The occurrence of eATP appears to be widespread across the plant kingdom—it has been directly or indirectly detected in many plant species including Arabidopsis thaliana,2–5 Zea mays,2 Phaseolus vulgaris,2 Nicotiana tabacum,1,2 Medicago truncatula6 and Salvia miltiorrhiza.7 The existence of eATP in all plant species investigated to date suggests that it likely has fundamental roles in plant growth and development. Although eATP is also found in vertebrate,8,9 Drosophila10 and microbial cells,11 transcriptional induction of genes coding for key enzymes of plant hormone biosynthesis by eATP4 points to the possibility that eATP-mediated signaling might have signaling functions unique to the plant kingdom. In our recent article1 we indeed demonstrated a unique function of plant eATP—it is a negative regulator of pathogen defense gene expression and disease resistance. Here, we propose a model that predicts the existence of a new class of genes/proteins mediating eATP effects on defense gene expression and cell death. We call these eARPs for extracellular ATP regulated proteins (eARPs). We also discuss the complex relationship between eATP and salicylic acid (SA)-mediated signaling in the context of eARP regulation.

Extracellular ATP has been detected on the surfaces of intact plant leaves on which cuticle development has been suppressed by high humidity,3 in the extracellular fluids of cell suspension cultures,1,2 germinating pollen5 and hydroponic plants,2,4,12,13 on surfaces of root cells,6 and in guttation fluids collected from intact leaves (Chivasa S, unpublished data). We previously demonstrated that enzymatic removal of eATP from these tissues or disruption of its signaling via competitive exclusion from target binding sites with the non-hydrolysable analogue β, γ-methyleneadenosine 5′-triphosphate (AMP-PCP), activates cell death.2 Our recent results reveal that this cell death response is strongly regulated by light intensity.1 This is reminiscent of the requirement of light for hypersensitive cell death induced by certain incompatible pathogens.14 This may indicate a possible involvement of chloroplasts in the cell death response as it has now been demonstrated that porphyrin/chlorophyll catabolites act as pro-death signals in plants15,16 and are the basis for the spontaneous cell death in some lesion-mimic mutants.17,18

Although the initial receptor or target protein that directly interacts with eATP in the extracellular matrix or plasma membrane is unknown, downstream second messengers recruited for intracellular signaling include Ca²⁺ ions,4,19,20 reactive oxygen species,21,22 and nitric oxide.20,23 A model of what we envisage to happen during eATP-mediated suppression of cell death and defense gene expression, based on our results in tobacco, is schematically depicted in Figure 1. We suggest that eATP signaling controls the expression of two sets of eARP genes, (1) eARP genes regulating cell death/viability (designated in our model as eARP^via) and (2) eARP genes negatively-regulating pathogen defense genes (designated eARP^def). Under normal growth conditions, eATP-mediated signaling controls eARP^via genes, whose expression is required to suppress a default cell death pathway. Removal of eATP using glucose-hexokinase treatment disrupts expression of these genes, thereby triggering the onset of cell death. Impairment of eATP signaling via competitive exclusion with AMP-PCP will have a similar effect to that activated by glucose-hexokinase treatment. However, even though exogenous SA treatments cause eATP depletion, cell death does not occur because SA activates a separate signaling that protects against cell death.1,15 The protective effects of SA may be mediated by positive regulation of eARP^via genes or different sets of genes/proteins that are not under eATP regulation.

Increases in SA via exogenous SA application, treatments with pathogen-derived elicitors,24 or inoculation with incompatible pathogens,25 will diminish the level of eATP.1 Ablation of eATP downregulates eARP^def gene expression, which switches on expression of defense genes such as pathogenesis-related (PR) genes. Likewise, depletion of eATP by glucose-hexokinase or impairment of eATP signaling using AMP-PCP also disrupts expression of eARP^def genes—this switches on defense gene expression. Remarkably, increasing eATP by addition of exogenous ATP enhances the expression of eARP^def genes, which hinders basal disease resistance and renders the treated tissues more susceptible to pathogens. However, we acknowledge the possibility that regulation of eARPs by both eATP and SA could be post-transcriptional. We also note that the relationship between SA and eATP is complex. Increasing eATP by infiltrating leaves with exogenous ATP triggers a decrease in the levels of SA. Exogenous SA also leads to a reduction in eATP. Therefore, these two signaling molecules mutually regulate their steady-state levels. Their effects on defense gene expression are antagonistic, while they both support cell viability.

We have investigated ATP-induced gene expression in Arabidopsis using whole genome DNA chips and identified several putative eARP genes. They fall into two categories with distinct expression profiles; (1) putative eARP^via, which respond similarly to both ATP or SA treatments, and (2) putative eARP^def, which respond to SA in the opposite direction to their response to ATP (Fig. 2A). Our hypothesis states that exogenous SA treatment depletes eATP and should dismantle eARP^def gene expression in order to activate defense genes. In accordance with this, the putative eARP^def is downregulated by SA (Fig. 2A) and its reduction precedes PR-1 transcript accumulation (Fig. 2B). Thus, Arabidopsis has putative eARPs whose gene expression profile fits our prediction of components involved in eATP regulatory functions. Current research in our group focuses on eARP identification and finding the subset that has a role in mediating eATP signaling in defense and cell death regulation.

In conclusion, our study has revealed a novel function of eATP—as a negative regulator of defense gene expression. The modulation of eATP levels by AMP-PCP, glucose-hexokinase, SA, and some pathogens mediates defense gene expression in tobacco. We have also proposed a model predicting the expression profiles of signaling genes downstream of eATP, which feed into inhibition of cell death or suppression of defense genes. We are currently testing this model in on-going research.