division would have

been lost during evolution from lower plants to

higher plants. Indeed, most key regulators of the

(supposedly more distant) animal cell cycle seem to

be conserved in higher plants. The occurrence of

cyclins and cyclin-dependent protein kinases in

higher plants similar to those found in animals is

now well established (Ferreira et al., 1991, 1994;

Hemerly et al., 1992; Reichheld et al., 1995, 1996),

and retinoblastoma protein homologues (Murray,

1997; Murray et al., 1997; Huntley et al., 1997) and

cdc25 homologues (Sabelli et al., 1998) have been

isolated recently. Given this overall conservation of

cell cycle machinery and circumstantial evidence

such as the fact that in a number of cyclins the

primary structure reveals signatures for post-trans-

lational modifications dependent on cAMP (anal-

ogous to cyclin D"), the observation made by Ehsan

et al. (1998) demands a further search for a role for

cAMP in the plant cell cycle.

There is also strong evidence that cAMP is

involved in the plant defence response that produces

phytoalexins. This stress response system has clear

analogies to the mammalian secondary-messenger

system, involving an extracellular signal, a receptor,

a signal transduction system and a metabolic re-

sponse (Smith, 1996). Several extracellular elicitors

have been identified, including polysaccharides,

oligosaccharides, oligogalacturonides, fatty acids,

proteins and glycoproteins; few receptors for these

elicitors have been identified, but those known so far

are proteins located on the plasma membrane. The

response to the stimulus at these receptors is the

activation of specific defence response genes, in-

cluding those leading to the production of enzymes

required for phytoalexin synthesis. The mechanism

by which the perceived signal at the plasma mem-

brane is relayed to the nuclear genes has been

variously cited to involve Ca#+, jasmonic acid, active

oxygen, diacylglycerol and inositol phosphates, and

cAMP (Smith, 1996).

The first evidence implicating the involvement of

cAMP in a cellular defence mechanism was the

activation of phytoalexin synthesis in Ipomoea

(Oguni et al., 1976); similar effects were later

reported in carrot (Kurosaki et al., 1987; Kurosaki &

Nishi, 1993) in which not only did cAMP enhance

phytoalexin accumulation, but also activators of

adenylyl cyclase, cholera toxin and forskolin, and a

phosphodiesterase inhibitor, theophylline, also en-

hanced phytoalexin synthesis with a concomitant

rise in intracellular [cAMP]. Compelling evidence

for the involvement of cAMP is available from a

stress reaction of Medicago, in which the phytoalexin

medicarpin is synthesized in response to a challenge

from a glycoprotein elicitor from the fungus Verti-

cillium albo-atrum (Smith, 1991). There is also a

significant increase in the activity of phenylalanine

ammonia-lyase (PAL), which catalyses an early

reaction in the committed phase of phytoalexin

synthesis. Treatment of Medicago seedlings with

dibutyryl-cAMP brought about an enhancement of

PAL activity and an induction of medicarpin

synthesis (Cooke et al., 1989, 1994); in cell sus-

pension cultures a 4–5-fold increase in cAMP,

unequivocally identified by FAB}MIKE spectrum

analysis, was observed after challenge by the elicitor.

Although in agreement with the trend of the data

from carrot cells (Kurosaki et al., 1987) the response

was much higher and more rapid in Medicago,

reaching a maximum 3–5 min after challenge, in

contrast to the 30 min required in carrot and the

15 min maximum response time in French bean

(Bolwell, 1992). In Medicago, the time course of

adenylyl cyclase stimulation and decay followed by a

phosphodiesterase elevation (Fig. 5a) is a classic

example of a switch-on}switch-off system; arguably

the most significant finding is that there is a dose-

dependent response to the elicitor (Fig. 5b) (Cooke

et al., 1994). This dose-dependent response of a

particulate adenylyl cyclase preparation in Medicago

indicates a direct effect by the elicitor either on the

cyclase itself or on an adjacent regulator. The effect

of cholera toxin in stimulating phytoalexin synthesis

in carrot (Kurosaki et al., 1987) and PAL activity

in bean cultures (Bolwell, 1992) suggests that a

G-protein might well be involved.

Although the data above give a clear indication

that cAMP is involved in the phytoalexin defence

response, it is not yet known how cAMP triggers off

a change in phytoalexin synthesis. In several plant

species inositol phosphates are also implicated in this

process (Smith, 1996): in Medicago a very rapid rise,

within 1 min of challenge, in [Ins(1,4,5)P$] in

response to the V. albo-atrum elicitor was observed,

returning to control levels within 3 min. This is

consistent with changes in [Ca#+] in elicitor-treated

cells (Walton et al., 1993). A release of Ins(1,4,5)P$

is potentially capable of releasing Ca#+ from internal

stores; there could therefore conceivably be a

mechanism involving Ca#+}calmodulin-dependent

protein kinases and protein kinase C. The great

sensitivity of the adenylyl cyclase to Ca#+ levels

(Cooke et al., 1994), together with the effect of

cAMP on ion fluxes (Kurosaki, 1997), suggests that



400

300

200

100

0

[cA

MP

] an

d e

nzy

me

acti

vity

(%

)

0 25 50 75 100

Duration of incubation (min)

(a)

0.7

0.6

0.5

0.4Ade

nyly

l cyc

lase

(pm

ol m

in–1

mg–1

)

0 150 300 750 900

Carbohydrate elicitor (lg)

(b)

Fig. 5. cAMP in the stress response of Medicago. (a) Response of cAMP (solid line), adenylyl cyclase (broken

line) and cyclic nucleotide phosphodiesterase (dotted line) in Medicago cell culture to challenge by fungal

elicitor. (Modified from Cooke et al. (1994).) (b) Dose-response curve of partly purified Medicago adenylyl

cyclase to fungal elicitor. (Modified from Cooke et al., 1994.)

at least part of the cAMP effect might result from

cross-talk between the adenylyl cyclase}cAMP and

phophoinositide}Ca#+ signalling pathways.Although

there is no apparent change in cAMP-responsive

protein kinase in response to elicitor in carrot cells,

there is a transient activation of Ca#+ and calmodulin-

dependent kinase activity in response to dibutyryl-

cAMP or forskolin, leading to a suggested mech-

anism in which the elicitor-induced synthesis of

cAMP leads to an influx of Ca#+, which in turn

stimulates a response to elicitor by increasing kinase

activity (Kurosaki & Nishi, 1993; Kurosaki, 1997).

VIII. c

When other cyclic nucleotides are considered it is

found that, unlike the comparisons between cAMP

in mammalian and higher-plant systems, there are

clear, established and apparently accepted parallels

between the functions of cGMP in plants and

mammals. Furthermore it could be argued that a

similar paucity of knowledge is available for cyclic

nucleotides other than cAMP and cGMP in mam-

mals and higher plants.

The first evidence for the occurrence of a second

cyclic nucleotide in a higher plant was the report that

changes in cGMP concentration took place during

cell enlargement and division in excised pith tissues

of Nicotiana (Lundeen et al., 1973). cGMP was

further reported in bean root tissue (Haddox et al.,

1974), pine pollen (Katsumata et al., 1978), in

tumour-prone Nicotiana amphyloid (Ames et al.,

1980) and in relatively high concentrations in Evodia

and Zizyphus (Cyong & Takahashi, 1982a,b,c;

Cyong et al., 1982). However, the means of identi-

fication of the putative cGMP was open to the same

criticism as was levelled at the identification of

cAMP discussed above.

To establish the identity of the putative plant

cGMP unequivocally, large-scale extracts from

French Dwarf bean seedlings were purified by

sequential chromatography and identified by physi-

cal analyses (Newton et al., 1984a). The UV

absorbance spectrum of the isolated compound had

the characteristic lmax

values of 255, 252 and 259 nm

at pH 1, 7 and 11 respectively; the NMR spectrum

was essentially identical to that of a cGMP standard,

with the large shift between the protons at positions

8 and 1 that is characteristic of a cyclic nucleotide.

These data strongly supported the identification of

the analyte as cGMP but did not establish it

unambiguously; however, the use of FAB-MS with

CID}MIKE spectrum scanning provided conclusive

evidence (Newton et al., 1984a). The protonated

molecule of the extracted putative cGMP (m}z 346)

yielded a CID}MIKE spectrum containing the

diagnostic fragments at m}z 152, 180 and 194

(Kingston et al., 1984), essentially identical to that of

the cGMP standard. It was clearly distinguishable

from that of the 2«,3«-cyclic GMP isomer, from

which the m}z 194 peak was absent and there were

different peaks at m}z 195, 214 and 217, the two

cGMP isomers undergoing analogous fragmenta-

tions to their cAMP counterparts. Evidence was

produced at the same time of cGMP in maize

seedlings by GC-MS (Janistyn, 1983).

The quantification of the cGMP extracted from

French Dwarf bean seedlings illustrated the pitfalls

of plant cyclic-nucleotide estimations, in that

whereas samples purified to remove non-nucleotide

and non-cyclic-nucleotide material gave the correct

linear binding curves in the RIA (radioimmune-

assay), less purified extracts gave falsely high positive

readings (Newton et al., 1984a). With a purified

extract the chloroplast was calculated to contain 3.3

pmol g−" cGMP, giving a ratio of substrate con-

centration to Km

for the chloroplast cGMP phos-

phodiesterase (see section VI) comparable to that

found in vertebrate cells. In addition, the chloroplast



contained guanylyl cyclase activity capable of con-

verting radiolabelled GTP to cGMP (Newton et al.,

1984a); the product of the enzyme activity has been

identified by its FAB}MIKE spectrum (Newton,

1996).

Although comparatively few investigations of the

function of cGMP in higher plants have been carried

out, they have met with significant success. As

detailed above, cGMP has a substantial role in the

responses to light within the mammalian eye, and

evidence has accumulated of an analogous role for

cGMP in the responses of higher plants to light.

Initially, dibutyryl-cGMP was found to induce

changes in the plastid terpenoid components of dark-

grown Spinacea that mimicked the effects of phyto-

chrome and were not induced by dibutyryl-cAMP

(Brown et al., 1989). Similarly, illuminating Lemna

with far-red light increased the level of cGMP, and

dibutyryl-cGMP stimulated flowering (Hasunuma

et al., 1988).

These data conflict with evidence from lichens

that suggests, for example, that in Cladonia verti-

cillaris the phytochrome-induced accumulation of

phenolics such as fumarprotocetraric acid is strongly

enhanced by red light or exogenous cAMP, with red

light increasing the endogenous synthesis of cAMP

(Mateos et al., 1993). These observations led to the

proposal of a mechanism involving the activation of

adenylyl cyclase by phytochrome, in which the

subsequent increase in intracellular [cAMP] initiated

a cascade process similar to that effected by rhodopsin

in the mammalian eye, with cAMP rather than

cGMP opening ion channels by the activation of a

protein kinase (Vicente, 1993). However, strong

evidence that cGMP is involved in the action of

phytochrome has been produced by single-cell assays

in a phytochrome-deficient tomato mutant (Bowler

et al., 1994a). After microinjection of putative

signalling molecules into hypocotyl cells of the

tomato mutant aurea, which lacks type 1 phyto-

chrome but has normal Phytochrome B activity

(Sharma et al., 1993), products of phytochrome

signalling could be monitored. Although these cells

did not develop chloroplasts or synthesize antho-

cyanins in response to light, microinjection of

exogenous type 1 phytochrome restored chloroplast

development and anthocyanin production (Neuhaus

et al., 1993) and it was possible to activate the

expression of a photoregulated reporter gene injected

with the photoreceptor (Bowler et al., 1994a).

Stimulation of anthocyanin biosynthesis and chloro-

plast development involved the participation of one

or more heterotrimeric G-proteins, together with

Ca#+ and calmodulin acting further down the

pathway. Although G-protein activation mediated a

full cellular response, in its absence Ca#+ and

calmodulin alone did not activate anthocyanin bio-

synthesis and produced incompletely developed

chloroplasts. It is therefore cGMP that triggers the

production of anthocyanins; together with Ca#+ it

activates the full development of chloroplasts and it

has been demonstrated by means of reporter genes

that cGMP and Ca#+ act primarily by modulating

gene expression (Bowler et al., 1994a). Subsequently

three signal transduction pathways, dependent on

cGMP or Ca#+, or both, were identified as the means

by which phytochrome controls the expression of

genes required for anthocyanin biosynthesis and

chloroplast development (Bowler et al., 1994b); chs

is controlled by a cGMP-dependent pathway, cab by

a Ca#+-dependent pathway and fur by a cGMP and

Ca#+-dependent pathway. Cross-talk occurs between

the pathways: cGMP concentration changes mediate

the induction and desensitization of the chs gene to

light, and high cGMP levels downregulate both the

Ca#+-dependent and the Ca#+}cGMP pathways

(Bowler et al., 1994a,b). A further gene, asparagine

synthetase, has been shown to be downregulated by

light, being expressed in the dark and repressed in

the light. The repression of asparagine synthetase in

the light seems to be controlled by the Ca#+}cGMP

pathway, which activates other light responses

(Neuhaus et al., 1997), indicating that the same

signal transduction pathway can both activate and

repress different responses to phytochrome. By the

use of complimentary loss-of-function and gain-of-

function studies a 17 base pair cis element within the

asparagine synthetase promoter has been identified

that is the target for a highly conserved phyto-

chrome-generated repressor regulated by both Ca#+

and cGMP (Neuhaus et al., 1997).

In addition to its action in mediating phytochrome

responses, cGMP has been proposed to have a role in

GA$-induced gene expression in barley aleurone

layers. In a manner consistent with the suggestion

that cGMP was involved in ion channel regulation

(Hoshi, 1995), the K+ channels KAT1 (Anderson

et al., 1992) and AKT1 (Sentenac et al., 1992) in

Arabidopsis and the cation channel HvCBT-1 in

barley aleurone layers (Schuurink et al., 1998) have

a cGMP-binding motif in the C-terminal region.

The barley aleurone layer, as a secretory tissue

metabolically regulated by hormones with many

examples of GA-ABA antagonism, has provided an

excellent system for exploring the role of cGMP in

plant signal transduction. Several studies have

demonstrated that the site of GA perception is at the

plasma membrane (Hooley et al., 1991; Gilroy &

Jones, 1994) and that elevations in cytosolic [Ca#+]

(Bush, 1996) and [calmodulin] (Schuurink et al.,

1996) are early events in signal transduction. GA$

was further shown to elevate [cGMP], which was

unaffected by ABA, and a guanylyl cyclase inhibitor

(LY83583) blocked both GA-induced [cGMP] el-

evation and α-amylase secretion (Penson et al.,

1996a,b). This effect of LY83583 on α-amylase

secretion could be reversed by cell-permeating

derivatives of cGMP, and whereas LY83583 in-



hibited the GA-induced synthesis of GAMyb, an

activator of the high-pI α-amylase promoter (Gubler

et al., 1995), it had little effect on the ABA-induced

accumulation of RAB21 (a characteristic response to

ABA) (Mundy & Chua, 1988), and led to the

conclusion that cGMP is an important early com-

pound in the response of cereal aleurone layer to

GA$

but not that to ABA (Penson et al., 1996a,b,

1997).

A further apparent interaction between Ca#+ and

cGMP in plants lies in the anaerobic signal trans-

duction pathway and provides another interesting

analogy with the role of cGMP in the mammalian

anaerobic response (Depre & Hue, 1994). Ca#+

mobilization from intracellular stores into the cyto-

plasm is necessary for the mRNA transcription of

anaerobic proteins such as alcohol dehydrogenase

and sucrose synthetase (Subbaiah et al., 1994a,b),

and in the anaerobic accumulation of γ-amino-

butyrate (Aurisano et al., 1995, 1996). Whereas

[cAMP] decreased, the imposition of anaerobic

conditions resulted in a rapid transient increase in

[cGMP] in both root and coleoptile of rice seedlings

(Reggiani, 1997), which led to a proposed role for

both cyclic nucleotides in shutting down ATP-

dependent ion channels during anoxia, a parallel to

that observed in hypoxia-tolerant animals (Hoch-

achka et al., 1988).

A final analogy between mammalian and higher-

plant cGMP systems might lie in the response to

nitric oxide. Nitric oxide was identified in mammals

as the long-observed but unidentified endothelium-

derived releasing factor, and has now been found to

have extensive physiological functions involving

cGMP (Moncada et al., 1992): in plants, nitric oxide

has been reported to stimulate cGMP formation in

spruce (Pfeiffer et al., 1994). Recent evidence shows

that nitric oxide potentiates the induction of hy-

persensitive cell death by reactive oxygen inter-

mediates produced in soybean cells in response to

pathogens and, independently of the oxygen inter-

mediates, nitric oxide induces genes for the synthesis

of protective natural products (Delledonne et al.,

1998). This leads to the conclusion that it has a key

role in plant cellular defence, a concept supported by

the compromising of disease resistance by inhibitors

of nitric oxide synthase. Among other effects, nitric

oxide has been shown to drive the activation of PAL

in a process, mediated by the guanylyl cyclase-

catalysed synthesis of cGMP, which can be further

stimulated by cADP-ribose (Durner et al., 1998).

Information relating to cyclic nucleotides other

than cAMP and GMP is almost as sparse in

mammals as in plants. In mammals, the occurrence

of cCMP, cIMP, cdTMP and cUMP has been

demonstrated, together with enzymes capable of

their synthesis and hydrolysis (Newton et al., 1984b,

1986): their function has not been established,

although initial evidence suggests that effects are not

solely due to action as agonists or antagonists of

cAMP or cGMP (Brus et al., 1984), and that these

four additional cyclic nucleotides have independent

functions, with cCMP effects and fluctuations in

concentration compatible with a role in the regu-

lation of cell proliferation (Newton, 1995). In plants,

as detailed above, cyclic nucleotides other than

cCMP and cGMP can bind to cyclic-nucleotide-

binding proteins; in lettuce, a phosphodiesterase

analogous to the mammalian multifunctional phos-

phodiesterase, which appears to function primarily

in cCMP metabolism, has been found (Chiatante

et al., 1986); for several of the phosphodiesterases a

major factor in the rate of cAMP hydrolysis is the

presence or absence of other competing cyclic

nucleotide substrates, so the presence of further

cyclic nucleotides in plant tissues can be seen to be of

considerable significance.

The occurrence of cCMP, cUMP, cIMP and

cdTMP in Pisum roots has been established by FAB-

MS and CID}MIKE technology (Newton et al.,

1989): the relative levels in meristematic and non-

meristematic tissues show significant differences. In

the meristems, cAMP, cCMP, cGMP and cUMP

are present, with cAMP, cGMP, cCMP, cIMP,

cUMP and cdTMP in the non-meristematic regions,

with a significantly higher concentration of cCMP in

the meristems and conversely a higher relative level

of cUMP in the non-meristematic region than the

meristem. The greater quantity of cCMP in the

meristem might reflect a role in the rapidly dividing

cells, analogous to that proposed for cCMP in

mammalian cells, whereas the presence of higher

relative concentrations of cUMP in non-meriste-

matic cells than in meristematic cells, taken together

with the presence of cIMP and cdTMP in the non-

meristematic regions and their absence from the

meristems, might also reflect the difference in

proliferation rate of the two types of tissue. Other

analogies suggest different roles for these cyclic

nucleotides, for example a possible role in inter-

cellular communication in higher plants, similar to

the functional extrusion of cCMP, cIMP, cAMP and

cdAMP from bacterial cells (Newton et al., 1998),

because similar extrusion processes have already

been reported for cAMP in lower plants (Francko &

Wetzel, 1980a,b).

IX.

This review has provided incontrovertible evidence

that cAMP, adenylyl cyclase and cyclic nucleotide

phosphodiesterase are indeed present in higher

plants. The mass spectrometric evidence in par-

ticular overcomes the objections raised about the

early reports over the identities of the putative

cAMP in tissue extracts and as the product of

adenylyl cyclase activity. This poses the question of

the correctness of the original observations and



Membrane

CaMATP

Cytoplasm

A-PK

ProteinPhosphatase

Protein-PO4

cAMP

Bindingprotein

Nucleotidase

(Cyclins)

Cell cyclemachinery

Nucleus

(Cyclins)

CRE sites

VBP1

CRE sites GAMyb chs fur ASI cab

Ca2+

Anthocyanins

GTP

Nitricoxide

cGMP

PhyA

CaM

G-proteinAdenylylcyclase

Fungalelicitor

Prostaglandin-likecompounds

(Jasmonic acid)Phytohormones

Ionchannel

Light GA3

RR

Anaerobiosis

Guanylylcyclase

G-protein

M+(+)

Phenylalanineammonia-lyase

AMP

PDE

GMPPDE

[G-PK]

Phytoalexin

a-amylase

Fig. 6. Cyclic nucleotides in higher plants. Diagram of cAMP and cGMP metabolism and function in higher

plants. Molecular conversions are represented by solid arrows, regulatory effects by dotted arrows. A-PK,

cAMP-dependent protein kinase; ASI, asparagine synthase; CaM, calmodulin; G-PK, cGMP-dependent

protein kinase; PDE, phosphodieasterase; PhyA, type 1 phytochrome; R, receptor.

theories, put forward in the 1970s, relating to cAMP

function. In our view this recent evidence does not

automatically render the earlier observations correct,

but equally they should not be rejected out of hand.

Instead there is an urgent need to re-examine them

with the use of the modern, specific, techniques of

affinity purification coupled with quantitative mass

spectrometric analysis to determine fluctuations in

cAMP and adenylyl cyclase in precise kinetic studies

of physiological events to assess the actions of cAMP.

Emphasis must be placed on the measurement of

fluctuations in concentrations and activities. A

consensus of the recent quantitative studies of cAMP

in plants suggests that levels are appreciably lower

than those reported in the 1970s. Although this

disparity partly reflects the ambiguities and cross-

reactivities intrinsic to the older methodologies and

procedures, it is our view that it also reflects that

there is a low basal level of cAMP in plant cells that

is significantly elevated for transient spells con-

comitant with a physiological event.

After the unequivocal demonstration of the exist-

ence of the enzymes and other key components of

cyclic nucleotide synthesis and breakdown, the

‘missing link’ is now the proof that cAMP is a

crucial element in the regulation of basic physio-

logical processes in plants. A comparison of Fig. 6,

showing the elements of cyclic nucleotide metab-

olism and action established in plants, with those of

animals and other cell types depicted in Figs 1 and 2

shows the progress made and the gaps remaining to

be filled.

The results discussed in section VIII detail the

evidence that leads to our opinion that cAMP has a

key role in, for example, the cell cycle and cellular

defence mechanisms in higher plants. Further eluci-

dation of cyclic nucleotide functions will require a

broad but systematic attack. In addition to moni-

toring of fluctuations of cyclic nucleotide and cyclic-

nucleotide-related enzymes during physiological

processes, the investigation of cyclic-nucleotide-

responsive protein kinases and cyclic-nucleotide-

binding proteins and the identification of their

targets will constitute a major step. Complementary

to mass spectrometric and biochemical analyses,

immunocytological techniques, incorporating high-

pressure freezing technology and molecular dis-

tillation, which permit the localization of small

diffusible molecules such as cyclic nucleotides, will

facilitate the identification of the subcellular dis-



tribution of cAMP and cAMP-binding sites; cyto-

chemical analyses will permit the localization of the

enzymes such as adenylyl cyclase. Molecular biology

techniques will permit the structural characterization

of the genes for adenylyl cyclase and phospho-

diesterase, and will be employed to prevent the

transient elevation of cAMP by anti-sense manipu-

lation of these enzymes in concert with the use of

cAMP analogues to modulate the kinase or cyclic

nucleotide-binding activities.

As with any scientific research field, the study of

cyclic nucleotides in higher plants contains particular

problems relating to the techniques and sample

sources involved; however, the major stumbling

block in this field has been the failure to attain a

‘critical mass’ in respect of the number of labora-

tories involved in such studies. A number of previous

‘ landmark’ or ‘seminal ’ reports have purported to

demonstrate the non-existence of cAMP regulatory

functions in plants and have had the effects of (i)

reducing the number of laboratories wishing to be

active in this area and (ii) dissuading the appropriate

funding bodies from supporting those that do. In

presenting the evidence above, which validates the

long-held opinions of the laboratories in the ‘small

coterie of believers’, it is our fervent hope that this

review will have the precisely opposite effect, and

stimulate a new momentum in cyclic nucleotide

research in higher plants.

Al-azzawi MJ, Hall JL. 1976. Cytochemical localization of

adenyl cyclase activity in maize roots. Plant Science Letters 6 :

285–289.

Ames IH, Richman RA, Weiss JP. 1980. Is cyclic GMP

involved in the regulation of tumorogenesis in the Nicotiana

genetic system? Plant Cell Physiology 21 : 367–372.

Amrhein N. 1974a. Evidence against the occurrence of cyclic

AMP in higher plants. Planta 118 : 241–258.

Amrhein N. 1974b. Cyclic nucleotide phosphodiesterases in

plants. Zeitschrift fuX r Pflanzenphysiologie 72 : 249–261.

Amrhein N. 1977. The current status of cyclic AMP in higher

plants. Annual Review of Plant Physiology 28 : 123–132.

Anderson JA, Huprikar SS, Kochian LV, Lucas WJ, GaberRF. 1992. Functional expression of a probable Arabidopsis

thaliana potassium channel in Saccharomyces cerevisiae. Pro-

ceedings of the National Academy of Sciences, USA 89 :

3736–3740.

Ashton AR, Polya GM. 1975. Higher plant cyclic nucleotide

phosphodiesterases. Biochemical Journal 149 : 329–339.

Ashton AR, Polya GM. 1977. Adenosine 3«,5«-cyclic mono-

phosphate in higher plants. Biochemical Journal 165 : 27–32.

Ashton AR, Polya GM. 1978. Cyclic AMP in axenic Rye grass

endosperm cell culture. Plant Physiology 61 : 718–722.

Assmann SM. 1995. Cyclic AMP as a second messenger in

higher plants. Plant Physiology 108 : 885–889.

Aurisano N, Bertani A, Reggiani R. 1995. Involvement of

calcium and calmodulin in protein and amino acid metabolism

in rice root under anoxia. Plant Cell Physiology 36 : 1525–1529.

Aurisano N, Bertani A, Reggiani R. 1996. Evidence for the

involvement of GTP-binding proteins in the process of

anaerobic g-aminobutyrate accumulation in rice roots. Journal

of Plant Physiology 149 : 517–519.

Balter M. 1999. Plant science - data in key papers cannot be

reproduced. Science 283 : 1987–1989.

Barlat I, Henglein B, Plet A, Lamb N, Fernadez A,McKenzie F, Pouysse! gur J, Vie! A, Blanchard JM. 1995.

TGF-beta 1 and cAMP attenuate cyclin A gene transcription

via a cAMP responsive element through independent pathways.

Oncogene 11 : 1309–1318.

Baroni MD, Monti P, Alberghina L. 1994. Repression of

growth-regulated G1 cyclin expression by cyclic AMP in

budding yeast. Nature 371 : 339–342.

Beavo JA. 1990. Multiple phosphodiesterase isoenzymes: back-

ground, nomenclature and implications. In: Beavo J, Houslay

MD, eds. Cyclic nucleotide phosphodiesterases : structure, regu-

lation and drug action. London, UK: John Wiley & Sons, 3–18.

Biermann B, Johnson EM, Feldman LJ. 1990. Characterization

and distribution of a maize cDNA encoding a peptide similar to

the catalytic region of second messenger dependent protein

kinases. Plant Physiology 94 : 1609–1615.

Bolwell GP. 1992. A role for phosphorylation in the down-

regulation of phenylalanine ammonia lyase in suspension

cultured cells of French bean. Phytochemistry 31 : 4081–4086.

Bolwell GP. 1995. Cyclic AMP, the reluctant messenger in

plants. Trends in Biochemical Science 20 : 489–492.

Bonnafous JC, Olive JL, Borgna JL, Mousseron-Cadet M.1975. L’amp cyclique dans les graines et les plantules d’orge et

la contamination bacte! rienne ou fongique. Biochimie 57 :

661–663.

Bowler C, Neuhaus G, Yamagata H, Chua NH. 1994a. Cyclic

GMP and calcium mediate phytochrome transduction. Cell 77 :

73–81.

Bowler C, Yamagata H, Neuhaus G, Chua NH. 1994b.Phytochrome signal transduction pathways are regulated by

reciprocal control mechanisms. Genes and Development 8 :

2188–2202.

Boynton AL, Whitfield JF. 1983. The role of cyclic AMP in cell

proliferation: A critical assessment of the evidence. Advances in

Cyclic Nucleotide Research 15 : 193–194.

Brennicke A, Frey HD. 1977. Properties of an adenosine cyclic

phosphate degrading enzyme in Nicotiana tabacum. Zeitschrift

fuX r Naturforschung 32 : 297–300.

Bressan RA, Ross CW, Vandepeute J. 1976. Attempts to detect

cyclic adenosine 3«,5«-monophosphate in higher plants by three

assay methods. Plant Physiology 57 : 29–34.

Brewin NJ, Northcote DH. 1973. Variations in the amount of

3«,5«-cyclic AMP in plant tissues. Journal of Experimental

Botany 24 : 881–888.

Brooker G, Harper JF, Terasaki WL, Moylan RD. 1979.Radioimmunoassay of cyclic AMP and cyclic GMP. Advances

in Cyclic Nucleotide Research 10 : 1–37.

Brown EG, Al-Najafi T, Newton RP. 1975. Partial purification

of adenosine 3«,5«-cyclic monophosphate phosphodiesterase

from Phaseolus vulgaris L.: associated activator and inhibitors.

Biochemical Society Transactions 3 : 393–395.

Brown EG, Al-Najafi T, Newton RP. 1977. Cyclic nucleotide

phosphodiesterase activity in Phaseolus vulgaris L. Phyto-

chemistry 16 : 1333–1337.

Brown EG, Al-Najafi T, Newton RP. 1979a. Adenosine 3«,5«-cyclic monophosphate, adenylate cyclase, and a cyclic AMP-

binding protein in Phaseolus vulgaris L. Phytochemistry 18 :

9–14.

Brown EG, Edwards MJ, Newton RP, Smith CJ. 1979b.Plurality of cyclic nucleotide phosphodiesterases in Spinacea

oleracea L.; subcellular distribution, partial purification and

properties. Phytochemistry 18 : 1943–1948.

Brown EG, Edwards MJ, Newton RP, Smith CJ. 1980a. The

cyclic nucleotide phosphodiesterases of spinach chloroplasts

and microsomes. Phytochemistry 19 : 23–30.

Brown EG, Newton RP, Evans DE, Walton TJ, Younis LM,Vaughan JM. 1989. Influence of light on cyclic nucleotide

metabolism in plants; effect of dibutyryl cyclic nucleotides on

chloroplast components. Phytochemistry 28 : 2559–2563.

Brown EG, Newton RP, Smith CJ. 1980b. A cyclic AMP-

binding protein from barley seedlings. Phytochemistry 19 :

2263–2266.

Brown EG, Newton RP. 1973. Occurrence of adenosine 3«,5«-cyclic monophosphate in plant tissues. Phytochemistry 12 :

2683–2685.

Brown EG, Newton RP. 1981. Cyclic AMP and higher plants.

Phytochemistry 20 : 2453–2463.

Brown EG, Newton RP. 1992. Analytical procedures for cyclic

nucleotides and their associated enzymes in plant tissues.

Phytochemical Analysis 3 : 1–13.



Brown PR. 1983. Current high-performance liquid chromato-

graphic methodology in analysis of nucleotides, nucleosides,

and their bases. I. Cancer Investigations 1 : 439–54.

Bru$ ggemann A, Pardo L, Stu$ hmer W, Pongs O. 1993. Ether-

a[ -go-go encodes a voltage-gated channel permeable to K+ and

Ca#+ and modulated by cAMP. Nature 365 : 445–448.

Brus R, Herman Z, Juraszczyk Z, Krzeminski T, TrzediakH, Juzcok A. 1984. Central action of cyclic-3«,5«-thymidine,

3«,5«-uridine and 3«,5«-cytidine monophosphates. Acta Medica

Polonica 25 : 1–4.

Bush DS. 1996. Effects of gibberellic acid and environmental

factors on cytosolic calcium in wheat aleurone cells. Planta 199 :

88–89.

Cailla HL, Racine-Weisbuch MS, Delaage MA. 1973. Adeno-

sine 3«,5« cyclic monophosphate assay at 10–15 mole level.

Analytical Biochemistry 56 : 394–407

Caprioli R, Malorni A, Sindona G. 1997. Selected topics in mass

spectrometry in the biomolecular sciences. Dordrecht, The

Netherlands: Kluwer Academic Publishers.

Carre! IA, Edmunds LN Jr. 1992. cAMP-dependent kinases

in the algal flagellate Euglena gracilis. Journal of Biological

Chemistry 267 : 2135–2137.

Carre! IA, Edmunds LN Jr. 1993. Oscillator control of cell

division in Euglena : cyclic AMP oscillations mediate the

phasing of the cell division cycle by the circadian clock. Journal

of Cell Science 104 : 1163–1173.

Carricarte VC, Bianchini GM, Muschietti JP, Te! llez-In4 o! n,Perticari A, Torres N, Flawia! MM. 1988. Adenylate cyclase

activity in a higher plant, alfalfa (Medicago sativa). Biochemical

Journal 249 : 807–811.

Chervet J-P, Ursem M, Salzmann JP. 1996. Instrumental

requirements for nanoscale liquid chromatography. Analytical

Chemistry 68 : 1507–1512.

Chiatante D, Balconi C, Newton RP, Brown EG. 1988.Immunoaffinity purification of cyclic nucleotide phosphodi-

esterase from Lactuca cotyledons. Phytochemistry 8 : 2477–2483.

Chiatante D, Newton RP, Brown EG. 1986. Partial purification

and properties of a multifunctional 3«,5«-cyclic nucleotide

phosphodiesterase from Lactuca cotyledons. Phytochemistry 25 :

1545–1551.

Chiatante D, Newton RP, Brown EG. 1987. Properties of

multifunctional 3«,5«-cyclic nucleotide phosphodiesterase from

Lactuca cotyledons: comparison with mammalian enzymes

capable of pyrimidine cyclic nucleotide hydrolysis. Phyto-

chemistry 26 : 1301–1306.

Chiatante D, Newton RP, Crignola S, Levi M, Brown EG.1990. The 3«,5«-cyclic nucleotide phosphodiesterase of meriste-

matic and differentiated tissues of pea roots. Phytochemistry 29 :

2815–2820.

Cohen P. 1989. Structure and regulation of protein phosphatases.

Annual Review of Biochemistry 58 : 453–508.

Conti M, Nemoz G, Sette C, Vicini E. 1995. Recent progress in

understanding the hormonal regulation of phosphodiesterases.

Endocrine Review 16 : 370–379.

Cook WH, Lipkin D, Markham R. 1957. The formation of

cyclic dianhydrodiadenylic acid (I) by the alkaline degradation

of adenosine-5«-triphosphoric acid (II). Journal of the American

Chemical Society 79 : 3607–3608.

Cooke CJ, Newton RP, Smith CJ, Walton TJ. 1989. Pathogenic

elicitation of phytoalexin in lucerne tissues; involvement of

cyclic AMP in intracellular mechanisms. Biochemical Society

Transactions 17 : 915–916.

Cooke CJ, Smith CJ, Walton TJ, Newton RP. 1994. Evidence

that cyclic AMP is involved in the hypersensitive response of

Medicago sativa to a fungal elicitor. Phytochemistry 35 : 899–995.

Curvetto N, Delmastro S. 1990. A biochemical and physio-

logical proposal for stomatal movement: possible involvement

of adenosine 3«,5«-cyclic monophosphate. Plant Physiology and

Biochemistry 28 : 367–378.

Cyong JC, Takahashi M, Hanabusa K, Otsuka V. 1982.Guanosine 3«,-5«-monophosphate in fruits of Evodia rutaecarpa

and E. officinalis. Phytochemistry 21 : 777–778.

Cyong JC, Takahashi M. 1982a. Guanosine 3«,5«-monophos-

phate activity in fruits of Zizyphus jujuba. Chemical and

Pharmacological Bulletin 30 : 1081–1083.

Cyong JC, Takahashi M. 1982b. Purification and identification

of guanosine 3«,5«-monophosphate from higher plants. Chemical

and Pharmacological Bulletin 30 : 2463–2466.

Cyong JC, Takahashi M. 1982c. Identification of guanosine

3«,5«-monophosphate in the fruit of Zizyphus jujuba. Phyto-

chemistry 21 : 1871–1874.

Delledonne M, Xia Y, Dixon RA, Lamb C. 1998. Nitric oxide

functions as a signal in plant disease resistance. Nature 394 :

585–588.

Depre C, Hue L. 1994. Cyclic GMP in the perfused rat heart.

Effect of ischaemia, anoxia and nitric oxide synthase inhibitor.

FEBS Letters 345 : 241–245.

Desdouets C, Matesic G, Molina CA, Foulkes NS, Sassone-Corsi P, Brechot C, Sobczak-The! pot J. 1995. Cell cycle

regulation of cyclin A gene expression by the cyclic AMP-

responsive transcription factors CREB and CREM. Molecular

and Cellular Biology 15 : 3301–3309.

Dhallan RS, Yau K-W, Schrader KA, Reed RR. 1990. Primary

structure and functional expression of a cyclic nucleotide-

activated channel from olfactory neurons. Nature 347 : 184–187.

Dumont JE, Jauniaux JC, Roger PP. 1989. The cyclic AMP-

mediated stimulation of cell proliferation. Trends in Biochemical

Science 14 : 67–71.

Dupon M, Van Onckelen HA, De Greef JA. 1987. Charac-

terization of cyclic nucleotide phosphodiesterase activity in

Phaseolus vulgaris. Physiologia Plantarum 69 : 361–365.

Durner J, Wendehenne D, Klessig D. 1998. Defence gene

induction in tobacco by nitric oxide, cyclic GMP and cyclic

ADP-ribose. Proceedings of the National Academy of Sciences,

USA 95 : 10328–10333.

Edmunds LN Jr. 1994. Clocks, cell cycles, cancer, and aging.

Role of the adenylate cyclase-cyclic AMP-phosphodiesterase

axis in signal transduction between circadian oscillator and cell

division cycle. Annals of the New York Academy of Science 719 :

77–96.

Ehrlich KC, Cary JW, Ehrlich M. 1992. A broad bean cDNA

clone encoding a DNA-binding protein resembling mammalian

CREB in its sequence specificity and DNA methylation

sensitivity. Gene 117 : 169–178.

Ehsan H, Reichheld J-P, Roef L, Witters E, Lardon F, VanBockstaele D, Van Montagu M, Inze! D, Van Onckelen H.1998. Effect of indomethacin on cell cycle dependent cyclic

AMP fluxes in tobacco BY-2 cells. FEBS Letters 442 : 165–169.

Endress R. 1979. Allosteric regulation of phosphodiesterase from

Portulaca callus by cGMP and papavarin. Phytochemistry 18 :

15–20.

Fedenko EP, Kasumov KK. 1993. Effect of GPP(NH)P on basal

and photoinduced activities of cyclic AMP phosphodiesterase

in maize sprouts. Izvestiya Akademii Nauk SSSR Seriya

Biologicheskaya 1 : 133–137.

Fedenko EP, Kasumov KK, Doman NG. 1992. Light-

dependent response of cyclic AMP phosphodiesterase to

calmodulin. Izvestiya Akademii Nauk SSSR Seriya Biologi-

cheskaya 1992 1 : 25–30.

Ferreira P, Hemerly A, De Almeida Engler J, BergouniouxC, Burssens S, Van Montagu M, Engler G, Inze! D. 1994.Three discrete classes of Arabidopsis cyclins are expressed

during different intervals of the cell cycle. Proceedings of the

National Academy of Sciences, USA 91 : 11313–11317.

Ferreira PCG, Hemerly AS, Villaroel R, Van Montagu M,Inze! D. 1991. The Arabidopsis functional homolog of the

p34cdc2 protein kinase. An extended family with diverse

functions. Annual Review of Physiology 58 : 395–426.

Fesenko EE, Kolesnikov SS, Lyubarsky AL. 1985. Induction

by cyclic GMP of cationic conductance in plasma membrane of

retinal rod outer segment. Nature 313 : 310–313.

Finn JT, Grunwald ME, Yau K-W. 1996. Cyclic nucleotide-

gated ion channels : an extended family with diverse functions.

Annual Review of Physiology 58 : 395–426.

Fink JS, Verhave M, Kasper S, Tsukada T, Mandel G,Goodman RH. 1988. The CGTCA sequence motif is essential

for biological activity of the vasoactive intestinal peptide gene

cAMP-regulated enhancer. Proceedings of the National Academy

of Sciences, USA 85 : 6662–6666.

Foulkes NS, Borelli E, Sassone CP. 1991. CREM gene: use of

alternative DNA-binding domains generates multiple antag-

onists of cAMP-induced transcription. Cell 64 : 739–749.

Francko DA. 1983. Cyclic AMP in photosynthetic organisms:

recent developments. Advances in Cyclic Nucleotide Research

15 : 97–117.

Francko DA, Wetzel RG. 1980a. Cyclic adenosine-3«,5«-



monophosphate: production and extracellular release from

green and blue-green algae. Physiologia Plantarum 49 : 65–67.

Francko DA, Wetzel RG. 1980b. Synthesis and release of cyclic

adenosine 3«,5«-monophosphate by aquatic macrophytes.

Physiologia Plantarum 52 : 33–36.

Gadeyne J. 1992. Het cyclisch 3« :5«-adenosine monofosfaat meta-

bolisme in Phaseolus vulgaris L. PhD thesis, University of

Antwerp (UIA), Belgium.

Gagelin C, Pierre M, Toru-Delbauffe D. 1994. Inhibition of

G1 cyclin expression and G1 cyclin-dependent protein kinases

by cAMP in an astrocytic cell line. Biochemical and Biophysical

Research Communications 205 : 923–929.

Galyov EE, Hakansson S, Forsberg A, Wolf-Watz H. 1993. A

secreted protein kinase of Yersinia pseudotuberculosis is an

indispensable virulence factor. Nature 361 : 730–732.

Gardner IC, McNally SF, Scott A. 1979. Electron histochemical

localisation of the adenyl cyclase activity in the root nodules of

Alnus glutinosa L. Plant Science Letters 16 : 387–395.

Gaymard F, Cerutti M, Horeau C, Lemaille G, Urbach S,Ravallec M, Dechauvelle G, Sentenac H, Thibaud J-B.1996. The baculovirus}insect cell system as an alternative to

Xenopus oocytes. First characterization of the AKT1 K+-

channel from Arabidopsis thaliana. Journal of Biological Chem-

istry 271 : 22863–22870.

Giannattasio M, Carratu’ G, Tucci GF. 1974a. Presence of a

cyclic-AMP-binding protein in Jerusalem artichoke rhizome

tissue. FEBS Letters 49 : 249–253.

Giannattasio M, Sica G, Macchia V. 1974b. Cyclic AMP

phosphodiesterase from dormant tubers of Jerusalem artichoke.


Gillespie PG. 1990. Phosphodiesterases in visual transduction by

rods and cones. In: Beavo J, Houslay MD, eds. Cyclic nucleotide

phosphodiesterases : structure, regulation and drug action, London,

UK: John Wiley & Sons, 163–184.

Gilroy S, Jones RL. 1994. Perception of gibberellins and abscisic

acid at the external face of barley aleurone protoplasts. Plant

Physiology 104 : 1185–1192.

Gonzalez GA, Montminy MR. 1989. cAMP stimulates somato-

statin gene transcription by phosphorylation of CREB at serine

133. Cell 59 : 675–680.

Gonzalez GA, Yamamoto KK, Fischer WH, Karr D, MenzelP, Biggs III W, Vale WW, Montminy MR. 1989. A cluster of

phosphorylation sites on the cyclic AMP-regulated nuclear

factor CREB predicted by its sequence. Nature 337 : 749–752.

Gubler F, Kalla R, Roberts JK, Jacobsen JV. 1995. Gibberellin-

regulated expression of a myb gene in barley aleurone cells :

evidence for myb transactivation of a high-pl α-amylase gene

promoter. The Plant Cell 7 : 1879–1891.

Guy HR, Durrell SR, Warmke J, Drysdale R, Ganetzky B.1991. Similarities in amino acid sequences of Drosophila eag and

cyclic-nucleotide gated channels. Science 254 : 730.

Haddox MK, Stevenson JH, Goldberg ND. 1974. Cyclic GMP

in meristematic and elongating regions of bean root. Federation

Proceedings 33 : 522.

Hai T, Liu F, Coukos WJ, Green MR. 1989. Transcription

factor ATF cDNA clones: an extensive family of leucine zipper

proteins able to selectively form DNA-binding heterodimers.

Genes and Development 3 : 2083–2090.

Hasunuma K, Funadera K, Furukawa K, Miyamoto-Shinoyama. 1988. Rhythmic oscillation of cyclic 3«,5«-adeno-

sine monophosphate and cyclic 3«,5«-guanosine monophosphate

concentrations and stimulation of flowering by cyclic GMP in

Lemna paucicotata 381. Photochemistry and Photobiology 48 :

89–92.

Hayashida N, Mizoguchi T, Shinozaki K. 1993. Cloning and

characterization of a plant gene encoding a protein kinase. Gene

124 : 251–255.

Hayashida N, Mizoguchi T, Yamaguchi-Shinozaki K,Shinozaki K. 1992. Characterization of a gene that encodes a

homologue of protein kinase in Arabidopsis thaliana. Gene 121 :

325–330.

Helfman DM, Shoji M, Kuo JF. 1981. Purification to hom-

ogeneity and general properties of a novel phosphodiesterase

hydrolysing cyclic CMP and cyclic AMP. Journal of Biological

Chemistry 256 : 6327–6334.

Hemerly A, Bergounioux C, Van Montagu M, Inze! D,Ferreira P. 1992. Genes regulating the plant cell cycle :

isolation of a mitotic-like cyclin from Arabidopsis thaliana.

Proceedings of the National Academy of Sciences, USA 89 :

3295–3299.

Hilton GM, Nesius KK. 1978. Localization of adenylyl cyclase

in meristems of young pea hypocotyls. Physiologia Plantarum

42 : 49–52.

Hochachka PW, Castellini JM, Hill RD, Hill S, Bengston J,Schneider RC, Liggins GC, Zapol WM. 1988. Seal liver

ischaemia: arrest of glycolytic and mitochondrial functions.

Molecular Cell Biochemistry 84 : 77–85.

Hofmann F, Dostmann W, Keilbach A, Landgraf W, Ruth P.1992. Structure and physiological role of cGMP-dependent

protein kinase. Biochimica et Biophysica Acta 1135 : 51–60.

Hong JC, Nagao RT, Key JL. 1987. Characterization and

sequence analysis of a developmentally regulated putative cell

wall protein gene isolated from soybean. Journal of Biological

Chemistry 262 : 8367–8376.

Hooley R, Beale MH, Smith RJ. 1991. Gibberellin perception

at the plasma membrane of Avena fatua aleurone protoplasts.

Planta 183 : 274–280.

Hopfgartner G, Bean K, Henion J. 1993. Ion mass spectro-

metric detection for liquid chromatography a concentration- or

mass-flow-sensitive device? Journal of Chromatography A 647 :

51–61.

Hoshi T. 1995. Regulation of voltage dependence of the KAT1

channel by intracellular factors. Journal of General Physiology

105 : 309–328.

Huber SC, Huber JL. 1991. Regulation of maize leaf sucrose-

phosphate synthase by protein phosphorylation. Plant Cell

Physiology 32 : 91–97.

Hunter T. 1991. Protein kinase classification. Methods in

Enzymology 200 : 3–37.

Huntley R, Healy S, Freeman D, Lavender P, de Jager S,Greenwood, J, Makker J, Walker E, Jackman M, Xie Q,Bannister AJ, Kouzarides T, Gutierrez C, Doonan JH,Murray JA. 1998. The maize retinoblastoma protein hom-

ologue ZmRb-1 is regulated during leaf development and

displays conserved interactions with G1}S regulators and plant

cyclin D (CycD) proteins. Plant Molecular Biology 37 :155–169

Hurst HC, Totty NF, Jones NC. 1991. Identification and

functional characterisation of the cellular activating transcrip-

tion factor 43 (ATF-43) protein. Nucleic Acids Research 19 :

4601–4609.

Ichikawa T, Suzuki Y, Czajaa I, Schommer C, Lessnick A,Schell J, Walden R. 1997. Identification and role of adenylyl

cyclase in auxin signalling in higher plants. Nature 390 :

698–701.

Ichikawa T, Suzuki Y, Czajaa I, Schommer C, Lessnick A,Schell J, Walden R. 1998. Identification and role of adenylyl

cyclase in auxin signalling in higher plants. Retraction of

Nature 390 : 698–701, 1997. Nature 396 : 6709.

Iguchi-Ariga SM, Schaffner W. 1989. CpG methylation of the

cAMP-responsive enhancer}promoter sequence TGACGTCA

abolishes specific factor binding as well as transcriptional

activation. Genes and Development 3 : 612–619.

Inamdar NM, Ehrlich KC, Ehrlich M. 1991. CpG methylation

inhibits binding of several sequence-specific DNA-binding

proteins from pea, wheat, soybean and cauliflower. Plant

Molecular Biology 17 : 111–123.

Ishiguro S, Tanaka M, Kojimoto A, Kato M, Iwabuchi M,Nakamura K. 1993. A nuclear factor that binds to a dyad-

symmetric sequence with a CGTCA motif in the 5«-upstream

region of the sweet potato beta-amylase gene. Plant Cell


Janda L, Tichy P, Spizek J, Petricek M. 1996. A deduced

Thermomonospora curvata containing serine}threonine protein

kinase and WD-repeat domains. Journal of Bacteriology 178 :

1487–1489.

Janistyn B. 1983. Gas chromatographic mass spectrometric

identification and quantification of cyclic guanosine 3«,5«-cyclic

monophosphate in maize seedlings. Planta 159 : 382–3886.

Janistyn B. 1988. Stimulation by manganese(II)sulphate of a

cAMP-dependent protein kinase from Zea mays seedlings.


Janistyn B. 1989. cAMP promoted protein phosphorylation of

dialysed coconut milk. Phytochemistry 28 : 329–331.

Josefsson L-G, Rask L. 1997. Cloning of a putative G-protein-

coupled receptor from Arabidopsis thaliana. European Journal

of Biochemistry 249 : 415–420.



Johnson LP, McLeod JK, Parker CW, Letham DS. 1981. The

quantitation of adenosine 3«-5«-cyclic monophosphate in cul-

tured tobacco tissue by mass spectrometry. FEBS Letters 124 :

119–121.

Kaneko T, Sato S, Kotani H, Tanaka A, Asamizu E,Nakamura Y, Miyajima N, Hirosawa M, Sugiura M,

Sasamoto S, Kimura T, Hosouchi T, Matsuno A, Muraki

A, Nakazaki N, Naruo K, Okumura S, Shimpo S, Takeuchi

C, Wada T, Watanabe A, Yamada M, Yasuda M, Tabata S.

1996. Sequence analysis of the genome of the unicellular

cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence

determination of the entire genome and assignment of potential

protein-coding regions. DNA Research 3 : 109–136.

Kasumov KK, Fedenko EP, Kritzsky K. 1991. Purification of

cyclic AMP phosphodiesterase of corn sprouts regulated by red

light. Izvestiya Akademii Nauk SSSR Seriya Biologicheskaya

5 : 661–668.

Katagiri F, Lam E, Chua N-H. 1989. Two tobacco DNA-

binding proteins with homology to the nuclear factor CREB.

Nature 340 : 727–730.

Kataoka T, Broek D, Wigler M. 1985. DNA sequence and

characterization of the S. cerevisiae gene encoding adenylate

cyclase. Cell 43 : 355–357.

Kato R, Uno I, Ishikawa T, Fujii T. 1983. Effects of cAMP on

the activity of soluble protein kinases in Lemna paucicostata.

Plant Cell Physiology 24 : 841–848.

Katsumata T, Takahashi N, Ejiri S. 1978. Changes of cyclic

AMP level and adenylate cyclase activity during germination of

pine pollen. Agricultural Biological Chemistry 42 : 2161–2162.

Keates RAB. 1973. Evidence that cyclic AMP does not mediate

the action of gibberellic acid. Nature 244 : 355–357.

Kennelly PJ, Krebs EG. 1991. Consensus sequences as substrate

specificity determinator for protein kinases and protein phos-

phatases. Journal of Biological Chemistry 266 : 15555–15558.

Kessler B, Levenstein R. 1974. Adenosine 3«,5«-cyclic mono-

phosphate in higher plants. Assay, distribution and age-

dependency. Biochimica et Biophysica Acta 347 : 156–166.

Kingston EE, Beynon JH, Newton RP. 1984. The identification

of cyclic nucleotides from living systems using collision-

induced dissociation of ions generated by fast atom bom-

bardment mass spectrometry. Biomedical Mass Spectrometry

11 : 367–374.

Kingston EE, Beynon JH, Newton RP, Liehr JG. 1985. The

differentiation of isomeric biological compounds using collision-

induced dissociation of ions generated by fast atom bom-

bardment. Biomedical Mass Spectrometry 12 : 525–534.

Koch KW, Kaupp UB. 1985. Cyclic GMP directly regulates a

cation conductance in membranes of bovine rods by a

cooperative mechanism. Journal of Biological Chemistry 260 :

6788–6800.

Komatsu S, Hirano H. 1993. Protein kinase activity and protein

phosphorylation in rice (Oryza sativa L.) leaf. Plant Science 94 :

127–137.

Kooijman R, de Wildt P, van den Briel W, Tan S-h,

Musgrave A, van den Ende H. 1990. Cyclic AMP is one of

the intracellular signals during the mating of Chlamydomonas

eugametos. Planta 181 : 529–537.

Kuo JF, Brackett NL, Shoji M, Tse J. 1978. Cytidine 3«,5«-cyclic

monophosphate phosphodiesterase in mammalian tissues.

Journal of Biological Chemistry 253 : 2518–2521.

Kurosaki F. 1997. Role of inward K+ channel located at carrot

plasma membrane in signal cross talking of cyclic AMP with

Ca#+ cascade. FEBS Letters 408 : 115–119.

Kurosaki F, Kaburaki H. 1995. Phosphodiesterase isoenzymes

in extracts of cultured carrot. Phytochemistry 40 : 685–689.

Kurosaki F, Nishi A. 1993. Stimulation of calcium influx and

calcium cascade by cyclic AMP in cultured carrot cells. Archives

of Biochemistry and Biophysics 302 : 144–151.

Kurosaki F, Tsurusawa Y, Nishi A. 1987. The elicitation of

phytoalexins by Ca#+ and cyclic AMP in cultured carrot cells.


Kwok RP, Lundblad JR, Chrivia JC, Richards JP, Bachinger

H, Brennan R, Roberts S, Green M, Goodman RH. 1994.Nuclear protein CBP is a coactivator for the transcription factor

CREB. Nature 370 : 223–226.

L’Allemain G, Lavoie JN, Rivard N, Baldin N, Pouysse! gur

J. 1997. Cyclin D1 expression is a major target of the cAMP-

induced inhibition of cell cycle entry in fibroblasts. Oncogene

14 : 1981–1990.

Lawson AM, Stillwell RN, Tacker MM, Tsuboyama K,

McCloskey JA. 1971. Mass spectrometry of nucleic acid

components. Trimethylsilyl derivatives of nucleotides. Journal

of the American Chemical Society 93 : 1014–1023.

Lawton MA, Yamamoto RT, Hanks SK, Lamb CJ. 1989.

Molecular cloning of plant transcripts encoding protein kinase

homologs. Proceedings of the National Academy of Sciences,

USA 86 : 3140–3144.

Levi M, Sparvoli E, Galli MG. 1981. Interference of the

phosphodiesterase inhibitor aminophylline with the plant cell

cycle. European Journal of Cell Research 25 : 71–75.

Li W, Luan S, Schreiber SL, Assmann SM. 1994. Cyclic AMP

stimulates K+ channel activity in mesophyll cells of Vicia faba

L. Plant Physiology 106 : 957–961.

Lin P-PC, Varner JE. 1972. Cyclic nucleotide phosphodiesterease

in pea seedlings. Biochimica et Biophysica Acta 276 : 454–474.

Lin P-PC. 1974. Cyclic nucleotides in higher plants. Advances in

Cyclic Nucleotide Research 4 : 439–461.

Lin X, Feng X-H, Watson JC. 1991. Differential accumulation

of transcripts encoding protein kinase homologs in greening pea

seedlings. Proceedings of the National Academy of Sciences,

USA 88 : 6951–6955.

Lundeen CV, Wood NH, Braun AC. 1973. Intracellular levels

of cyclic nucleotides during cell enlargement and cell division in

excised tobacco piths. Differentiation 1 : 255–260.

Lusini P, Trabalzini L, Franchi GG, Bovalini L, Martelli P.

1991. Adenylate cyclase in roots of Ricinus communis ; stimu-

lation by GTP and Mn#+. Phytochemistry 30 : 109–111.

Magnaldo I, Pouysse! gur J, Paris S. 1989. Cyclic AMP inhibits

mitogen-induced DNA synthesis in hamster fibroblasts, re-

gardless of the signalling pathway involved. FEBS Letters 245 :

65–69.

Manganiello VC, Tanaka T, Murashima S. 1990. Cyclic

GMP-stimulated cyclic nucleotide phosphodiesterases. In:

Beavo J, Houslay MD, eds. Cyclic nucleotide phosphodiesterases :

structure, regulation and drug action. London, UK: John Wiley

& Sons, 61–86.

Masson P, Jacquemin J-M, Culot M. 1984. Molecular cloning

of the tsm0185 gene responsible for adenylate cyclase activity in

Saccharomyces cerevisiae. Annales de Microbiologie 135A:

343–351.

Mateos JL, Pedrosa MM, Molina MD, Pereira EC, VicenteC, Legaz ME. 1993. Involvement of phytochrome-mediated

cyclic AMP in the synthesis and deposition of fumarproto-

cetraric acid on cortical hyphae of Cladonia verticilaris. Plant

Physiology and Biochemistry 31 : 667–674.

Mayer D, Ehemann V, Hacker H-J, Klimek F, Bannasch P.

1985. Specificity of cytochemical demonstration of adenylate

cyclase in liver using adenylate (ab methylene) diphosphate as

substrate. Histochemistry. 82 :135–140.

Messiaen J, Van Cutsem P. 1996. Cloning of a cAMP pathway

element in chicory. Plant Physiology and Biochemistry, special

issue, 195.

Mizoguchi T, Hayashida N, Shinozaki KY, Harada H,

Shinozaki K. 1992. Nucleotide sequence of cDNA encoding

a protein kinase homologue in Arabidopsis thaliana. Plant

Molecular Biology 18 : 809–812.

Moncada S, Marletta MA, Hibbs JB, Higgs EA. 1992. The

biology of nitric oxide. London, UK: Portland Press.

Montminy MR, Sevarino KA, Wagner JA, Mandel G,

Goodman RH. 1986. Identification of cAMP responsive

element in within the rat somatostatin gene. Proceedings of the


Mumby H, Walter G. 1993. Protein serine threonine phopsha-

tases : structure, regulation and functions in cell growth.

Physiological Reviews 73 : 673–699.

Mundy J, Chua NH. 1988. Abscisic acid and water stress

introduce the expression of a novel rice gene. EMBO Journal 7 :

2279–2286.

Mun4 oz-Dorado J, Inouye S, Inouye M. 1993. Eukaryotic-like

protein serine}threonine kinases in Myxococcus xanthus, a

developmental bacterium exhibiting social behaviour. Journal

of Cell Biochemistry 51 : 29–33.

Murray JAH. 1997. The retinoblastoma protein is in plants!

Trends in Plant Science 2 : 82–84.



Murray JAH, Freeman D, Greenwood J, Huntley R,Makkerh J, Riou-Khamlichi C, Sorrell DA, Cockcroft C,Carmichael JP, Soni R, Shah ZH. 1997. Plant D cyclins and

retinoblastoma (Rb) protein homologues. In: Francis D, Dudits

D, Inze D, eds. Plant cell division, vol. 10. London, UK:

Portland Press, 99–127.

Mutzel R, Lacombe ML, Simon MN, de Gunzburg J, Ve! ronM. 1987. Cloning and cDNA sequence of the regulatory

subunit of cAMP-dependent protein kinase from Dictyostelium

discoideum. Proceedings of the National Academy of Sciences,

USA 84 : 6–10.

Narayanan A, Vermeersch J, Pradet A. 1970. Dosage

enzymatique de l’acid adenosine 3«,5«-monophosphate cyclique

dans les semences de laitue, variete! Reine de Mai. Comptes

Rendues de l’AcadeUmie des Sciences 271 : 2406–2407.

Negishi M, Ichikawa A, Oshio N, Yatsunami K, Tomita K.1982. Cell cycle specific fluctuations of adenosine 3«,5«-monophosphate and prostaglandin binding in synchronized

mastocytoma P-815 cells. Biochemical Pharmacology 31 : 173–

179.

Neuhaus G, Bowler C, Hiratsuka K, Yamagata H, Chua NH.1997. Phytochrome-regulated repression of gene expression

requires calcium and cyclic GMP. EMBO Journal 16 :

2554–2564.

Neuhaus G, Bowler C, Kern R, Chua NH. 1993. Calcium}calmodulin-dependent and -independent phytochrome signal

transduction pathways. Cell 73 : 937–952.

Newton RP. 1995. Cytidine 3«,5«-cyclic monophosphate: a third

cyclic nucleotide secondary messenger? Nucleosides and Nucleo-

tides 14 : 743–747.

Newton RP. 1996. Mass spectrometric analysis of cyclic

nucleotides and related enzymes. In: Newton RP, Walton TJ,

eds. Applications of modern mass spectrometry in plant science

research. Oxford, UK: Oxford Science Publications, 159–181.

Newton RP, Brenton AG, Ghosh D, Walton TJ, Langridge J,Harris FM, Evans AM. 1991. Qualitative and quantitative

mass spectrometric analysis of cyclic nucleotides and related

enzymes. Analytica Chimica Acta 247 : 161–175.

Newton RP, Brown EG. 1986. The biochemistry and physiology

of cyclic AMP in higher plants. In: Garrod DR, Chadwick CM,

eds. Receptors in plants and cellular slime moulds. Cambridge,

UK: Cambridge University Press, 115–153.

Newton RP, Chiatante D, Ghosh D, Brenton AG, Walton TJ,Harris FM, Brown EG. 1989. Identification of cyclic nucleo-

tide constituents of meristematic and non-meristematic tissues

of Pisum sativum roots. Phytochemistry 28 : 2243–2254.

Newton RP, Evans AM, van-Geyschem J, Diffley PJ,Hassam HG, Hakeem NA, Moyse CD, Cooke R, SalvageBJ. 1994. Radioimmunoassay of cytidine 3«,5«-cyclic mono-

phosphate: unambiguous assay by means of an optimized

protocol incorporating a trilayer column separation to obviate

cross-reactivity problems. Journal of Immunoassay 15 : 317–337.

Newton RP, Gibbs N, Moyse CD, Wiebers JL, Brown EG.1980. Mass spectrometric identification of adenosine 3«,5«-cyclic monophosphate isolated from a higher plant tissue.


Newton RP, Kingston EE, Evans DE, Younis LM, Brown EG.1984a. Occurrence of guanosine 3«,5«-cyclic monophosphate

(cyclic GMP) and associated enzyme systems in Phaseolus

vulgaris L. Phytochemistry 23 : 1367–1372.

Newton RP, Kingston EE, Hakeem NA, Salih SG, BeynonJH, Moyse CD. 1986. Extraction, purification, identification

and metabolism of 3«,5« cyclic-UMP, 3«,5«-cyclic IMP and

3«,5«-cyclic dTMP from rat tissues. Biochemical Journal 236 :

431–439.

Newton RP, Kingston EE, Overton A. 1995. Identification of

novel nucleotides found in the red seaweed Porphyra umbilicalis.

Rapid Communications in Mass Spectrometry 9 : 305–311.

Newton RP, Kingston EE, Overton A. 1998. Mass spectro-

metric identification of cyclic nucleotides released by the

bacterium Corynebacterium murisepticum into the culture me-

dium. Rapid Communications in Mass Spectrometry 12 : 729–735.

Newton RP, Salih SG, Khan JA. 1990. Cyclic CMP-specific

phosphodiesterase activity. In: Beavo J, Houslay MD, eds.

Cyclic nucleotide phosphodiesterases : structure, regulation and

drug action. London, UK: John Wiley & Sons, 141–162.

Newton RP, Salih SG, Salvage BJ, Kingston EE. 1984b.Extraction, purification and identification of cytidine 3«,-5«-

cyclic monophosphate from rat tissues. Biochemical Journal

221 : 665–673.

Newton RP, Walton TJ. 1996. Applications of modern mass

spectrometry in plant science research. Oxford, UK: Oxford

Science Publications.

Niles RM, Mount MS. 1973. Failure to detect cyclic adenosine

3«,5«-monophosphate in healthy and crown gall tumour tissues

of Vicia faba. Plant Physiology 54 : 372–373.

Niles RM, Mount MS. 1974. Cyclic nucleotide phosphodi-

esterase from carrot. Phytochemistry 13 : 2735–2740.

Nougare' de AP, Landre! P, Rembur J, Hernandez MN. 1984.

Des variations d’activite! s de la 5«-nucle!otidase et de l’ade!nylate-

cyclase sont-elles des composantes de la levee! d’inhibition du

bourgeon cotyle!donaire du pois? Canadian Journal of Botany

63 : 309–323.

Oguni I, Suzuki K, Uritani I. 1976. Terpenoid induction in

sweet potato roots by adenosine 3«,5«-cyclic monophosphate.

Agricultural Biological Chemistry 40 : 1251–1252.

Oota Y. 1972. A possible mechanism for sugar inhibition of

duckweed flowering. Plant Cell Physiology 13 : 195–199.

Pacini B, Petrigliano A, Diffley P, Paffetti A, Brown EG,

Martelli P,Trabalzini L, Bovalini L, Lusini P, Newton RP.

1993. Adenylyl cyclase activity in roots of Pisum sativum.


Parent CA, Devreotes PN. 1995. Isolation of inactive and

G-protein resistant adenylyl cyclase mutants using random

mutagenesis. Journal of Biological Chemistry 270 : 22693–22696.

Parker D, Ferreri K, Nakajima T, LaMorte VJ, Evans R,

Koerber SC, Hoeger C, Montminy MR. 1996. Phosphory-

lation of CREB at Ser-133 induces complex formation with

CREB-binding protein via a direct mechanism. Molecular and

Cellular Biology 16 : 694–703.

Penson SP, Schuurink RC, Fath A, Gubler F, Jacobsen JV,

Jones RL. 1996a. cGMP is required for gibberellic acid-

induced gene expression in barley aleurone. The Plant Cell 8 :

2325–2333.

Penson SP, Schuurink RC, Fath A, Gubler F, Jacobsen JV,

Jones RL. 1996b. cGMP is required for gibberellic acid-

induced gene expression in barley aleurone. The Plant Cell 9 :

271.

Penson SP, Schuurink RC, Shartzer SF, Jones RL. 1997.

Cyclic GMP and its possible targets in barley aleurone cells.

Plant Physiology 114 : 1392.

Pfeiffer S, Janistyn B, Jessner G, Pichorner H, EbermannRE. 1994. Gaseous nitric oxide stimulates guanosine 3«,5«-cyclic

monophosphate formation in spruce needles. Phytochemistry

36 : 259–262.

Plakidou-Dymock S, Dymock D, Hooley R. 1997. A higher

plant seven-transmembrane receptor that influences sensitivity

to cytokinins. Current Biology 8 : 315–324.

Pollard SJ. 1970. Influence of gibberellic acid on the incor-

poration of [8-"%C]-adenine into adenosine 3«,-5«-cyclic mono-

phosphate in barley aleurone layer. Biochimica et Biophysica

Acta 210 : 511–512.

Polya GM. 1975. Purification and characterization of a cyclic

nucleotide-regulated 5«-nucleotidase from potato. Biochimica et

Biophysica Acta 384 : 443–457.

Polya GM, Ashton AR. 1973. Inhibition of wheat seedling 5«(3«)-ribonucleotide phosphohydrolase by adenosine 3«,5«-cyclic

monophosphate. Plant Science Letters 1 : 349–357.

Polya GM, Chung R, Menting J. 1991. Resolution of a higher

plant protein kinase similar to the catalytic subunit of cyclic

AMP-dependent protein kinase. Plant Science 79 : 37–45.

Polya GM, Hunziker K. 1987. Purification and properties of a

high affinity guanosine 3« :5«-cyclic monophosphate-binding

phosphatase from silver beet leaves. Plant Science 50 :117–123.

Polya GM, Wettenhall REH. 1992. Rapid purification and N-

terminal sequencing of a potato tuber cyclic nucleotide binding

phosphatase. Biochimica et Biophysica Acta 1159 : 179–184.

Rall TW, Sutherland EW, Berthet J. 1957. The relation of

epinephrine and glucagon to liver phosphorylase. Journal of

Biological Chemistry 224 : 1987–1995.

Reggiani R. 1997. Alteration of levels of cyclic nucleotides in

response to anaerobiosis in rice seedlings. Plant Cell Physiology

38 : 740–742.

Reichheld J-P, Chaubet N, Shen WH, Renaudin J-P, Gigot C.

1996. Multiple A-type cyclins express sequentially during the



cell cycle in Nicotiana tabacum BY2 cells. Proceedings of the


Reichheld J-P, Sonobe S, Cle! ment B, Chaubet N, Gigot C.

1995. Cell cycle-regulated histone gene expression in syn-

chronized plant cells. Plant Journal 7 : 245–252.

Richards PA, Richards PDG. 1998. Microscopical localization

of adenylate cyclase: a historical review of methodologies.

Microscopy Research and Technique 40 : 434–439.

Rickenberg HV. 1974. Cyclic AMP in prokaryotes. Annual

Review of Microbiology 28 : 353–369.

Roef L, Witters E, Gadeyne J, Marcussen J, Newton RP, Van

Onckelen HA. 1996. Analysis of 3«,5«-cAMP and adenylyl

cyclase activity in higher plants using polyclonal chicken egg

yolk antibodies. Analytical Biochemistry 233 : 188–197.

Roef L. 1997. Het 3«,5«-cAMP metabolisme in hogere planten :

bijdrage tot de karakterisatie van adenylyl cyclase en cAMP-

afhankelijk proteıXne kinase. PhD thesis, University of Antwerp

(UIA), Belgium.

Roger PP, Reuse S, Maenhaut C, Dumont JE. 1995. Multiple

facets of the modulation of growth by cAMP. Vitamins and

Hormones 51 : 59–191.

Rougier M, Jnoud N, Dumas C. 1988. Cytochemical study of

adenylate cyclase in pollen-pistil interactions and its relation to

incompatibility. In: Cresti M, Gori P, Pacini E, eds. Sexual

reproduction in higher plants. Berlin, Germany: Springer-

Verlag, 363–368.

Sabelli PA, Burgess SR, Kush AK, Shewry PR. 1997. DNA

replication initiation and mitosis induction in eukaryotes : the

role of MCM and Cdc25 proteins. In: Francis D, Dudits D,

Inze D, eds. Plant cell division. Colchester, UK: Portland

Press, 243–268

Saito T, Small L, Goodenough UW. 1993. Activation of

adenylyl cyclase in Chlamydomonas reinhardtii by adhesion and

heat. Journal of Cell Biology 122 : 137–147.

Sakuanrungsirikul S, Hocart CH, Harper JDI, Parker CW,

John PCL. 1996. Temperature conditional cAMP-requiring

mutant strains of Chlamydomonas reinhardtii arrest in G"

and

are rescued by added cAMP. Protoplasma 192 : 159–167.

Salomon D, Mascarenhas JP. 1971. Auxin induced synthesis of

cyclic 3«,5«-AMP in Avena coleoptiles. Life Science 10 : 879–885.

Schimz A, Hinsch K-D, Hildebrand E. 1989. Enzymatic and

immunological detection of a G-protein in Halobacterium

halobium. FEBS Letters 249 : 59–61.

Schuurink RC, Chan PV, Jones RL. 1996. Modulation of

calmodulin mRNA and protein levels in barley aleurone. Plant


Schuurink RC, Shartzer SF, Fath A, Jones RL. 1998.

Characterization of a calmodulin-binding transporter from the

plasma membrane of barley aleurone. Proceedings of the

National Academy of Sciences, USA 19A : 153.

Sentenac H, Bonneaud N, Minet M, Lacroute F, Salmon

JM, Gaynard F, Grignon C. 1992. Cloning and expression in

yeast of a plant potassium ion transport system. Science 256 :

663–665.

Sewing A, Mu$ ller R. 1994. Protein kinase A phosphorylates

cyclin D"at three distinct sites within the cyclin box and at the

C-terminus. Oncogene 9 : 2733–2736.

Sharma R, Lopez-Juez E, Nagatani A, Furuya M. 1993.

Identification of photoinactive phytochrome A in etiolated

seedlings and photoactive phytochrome B in green leaves of

aurea mutant of tomato. Plant Journal 4 : 1035–1042.

Shenolikar S. 1994. Protein serine}threonine phosphatases: new

avenues for cell regulation. Annual Review of Cell Biology 10 :

55–86.

Shimoyama M, Sakamoto M, Nasu N, Shigehisa S, Ueda I.

1972. Identification of the 3«,5«-cyclic AMP phosphodiesterase

inhibitor in potato. Biochemical and Biophysical Research

Communications 48 : 235–241.

Smith CJ, Brown EG, Newton RP. 1979. Isolation of an

adenosine 3«,5«-cyclic monophosphate binding protein from the

tissues of higher plants. Biochemical Society Transactions 6 :

1268–1269.

Smith CJ. 1991. Signal transduction in elicitation of phytoalexin

synthesis in Medicago sativa L. In: Smith CJ, ed. Biochemistry

and molecular biology of plant pathogen interactions. Oxford,

UK: Oxford University Press, 255–270.

Smith CJ. 1996. Accumulation of phytoalexins: defence mech-

anisms and stimulus response system. New Phytologist 132 :

1–45.

Spiteri A, Viratelle OH, Raymond P, Rancillac M,

Labouesse J, Pradet A. 1989. Artefactual origins of cyclic

AMP in higher plant tissues. Plant Physiology 91 : 624–628.

Stryer L. 1986. Cyclic GMP cascade of vision. Annual Review of

Neuroscience 9 : 87–119.

Subbaiah CC, Bush DS, Sachs MM. 1994a. Elevation of

cytoplasmic calcium precedes anoxic gene expression in maize

suspension-cultured cells. The Plant Cell 6 : 1747–1762.

Subbaiah CC, Zhang JJ, Sachs MM. 1994b. Involvement of

intracellular calcium and calmodulin in anaerobic gene ex-

pression and survival of maize seedlings. Plant Physiology 105 :

369–376.

Tarantowicz-Marek E, Kleczkowski K. 1978. Effect of

gibberellic acid on the adenosine 3«,5«-cyclic monophosphate

content in dwarf maize shoots. Plant Science Letters 13 :

121–124.

Taussig R, Gilman AG. 1995. Mammalian membrane-bound

adenylyl cyclases. Journal of Biological Chemistry 270 : 1–4.

Taussig R, Iniguez-Lluhi JA, Gilman AG. 1993. Inhibition of

adenylyl cyclase by Gi alpha. Science 261 : 218–221.

Terada K, Kai T, Okuno S, Fujisawa H, Katsura I. 1990.

Maize leaf phosphoenolpyruvate carboxylase: phosphorylation

of Ser"& with a mammalian cyclic AMP-dependent protein

kinase diminishes sensitivity to inhibition by malate. FEBS

Letters 259 : 241–244.

Toda T, Uno I, Ishikawa T, Powers S, Kataoka T, Broek D,

Cameron, S, Broach J, Matsumoto K, Wigler M. 1987. In

yeast, RAS proteins are controlling elements of adenylate

cyclase. Cell 40 : 27–36.

Tokiwa G, Tyers M, Volpe T, Futcher B. 1994. Inhibition of

G1 cyclin activity by the Ras}cAMP pathway in yeast. Nature

371 : 342–345.

Tong J, Carre! IA, Edmunds LN. 1991. Circadian rhythmicity in

the activities of adenylate cyclase and phosphodiesterase in

synchronously dividing and stationary-phase cultures of the

achlorophyllous ZC mutant of Euglena gracilis. Journal of Cell

Science 100 : 365–369.

Tong J, Edmunds LN. 1993. Role of cGMP in the mediation of

circadian rhythmicity of the adenylate cyclase-cyclic AMP-

phosphodiesterase system in Euglena. Biochemical Pharma-

cology 45 : 2087–2091.

Trewavas AJ. 1997. Plant cyclic AMP comes in from the cold.

Nature 390 : 657–658.

Truelsen TA, De Langhe E, Verbeek-Wyndaele R. 1974.

Cyclic AMP induced growth promotion in sun flower callus

tissue. Archives Internationales de Physiologie et Biochimie 82 :

109–114.

Tu JC. 1978. Biochemical and histochemical investigation of

diurnal variation in adenosine 3«,5«-cyclic monophosphate

concentration and adenylate cyclase activity in white dutch

clover. Protoplasma 99 : 139–146.

Urabe H, Ogawara H. 1995. Cloning, sequencing and expression

of serine}threonine kinase-encoding genes from Streptomyces

coelicolor A3 (2). Gene 153 : 99–104.

Van Onckelen HA, Dupon M, De Greef JA. 1982. High

performance liquid chromatographic identification and quanti-

tation of cyclic adenosine 3« :5«-monophosphate in higher

(Phaseolus vulgaris L.) and lower (Chlorella sp.) plants.

Physiologia Plantarum 55 : 93–97.

Vandepeute J, Huffaker RC, Alvarez R. 1972. Cyclic nucleo-

tide phosphodiesterase activity in barley seeds. Plant Physiology

52 : 278–282.

Vanhoutte K, Van Dongen W, Hoes I, Esmans EL, Van

Onckelen HA, Van den Eeckhout E, van Soest REJ,

Hudson AJ. 1997. Development of a nanoscale liquid chroma-

tography-electrospray mass spectrometry method for the de-

tection and identification of DNA adducts. Analytical Chemistry

69 : 3161–3168.

Venere RJ. 1972. Nucleoside phosphotransferase, phosphomono-

esterase and cyclic nucleotide phosphodiesterase activities of

carrot leaves. Dissertation Abstracts International B, 1369.

Vicente C. 1993. Evolutionary convergence of light perceiving

systems: vision-like cascade process started by phytochrome in

lichens. Endocytobiosis Cell Research 9 : 255–267.

Wachstein M, Meisel E. 1957. Histochemistry of hepatitic



phosphatases at a physiological pH with a special reference to

the demonstration of bile canaliculi. American Journal of

Clinical Pathology 27 : 13–23.

Walton TJ, Cooke CJ, Newton RP, Smith CJ. 1993. Evidence

that generation of inositol 1,4,5-trisphosphate and hydrolysis of

phosphatidyl inositol 4,5-bisphosphate are rapid responses

following addition of fungal elicitor which induces phytoalexin

synthesis in lucerne suspension culture cells. Cell Signalling 5 :

345–356.

Wang T, Sheppard JR, Foker JE. 1978. Rise and fall of cyclic

AMP required for onset of lymphocyte DNA synthesis. Science

201 : 155–157.

Ward AC, Csar XF, Hoffmann BW, Hamilton JA. 1996. Cyclic

AMP inhibits expression of D-type cyclins and cdk4 and

induces p27Kip1 in G-CSF-treated NFS-60 cells. Biochemical

and Biophysical Research Communications 224 : 10–16.

Warmke J, Drysdale R, Ganetzky B. 1991. A distinct potassium

channel polypeptide encoded by the Drosophila eag locus.

Science 252 : 1560–1562.

Witters E, Roef L, Newton RP, Van Dongen W, Esmans EL,Van Onckelen HA. 1996. Quantitation of cyclic nucleotides in

biological samples by negative electrospray tandem mass

spectrometry coupled to ion suppression liquid chroma-

tography. Rapid Communications in Mass Spectrometry 10 :

225–231.

Witters E, Van Dongen W, Esmans EL, Van Onckelen HA.1997a. Ion pair liquid chromatography-electrospray mass

spectrometry for the analysis of cyclic nucleotides. Journal of

Chromatography B 694 : 55–63.

Witters E, Vanhoutte K, Dewitte W, Macha! ckova! I, Benkova!

E, Van Dongen W, Esmans EL, Van Onckelen HA. 1998.Analysis of cyclic nucleotides and cytokinins in minute plant

samples using phase-system switching capillary electrospray-

LC-MSMS. Phytochemical Analysis 10 : 143–151.

Witters E, Vanhoutte K, Van Dongen W, Esmans EL, VanOnckelen HA. 1997b. Qualitative analysis of cyclic nucleotides

and cytokinins using capillary column switching ES-LC-

MSMS. Proceedings of the 45th ASMS conference on mass

spectrometry and allied topics, 1–5 June 1997, Palm Springs, CA,

USA, 163.

Wong Y-S, Cheng HC, Walsh DA, Lagarias JC. 1986.Phosphorylation of Avena phytochrome in vitro as a probe of

light-induced conformational changes. Journal of Biological

Chemistry 261 : 13321–13328.

Wood HN, Lin MC, Braun AC. 1972. The inhibition of plant

animal adenosine 3«,5«-cyclic monophosphate phosphodi-

esterase by a cell division promoting substance from tissues of

Pinus radiata. Biochemical Journal 175 : 931–936.

Yamamoto KK, Gonzalez GA, Biggs WH, Montminy MR.

1988. Phosphorylation-induced binding and transcriptional

efficiency of nuclear factor CREB. Nature 334 : 494–498.

Yount RG, Babcock D, Ballentyne W, Ohala D. 1971.

Adenylyl-imidodiphosphate: an adenosine triphosphate analog

containing a P-N-P linkage. Biochemistry 10 : 2484–2489.

Zhang Y, Snell WJ. 1993. Differential regulation of adenylyl

cyclases in vegetative and gametic flagella of Chlamydomonas.

Journal of Biological Chemistry 268 : 1786–1791.

Zhang Y, Snell WJ. 1994. Flagellar adhesion-dependent regu-

lation of Chlamydomonas reinhardtii adenylyl cyclase in vitro : a

possible role for protein kinases in sexual signalling. Journal of

Cell Biology 125 : 617–624.

Zufall F, Firestein S, Sheperd GM. 1994. Cyclic nucleotide

gated ion channels and sensory transduction in olfactory

receptor neurons. Annual Review of Biophysics and Biomolecular

Structure. 23 : 577–607.

Note added in proof

The presence of both adenylyl cyclase and guanylyl-

cyclase has now been demonstrated in spinach

chloroplasts by quantitative mass spectrometry.

Initial evidence indicates crosstalk between the

enzymes (Newton RP, Bayliss MA, Langridge

JA, Wilkins ACR, Games DE, Walton TJ, Bren-

ton AG, Diffley PE, Harris FM, Smith CJ. 1999.

Product identification and kinetic studies of nucleo-

tidyl cyclase activity in isolated chloroplasts by

quantitative fast-atom bombardment mass spectro-

metry. Rapid Communications in Mass Spectrometry

13 : 979–985.

Tansley Review No. 106

Documents