Plant Physiol. (1999) 120: 1-6

UPDATE ON REGULATION

Le Calcium, C'est la Vie: Calcium Makes Waves¹

Anthony Trewavas^*

Institute of Cell and Molecular Biology, University of Edinburgh, Kings Building, Mayfield Road, Edinburgh EH9 3JH, Scotland, United Kingdom

	INTRODUCTION

Top Introduction References

Every reader of this article knows that he or she started life as a single cell. Less familiar is the debt we owe to calcium in our earliest seconds. Penetration of the egg by the paternal sperm initiated an epigenetic calcium wave that moved quickly as a hollow band across the cytoplasm. In the wake of this calcium wave, processes were activated that led to cell division, differentiation, growth, and our eventual appearance as mature adults. As it is said in France, "Le calcium, c'est la vie." A calcium wave marked the onset of our existence, and will quite probably mark our demise: Irreversible failure of calcium-wave generation in the heart is the most common cause of death.

Therefore, calcium waves are a life-and-death issue. So much depends on the intricacies and cellular patterning of the simple ion that is calcium. How much is understood about calcium waves? What is the mechanism of calcium-wave formation? What is their significance? What specific role could calcium waves perform in plant cells? Some profound insights come from considering these questions. The true significance of calcium waves may lie in constructing a cellular "intelligence."

	CALCIUM IS NOT A MOBILE ION IN THE CYTOPLASM

In 1957, Hodgkin and Keynes performed a simple but seminal experiment, injecting small portions of the recently developed Ca and ⁴²K into squid axons. Several hours later, they transversely sectioned the axon and examined the distribution of these isotopes. Whereas K had uniformly diffused throughout the axon, ⁴⁵Ca had remained at the site of injection (Fig. 1), showing that calcium is not freely mobile in the cytoplasm.

View larger version (23K): [in this window] [in a new window]   Figure 1. Diagrammatic summary of the Hodgkin and Keynes (1957) experiment. Solid line, 45Ca; dotted line, 42K.

If calcium does move within the cytoplasm, mechanisms other than diffusion are responsible.

Two facts can account for the above observations. Calcium binds to many proteins that are attached to the cytoskeleton or to membrane surfaces. Estimates suggest a cytoplasmic-binding capacity per cell of 0.1 to 0.5 mM calcium (Mahlo et al., 1998). Calcium-dependent calcium-ATPases rapidly pump excess calcium into organelles, vesicles, and, in plant cells, out into the cell wall. The vacuole is a major repository of plant-cell intracellular calcium and must make an unusual contribution to cell-calcium signaling. Other important intracellular calcium stores are the ER, the mitochondria, and possibly the Golgi vesicles. Intracellular stores and the cell wall may contain 0.1 to 10.0 mM calcium. The calcium remaining in the cytoplasm after protein binding and organelle uptake is between 10 and 200 nM. This is called "resting" calcium. There are pronounced electrochemical concentration gradients of calcium between the cytoplasm and the cell wall and between the cytoplasm and the intracellular vesicles and organelles.

	CALCIUM CHANNELS CONNECT THE STORES OF CALCIUM WITH THE CYTOPLASM

Calcium-channel proteins permit the flow of calcium between the cytoplasm (with its very low concentrations) and the other cellular compartments (with much higher concentrations). Channels open when cells are signaled, and calcium enters the cytoplasm down its electrochemical gradient. Families of channels are known to exist, some activated by membrane potential, others by membrane stretch, and still others by various kinds of signals, including second messengers. A single channel can transmit 10⁶ atoms of calcium per second. Elevation of calcium at the cytoplasmic channel mouth can be rapid: calcium concentrations there can reach 0.1 mM. However, channels rapidly close when the cytoplasmic level increases and the activation of calcium-dependent calcium-ATPases causes calcium to move back to the intracellular stores and into the cell wall. Transient elevations of calcium (sometimes called spikes) may last anywhere from a few seconds to many minutes in plant cells, depending on the characteristics of the stimulating signal.

	HOW TO MAKE CALCIUM MORE MOBILE IN THE CYTOPLASM

The observations by Hodgkin and Keynes (1957) make it difficult to understand how calcium waves can move. If calcium is not mobile in the cytoplasm, then channel opening would limit calcium elevations to the domain of the channel mouth. Although a wave might be initiated, it could not move. The discovery of the role of inositol phosphates in the generation of calcium waves provided a partial answer to this dilemma.

Phospholipase C, which is bound to the plasma membrane and activated by signals, is present as several isoforms in plant cells and many isoforms in animal cells. The substrate for phospholipase C, PIP₂, is synthesized from precursors in the plasma membrane. When phospholipase C is activated, PIP₂ is hydrolyzed to diglyceride and IP₃. Unlike calcium, IP₃ is freely mobile in the cytoplasm. IP₃-sensitive channels located in the membranes of organelles, vesicles, and the plasma membrane are opened, enabling calcium to enter the cytoplasm. Alexandre et al. (1990) and Gilroy et al. (1990) established the IP₃ mobilization of calcium in plant cells a decade ago; more recently, Allen et al. (1995) demonstrated IP₃-binding (and heparin-sensitive) channels.

	HOW TO MAKE A CALCIUM WAVE

Although these observations helped our understanding, they did not resolve the question of calcium-wave construction and movement (Fig. 2A). Elevation of IP₃ and its rapid diffusion throughout the cytoplasm would simply open all intracellular IP₃-dependent channels; no spatial or kinetic distinction between different regions of the cell would be detected and no calcium wave would be observed. Only with purification and kinetic analysis of the IP₃-sensitive channel was the conundrum resolved: Marchant and Taylor (1997) found that the IP₃ channel needs both calcium and IP₃ to open.

View larger version (18K): [in this window] [in a new window]   Figure 2. The effect of IP3 and calcium on the opening of IP3-sensitive channels. A, Closed-channel binding only IP3 and an open-channel binding both IP3 and calcium. B, Illustration of how an open channel opens an adjacent channel by locally increasing calcium.

When IP₃ binds to the channel, a calcium-binding site is briefly exposed. If calcium is absent, the channel rapidly inactivates (Fig. 2A); if calcium is present, the calcium-binding site is occupied and the channel is opened. But the binding of calcium serves only to delay eventual IP₃-induced channel inactivation. Protein phosphorylation probably inactivates the channel, but the mechanism is not yet understood. Mobilization of calcium through a single IP₃-dependent channel is therefore brief and self-limiting. However, IP₃-sensitive channels act as a kind of coincidence counter of local calcium and IP₃ concentrations, safeguarding against spontaneous opening.

Kinetic analysis of the IP₃ channel indicates flexibility in the requirements for calcium and IP₃. High concentrations of calcium need only low concentrations of IP₃ to open the channel; conversely, high concentrations of IP₃ need only low concentrations of calcium.

	HOW TO MAKE THE CALCIUM WAVE MOVE

When an IP₃-dependent channel opens, the calcium concentration surrounding adjacent channels increases (Fig. 2B). The calcium-binding sites of these channels will be occupied and will briefly open, enabling the opening of others. Calcium is therefore responsible for inducing further calcium release, and this release underpins wave movement. This wave is not a forward movement of calcium, but a forward movement of calcium release. IP₃-induced inactivation of channels causes the direction of calcium-wave movement to be away from the point of origin (Fig. 3A).

View larger version (31K): [in this window] [in a new window]   Figure 3. Comparison of the information flow through a calcium cloud and a simple neuronal network. A, Generation of a wavefront of calcium release through a number of IP3-sensitive channels from a single activated channel (black arrows). B, Diversification of a single electrical signal through a neuronal network with the signal impacting on a single neuron (black arrows).

IP₃-sensitive channels are located in membranes, and the calcium wave thus moves across the cytoplasmic membrane face. Many calcium-dependent proteins (e.g. calmodulin and calcium/calmodulin-dependent protein kinases) are bound to the membrane surface and will therefore be activated (Collinge and Trewavas, 1989). Processes downstream from calcium signaling (e.g. opening of potassium and chloride channels and calcium-dependent gene expression) will be induced. The domains on the membrane surface are difficult to examine, but new technology with targeted calcium probes is providing new insights (Knight et al., 1996).

Other second messengers (e.g. ADP-Rib) can also mobilize calcium from intracellular stores (Allen et al., 1995). Another layer of complexity in calcium signaling is therefore added.

	CONSTRAINTS ON CALCIUM-WAVE FORMATION AND LIFETIME

In the presence of optimal cellular IP₃ concentrations, a full calcium wave is likely. The wave will move from its initiation site progressively throughout the cell. The specific characteristics of initiation sites are not known but might reflect the clustering of receptors or an unusual density of channels. However, there are many situations in which the wave will be truncated and elevation of calcium will be limited to particular regions of the membrane surface.

First, IP₃ is rapidly hydrolyzed by specific phosphatases. If the initial stimulating signal is weak, IP₃ elevations will be low and will rapidly decline below the critical threshold. The calcium wave will be confined to specific cellular regions and can be described as "clouds" or "puffs."

Second, spatial separation of phospholipase C from IP₃-specific phosphatases will generate transient gradients of IP₃. Only where IP₃ exceeds a critical threshold will spatially constrained increases of calcium be detected.

Third, the calcium contents of regions of the ER are known to vary. In regions that contain insufficient stored calcium, wave propagation will cease. The final wave will have a complex topology composed of high and low calcium concentrations and regions where no calcium elevation is detectable. Spatial differentiation of calcium-dependent proteins on the membrane surface enables many different combinations of downstream proteins to be activated.

Fourth, if calcium-dependent calcium-ATPases are clustered on the membrane surface, a complex topological calcium wave will again result.

And, finally, whether waves jump from one membrane surface to another depends on the distance between them and the strength of the initial signal. The wave-initiation site determines the region in which the wave starts. Figure 4 illustrates the expected spatial distribution of calcium elevation as a result of truncated waves induced by different signals.

View larger version (34K): [in this window] [in a new window]   Figure 4. Anticipated changes in cytosolic calcium by a truncated calcium wave initiated by different signals in different parts of the cell. Spatial differences should initiate different molecular and physiological changes.

	CALCIUM WAVES IN PLANT CELLS

IP₃ channels are located in the vacuole membrane (tonoplast) and almost certainly in the ER and plasma membrane of plant cells. Calcium waves can be induced in pollen tubes by photolytic release of IP₃ from loaded, caged IP₃ (Franklin-Tong et al., 1996). The wave initiates in a cellular region containing the nucleus and ER and moves to the tip within 1 min via a relay of IP₃-sensitive channels. The absence of a vacuole in this region of the pollen tube indicates that the tonoplast is not involved in the transmission of this calcium wave. Some of these IP₃-sensitive channels are found in the plasma membrane. Caged IP₃ photolysis in pollen tubes containing heparin (to inhibit binding of IP₃ to the channel) permits wave initiation; however, the wave is rapidly truncated and fails to reach the tip.

A calcium wave in pollen tubes can also be induced by localized photolytic release of calcium in just the cellular region containing the nucleus and ER (Mahlo and Trewavas, 1996). These observations confirm that calcium and IP₃ can separately activate plant IP₃-sensitive channels. Furthermore, a wave-initiation site has been clearly identified. Calcium waves have also been induced by suddenly reducing the plasma membrane potential of guard cells and subjecting Fucus rhizoids to hypo-osmotic shock, and in both cases the wave was initiated near the cell periphery (Taylor et al., 1996; Grabov and Blatt, 1998).

"Clouds" of calcium confined to discrete cellular regions have been imaged in guard cells induced to close with different signals. The spatial location of the cloud differed according to the signal used (Gilroy et al., 1991).

	DIFFERENT SIGNALS WILL INDUCE CALCIUM WAVES OF DIFFERENT SHAPE, FORM, AND KINETICS IN PLANT CELLS

Different signals do not uniformly activate phospholipase C. Furthermore, the strength of any signal should quantitatively modify cellular IP₃ concentrations. Variable kinetics in the development of the calcium wave will result. Because a number of channels must open within the vicinity of each other and within a certain time period for calcium-wave initiation, a lag period may also be detected. Lag periods may be common when IP₃ concentrations are low. Many factors (e.g. channel density, channel state, intracellular store replenishment, other second messengers, and the age and state of development of the cell) may also contribute to the final kinetics of the wave.

Table I summarizes the measured kinetics of calcium transients in tobacco seedlings after initiation by different signals. Each signal does indeed induce unique kinetics and therefore should induce a unique topology of calcium elevation (the so-called fingerprint) and activate unique combinations of calcium-dependent proteins.

Some cellular states and signals will enable repetitive wave initiation. Repetition may lead to oscillations of calcium, which have been observed in guard cells, root hairs, pollen tubes, and tobacco seedlings (for refs., see Trewavas and Mahlo, 1997).

	NEURAL NETWORKS PROVIDE ORGANISMS WITH INTELLIGENCE

Figure 3B indicates that there is a formal equivalence in character, structure, and information flow between a neural network and a calcium wave. Neural networks, both real and artificial, have five important properties that enable them to act intelligently: (a) they are spatially structured; (b) individual neurons can act as coincidence controllers, passing or blocking specific signals arriving coincidentally from different neurons and thus from different signals (protein kinases control information flow through neurons; Abel et al., 1998); (c) synchronization among the membrane potentials of many neurons leads to rhythmic oscillations (e.g. -rhythms) in membrane potential, which are modified by further signaling; (d) neural networks "learn"; and (e) memory of previous signals can be accessed and used to modify current signal information.

Learning and memory are interlinked. Networks can learn by reinforcing a neural pathway connecting the signal and the response. Reinforcement can occur either by increasing the numbers of dendrites (connections) or by increasing the strength of existing connections between adjacent neurons in the pathway. The latter process is thought to involve integrins (Grotweil et al., 1998). The consequence of learning is an enormous acceleration in information flux between the signal and the elements initiating the desired response.

	THE CALCIUM WAVE IS ANALOGOUS TO AN OPERATING NEURAL NETWORK

If the calcium signaling system has a formal equivalence to a neural network, it should be able to compute, remember, and learn even though it is confined to single cells. Although admittedly more limited than a complex neural network, calcium waves most certainly provide for some aspects of "intelligent" behavior by plants. Even simple neural nets involving no more than 12 to 14 neurons have good computational properties (Lewis and Kristan, 1998), and a calcium system should do better.

The parallels between neural networks and calcium waves are striking: the calcium wave is spatially structured. The IP₃-sensitive channel can act as a coincidence counter (specific directions of calcium release can be propagated whereas others are inhibited) and channel inactivation can be regulated by protein kinases that regulate coincidence signaling. Oscillations in cellular calcium represent the cooperative integration of the behavior of many individual IP₃-sensitive channels. When plant cells are signaled, they increase the cellular constituents concerned with calcium signaling. The cellular content of phospholipase C, calmodulin, calcium-activated protein kinases and other kinases, cADP-Rib, PIP₂, and IP₃ are all increased by signaling (Braam and Davis, 1990; Drobak and Watkins, 1994; Urao et al., 1994; Hirayama et al., 1995; Wu et al., 1997). However else these data might be regarded, this is a clear example of cellular learning. The consequence of these changes is an accelerated information flux through calcium-dependent pathways and is analogous to reinforcement.

So far as I am aware, this property of increasing the cellular content of signal components is unique to plants. Some of the early difficulties in detecting inositol phosphates such as PIP₂ and IP₃ can be traced to using unstimulated cells for investigation. ABA increases the cellular accumulations of cADP-Rib, and the wall removal that forms protoplasts also elevates IP₃. The changes in these constituents are often slow, taking many hours, which of course emphasizes the similarity to neural network learning (Drobak and Watkins, 1994; Wu et al., 1997). In an animal cell, for example, complete elevations of inositol phosphates tend to occur in 0.1 s or less.

Another similarity between neural networks and calcium waves is that cells are able to access a memory because a similar calcium signal in different cell types initiates different effects. Cells know what they are and where they are. They are able to "remember" previous experience and incorporate this information to modify current signals. Important aspects of this memory reside in the cell wall because cell wall removal resets many aspects of development in algae and higher plants (Pennell, 1998).

A calcium-based memory can result from exhaustion of calcium stores from certain areas of the ER. The final topology of any calcium wave passing over this ER surface would reflect variations in ER-stored calcium. A unique combination of effector proteins would thus be activated. However, this memory would be time-limited because cellular membranes are dynamic in characteristics and structure.

In the cell wall, on the other hand, calcium-pectate secretion could be targeted to discrete wall areas during prior signaling, growth, and development. Calcium concentrations outside the plasma membrane would be clustered in a spatially unique pattern. The development of a calcium wave on the plasma membrane surface would be propagated only in those regions where calcium was sufficiently high. The final geometry of the wave would represent the integration of the present signal modified by a stable, long-term memory. Protein distribution on the plant plasma membrane has been observed to be highly differentiated. If plasma membrane effector proteins (e.g. calmodulin) are clustered in specific regions, novel physiological outcomes result.

Verdus et al. (1997) observed that exogenous calcium can modify a long-term memory controlling regeneration in plant cells. Bands of wall material have been observed to be laid down in growing pollen tubes (Pierson et al., 1995), suggesting that localized pectin secretion is feasible.

	WHY DO PLANTS NEED AN INTELLIGENT COMPUTATIONAL SYSTEM?

Intelligence is not usually associated with plant behavior. But intelligence is not the same as sentience. Both Darwin and Sachs have commented on the similarity between plant cell signal transduction and neural network behavior: "In several respects light seems to act on plants in nearly the same manner as it does on animals by means of the nervous system" (Darwin, 1880); and "I have repeatedly had cause to refer to certain resemblances between the phenomena of irritability in the vegetable kingdom and those of the animal body" (Sachs, 1881).

What particular problems faced by plants require intelligent behavior? Wild seedlings must grow where they land. The external environment is composed of probably 17 distinct constituents and, being variable in intensity even from minute to minute, generates an almost infinite variety of environmental states. There are probably as many internal plant signals that either pass through or are perceived at the plasma membrane (Trewavas and Mahlo, 1997).

In responding intelligently to this multiplicity of signals, plants have become masters of phenotypic and physiological plasticity, which allows them to cope with the variable circumstances that surround them. Behavioral plasticity surely demands a cellular system of considerable computing power if plants are to survive the signal morass in which they find themselves. The ubiquity of calcium involvement in plant-cell signal transduction suggests that calcium forms the basis of the intelligent system controlling plasticity.

Many examples of phenotypic plasticity exist. I have chosen one example, root branching, to illustrate this article; however, the reader can construct additional scenarios for any of the multiple phenotypically plastic responses to water, light, herbivory, pests, wind, temperature, minerals, soil factors, and so on. Leaf abscission and herbivory provide excellent examples (Addicott, 1982; Karban and Baldwin, 1997). There are far too many environmental situations for simple genetic control to work; so epigenetic control must be the ultimate regulator.

In the laboratory Drew et al. (1973) have shown that the primary roots of cereals produce a profusion of lateral roots when growing in layers of soil rich in nitrate or phosphate. Some plants can efficiently mine the soil, much as humans would. But in wild grasses much other information that will modify the outcome has to be taken into account (Trewavas, 1992): the age of the root, the state of development of the pericycle, whether other laterals have already formed (grasses produce a mat of secondary roots that efficiently exploit the surrounding soil), the quality of the soil (e.g. balance, excess of other minerals, or compactness), the availability of water (which may override other factors), the state of the shoot (etiolated, shaded or green, damaged, or eaten), nearness of the shoot, wind sway, carbohydrate levels, availability of growth factors and amino acids, nearness of or damage by parasites, ethylene, carbon dioxide and oxygen in the soil, flowering, and so on. Each of these variables known to modify root branching does so quantitatively. The combination of possible environmental circumstances that pericycle cells must assess is enormous, and the amount of information that has to be computed rapidly mounts up; but after assessment, the numbers of pericycle cells to be induced or root primordia to be formed will be decided. The decision will be intelligent, geared in some way to an optimal solution for survival of the particular individual plant. As plants represent 99% or more of the biomass of the earth, the computing power has some degree of accuracy!

	CALCIUM SIGNALING IS AN EMERGENT PROPERTY

Cellular molecules, cells, tissues, whole plants, populations, and ecosystems are frequently arranged in a hierarchy. At each level of the hierarchy, the connections between the constituents generate the level above. The 50,000 or so cellular molecular species represent the lowest level. Interactions between these molecules construct a complex network perhaps best described as analogous to a very large, badly woven fishing net. Like all networks, the connections form molecular collectives that generate new emergent properties. Cellular behavior is emergent behavior, as are cytoskeletal dynamics, cell division, cell growth, and cell development. Emergent properties are not predictable from the most detailed analysis of the constituents, because by definition it is the interactions between the molecules that are critical. Emergent properties can be observed, but understanding is another matter.

Undoubtedly, the spatial calcium signal is a primary emergent characteristic in its control of metabolism and gene expression. However, every cell is unique in certain molecular characteristics, so the internal calcium signal in each cell is represented by a unique topology (Trewavas and Malho, 1997). We may never fully understand the complex information implicit in the topological calcium signal that is generated. But, then again, 20 years ago we understood nothing at all of calcium in plant cells and now knowledge increases on a daily basis. Life is complex, and "le calcium, c'est la vie" suggests that calcium signaling is as infinitely complex as life itself.

	FOOTNOTES

   Received January 15, 1999; accepted February 17, 1999.

	ABBREVIATIONS

Abbreviations: IP₃, inositol-1,4,5-trisphosphate. PIP₂, phosphatidylinositol-4,5-bisphosphate.

	ACKNOWLEDGMENTS

Many people have contributed to this discussion. Errors and omissions are my own responsibility.

	LITERATURE  CITED

Top Introduction References

Abel T, Martin KC, Batsch D, Kandel ER (1998) Memory suppressor genes: inhibitory constraints on the storage of long-term memory. Science 279: 339-341

Addicott FT (1982) Abscission. University of California Press. Berkeley, CA

Alexandre J, Lassales JP, Kado RT (1990) Opening of calcium channels in isolated red beet root vacuole membrane by inositol 1,4,5 trisphosphate. Nature 343: 567-570 [CrossRef]

Allen GJ, Muir SR, Sanders D (1995) Release of calcium from individual plant vacuoles by both InsP₃ and cyclic ADP-ribose. Science 268: 735-737 [Abstract/Free Full Text]

Braam J, Davis RW (1990) Rain, wind and touch induced expression of calmodulin and calmodulin related genes in Arabidopsis thaliana. Cell 63: 357-364

Collinge M, Trewavas AJ (1989) Location of calmodulin in the pea plasma membrane. J Biol Chem 364: 8865-8872

Darwin C (1880) The Power of Movement in Plants. John Murray, London

Drew MC, Saker LR, Ashley TW (1973) Nutrient supply and the growth of the seminal root system in barley. J Exp Bot 24: 1189-1202 [Abstract/Free Full Text]

Drobak BK, Watkins PAC (1994) IP₃ production in plant cells: stimulation by the venom peptides mellitin and mastaporan. Biochem Biophys Res Commun 205: 739-745 [CrossRef][ISI][Medline]

Franklin-Tong VE, Drobak BK, Allan AC, Watkins PAC, Trewavas AJ (1996) Growth of pollen tubes of Papaver rheoas is regulated by a slow-moving calcium wave propagated by inositol triphosphosphate. Plant Cell 8: 1305-1321 [Abstract]

Gilroy S, Fricker M, Read ND, Trewavas AJ (1991) Role of calcium in signal transduction of Commelina guard cells. Plant Cell 3: 333-344 [Abstract/Free Full Text]

Gilroy S, Read ND, Trewavas AJ (1990) Elevation of stomatal cytosol calcium by photolysis of loaded caged probes initiates stomatal closure. Nature 346: 769-771 [CrossRef][Medline]

Grabov A, Blatt MR (1998) Membrane voltage initiates calcium waves and potentiates calcium increase with abscisic acid in guard cells. Proc Natl Acad Sci USA 95: 4778-4783 [Abstract/Free Full Text]

Grotwiel MS, Beck CDO, Wu KH, Zhu X, David RL (1998) Integrin-mediated short term memory in Drosophila. Nature 391: 455-460 [CrossRef][Medline]

Hirayama T, Ohto C, Mizoguchi T, Shinozaki K (1995) A gene encoding a phosphatidylinositol-specific phospholipase C is induced by dehydration and salt stress in Arabidopsis thaliana. Proc Natl Acad Sci USA 92: 3903-3907 [Abstract/Free Full Text]

Hodgkin AL, Keynes RD (1957) Movements of labeled calcium in giant squid axons. J Physiol 138: 253-281

Karban R, Baldwin IT (1997) Induced Responses to Herbivory. The University of Chicago Press, Chicago, IL

Knight H, Trewavas AJ, Knight MR (1996) Cold calcium signalling in Arabidopsis involves two cellular pools and a change in calcium signature after acclimation. Plant Cell 8: 489-503 [Abstract]

Lewis JE, Kristan WB (1998) A neuronal network for computing population vectors in the leech. Nature 391: 76-78 [CrossRef][Medline]

Malho R, Moutinho A, Van der Luit A, Trewavas AJ (1998) Spatial characteristics of calcium signalling: the calcium wave as a basic unit in plant cell calcium signalling. Phil Trans R Soc Lond B 3: 1463-1473

Malho R, Trewavas AJ (1996) Localized apical increases of cytosolic free calcium control pollen tube orientation. Plant Cell 8: 1935-1949 [Abstract]

Marchant JS, Taylor CW (1997) Cooperative activation of IP₃ receptors by sequential binding of IP₃ and calcium safeguards against spontaneous activity. Curr Biol 7: 510-518 [CrossRef][ISI][Medline]

Pennell R (1998) Cell walls: structures and signals. Curr Opin Plant Biol 1: 504-510 [CrossRef][ISI][Medline]

Pierson ES, Li YQ, Zhang HQ, Willemse MTM, Linskens HF, Cresti M (1995) Pulsatory growth of pollen tubes: investigation of a possible relationship with the periodic distribution of wall components. Acta Bot Neerl 44: 121-128

Sachs JV (1881) Lectures on the Physiology of Plants. Clarendon Press, Oxford, UK

Taylor AR, Manison NFH, Fernandez C, Wood J, Brownlee C (1996) Spatial organisation of calcium signalling involved in volume control of the Fucus rhizoid. Plant Cell 8: 2015-2031 [Abstract]

Trewavas AJ (1992) Growth substances in context: a decade of sensitivity. Biochem Soc Trans 20: 102-108 [ISI][Medline]

Trewavas AJ, Malho R (1997) Signal perception and transduction, the origin of the phenotype. Plant Cell 9: 1181-1195 [CrossRef][ISI][Medline]

Urao T, Katagiri T, Mizoguchi T, Yamaguchi-Shinozaki K, Hayashida N, Shinozaki K (1994) Two genes that encode calcium-dependent protein kinases are induced by drought and high salt stresses in Arabidopsis thaliana. Mol Gen Genet 244: 331-340 [ISI][Medline]

Verdus MC, Thellier M, Ripoli C (1997) Storage of environmental signals in flax: their morphogenetic effect as enabled by a transient depletion of calcium. Plant J 12: 1399-1410 [CrossRef]

Wu Y, Kuzma J, Marechal E, Graeff R, Lee HC, Foster R, Chua NM (1997) Abscisic acid signalling through cyclic ADP-ribose in plants. Science 278: 2127-2130

Copyright Clearance Center:   0032-0889/99/120//06
 
© 1999 American Society of Plant Physiologists