Table Of ContentsNext Page

Genomic approaches to understanding allelochemical modes of action and defenses against allelochemicals

Stephen O. Duke1, Scott R. Baerson1, Zhiqiang Pan1, Isabelle A. Kagan1, Adela Sánchez-Moreiras2, Manuel J. Reigosa2, Nuria Pedrol-Bonjoch3 and Margot Schulz4

1P. O. Box 8048, Natural Products Utilization Research, ARS, USDA, University, MS 38677, USA; http://www.olemiss.edu/depts/ncnpr/usda/; sduke@olemiss.edu
2
Laboratorio de Ecofisioloxia Vexetal, Facultade de Ciencias, Universidade de Vigo, Vigo, Spain; Email mreigosa@uvigo.es
3
Area de Nutrición, Pastos y Forrajes, SERIDA, Estación Experimental "La Mata", Grado, Spain; npedrol@serida.org
4
Institut für Molekulare Physiologie und Biotechnologie der Pflanzen, Universität Bonn, Biozentrum Karlrobert Kreiten Str. 13, 53115 Bonn, Germany; ulp509@uni-bonn.de

Abstract

Little is known concerning the mode of action of allelochemicals or plant defense responses mounted against them. Theoretically, changes in the expression of genes encoding the primary target or other proteins in the same pathway should occur soon after phytotoxin exposure. Defense responses, such as the induction of genes involved in chemical detoxification, may occur later, depending on the nature of the chemosensors which presumably exist in plant cells. We first used yeast (Saccharomyces cerevisiae) to test the concept of transcriptome profiling of toxicant modes of action. Characteristic gene induction profiles related to specific molecular target sites were verified with several fungicides. A battery of xenobiotic defense-associated genes were found to be dramatically induced in Arabidopsis following exposure to an array of structurallyunrelated xenobiotics, including a herbicide, an allelochemical, and herbicide safeners. These genes are unlikely to be strongly linked to the mode of action of a specific phytotoxin, but rather constitute a coordinately-controlled xenobiotic defense gene network. Transcriptional profiling experiments using microarrays are being conducted to examine the effects of various herbicides and natural phytotoxins on the Arabidopsis transcriptome.

Media summary

Using genomic methods to determine how allelochemicals work and how plants protect themselves from them is discussed.

Key Words

mode of action, BOA, microarray, fungicide, herbcide

Introduction

Advances in molecular biology have provided powerful new tools that can be used to understand complex processes in allelopathy. In this paper, we provide examples of how whole genome transcriptome analyses with DNA microarrays might be used to provide clues about the mode of action of allelochemicals, as well as the mechanisms of defense against allelochemicals and the biosynthesis of allelochemicals. We will summarize our research in this area and discuss it in the context of relevant research.

Transcriptome profiling of the mode of action of allelochemicals

Detection of the global expression response of plant genomes after treatment with phytotoxins is possible with DNA microarrays. Theoretically, at a given dose of a phytotoxin, at a specific time after exposure to the phytotoxin, one might expect changes in the transcriptome that would be specific for phytotoxins with the same molecular target site (Eckes et al. 2004, Duke et al. 2005). Thus, one can generate a library of transcriptome profiles for phytotoxins with different molecular target sites that would be useful in the determination of the molecular targets of phytotoxins with unknown sites of action. This approach has been used in pharmaceutical research.

For example, Boshoff et al. (2004) generated 430 transcription microarray profiles of inhibitors of Mycobacterium tuberculosis. The profiles of compounds with known modes of action were useful in determination of several compounds with unknown modes of action. Analysis of these data revealed 150 clusters of coordinately regulated genes, and a signature subset of these gene clusters was sufficient to classify all known agents as to mechanism of action of anti-tuberculosis drugs. Transcriptional profiles generated by a crude marine natural product generated the same prediction of a mode of action as the pure active component.

There are several potential problems with this approach. Any toxicant that kills an organism will potentially affect a huge number of genes at some doses and times after exposure. Consequently, determining the proper dose and time after treatment for best seeing effects on transcription of genes directly related to the molecular target site is important, and not a simple endeavor. Toxicants also induce genes associated with stress and protection from xenobiotics. Sometimes these effects can be very dramatic compared to effects on genes associated with the molecular target site. Many target sites are associated with genes that are normally well expressed, encoding gene products necessary for fundamental cellular functions. Thus, there may not much latitude for dramatic changes in expression of these genes. Sifting through the huge amount of data generated from microarray experiments to find effects on genes related to the target site of the toxicant can be challenging.

We began our studies of mode of action with microarrays by examining the effects of agricultural fungicides on yeast (Saccharomyces cerevisiae), using whole genome cDNA chips (Kagan et al. 2005). There are several advantages of S. cerevisiae over plants for this type of study. First, there is only one cell type, so effects of a toxin on tissue- or cell-type specific genes are not diluted by lack of effect on these genes in other tissues or cell types. Second, all cells can be treated rapidly and uniformly with the toxicant, unlike the situation with whole plants. Third, the number of genes in the yeast genome is significantly smaller than in that of Arabidopsis thaliana, a plant with a very small genome compared to most other plants. Lastly, the functions of yeast genes are better annotated in S. cerevisiae than those of any higher plant, making it more likely that effects of toxins on this organism’s transcriptome can be more readily understood.

Our strategy has been to find reproducible effects on specific genes or groups of genes that can be linked to a molecular target site. We tested the effects of eight fungicide inhibitors of ergosterol synthesis, representing three classes of these inhibitors targeting three different target sites of the pathway (Figure 1). A putative inhibitor of methionine synthesis, cyprodinil, was also examined. Characteristic changes in gene transcription for the genes of the ergosterol pathway were seen for Class I and Class II inhibitors.

This pattern was not found for the Class III ergosterol inhibitor, nor was it found with cyprodinil. Cyprodinil causes upregulation of three genes involved in methionine metabolism, and there were essentially no effects of ergosterol inhibitors on methionine synthesis genes. From these results, the effects of the Class III ergosterol inhibitor on this pathway in S. cerevisiae are questionable. Using oligonucleotide microarrays, we have generated unpublished results that link fungicides with known modes of action to genes related to their target sites. Although we still have much work to do with yeast, the concept of transcriptome profiling appears to be valid for fungicides that inhibit growth of yeast.

Similarly, Gutteridge et al. (2005), working with S. cerevisiae, in a search of the mode of action of a potential agricultural fungicide, found that with some compounds specific gene clusters were affected in ways that provided clues to their mode of action. Agarwal et al. (2003) found that S. cerevisiae responded to pharmaceutical fungicides with several different molecular target sites with drug-specific effects on gene transcription.

Companies involved in herbicide discovery apparently have extensive transcriptome profile libraries for herbicides with different modes of action, although no details of their results have been published. A very few publications exist on transcriptome profiles for individual phytoxins (2,4-D, Raghavan et al. 2005; thaxtomin A, Scheible et al. 2004; and flufenacet, Lechelt-Kunze et al. 2003). There is one report of DNA microarray methods leading to the discovery of the mode of action of an allelochemical. Bais et al. (2004) reported that (–)-catechin is phytotoxic to Arabidopsis due to an effect on a calcium ion signaling cascade. This result has not yet been confirmed by other laboratories.

Figure 1. Effects of Class I (A), II ( B) and III (C) ergosterol biosynthesis inhibitors, and a putative methionine biosynthesis inhibitor (D) applied at the I50 concentrations for 2 h on expression levels of genes in the ergosterol pathway. Standard errors are shown in A and B, and standard deviations are shown in C and D. Genes are listed on the x-axis from left to right in the order in which they appear in the pathway. The transcription relative to untreated controls is shown on the y-axis. Dashed horizontal lines on the graphs indicate the level of expression at which no change is seen relative to the control. Arrows indicate gene(s) targeted by the inhibitor. (from Kagan et al., 2005)

Our laboratory has initiated work on establishing a data base of phytotoxin-related transcriptome profiles. Our first experiment in this endeavor was to examine the effect of the allelochemical benzoxazolin-2(3H)-one (BOA) on gene expression in Arabidopsis. Careful dose response experiments (Figure 2) allowed us to determine I50 and I80 concentations for root growth inhibition. Then, plants were grown for about 10 days and exposed to these concentrations of BOA, after which mRNA was extracted and analyzed with Affymetrix Arabidopsis ATH1 Genome Arrays (Baerson et al. 2005). One hundred-fifty eight genes were significantly induced, and 30 were repressed in both the I50 and I80 treatments, totaling approximately 0.8% of all genes represented on the ATH1 gene chip. The breakdown of categories of genes that were affected is provided in Figure 3. Unfortunately, the exact mode of action of BOA is unknown, so we did not know what genes to focus on regarding the mode of action. We will need a more complete library of transcriptome responses to phytotoxins with different modes of action in order to better interpret the results of this experiment. Nevertheless, our results with BOA revealed a considerable amount of information about the responses of Arabidopsis to an allelochemical, in terms of how the plant protects itself from such a chemical threat.

Figure 2. Dose-response experiments on effects of BOA on root length of Arabidopsis. Each data point represents mean root length from two independent replicates + 1 SD. (Adapted from Baerson et al. 2005)

Figure 3. Distribution of BOA-induced genes into functional categories.

Using transcription responses to understand plant/plant interactions

Defenses against allelochemicals

Non-phytotoxic compounds can also induce genes that provide defences against phytotoxins. This is the principle of crop safeners that are used to protect crops from herbicide injury. Microarray technology has been used to probe the mechanism of action of safeners. Using a cDNA microarray, genes of hybrid poplar ((Populus nigra x Populus maximowiczii), Rishi et al. (2004) found differentially transcribed genes in response to a safener. Genes encoding enzymes involved in oxidation, conjugation, and sequestration of xenobiotics were found to be upregulated by the safener. Little has been done with molecular biology to determine how plants protect themselves from allelochemicals. Matvienko et al. (2004) found genes encoding quinone oxidoreductases to be upregulated by treatment of plants with allelopathic 2,6-dimethoxybenzoquinone, indicating that this enzyme is involved in detoxification of the compound. In earlier work examining the upregulated genes in response to this quinone, several genes encoding enzymes predicted to detoxify the quinone were found to be upregulated (Matvienko et al. 2001).

Our work with BOA effects on the Arabidopsis transcriptome was even more informative (Baerson et al. 2005). As shown in Figure 3, genes encoding proteins related to cellular defense were the second largest category of genes induced by BOA. Table 1 lists some of the most affected genes in this category.

The effect of BOA on transcription of these genes was more quantitatively determined with quantitative real time RT-PCR (Figure 4). In most cases, the level of induction was similar, however, in a few cases the microarray method underestimated the level of upregulation. We were amazed that the massive detoxification response of this plant to this allelochemical. This led us to try to determine whether any of the metabolic detoxification products known to occur in plants (Figure 5) were present. We found three of these metabolites in BOA-treated Arabidopsis (Table 2), indicating that at least three detoxification enzymes that were induced helped to detoxify this phytotoxin. The primary metabolites were BOA-6-OH and its glucoside, which most likely requires a cytochrome P450 and a UDP glucosyltransferase to be produced. Several representatives of each of these enzymes were upregulated in response to BOA exposure. Further experiments, using quantitative real time RT-PCR, revealed that most of the genes of Table 1 and Figure 4 are induced by a wide range of xenobiotics, including 2,4-D, two herbicide safeners, and phenobarbitol (Baerson et al. 2005). Our results suggest that allelochemicals induce a wide range of genes involved in detoxification of potential phytotoxins.

Table 1. Selected genes of the Arabidopsis associated with detoxification of xenobiotic that were most highly induced by BOA as measured by microarray. (adapted from Baerson et al. 2005)

Locus ID

Gene description

fold increase

   

I50

I80

At1g15520

ABC transporter

2.1

10.8

At3g04000

short-chain type dehydrogenase/reductase-related

5.1

15.0

At1g17170

glutathione transferase, putative

6.9

18.1

At5g13750

MFS antiporter

2.5

3.9

At5g16980

quinone oxidoreductase, putative

3.2

9.4

At4g20860

FAD-linked oxidoreductase family

8.4

9.3

At2g19190

light repressible receptor protein kinase, putative

8.4

7.6

At5g27420

ING-H2 zinc finger protein-related

4.6

6.6

At4g34135

glucosyltransferase-related protein

9.4

18.3

At1g05560

UDP-glucose transferase (UGT1)

6.4

14.6

At2g15480

glucosyltransferase-related protein

5.2

12.0

At2g15490

glucosyltransferase-related protein

3.3

10.4

At5g39050

malonyl transferase

3.0

5.5

At4g12490

PEARL1 related

30.1

26.6

At4g12500

PEARL1 related

16.5

10.5

Figure 4. Sixteen representative genes identified as differentially expressed by microarray analysis (Table 1) were analyzed by quantitative real-time RT-PCR using gene-specific primer pairs. A, Transcriptional profiling results. Shown are relative gene expression values obtained from microarray experiments with BOA-treated Arabidopsis seedlings. The data represent selected genes up-regulated in both I50 (open bars) and I80 (closed bars) BOA treatments. B, Quantitative real-time RT-PCR results. The RNA samples used were identical to those used for transcriptional profiling results shown in panel A. Data were normalized to an internal 18S ribosomal RNA control. Data are shown as means ± 1 SD. (Adapted from Baerson et al. 2005).

Elicitation of allelochemical defenses

Use of whole-genome microarrays should be useful in the future for elucidating the genetics and enzymology of allelochemical biosynthesis. Kong et al. (2004) recently found that weeds can induce biosynthesis of allochemicals in allelopathic rice, in much the same way that pathogens induce phytoalexin production by plants. One of the genes for an enzyme involved in biosynthesis of momilactone B, a rice allelochemical and phytoalexin (Kato-Noguchi and Ino 2004), was induced in rice leaves (Xu et al. 2004). The whole-genome rice microarray was not used in this study. Since high quality, whole-genome microarrays are available for rice, this technology could be very useful in identifying the entire biochemical pathway for all of the induced alleleochemcals of rice. Furthermore, this approach could provide valuable information on how these pathways are regulated.

Figure 5. Two metabolic detoxification of BOA schemes known to occur in plants.

Table 2. Metabolites of BOA were quantified by HPLC for three-week-old plants exposed to 10, 100, 250, and 500 μM concentrations of BOA for a period of 24 h. A minimum of 30 plants were used per treatment. Each datum represents the mean from three replicates ± S.D. (Adapted from Baerson et al., 2005)

 

Metabolite (nmol/g fresh weight)

[BOA]

BOA-6-OH

BOA-6-O-glucoside

Glucoside carbamate

10 μM

20.3 ± 4.0

23.8 ± 5.3

n.d.

100 μM

66.0 ± 10.0

121.5 ± 41.8

33.5 ± 32.8

250 μM

174.8 ± 58.1

249.3 ± 90.0

38.7 ± 20.0

500 μM

212.3 ± 35.6

381.3 ± 147.7

54.8 ± 10.0

Summary

Allelopathy is one of the last areas of plant science to employ molecular biology as a tool in understanding the phenomena. We and others have provided a tiny glimpse of what might be done with the powerful technique of transcriptional profiling using whole genome microarrays. In the future, we hope to generate a much more complete transcriptional profile for phytotoxins with different modes of action. As such a data base becomes more complete, its use in providing clues to the modes of action of allelochemicals will become more robust. We also intend to use this technology to examine biosynthesis of allelochemicals and to further study plant defenses against allelochemicals.

References

Agarwal AK, Baerson SR, Rogers PD, Jacob MR, Barker KS, Cleary JD, Walker LA, Nagle DG and Clark AM (2003) Genome-wide expression profiling of the response to polyene,pyrimidine, azole, and echinocandin antifungal agents in Saccharomyces cervisiae. Journal of Biological Chemistry 278, 34998-35015.

Baerson SR, Sánchez-Moreiras A, Pedrol-Bonjoch N, Schulz M, Kagan IA, Agarwal AK, Reigosa MJ and Duke SO (2005) Detoxification and transcriptome reponse in Arabidopsis seedlings exposed to the allelochemical benzoxazolin-2(3H)-one (BOA). Journal of Biological Chemistry (In press).

Bais HP, Vepachedu R, Gilroy S, Callaway RM and Vivanco JM (2003) Allelopathy and exotic plant invasion: From molecules and genes to species interactions. Science 301, 1377-1380.

Boshoff HIM, Myers TG, Copp BR, McNeil MR, Wilson BA and Barry CE III (2004) The transcriptional responses of Mycobacterium tuberculosis to inhibitors of metabolism: Novel insights into drug mechanisms of action. Journal of Biological Chemistry 279, 40174-40184.

Duke SO, Dayan FE, Kagan IA and Baerson SR (2005) New herbicide target sites from natural compounds. American Chemical Society Symposium Series 892, 132-141.

Eckes P, van Almsick C and Weidler M (2004) Gene expression profiling, a revolutionary tool in Bayer CropScience herbicide discovery. Pflanzenschutz-Nachrichten Bayer 57, 62-77.

Gutteridge S, Pember, SO, Wu L, Tao Y and Walker M (2005) Fungal site of action determination: Integration with high-volume screening and lead progression. American Chemical Society Symposium Series 892, 132-141.

Kagan IA, Michel A, Prause A, Scheffler BE, Pace P and Duke SO (2005) Gene transcription profiles of Saccharomyces cerevisiae after treatment with plant protection fungicides that inhibit ergosterol biosynthesis. Pesticide Biochemistry and Physiology 82, 133-153.

Kong C, Xu X, Zhou B, Hu F, Zhang C and Zhang M (2004) Two compounds from allelopathic rice accession and their inhibitory activity on weeds and fungal pathogens. Phytochemistry 65, 1123-1128.

Lechelt-Kunze C, Meissner RC, Drewes M and Tietjen K (2003) Flufenacet herbicide treatment phenocopies the fiddlehead mutant in Arabidopsis thaliana. Pest Management Science 59, 847-856.

Matvienko M, Torres MJ and Yoder JI (2001) Transcriptional responses in the hemiparasitic plant Triphysaria versicolor to host plant signals. Plant Physiology 127, 272-282.

Matvienko M, Wojtowicz A, Wrobel R, Jamison D, Goldwasser Y and Yoder JI (2004) Quinone oxidoreductase message levels are differentially regulated in parasitic and non-parasitic plants exposed to allelopathic quinones. Plant Journal 25, 375-387.

Raghavan C, Ong EK, Dalling MJ and Stevenson TW (2005) Effect of herbicidal application of 2,4-dichlorophenoxyacetic acid in Arabidopsis. Functional and Integrative Genomics 5, 4-17.

Rishi AS, Munir S, Kapur V, Nelson ND and Goyal A (2004) Identification and analysis of safener-inducible expressed sequence tags in Populus using a cDNA microarray. Planta 220, 296-306.

Scheible W-R, Fry B, Kochevenko A, Schindelasch D, Zimmerli L, Somerville S, Loria R and Somerville CR (2003) An Arabidopsis mutant resistant to thaxtomin A, a cellulose synthesis inhibitor from Streptomyces species. Plant Cell 15, 1781-1794.

Xu M, Hillwig ML, Prisic S, Coates RM and Peters RJ (2004) Functional identification of rice syn-copalyl diphosphate synthase and its role in initiating biosynthesis of diterpenoid phytoalexin/allelopathic natural products. Plant Journal 39, 309-318.

Top Of PageNext Page