Ins and Outs of Programmed Cell Death and Toxin Action

Programmed cell death is a primary characteristic of plant response to both incompatible and compatible plant–pathogen interactions (reviewed in Greenberg and Yao, 2004). Plant disease resistance is often manifested by a localized cell death response (hypersensitive response) that prevents the spread and establishment of the pathogen and depends on interaction between the products of plant resistance and pathogen avirulence genes (classical gene-for-gene resistance). A number of bacterial and fungal plant pathogens also produce toxins that act as virulence factors and elicit a programmed cell death response contributing to the development of disease. Some pathogens produce host-selective toxins that are the principal determinants both of host range and of the progression of disease in sensitive plant species or varieties.

In this issue of The Plant Cell, three articles report significant new information on the activity of two different toxins, Ptr ToxA, a proteinaceous pathogenicity factor secreted by Pyrenophora tritici-repentis, the causal agent of tan spot in wheat (Manning and Ciuffetti, pages 3203–3212; Sarma et al., pages 3190–3202), and fumonisin B1 (FB1), a sphingoid-like elicitor of cell death produced by the maize pathogen Fusarium moniliforme (Chivasa et al., pages 3019–3034). Although little is known about their targets and the nature of their activity in plant tissues, these toxins likely induce cell death via distinct modes of action. Both have been utilized as the basis of model systems to investigate cell death in pathogen-free conditions, which is advantageous as it avoids both the confounding effects of pathogen growth in experimental media and tissues and the necessity of maintaining and propagating pathogenic organisms.

Manning and Ciuffetti show that the ToxA protein must cross the plasma membrane and be internalized to produce a toxic effect, and resistant wheat cultivars avoid toxin internalization, suggesting that resistance is related to protein import. This is significant because it shows that the presence of the pathogen is not necessary for internalization and suggests that the host plant must have a toxin receptor and mechanism for transporting the protein into cells. Sarma et al. present a high-resolution crystal structure of ToxA, the topology of which supports the idea that the toxin binds to an integrin-like receptor to gain entry to the host plant.

Chivasa et al. show that FB1 exerts a major effect outside of cells, and FB1-induced cell death is associated with a depletion of extracellular ATP. This work is notable in that it shows that extracellular ATP is essential for maintaining plant cell viability and that FB1-induced cell death is mediated by depletion of extracellular ATP. In animals, extracellular ATP is known to be an essential signaling molecule involved in neutotransmission, immune responses, cell growth, and other processes (reviewed in Gordon, 1986). A few studies have suggested that extracellular ATP may have physiological significance in plants (e.g., Demidchik et al., 2003; Jeter et al., 2004), but it has not been a subject of intensive investigation among plant biologists.

ToxA SENSITIVITY AND PROTEIN IMPORT

Manning and Ciuffetti sought to determine the site of ToxA action in sensitive wheat cultivars. First, they found that ToxA was protected from treatment with proteinase K, an extracellular protease, in leaves of sensitive but not insensitive cultivars. Intact leaves of sensitive and insensitive cultivars were treated with a solution of heterologous His-tagged ToxA followed by treatment with proteinase K and ToxA retrieved from whole-cell lysates of treated leaves and detected with anti-ToxA antibody. ToxA was present in the lysates from sensitive cultivars, but almost undetectable in lysates from insensitive cultivars, indicating that the toxin was protected from the extracellular protease in sensitive but not insensitive plants. They next showed that ToxA was internalized to the cytoplasm and chloroplasts of sensitive wheat leaves, using immunolocalization in leaves treated with purified ToxA and fluorescence detection in leaves treated with green fluorescent protein-ToxA fusion protein purified from Escherichia coli. The results further indicated that ToxA enters the cells of sensitive leaves without disrupting the plasma membrane or cell integrity. It was also demonstrated that ToxA-induced cell death is light dependent, suggesting either that active photosynthesis or light activation of some component is necessary for toxin activity. Finally, the authors show that internalization of ToxA determines susceptibility in wheat because transient expression of ToxA in insensitive cultivars was sufficient to cause a cell death response.

Taken together, the results show that (1) light-dependent internalization of ToxA is necessary and sufficient for toxicity and cell death, (2) the toxin crosses the plasma membrane without noticeable disruption, and (3) sensitivity to ToxA in wheat is based on the ability to internalize the toxin. Sensitivity of wheat cultivars to ToxA is dependent on the Tsn1 locus (Anderson et al., 1999). The work of Manning and Ciuffetti suggests that Tsn1 function may be related to ToxA internalization (i.e., perception at the cell surface and/or transport across the plasma membrane).

In a companion article, Sarma et al. present the high-resolution crystal structure of ToxA in two different crystal forms. P. tritici-repentis ToxA encodes a pre-pro-protein: the pre-region (residues 1 to 22) is a signal peptide that targets the protein to the secretory pathway, and the pro-region (the N-domain residues 23 to 60) is cleaved prior to secretion of the mature 13.2-kD toxin (C-domain residues 61 to 178). Tuori et al. (2000) showed that the pro-region likely is required for proper folding of the protein, including formation of a critical disulfide bond that stabilizes the mature active toxin. Database searches have not yielded any potential homologs of ToxA. Protein motif searches showed potential sites for phosphorylation and myristoylation and an RGD-containing motif that may have functional importance (Tuori et al., 2000; Meinhardt et al., 2002; Manning et al., 2004). In animals, interaction of RGD-containing proteins, including certain viral proteins, with integrin receptors leads to internalization. The structure of the RGD-containing loop in ToxA adds support for the notion that integrin-like receptors exist in plants, although they have not been described.

FB1 ACTION AND A REQUIREMENT FOR EXTRACELLULAR ATP IN PLANTS

Chivasa et al. conducted experiments with radiolabeled ATP and cell-impermeant ATP traps to show that Arabidopsis thaliana cells tightly regulate extracellular ATP levels and that maintenance of extracellular ATP concentrations is essential for cell viability. Depletion of extracellular ATP was found to induce cell death in both cell cultures and intact plant tissues in a number of plant species, suggesting that extracellular ATP plays a vital physiological role in plants. The authors then show that treatment of Arabidopsis cell cultures with FB1 toxin results in rapid depletion of extracellular ATP and that exogenous application of ATP significantly attenuated the FB1-induced cell death response. They further showed that intact ATP was necessary for this effect, as ADP, AMP, or inorganic phosphate had no effect.

FB1 is a small molecule with a sphingoid-like long chain base structure, similar to the AAL-toxin produced by the tomato pathogen Alternaria alternata. FB1 is also selectively toxic to AAL-sensitive tomato varieties (Gilchrist et al., 1992). FB1 is toxic to animal cells, which has been studied because animals can be affected by consumption of grain infected with F. moniliforme. In animals, both FB1 and AAL-toxin are known to inhibit sphingolipid biosynthesis by competitive inhibition of ceramide synthase, which can lead to cell death. A number of observations support the hypothesis that AAL-toxin also inhibits ceramide synthase in plants. First, the Alternaria stem canker (Asc1) gene, which confers resistance to AAL-toxin in tomato, may be related to sphingoid metabolism (Brandwagt et al., 2000). Asc1 is a homolog of Saccharomyces cerevisiae LAG1, which has been associated with yeast life span. LAG1 functions to facilitate transport of GPI-anchored proteins from the endoplasmic reticulum to the Golgi apparatus. Sphingolipids and GPI-anchored proteins are both major components of lipid rafts, which are involved in membrane trafficking and endocytosis. Brandwagt et al. (2000) proposed that Asc1 has a similar function to LAG1 and prevents cell death by restoring ER-to-Golgi transport of GPI-anchored proteins. Secondly, Abbas et al. (1994) showed that both FB1 and AAL-toxin cause an accumulation of free sphingoid bases in duckweed, tomato plants, and tobacco callus.

FB1 induces cell death reminiscent of the hypersensitive response in Arabidopsis. FB1 has been used in an Arabidopsis protoplast system to investigate plant defense–related cell death signaling events (Asai et al., 2000), and a number of FB1 resistant (fbr) mutants have been isolated (Stone et al., 2000). There is some suspicion that FB1 may act in a different manner in Arabidopsis relative to other plant species (S. Chivasa and A. Slabas, personal communication). However, the precise targets of FB1 in Arabidopsis and other plants remain unknown. Chivasa et al. now show that FB1 mediates depletion of extracellular ATP in Arabidopsis. It would be of interest to know if this effect extends to other plant species and whether or not the effect is related to sphingoid metabolism. Liang et al. (2003) identified a ceramide kinase in Arabidopsis from the mutant accelerated cell death 5, the characterization of which supports a role for ceramide phosphorylation in regulating cell death in plants. It is possible that ceramide phosphorylation is related in some manner to the maintenance of extracellular ATP. Chivasa et al. suggest that depletion of extracellular ATP may be a general feature of hypersensitive cell death. Thus, it may be of interest to investigate extracellular ATP responses under other conditions associated with the hypersensitive response.

Chivasa et al. also found FB1 sensitivity to be light dependent, as reported previously (Asai et al., 2000). Asai et al. (2000) suggested that reactive oxygen species produced during photosynthesis and/or light-dependent activation of PAL activity might be related to FB1 activity, the latter because PAL is a key enzyme in salicylic acid biosythesis, which was found to be required for FB1-induced cell death. It is interesting to note that Manning and Ciuffetti also found ToxA sensitivity to be light dependent. This could be due to a general requirement for light-induced reactive oxygen species in cell death responses or to completely unrelated light-dependent requirements in the responses to these distinct toxins.

These three articles show that controlling cell death depends on complex processes taking place both inside cells and in the extracellular milieu. In addition, the work opens up some relatively unexplored areas in plant biology: the existence and importance of integrin-like receptors and the function of extracellular ATP in controlling hypersensitive cell death responses and other processes.