Expression of Uroporphyrinogen Decarboxylase or Coproporphyrinogen Oxidase Antisense RNA in Tobacco Induces Pathogen Defense Responses Conferring Increased Resistance to Tobacco Mosaic Virus*

Transgenic tobacco plants with reduced activity of either uroporphyrinogen decarboxylase or coproporphyrinogen oxidase, two enzymes of the tetrapyrrole biosynthetic pathway, are characterized by the accumulation of photosensitizing tetrapyrrole intermediates, antioxidative responses, and necrotic leaf lesions. In this study we report on cellular responses in uroporphyrinogen decarboxylase and coproporphyrinogen oxidase antisense plants, normally associated with pathogen defense. These plants accumulate the highly fluorescent coumarin scopolin in their leaves. They also display increased pathogenesis-related protein expression and higher levels of free and conjugated salicylic acid. Upon tobacco mosaic virus inoculation, the plants with leaf lesions and high levels of PR-1 mRNA expression show reduced accumulation of virus RNA relative to wild-type controls. This result is indicative of an increased resistance to tobacco mosaic virus. We conclude that porphyrinogenesis as a result of deregulated tetrapyrrole synthesis induces a set of defense responses that resemble the

All organisms contain tetrapyrroles that play important roles in the transfer of energy, redox sensing, or catalysis. Chlorophylls are the most abundant tetrapyrroles in plants and are involved in the harvesting of light and its subsequent conversion to chemical energy. In response to varying environmental conditions the coordinated synthesis and stoichiometric assembly of the components of the photosynthetic complexes ensure an adjusted electron flow and avoid exacerbated oxidant production (1). In addition to the regulatory mechanisms, plants have evolved effective protection systems against photooxidative stress consisting of enzymic and low molecular weight antioxidants (2).
Genetic or chemical impairment of chlorophyll or heme bio-synthesis results in the accumulation of photosensitizing tetrapyrroles, e.g. by feeding of the precursor 5-aminolevulinate or by application of diphenyl ether herbicides acting on protoporphyrinogen oxidase (3). Formation of reactive oxygen species such as singlet oxygen is essential for the photodynamic action of tetrapyrroles (3). Expression of antisense or additional sense genes encoding enzymes of the tetrapyrrole metabolism leads to modified enzymic activities and, in consequence, to an imbalance of the substrate flow. Transgenic tobacco plants with reduced uroporphyrinogen decarboxylase (UROD) 1 or coproporphyrinogen oxidase (CPO) were characterized by light-dependent formation of leaf lesions and by stunted growth (4,5). Reduced activity of either of the two enzymes may increase the level of porphyrinogen substrates up to 500-fold in transformants relative to control plants. These compounds, predominantly uroporphyrinogen or coproporphyrinogen in UROD or CPO antisense plants, respectively, can be photooxidized, thus triggering photodynamic cellular destruction. Cellular stress responses observed upon accumulation of these tetrapyrroles included increased steady state levels of antioxidant mRNAs and increased activity of enzymes involved in stress defense. In particular, higher activities of superoxide dismutase in younger leaves and of ascorbate peroxidase in older leaves were observed (6). In addition to the plastidic superoxide dismutase isoform, mitochondrial and cytosolic isoforms contributed to the increased total enzyme activity suggesting that tetrapyrroles initially synthesized in the plastids may spread to other compartments where they induce local antioxidative defense responses (6). However, limitation of the detoxifying system was indicated by a decrease in the total content and a higher percentage of the oxidized form of ascorbate, reduced content of glutathione, and the loss of tocopherol (4,6). In addition to the well established roles of these protective components, phenolic constituents also contribute to the antioxidative capacity of cells (7,8). Different phenylpropanoids are formed in response to various stresses such as wounding, high light, UV radiation, or pathogen attack. These molecules often exert antimicrobial and UV protecting or signaling functions (9). We have continued to examine the protective responses to oxidative stress following the accumulation of tetrapyrroles in UROD and CPO antisense plants with the analysis of phenolic constituents and found an increased content of the coumarin scopolin which is also synthesized in tobacco upon TMV infection (10,11). We demonstrate that other pathogen defense responses are also activated indicating that these cellular reactions upon porphyrinogenesis resemble a hypersensitive reaction after pathogen attack.

EXPERIMENTAL PROCEDURES
Plant Material-Tobacco (Nicotiana tabacum var. Samsun NN) plants expressing antisense RNA for CPO (4) or UROD (5) and control plants (Samsun NN) were grown under controlled environmental conditions (24°C, 300 mol of photons m Ϫ2 s Ϫ1 , 16/8-h light/dark cycles) that favor development of necrotic leaf lesions in the transformants (5). For biochemical analysis plant material from progenies of line 2 of UROD and line 3 of CPO antisense plants was harvested as described previously (6). Additional lines were used for the TMV infection experiment (Fig. 7), and these have also been characterized previously (4,5). Necrotic leaves from individual plants with lesion formation and leaves of the same developmental stage of transformants without lesion formation as well as wild-type plants were used for inoculation. For comparative studies of necrotic and non-necrotic leaf areas of transformants, tissue was separately harvested from the same leaf, and for each single experiment, material was pooled from several leaves. The plant material used for biochemical analysis was frozen in liquid nitrogen, ground to a fine powder with a mortar and a pestle, and stored at Ϫ80°C until further processing.
Extraction and Analysis of Phenolic Constituents-Aliquots of powdered plant material were extracted repeatedly with methanol in Eppendorf tubes (final volume 1 ml). Homogenization was accomplished by a motor-driven pestle which fitted into the Eppendorf tubes. Extracts were centrifuged at 13,000 ϫ g for 10 min at 4°C.
Phenolic constituents were analyzed on an HPLC system consisting of a Waters 600 controller and pump unit and a Waters 996 photodiode array detector (Waters, Eschborn, Germany) combined with a fluorescence detector (FP-920, JASCO, Groß-Umstadt, Germany). Data acquisition and processing were performed with the Millenium software package (Waters, Eschborn, Germany). Methanolic extracts were diluted with 0.25 volume of water and centrifuged again. Aliquots of the supernatants were injected and separated on a RP-18 phase column (LiChrospher, 5 m, 125-4; Merck, Darmstadt, Germany). The mobile phase used consisted of 0.1% ammonium formate in 2% formic acid (solvent A) and 90% methanol in solvent A (solvent B) as described previously (12) but with a slightly modified gradient profile. After 5 min elution with 100% solvent A, a gradient from 0 to 40% solvent B over 25 min was applied followed by a gradient from 40% solvent B to 100% B over further 20 min. Spectra from 250 to 400 nm were recorded (1/s) at maximal spectral resolution of the photodiode array detector (1.2 nm), and chromatograms of absorbance at 280 nm were extracted from these data sets. The wavelength settings for the fluorescence detector were 300 nm for excitation and 400 nm for emission.
Chlorogenic acid was identified as the major peak in the UV chromatograms (280 nm) in methanolic extracts of wild-type leaves or roots. This peak eluted with the same retention time and displayed an identical diode array spectrum as authentic chlorogenic acid (CGA) in the HPLC analysis. For subsequent MS analysis of the major constituent, appropriate fractions were collected after HPLC and monitoring at 280 nm, evaporated to dryness, and redissolved in 50% methanol, 2% acetic acid. The purity of the preparation was confirmed by HPLC. ES-MS spectra were recorded on a Bruker Esquire ion trap system (Bruker Franzen, Bremen, Germany) equipped with an API interface (Analytica, Branford, CT). The mass spectrometer was operated in the positive ion mode with a resolution of 0.4 atomic mass units. Samples were infused at a flow rate of 80 l h Ϫ1 . A temperature of 150°C was set for the drying gas (nitrogen). Helium was used as a gas for collisioninduced dissociation. Authentic CGA and purified plant extracts with tentatively identified CGA both exhibited a peak at m/z ϭ 355 corresponding to the (M ϩ H ϩ ) ion.
For the identification of scopolin, 3 g of powdered plant material was extracted in 50 ml of methanol (30 min, 4°C). The extract was centrifuged (15 min, 6,000 ϫ g). The supernatant was evaporated to dryness, redissolved in 500 l of methanol, and adjusted to 20% methanol by addition of water. After centrifugation (15 min, 6,000 ϫ g) 200-l aliquots were injected for HPLC. To allow better separation of scopolin and chlorogenic acid, the mobile phase was as follows: 5% methanol (solvent A1) and 80% methanol (solvent B1) and a gradient from 0 to 50% B1 over 30 min. Elution was monitored by UV absorbance at 280 nm. Scopolin eluting at 19 min was collected from several injections, evaporated to dryness, and redissolved either in acetate buffer (50 mM, pH 5) for subsequent digestion with ␤-glucosidase, or in 50% methanol, 2% acetic acid for MS analysis. The purity of the scopolin preparation was confirmed by HPLC using the standard separation system described for phenolic constituents.
For ␤-glucosidase digestion, aliquots of isolated scopolin were redissolved in acetate buffer (50 mM, pH 5) and incubated at 37°C for 30 min in the presence of 1 unit of almond ␤-glucosidase (Sigma, Deisenhofen, Germany) dissolved in the same buffer. Control experiments with buffer alone were run in parallel. Samples were then transferred to ice, and 4 volumes of methanol were added. After centrifugation, aliquots of the supernatant were spiked with either scopoletin or isoscopoletin dissolved in 80% methanol or 80% methanol alone and analyzed by HPLC using the standard separation system and fluorescence detection.
The chemical synthesis of scopolin was essentially according to Ref. , and 1 mg of 1,4-diazabicyclo(2,2,2)octane in 1 ml of dry pyridine were mixed in an Eppendorf tube and stirred overnight in the dark. After centrifugation the supernatant was separated and the solvent removed under a stream of nitrogen. The residue was dissolved in 0.5 ml of trichloromethane and applied to a preparative silica gel plate purified twice before use with trichloromethane/methanol 1:1 (v/v) containing 2% (v/v) acetic acid. The plate was developed with trichloromethane/ glacial acetic acid/water, 70:28:2 (v/v/v), and air-dried. The band at R f ϭ 0.8 -0.9 ahead of the residual scopoletin was scraped off, the scopolin tetraacetate eluted with trichloromethane/methanol, 8:2 (v/v), and the solvent removed in vacuo. The crude scopolin tetraacetate was treated with 5 ml of 0.5 M barium hydroxide at 80°C for 15 min, and the reaction mixture was then cooled to room temperature and brought to pH 7 with 2 M sulfuric acid. The barium sulfate formed was precipitated by centrifugation, the supernatant separated, and the solvent removed in vacuo. The residue was chromatographed on Sephadex LH2O (Pharmacia, Freiburg, Germany) with methanol, and the scopolin characterized by HPLC combined with diode-array spectroscopy and by mass spectrometry as described under "Results." Analysis of conjugated and free SA was performed essentially according to Ref. 14, with the following minor modifications: 200 mg of powdered leaf material was used for extraction, and trichloroacetic acid was replaced by 20% formic acid. The preparations were finally resuspended in 20% methanol prior to HPLC analysis which was performed using the HPLC system described above with fluorescence detection. The mobile phase consisted of solvent A and solvent B2 (80% methanol and 10% acetonitrile in solvent A), and the column was kept at 30°C. Elution was performed at a flow rate of 1 ml min Ϫ1 with a hyperbolic gradient from 10% solvent B to 45% B in 25 min using curve 5 offered by the Millenium software. The column was then washed with 100% solvent B and equilibrated. Salicylic acid eluted at 16.4 min. Recovery rates for each individual experiment were determined by spiking parallel samples with appropriate amounts of authentic SA.
Fluorescence Microscopy-Hand-cut cross-sections of leaves were mounted in water and viewed under a Zeiss Axiovert microscope with a Fluar 20 ϫ 0.75 UV. The filter combination used was as follows: excitation BP 365, DBS 395; emission LP 397. Photos were taken using a Kodak Elite 400 slide film.
Western Blot Analysis-Protein extraction, SDS-polyacrylamide gel electrophoresis, and immunoblotting were performed as described previously (4). Protein was quantified according to Ref. 15 using bovine serum albumin as a standard. Antisera raised against tobacco PR proteins PR-1c, -2, -Q, and -S were a gift of Drs. P. Geoffroy and B. Fritig (Université Louis Pasteur, Straßbourg, France). The immunospecificity of the antisera has been described (see Ref. 16 and references therein).
TMV Inoculation and RNA Analysis-Samsun NN tobacco plants, a local lesion host for TMV, UROD, and CPO antisense plants were grown under controlled conditions favoring development of necrotic lesions (6). Ten-week-old plants were inoculated with a suspension of the Flavum strain of TMV (0.5 mg/liter) in a solution of 10 mM sodium phosphate (pH 7) containing carborundum as described previously (17). In parallel, other plants were mock-inoculated with buffer and carborundum only. At least two leaves per plant and five plants per transgenic line or wild-type were used for inoculation. Leaves were harvested directly before and 5 days after TMV inoculation, and total RNA was isolated by phenol/chloroform extraction followed by lithium chloride precipitation (18). Total RNA (7.5 g) was subjected to electrophoresis on formaldehyde/agarose gels and blotted to Hybond-N membranes (Amersham, Braunschweig, Germany) as described (19). Equal loading of samples was confirmed by including 40 g/ml ethidium bromide in the sample loading buffer allowing visualization of RNA under UV light. 32 P-Labeled probes were synthesized using the random primers DNA-labeling system (Life Technologies, Inc.). The cDNA probes for tobacco PR-1 and reverse-transcribed TMV-RNA have been previously described (20). Hybridization and washing were carried out at 65°C as described (21). Relative amounts of TMV transcripts were quantified using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) according to the manufacturer's instructions. Fig. 1 shows chromatograms obtained from methanolic extracts of leaf 9 of a UROD antisense plant. The main peak observed by UV detection at 280 nm (24.6 min) (Fig. 1A) was identified as chlorogenic acid (CGA) by co-chromatography with authentic CGA, combined with diode array and electrospray mass spectrometry (ES-MS) as described under "Experimental Procedures." In leaf extracts of transgenic plants the chromatograms recorded by using fluorescence detection showed a dominant peak (24.4 min) eluting very close to CGA (24.9 min) (Fig. 1B). This peak corresponded to the small signal observed at 24.1 min in the UV chromatogram (Fig. 1A, see arrow). The insert in Fig. 1A shows the diode array spectrum of this peak. This compound was absent in UV chromatograms in extracts of wild-type plants (data not shown), and only a small signal was found by fluorescence detection (Fig. 1C). Comparing the integrated peak areas obtained from control and transformed plants on a fresh weight basis, the content of this compound was increased (up to 23-fold) in leaves of the transformants (see Table I below). The content of other constituents remained unchanged or differed only slightly in leaves of control and transformed plants.

A Highly Fluorescent Phenolic Compound Accumulates in Leaves of UROD and CPO Antisense Plants-
Isolation and Identification of the Fluorescent Compound, Scopolin-Based on the strong fluorescence (22) and on the UV spectrum (see inset, Fig. 1A (23)), we tentatively identified the unknown substance as a coumarin. The metabolite was isolated from necrotic leaves of UROD antisense plants by HPLC as described under "Experimental Procedures." It was identified as scopolin based on HPLC analysis of a ␤-glucosidase digest, mass spectrometry, and comparison with authentic standards. The isolated compound was incubated with or without ␤-glucosidase and separated by HPLC (Fig. 2). After glucosidase digestion the fluorescent peak at 24 min completely disappeared, and one new peak of approximately the same fluorescence intensity became apparent. The retention time of this product was identical to that of scopoletin, the aglycon of scopolin, as demonstrated by cochromatography with authentic scopoletin (Fig. 2B). The product released by ␤-glucosidase also displayed the same diode array spectrum as scopoletin (data not shown). Isoscopoletin clearly differed in its elution time (28.9 min) under the same chromatographic conditions and showed a much lower ratio of peak areas from fluorescence to UV detection (data not shown). Analysis of the isolated metabolite by electrospray-mass spectrometry (ES-MS) gave a mono-isotopic mass of 354.1 identical with the value of 354.1 calculated for scopolin, C 16 H 18 O 9 . The collision-induced fragmentation of the protonated molecular ion produced one intense fragment ion at m/z 193.0 (Fig. 3) resulting from the loss of the glucose moiety. The appearance of this ion and a simultaneous reduction in intensity of the protonated molecular ion at 355.1 was observed when the temperature of the drying gas was elevated (data not shown). The chemical structure was further confirmed by chemical synthesis of scopolin from scopoletin and acetobromoglucose. The product obtained after alkaline hydrolysis of the tetra-acetyl intermediate was identical to the compound isolated from necrotic leaves of UROD antisense plants when analyzed by the methods described. In addition, both preparations gave an identical pattern of product ions when the molecular ions at m/z 193 were further fragmented (data not shown). The content of scopolin in leaves of transformants and controls was quantified using the chemically synthesized compound for calibration (Table I).
Scopolin Accumulates in Necrotic Leaf Areas of UROD and CPO Antisense Plants-The content of scopolin was quantified FIG. 1. Separation of methanolic leaf extracts by HPLC. Methanolic extracts were prepared from powdered plant material and analyzed by HPLC as described under "Experimental Procedures." UV detection with a diode array detector (250 -400 nm) was immediately followed by fluorescence detection ( EX , 300 nm; EM , 400 nm). The injected volume corresponds to 0.8 mg of leaf material. A, extracts from leaf 9 of UROD antisense plants monitored by absorbance at 280 nm. Inset shows the spectrum of the peak eluting at 24.1 min. B, same injection monitored by fluorescence emission at 400 nm following excitation at 300 nm. C, fluorescence profile of extracts from leaf 9 of wild-type tobacco. in leaf tissue with or without visible necrotic lesions from UROD (Fig. 4A) and CPO (Fig. 4B) antisense plants. Necrotic and non-necrotic leaf tissues were defined as indicated in Fig.  4C. In both transgenic lines the amount of scopolin was more than 10-fold higher in the necrotic tissue compared with nonnecrotic areas. The high content of scopolin and its predominance among the fluorescent compounds found in methanol extracts prompted us to investigate the subcellular localization of this compound in leaf tissues of transformants. Fig. 5 shows leaf sections of a necrotic (Fig. 5A) and a non-necrotic area (Fig.  5B) from the same leaf of a UROD antisense plant visualized by fluorescence microscopy. A white to bright-bluish fluorescence in the sections of necrotized leaves coincided with dead tissue. Around this tissue, a strong deep-blue fluorescence surrounded by a layer of red fluorescence was visible in cells of both the palisade and spongy parenchyma. The deep-blue fluorescence indicated the accumulation of scopolin in the vacuoles. Nonnecrotic sections of necrotized leaves from UROD antisense plants (Fig. 5B) displayed similar intensity of the red fluorescence, whereas the deep-blue fluorescence was considerably lower, consistent with the quantitative analysis of scopolin (Fig. 4). Fluorescence analysis of necrotic and non-necrotic leaf sections of CPO antisense plants gave similar results (data not shown).
Expression of Pathogenesis-related (PR) Proteins Is Strongly Induced in UROD and CPO Antisense Plants-We were next interested in determining if porphyric plants showed other defense reactions which are normally observed after pathogen attack. TMV infection of tobacco induces the expression of PR proteins including ␤-glucanases (class two of tobacco PR proteins), chitinases (e.g. PR Q), as well as other proteins of unknown function (24). Western blot analyses of PR proteins representing these different classes showed stronger accumulation of all PR proteins in older leaves of transformants relative to control plants (Fig. 6A). A similar pattern was obtained for CPO antisense plants (data not shown). We also compared the amount of PR proteins in necrotic and non-necrotic areas of the same leaf from UROD and CPO antisense plants, respectively (Fig. 6B). In both transformants PR proteins mainly accumulated in the leaf area with necrotic lesions and were found only to a minor extent in the area devoid of lesions.
UROD and CPO Antisense Plants Accumulate Elevated Levels of Free and Conjugated Salicylic Acid (SA)-Leaf lesion formation in tobacco infected with TMV is accompanied by increased levels of SA (25). We determined the levels of free and conjugated SA in leaves of UROD and CPO antisense and control plants (Table II). The contents of free SA was increased in all leaves of transformants relative to controls but was more enhanced in UROD antisense plants (up to 5-fold) compared with CPO antisense plants. The amount of conjugated SA was severalfold higher in leaves of both transformants, relative to wild-type, and the increase was more pronounced (up to 25fold) for CPO antisense plants. We next determined if SA was uniformly distributed in leaves or accumulated preferentially in tissue adjacent to the necrotic lesions as was observed after TMV infection (25). In both UROD and CPO antisense plants free SA was equally distributed between necrotic and nonnecrotic leaf areas, whereas conjugated SA preferentially accu-  2. HPLC analysis of isolated scopolin incubated with ␤-glucosidase. Scopolin was isolated from methanolic leaf extracts of UROD antisense plants as described under "Experimental Procedures." Incubations in acetate buffer (pH 5) alone or with the addition of ␤-glucosidase were performed at 37°C for 30 min. Assays were then brought to 80% methanol. HPLC analysis was performed as described in Fig. 1 (Table III). For comparison, in the TMV inoculation experiment (see Fig. 7) the amount of free and conjugated SA in Samsun NN controls 5 days after inoculation was 3.6 g/g fresh weight and 25 g/g fresh weight, respectively.
Tobacco Mosaic Virus (TMV) Infection Is Reduced in UROD and CPO Antisense Plants-It has previously been shown that TMV infection of local lesion hosts leads to a hypersensitive reaction and to increased resistance against subsequent infections with either TMV as well as other viruses (26), fungi, and bacteria (27,28). The development of the hypersensitive reaction and systemic acquired resistance (29) tightly correlates with increased PR protein expression and SA levels (17). We next investigated if the strong induction of PR proteins and SA formation would result in resistance to TMV in UROD and CPO antisense plants. Fig. 7 shows an inoculation experiment with TMV of wild-type tobacco and several transgenic lines of UROD and CPO antisense plants (4,5). Leaves were harvested before TMV inoculation to analyze PR-1 gene expression levels. As expected from the Western blot analysis (Fig. 6), necrotic UROD (2L and 15L) and CPO (3L) antisense plants had high levels of PR-1 mRNA (Fig. 7A). Wild-type controls (WT) and UROD (2,12,15) or CPO (17) antisense plants without visible leaf necrosis showed no expression of PR-1 (Fig. 7A). Five days after inoculation with TMV total RNA was extracted from leaf tissue, and the amount of virus RNA was analyzed by Northern blot using a 32 P-labeled probe derived from reverse-transcribed TMV RNA (Fig. 7B). Viral RNA accumulation was reduced in transgenic plants that were characterized by necrotic lesions and high PR-1 mRNA expression before inoculation. No significant differences in the amount of TMV RNA were detected TABLE III Comparison of SA content in necrotic and non-necrotic leaf areas of UROD and CPO antisense plants Necrotic and non-necrotic leaf areas of UROD (URODAS) and CPO (CPOAS) antisense plants were separated as described for Fig. 4. Content of free and conjugated SA was analyzed as given in Table II

Tetrapyrrole-induced Pathogen Defense Responses in Tobacco
between wild-type plants and UROD or CPO antisense plants without a necrotic phenotype. This finding is correlated with the lack of PR-1 protein expression in these plants (Fig. 7A). Quantification of hybridization signals corresponding to the two viral RNAs (Fig. 7B) revealed a 70 -90% reduction of TMV in UROD and CPO antisense plants showing necrotic lesions in comparison to wild type (Fig. 7C).

DISCUSSION
Accumulation and Localization of Scopolin-Transgenic plants with reduced activity of either UROD or CPO accumulated considerable amounts of a fluorescent compound in close proximity to necrotic leaf areas (Fig. 4). This compound was identified as scopolin based on its spectroscopic and fluorescence properties, HPLC analysis, by mass spectrometry, and by comparative analysis of the aglycone obtained by ␤-glucosidase digestion. The structure of the product was further confirmed by chemical synthesis of scopolin which gave identical results. The data excluded isoscopolin which has been detected in leafy galls of tobacco (30). Results from fluorescence microscopy suggested that scopolin was localized in the vacuoles (Fig. 5). Scopoletin and scopolin accumulation has previously been demonstrated in cell suspension cultures of tobacco (31). Although scopoletin has mainly been recovered from the culture filtrate, scopolin accumulated within the cells (31). Vacuolar localization is consistent with the rapid uptake of the coumarin glucoside esculin into isolated barley leaf vacuoles (32).
There is substantial evidence that scopolin and scopoletin play an important role in disease resistance. It has been shown that scopoletin possesses antimicrobial activity (34 -37). In incompatible plant-pathogen interactions a rapid and pronounced synthesis of scopoletin was observed, and a slower and reduced formation was found in compatible interactions (36 -38). Elevated constitutive levels of scopoletin and scopolin were found in a disease-resistant Nicotiana hybrid (35). Spraying of tobacco plants with scopoletin but not with scopolin prior to TMV inoculation reduced lesion formation (35). In potato tubers infected with Phytophthora infestans a ring of blue fluorescence derived from accumulated scopolin was observed around the site of infection (39). Scopolin also accumulated during the hypersensitive reaction of TMV-infected local lesion host tobacco varieties (10,11). Its low toxicity compared with the aglycon scopoletin and its putative vacuolar localization classifies scopolin as a preformed antimicrobial compound (40). Upon tissue disruption following pathogen attack, scopoletin is released by the action of ␤-glucosidases originating from the plant itself or from the attacking microorganism and exerts its inhibitory effects. In analogy, the action of a ␤-glucosidase is required to exert the antimicrobial effects of the cyanogenic glucoside dhurrin or the isoflavanoid derivative maackiain, which is stored as a glucoside in roots of red clover (41,42).
Lesion Formation in Leaves of UROD and CPO Antisense Plants Resembles the Hypersensitive Reaction Observed after TMV Infection-The strong accumulation of scopolin in leaves of UROD and CPO antisense plants suggests similarities between the formation of necrotic lesions upon accumulation of photosensitizing porphyrins and the hypersensitive reaction following TMV infection. Tobacco leaves infected with TMV similarly show substantial increase in the levels of both free and conjugated SA (25) and have stimulated PR protein expression (17, 43) with a gradual distribution and a maximum near the center of the lesions (25,16). It is widely accepted that in infected plants concentration of SA and induction of PR proteins correlate with enhanced resistance against pathogens (29, 44 -47). It has been shown that treatment of tobacco with SA or analogs of SA, such as 2,6-dichloroisonicotinic acid or benzo(1,2,3)-thiadiazole-7-carbothioic acid (S)-methyl ester, induce PR gene expression and activate resistance against TMV as well as bacterial and fungal pathogens (17,20). TMV inoculation leads to local leaf lesions in tobacco cultivars bearing the N gene limiting the spread of the virus (48). In UROD and CPO antisense plants of the cultivar Samsun NN, the activation of defense responses, e.g. the accumulation of scopolin (Table I), the local induction of PR proteins (Fig. 6), and increased content of free and bound SA (Tables II and III) clearly correlated with a reduction in TMV infection (Fig. 7, B and C). This result obtained with the transgenic local lesion host is in analogy to the induced TMV resistance in the same cultivar by treatment with ethrel (49). Future experiments will reveal if accumulation of tetrapyrroles in a susceptible tobacco variety lacking the N gene contributes a systemic pathogen resistance in uninoculated tissue.
Deregulation of Chlorophyll Biosynthesis Can Initiate Defense Responses Against Pathogen Attack-Cellular responses in plants similar to those observed after pathogen attack can be provoked by a wide range of other stress situations in plants. UV light, heavy metal, or ozone treatment, or genetic modification of metabolic pathways can lead to the induction of PR protein expression (50 -53). In addition, increased resistance to some pathogens has been observed in transgenic plants expressing cholera toxin (54), yeast invertase (55), bacterio-opsin (56) as well as in response to reduced catalase activity (57,58). Our results are consistent with these and other examples of altered cell death control by transgene expression (53,59).
Free porphyrins are potent photosensitizers and have been used in photodynamic therapy for the treatment of cancer (60). They are also involved in the mode of action of diphenyl ether type herbicides (e.g. acifluorfen), which act on protoporphyrinogen oxidase (61). The phototoxicity of tetrapyrroles is exerted via reactive oxygen species. Singlet oxygen is generated from molecular oxygen after UV irradiation of coproporphyrin (62). By using the electron spin resonance technique a range of free radicals is observed when porphyrin solutions are illuminated (63). Reduced activities of UROD or CPO led to the accumulation of tetrapyrroles (4,5). The limited capacity of the antioxidative defense mechanisms in UROD and CPO was indicated by decreased levels of low molecular weight antioxidants (6). Reactive oxygen species derived from excessive tetrapyrroles could mimic signaling molecules formed after pathogen attack in triggering defense responses (64 -66). However, other cellular processes resulting from tetrapyrrole-induced oxidative stress could contribute to the induction of antimicrobial defense responses.
Evidence was recently provided that lesion formation after TMV inoculation of tobacco can be attributed to programmed cell death (67). Programmed cell death was inhibited in plants grown under low oxygen pressure suggesting that ambient oxygen levels are required for this response (68). Based on the similarity of antimicrobial defense responses triggered in UROD or CPO antisense plants and TMV-infected tobacco, one could speculate that lesion formation in the porphyric transformants is exerted via programmed cell death (59,68). In a model system for the photodynamic therapy of cancer, the application of a mixture of porphyrins could provoke either apoptotic or necrotic cell death depending on the incubation protocol (69). It remains a future challenge to prove concepts of cell death caused by accumulated tetrapyrroles in plants.