425 Significant advances in understanding plant cyclic nucleotide signalling have been made in the past two years. The roles of these molecules in the regulation of ionic channels, defence responses and the apical growth of cells are being uncovered. Addresses *Institute of Cell and Molecular Biology, University of Edinburgh, Daniel Rutherford Building, Edinburgh EH9 3JH, United Kingdom; e-mail: trewavas@ed.ac.uk † Departamento de Biologia Vegetal, Faculdade de Ciências de Lisboa, R. Ernesto de Vasconcelos, Bloco C2, 1749-016 Lisboa, Portugal Current Opinion in Plant Biology 2002, 5:425–429 1369-5266/02/$ —see front matter © 2002 Elsevier Science Ltd. All rights reserved. Published online 30 July 2002 Abbreviations ABA abscisic acid cAMP cyclic adenosine monophosphate cGMP cyclic guanosine monophosphate CN cyclic nucleotide CNGC cyclic-nucleotide-gated channel GA gibberellin PKA protein kinase A Introduction The study of plant cyclic nucleotides (CNs) has had a very chequered history. Although the initial doubts and uncertainties over the role of CNs in plant cell signalling have now dissolved, there are still many areas of CN biology that are not being adequately investigated in plants. A detailed review of plant CNs has been published that covers research up to 1999 [1]. We refer to evidence published before this only where it is necessary to provide continuity or to illustrate less investigated areas (e.g. gene expression). CNs are synthesised by adenylyl/guanylyl cyclases and degraded by specific phosphodiesterases. Phosphodiesterases are usually regulated by the Ca 2+ -binding protein calmodulin. Thus, an intimate relation between CN-signalling and Ca 2+ signalling is to be expected and this is slowly being uncovered [2]. Adenylyl cyclase and phosphodiesterase Although there is good evidence for the presence of specific CN phosphodiesterases in plant cells [1], there is only indirect evidence for the presence of adenylyl cyclases [3 •• ]. Nevertheless, the presence of CNs in plant tissues has been established chemically [1], and so CN-synthetic enzymes must be present in plants. The purification of plasma-membrane-bound adenylyl cyclase from animal cells has been achieved only recently, and there is no obvious equivalent to these large 120-kiloDalton enzymes in the Arabidopsis database. Recently, a new form of mammalian adenylyl cyclase was detected in rat. This protein has a catalytic domain that is very similar to adenylyl cyclase from cyanobacteria and myxobacteria [4,5], indicating great diversity in the adenylyl cyclase family. Soluble adenylyl cyclases whose sequences differ from those of the classic mammalian enzymes have been reported in algae and fungi [6] and it is likely that other cyclases might emerge. Cloning of a putative adenylyl cyclase from pollen revealed that this cDNA had common motifs not only with its fungal counterpart [3 •• ] but also with genes encoding proteins that are involved in disease responses. Cyclic adenosine monophosphate (cAMP) is believed to be involved in such responses [7], and parallels between pollen-tube growth in the style and infection by fungal hyphae are frequently drawn. The pollen cDNA, although not full length, caused accumulations of cAMP when expressed in Escherichia coli and complemented a catabolic defect in carbohydrate fermentation in an E. coli cyaA mutant [3 •• ]. Transformation with antisense oligos directed against the pollen cDNA or treatment with antagonists of adenylyl cyclase disrupted pollen-tube growth (Figure 1), suggesting that this requires continued synthesis of cAMP. This hypothesis was supported by the imaging of cAMP in growing pollen tubes. There, forskolin, an activator of adenylyl cyclase, transiently increased cAMP levels, whereas dideoxyadenosine, an inhibitor of adenylyl cyclases, caused a temporary decline in cAMP [3 •• ]. Transient changes in cAMP levels also indicated that cAMP-degrading enzymes, such as phosphodiesterases, must be present in pollen tubes. These exciting results require confirmation in other systems, most notably in guard cells. Furthermore, cDNA libraries need to be screened in the cyaA bacterial line to aid the identification of other adenylyl cyclases in plants. Cyclic nucleotides and ion channels It is well established that plant cells possess cyclic- nucleotide-gated channels (CNGC) that transport several ions such as sodium, potassium or calcium [8–10]. Experiments using patch clamp on mesophyll cells showed that an outward K + current could be increased by step-wise elevation of cAMP (but not cyclic guanosine monophosphate [cGMP]), but only when an inhibitor of phosphodiesterase was present [9]. A specific competitive peptide inhibitor of protein kinase A inhibited the outward current. An inward K + channel that was modulated by cAMP was also reported [10], and the highly specific inwardly rectifying K + channels KAT1 and AKT1 of Arabidopsis have their activity modulated by cGMP [8]. Cyclic nucleotides: the current dilemma! Anthony J Trewavas*, Cecília Rodrigues † , Cláudia Rato † and Rui Malhó †