Cellular toxicity of elsinochrome phytotoxins produced by the pathogenic fungus, Elsinoë fawcettii causing citrus scab


  • Hui-Ling Liao,

    1. Citrus Research and Education Center, and Department of Plant Pathology, Institute of Food and Agricultural Sciences (IFAS), University of Florida, 700 Experiment Station Road, Lake Alfred, Florida 33850, USA
    Search for more papers by this author
  • Kuang-Ren Chung

    1. Citrus Research and Education Center, and Department of Plant Pathology, Institute of Food and Agricultural Sciences (IFAS), University of Florida, 700 Experiment Station Road, Lake Alfred, Florida 33850, USA
    Search for more papers by this author

Author for correspondence:
Kuang-Ren Chung
Tel: +1 863 956 1151 ext. 1369
Fax: +1 863 956 4631
E-mail: krchung@ufl.edu


  • • Elsinochromes are the red/orange pigments produced by many Elsinoë fungal species and are structurally similar to the phytotoxin, cercosporin. Here, pigments were extracted from cultures of a citrus pathogen, Elsinoë fawcettii and tested for cellular toxicity.
  • • On irradiation with light, elsinochromes rapidly killed suspension cultured citrus and tobacco cells. The toxicity was decreased by adding the singlet oxygen (1O2) quenchers (bixin (carotenoid carboxylic acid), DABCO (1, 4-diazabicyco octane), ascorbate or reduced glutathione). Application of elsinochromes onto rough lemon leaves resulted in necrotic lesions, whereas lesion development was inhibited by the addition of bixin, DABCO or ascorbate, but not a-tocopherol. Incubation of rough lemon leaf discs with elsinochromes in the light induced a steady increase of electrolyte leakage.
  • • Compared with two photosensitizing compounds, hematoporphyrin and cercosporin, the accumulation of 1O2 induced by elsinochromes after irradiation was indicated by successful detection of the cholesterol oxidation product, 5a-hydroperoxide. Addition of a potent quencher, b-carotene prevented 5a-hydroperoxide production. Elsinochromes generated superoxide ions (inline image), whereas accumulation of inline image was blocked by addition of the superoxide dismutase, a scavenger of inline image, but not the 1O2-quencher, DABCO.
  •  Our study indicated that elsinochromes are functioning as photosensitizing compounds that produce 1O2 and inline image, and exert toxicity to plant cells.


Elsinochrome pigments containing a phenolic quinone chromophore consist of at least four derivatives (A, B, C and D) (Fig. 1a). Elsinochromes were first isolated and chemically characterized by Weiss et al. in 1957 from cultures of Elsinoë randii (anamorph: Sphaceloma randii). The structures and chemical properties of elsinochromes A, B, and C have been well established (Weiss et al., 1957, 1965; Chen et al., 1966; Lousberg et al., 1969a,b, 1970; Mebius et al., 1990). Elsinochrome D, likely derived from elsinochrome C by forming a methylenedioxy ring (Fig. 1a), also was identified and characterized from the pigment mixture of Elsinoë annonae (Lousberg et al., 1970; Kurobane et al., 1981; Shirasugi & Misaki, 1992). Elsinochromes have since been identified or observed for production in cultures from a large number of phytopathogenic Elsinoë species (reviewed by Weiss et al., 1987). However, the elsinochromes have never been isolated from Elsinoë-induced lesions.

Figure 1.

Elsinochromes are the red/orange pigments produced by many Elsinoë species. (a) Structure of elsinochromes with various side-groups (R) (redrawn based on the work of Weiss et al., 1987). (b) Thin-layer chromatography analysis of elsinochromes produced by Elsinoë fawcettii. Elsinochromes were extracted from agar plugs containing fungal mycelium with acetone, spotted onto a silica gel plate, and separated with chloroform and ethyl acetate, resulting in five distinct (red/orange) bands with differing Rf values (1–5). (c) Absorption spectrum of the acetone crude extracts, showing three major peaks at 460, 530 and 570 nm.

Elsinochromes are structurally related to several perylenequinones such as altertoxin I produced by Alternaria alternata, cercosporin produced by many Cercospora spp., hypericin produced by Hypericum spp., hypocrellin A produced by Hypocrella bambusae, and phleichrome produced by Cladosporium spp. (Yoshihara et al., 1975; Assante et al., 1977; Daub, 1982a; Duran & Song, 1986; Stierle & Cardellina, 1989; Daub et al., 2005). All of the compounds have a common 4,9-dihydroxyperylene-3,10-quinone chromophore and only differ in attached side-chains (Daub et al., 2005). In addition, these compounds are grouped as photosensitizers based on their ability to sensitize cells to visible light and produce reactive oxygen species (ROS) (Yamazaki et al., 1975; Daub, 1982a). Light and oxygen are absolutely required for photodynamic function and toxicity of these compounds. Photosensitizing compounds absorb light energy and convert to a stable electronically excited state (triplet state) which in turn reacts with oxygen, mainly in two different ways, to produce toxic ROS such as superoxides (inline image), hydrogen peroxide (H2O2), hydroxyl radical (OH), and/or singlet oxygen (1O2) (Dobrowolski & Foote, 1983; Girotti, 1990). In the Type I reaction, the activated triplet photosensitizers can react with oxygen molecules directly by transferring a hydrogen atom or electron from reducing substrates (such as NADPH, ascorbate, and l-cysteine), resulting in reduced oxygen species including inline image (Girotti, 1990). In the Type II reaction, the activated photosensitizers react with oxygen molecules by an energy-transfer process, producing the electronically reactive 1O2 (Spikes, 1989).

Both 1O2 and inline image are toxic to cells, causing the oxidation of various biological components including fatty acids, membranes, proteins/enzymes, sugars and nucleic acids, and often resulting in cell death (reviewed by Daub et al., 2005). Many of the perylenequinone pigments produced by fungi have been shown to be toxic to mice, bacteria, and many fungi, to be cytotoxic to animal tumors, to be potent antiviral agents, and to inactivate protein kinase C (Tamaoki & Nakano, 1990; Hudson & Towers, 1991; Diwu, 1995; Hudson et al., 1997). Except for cercosporin produced by many members of the fungal genus Cercospora (Callahan et al., 1999; Choquer et al., 2005, 2007; Chen et al., 2007; Dekkers et al., 2007), none of the perylenequinones of fungal origin have been demonstrated to act as a crucial factor in plant diseases caused by the producing pathogen.

Citrus scab, caused by Elsinoë fawcettii, is one of the most important foliage fungal pathogens in many citrus-producing areas worldwide. Citrus scab affects the fruit, leaves, and twigs of many susceptible cultivars of citrus, including lemons, grapefruit, many tangerines and their hybrids, causing external blemishes and reducing acceptability of the fruit for the fresh market (Timmer et al., 1996). Isolates of E. fawcettii grow slowly in culture, forming < 10-mm colony size in 30 d. We have observed that many E. fawcettii isolates from field-grown citrus accumulated the red/orange pigments in culture, resembling elsinochromes that have been characterized (Weiss et al., 1957). In the present study we extracted the elsinochrome-like pigments from one of the E. fawcettii isolates and demonstrated that the pigments were toxic to host and nonhost plant cells. We also provide evidence that elsinochromes function as photosensitizing agents in culture and in planta, by producing toxic reactive oxygen species, mainly 1O2 and inline image.

Materials and Methods

Biological materials and cultural conditions

Elsinoë fawcettii Bitancourt Jenk. (anamorph: Sphaceloma fawcettii Jenk.) CAL WH-1 isolate used in this study was single-conidium isolated from scab affected calamondin (Citrus madurensis Lour) fruit in Florida and was kindly provided by L. W. Timmer (University of Florida, Citrus Research and Education Center, Lake Alfred, FL, USA). Fungus was grown on a sterilized filter paper, and stored at –20°C for long-term storage. Fungal cultures were routinely maintained on potato dextrose agar (PDA; Difco, Becton, Dickinson and Company, Sparks, MD, USA). For toxin production, fungal mycelia were ground with a sterile blender, spread on PDA and incubated under continuous fluorescent light for 5 d.

Tobacco (Nicotiana tabacum cv. Xanthi) cell suspension was kindly provided by D. J. Lewandowski (Ohio State University, Columbus, OH, USA) and maintained in a Murashige and Skoog medium (Murashige & Skoog, 1962) with gentle agitation under a daily regime of 16-h light and 8-h dark. Plant cells were subcultured weekly in a freshly prepared medium for toxicity assays. For protoplast isolation, 4-d-old tobacco cells were harvested and digested with cell wall degrading enzymes (a mixture of 1.5% cellulase and 0.15% pectolyase) as described by Lewandowski & Dawson (1998). Sweet orange (Citrus sinensis (L.) Osbeck) suspension cells, kindly provided by J. W. Grosser (University of Florida, CREC, Lake Alfred, FL, USA), were maintained in a Murashige & Tucker medium (1969). Citrus protoplasts were prepared by incubating with 10% of cell wall degrading enzyme complex (Sigma-Aldrich, St Louis, MO, USA) for 2 h as described by Grosser et al. (1988). Rough lemon (Citrus jambhiri Lush) trees were maintained in glasshouse and leaves of 4–7 d after flushing, approx. 13–20 mm long and 4–7 mm wide, were harvested for toxicity assays.

Isolation and analysis of fungal toxins

Elsinochromes are orange/red pigments and can be easily visualized when secreted in culture. Elsinochromes were extracted twice from dried agar medium bearing fungal mycelia with acetone for 16 h. Organic solvent was combined and evaporated with a Model R110 of Rotavapor (Brinkmann, Buchi, Switzerland). For thin-layer chromatography (TLC) analysis, elsinochromes dissolved in acetone were separated and visualized on TLC plates coated with a 60 F254 fluorescent silica gel (5 × 20 cm; Selecto Scientific, Suwanee, GA, USA) with a solvent system containing chloroform and ethyl acetate (1 : 1, v : v). The crude extracts separated by TLC were examined by a hand-held long wavelength UV lamp (UVP, San Gabriel, CA, USA). Elsinochrome derivatives were scraped from the TLC plate, dissolved in acetone, and separated from silica gel by low-speed centrifugation. The acetone was dried and the amounts of elsinochromes recovered were determined by weight. Elsinochromes dissolved in acetone were examined by spectrophotometry at absorbance between 400 and 650 nm. All bands recovered from the TLC plate have similar absorption spectra. Thus, elsinochrome derivatives were combined and directly mixed with plant cells for toxicity assays. The concentration of elsinochromes was calculated by reference to a regression line that was established using pure cercosporin (Sigma-Aldrich) as standard and was expressed as cercosporin equivalents. For toxicity assays, cercosporin was extracted from cultures of Cercospora nicotianae in separate studies (Choquer et al., 2005; Chen et al., 2007). Unless otherwise indicated, all chemicals were purchased from Sigma-Aldrich.

Toxicity assays to plant cells

Citrus or tobacco suspension cells or protoplasts at density 1 × 106 ml−1 were mixed with various concentrations of elsinochromes (dissolved in acetone) with or without singlet oxygen (1O2) quenchers, and placed on the top of 1% agarose (in a 35 × 10 mm Petri dish). Cell cultures were illuminated with fluorescent lights at an intensity of 3.5 J m−2 s−1 at room temperature (c. 25°C) for citrus cells or at 32°C for tobacco cells. Light intensity was determined by a Dual-Display light meter (Control Company, Friendswood, TX, USA). Dark-grown cultures were wrapped in aluminum foil and incubated in the same conditions. Cells were examined over time by staining with 1% Evan's blue (Taylor & West, 1980). Dead cells were stained blue as they cannot exclude Evan's blue, whereas live cells remained clear. Cell viability was determined with the aid of a hemocytometer using a microscope at ×100 magnification. The percentage of cell viability was calculated by the number of live cells divided by the total number of cells examined. Control cultures were untreated or cells treated with equal volumes of acetone and/or other solvents (final concentration < 1%) as appropriate.

The toxicity of elsinochromes on host plants was determined on detached rough lemon leaves (4–7 d after emergence). Elsinochromes (1 mm, cercosporin equivalent) with or without antioxidants were applied onto the surface of rough lemon leaves and incubated in a moist chamber under fluorescent light (3.5 J m−2 s−1), and monitored daily for development of necrotic lesions. Antioxidants used in this study include: bixin (carotenoid carboxylic acid), β-carotene, DABCO (1,4-diazabicyco octane), α-tocopherol (vitamin E), reduced glutathione, l-cysteine, and (+)-sodium l-ascorbate (vitamin C). Bixin, β-carotene, and α-tocopherol were dissolved in 95% ethanol and others were dissolved in water to make a stock solution.

Determination of electrolyte leakage

Electrolyte leakage was measured by the method of Alferez et al. (2006) with some modifications. Leaf discs (0.5 cm in diameter) were cut from 4-d-old rough lemon leaves, placed in a 96-well microtiter plate, and incubated with elsinochromes (2 mm; dissolved in 7% acetone) under constant fluorescent light (3.5 J m−2 s−1) or in complete darkness at room temperature. Control leaf discs were treated with equal amounts of acetone or deionizer water as appropriate. Leaf discs (10) were randomly collected at 12-h intervals, soaked in water for an additional 1 h on a rotary shaker (60 r.p.m), and measured for initial conductivity (IC). Leaf samples were immediately frozen in liquid nitrogen and kept at –80°C for at least 12 h. Total conductivity (TC) of leaves was determined after leaf discs were thawed at room temperature for 10–15 min. Conductivity was determined by a Model 115 Orion conductivity meter equipped with a Pentrode probe (Thermo Electron, Boston, MA, USA). Electrolyte leakage, expressed as percentage of total conductivity, was calculated by dividing IC by TC.

Detection of singlet oxygen and superoxide ions

Production of 1O2 by elsinochromes upon irradiation was determined by their ability to oxidize cholesterol as described (Daub & Hangarter, 1983a) with modification. The 1O2-generating photosensitizers, hematoporphyrin and cercosporin, were used as the positive controls. Photosensitizing compounds (8 mg each) in 20 ml pyridine were mixed with cholesterol (200 mg), and irradiated under a fluorescent light at intensity of 2.2 J m−2 s−1 with gentle bubbling for 5 h. After solvent was removed, the oxidized products were suspended in 20 ml of hot methanol, passed through a filter to remove precipitation after cooling, and analysed by TLC with a solvent system containing hexane–isopropanol (9 : 1 or 24 : 1, v : v). The oxidized products of cholesterol were visualized as bands after staining with a chromogenic reagent, N,N-dimethyl-p-phenylenediamine (1%) dissolved in methanol–H2O–glacial acetic acid (5 : 5: 0.1, v : v : v) (Smith & Hill, 1972; Smith et al., 1973). The cholesterol 5α-hydroperoxide standard was prepared by photo-oxidation with hematoporphyrin as described by Ramm & Caspi (1969).

Photochemical reduction of nitrotetrazolium blue chloride (NTB) was used to determine inline image production by elsinochromes (Daub & Hangarter, 1983). Reactions were carried out in a solution containing 5 mm 3-(N-morpholino) propanesulfonic acid (MOPS), 10 mm methionine (as a reducing agent), 2.5 mm NTB, 2 µm riboflavin or 10 µm of elsinochromes or cercosporin, and/or superoxide dismutase (SOD, 1 mg ml−1) or DABCO (1 mm). Nitrotetrazolium blue chloride was added into the buffer before the addition of photosensitizers and the reaction mixture was irradiated under a constant light at intensity of 4.7 J m−2 s−1. Superoxide accumulation, as shown by the increased absorbance at 560 nm as a result of the reduction of NTB, was measured spectrophotometrically (Beruchamp & Fridovich, 1971).


Characteristics of the red/orange pigments from E. fawcettii

The acetone extracts obtained from the cultures of an E. fawcettii isolate resulted in five distinct bands after TLC separation (Fig. 1b). Bands 1, 2 and 3 (in order of decreasing Rf value (the ratio of the distance migrated by a substance compared with the solvent front)) appeared to be major compounds of the extracts based on band width, whereas bands 4 and 5 appeared to be minor compounds. Spectrophotometric scanning revealed that the acetone extracted pigments displayed a strong absorbance at 460 nm with two minor peaks at 530 and 570 nm (Fig. 1c) markedly resembling the elsinochromes previously isolated from several Elsinoë species (Weiss et al., 1957, 1965). Further analysis indicated that each band recovered from TLC plates had similar absorption spectra, also displaying three major absorption peaks at 460, 530 and 570 nm (data not shown). The red/orange pigments from E. fawcettii had several typical characteristics of elsinochromes containing phenolic quinones. Similar to elsinochromes, the extracted red/orange pigments became bright green in color when dissolved in aqueous KOH or sodium carbonate. Further, the green color reverted to pink when alkaline sodium dithionite was added. The green color changed to a distinct leuco compound, showing yellow/greenish fluorescence when treated with zinc dust (Weiss et al., 1957, 1987). These results suggested that the red/orange pigments extracted from E. fawcettii were elsinochromes.

Toxicity of elsinochromes from E. fawcettii to host and nonhost cells

The toxicity of elsinochromes was assayed using citrus protoplasts. As shown in Fig. 2a, elsinochromes exhibited dose–response toxicities with respect to citrus protoplasts. At 10 µm, no viable citrus cells remained after 4 h in the light. At 5 µm, there was a decrease in the rate of cell death. By contrast, untreated citrus cells or cells treated with acetone alone remained viable throughout the assay period. Elsinochromes exerted no obvious toxicity when cells were incubated in complete darkness.

Figure 2.

Reduction of live cells (%) of citrus protoplasts (a), tobacco protoplasts (b), or suspension-cultured tobacco cells (c) treated with elsinochromes (ESC), phytotoxin produced by Elsinoë fawcettii or cercosporin (CR) toxin extracted from a tobacco pathogen, Cercospora nicotianae and incubated under constant fluorescent light or in the darkness. Cell viability of citrus protoplasts in the darkness was determined only at 6 h after incubation (a). Insets indicate viability of plant cells incubated in the dark. Data represent the means of two different experiments with at least three replicates.

The toxic effect of elsinochromes was also evaluated with suspension-cultured tobacco cells and protoplasts. Similar to cercosporin produced by a tobacco pathogen, C. nicotianae (Daub, 1982a), elsinochromes caused rapid death of tobacco protoplasts (Fig. 2b) or cultured suspension cells (Fig. 2c), in a dose–response manner within 1 h after irradiation with light. Untreated or acetone-treated tobacco cells in the light or cells incubated in the darkness remained viable for the duration of the experiment.

Antioxidants reduce elsinochrome toxicity

The toxicity of elsinochromes was alleviated to various degrees by adding 400 µm bixin (Fig. 3a), 2 mm DABCO, or 4 mm of ascorbate or reduced glutathione (Fig. 3b–d). Compared with bixin, DABCO, and ascorbate, reduced glutathione had less effect on the protection against toxicity of elsinochromes, by showing an extended lag period. Application of antioxidants with lower concentrations had little or no effect on toxicity reduction of elsinochromes (data not shown). Addition of α-tocopherol or l-cysteine (4 mm each) appeared to enhance elsinochrome phytotoxicity as duration of incubation increased (Fig. 3e,f).

Figure 3.

Cell viability (% live cells) of citrus protoplasts after treatment with elsinochromes (ESC) isolated from Elsinoë fawcettii with and without antioxidant compounds: (a) 400 µm bixin (carotenoid carboxylic acid); (b) 2 mm DABCO (1,4-diazabicyco octane); (c) 4 mm l-ascorbic acid; (d) 4 mm reduced glutathione; (e) 4 mmα-tocopherol; and (f) 4 mm l-cysteine. Citrus protoplasts (1 × 106) were mixed with or without elsinochromes (5 µm, cercosporin equivalent) and antioxidants as indicated under constant fluorescent light. The controls were treated with acetone–H2O or acetone–95% ethanol as appropriate. Data represent the means of two different experiments with at least three replicates. Open circles, control; closed circles, antioxidant; open triangles, ESC; closed squares, ESC + antioxidant.

Similar to citrus protoplasts, the cellular toxicity of elsinochromes to suspension-cultured tobacco cells could also be alleviated by the addition of antioxidants such as bixin, DABCO, ascorbate and α-tocopherol (Fig. 4). In the presence of bixin or DABCO, over 40% of tobacco cells were viable after incubation with elsinochromes for 24 h in the light. Both ascorbate and α-tocopherol significantly delayed photo-induced cell death (by at least 18 h), yet neither compound was able to protect against cell death beyond 24 h. Tobacco cells incubated in the dark were healthy for the duration of the experiment (data not shown).

Figure 4.

Cell viability (% live cells) in suspension-cultured tobacco cells by 1 µm elsinochromes (ESC) with antioxidants (400 µm bixin (carotenoid carboxylic acid), 2 mm DABCO (1,4-diazabicyco octane), 4 mm of l-ascorbic acid, or α-tocopherol). Tobacco cells (1 × 106) were treated with elsinochromes in the presence or absence of antioxidants, as indicated, and incubated under constant fluorescent light. Data represent the means of two different experiments with at least three replicates. Open circles, control (1 µl acetone–ethanol); closed circles, antioxidant; open triangles, ESC; closed squares, ESC + antioxidant.

Elsinochromes are toxic to citrus leaves

To determine if elsinochromes were toxic to the host, the crude extracts were applied onto detached rough lemon leaves. After incubation for 10 d, the elsinochrome-treated spot on rough lemon leaves developed noticeable necroses in the light (Fig. 5). Neither leaves treated with the solvent alone nor leaves incubated in complete darkness developed necrotic lesions (Fig. 5, and data not shown). Co-inoculation of elsinochromes with bixin, DABCO, or ascorbate onto the leaves prevented development of necrotic lesions on detached rough lemon leaves (Fig. 5a–c). By contrast, superoxide dismutase (SOD) or peroxidase had little effect on protection against the toxicity of elsinochromes on rough lemon leaves (data not shown). Bixin showed a light brown color when dissolved in 95% ethanol. Application of α-tocopherol alone, however, resulted in a brownish lesion and failed to alleviate the toxicity of elsinochromes (Fig. 5d).

Figure 5.

Development of necrotic lesions on detached rough lemon leaves after a 10-d application of 1 mm elsinochromes (ESC) from Elsinoë fawcettii, and prevention of necrosis by coapplication of antioxidants (100 mm bixin (carotenoid carboxylic acid), 300 mm of DABCO (1,4-diazabicyco octane), 300 mm l-ascorbic acid, or 500 mmα-tocopherol). Citrus leaves were treated with or without ESC and antioxidants as indicated and incubated in a moist chamber under a constant fluorescent light. The mocks were treated with equal volume (3 µl) of water, 95% ethanol, and/or acetone as appropriate.

Production of reactive oxygen species by elsinochromes

Production of inline image by elsinochromes was evaluated with a superoxide scavenging assay (Beruchamp & Fridovich, 1971; Daub & Hangarter, 1983) based on its ability to reduce nitrotetrazolium blue chloride (NTB), and compared with the levels of inline image induced by other photosensitizing compounds such as cercosporin and riboflavin known to generate inline image (Oster et al., 1962; Daub & Hangarter, 1983). In the absence of NTB, photosensitizers dissolved in the MOPS buffer produced low absorbance values at 560 nm (Fig. 6a and data not shown), representing the reaction baseline. Mixing NTB with the crude extracts of elsinochromes significantly elevated absorbance values because of reduction of NTB (Fig. 6a). However, the values, were repressed in the presence of the inline image scavenging enzyme, SOD, indicating the production of superoxides. Compared with cercosporin and riboflavin, elsinochromes appeared to induce high levels of inline image after irradiation with light (Fig. 6b). Addition of SOD, but not DABCO, drastically reduced the accumulation of superoxides from the actions of elsinochromes (Fig. 6c).

Figure 6.

Superoxide production assays based on the reduction of nitrotetrazolium blue chloride (NTB) as substrate. (a) Spectrophotometric absorbance of elsinochromes (ESC) alone (open bars) or ESC treated with NTB with (hatched bars) or without (closed bars) superoxide dismutase (SOD) in the 3-(N-morpholino) propanesulfonic acid (MOPS) buffer at 560 nm. (b) Accumulation of superoxide ions (inline image) by photosensitizing compounds riboflavin (circles), cercosporin (squares) extracted from Cercospora nicotianae, and elsinochromes (triangles) extracted from cultures of Elsinoë fawcettii after irradiation with light. The respective photosensitizers were dissolved in buffer and used as the blanks. (c) Suppression of elsinochrome-induced superoxide accumulation by addition of SOD but not DABCO (1,4-diazabicyco octane). Reactions were carried out in the MOPS buffer solution containing 2.5 mm NTB, 10 mm methionine, and 10 µm of cercosporin or ESC, or 2 µm riboflavin, and/or SOD (1 mg ml−1) or DABCO (1 mm), and absorbance at 560 nm (A560) measured.

Production of cholesterol 5α-hydroperoxide from cholesterol is one of the best and simplest ways to test for the presence of 1O2 (Kulig & Smith, 1973). To assess the production of 1O2in vitro, elsinochromes were mixed with cholesterol and illuminated. The resulting products were chromatographed in two different solvent systems and detected as distinct bands after staining with a chromogenic reagent, N,N-dimethyl-p-phenylenediamine. Reaction of cholesterol 5α-hydroperoxide with dimethyl-p-phenylenediamine resulted in distinct pink (changed to dark-green) pigments on the TLC plate. As with hematoporphyrin (positive control) and cercosporin photosensitizers, elsinochromes converted cholesterol by 1O2-induced photodynamic oxidation into the 5α-hydroperoxides of cholesterol in the light (Rf 0.4 and 0.1 in Fig. 7a and b, respectively). No visible band was detected from the untreated cholesterol. Radiation of cholesterol to UV light (240 nm) for 24 h yielded several bands on TLC, but no cholesterol 5α-hydroperoxide was detected (Fig. 7c). A faint band with slightly faster movement of unknown identity was detected in products generated by cercosporin and elsinochromes. The cholesterol 5α-hydroperoxides and the faint bands were undetectable when β-carotene, a potent 1O2 quencher, was added to the reaction mixture.

Figure 7.

Oxidation of cholesterol by photosensitizers: hematoporphyrin (HP), cercosporin (CR), and elsinochromes (ESC) under light. Cholesterol (200 mg) was dissolved in pyridine, mixed with photosensitizers with or without β-carotene, as indicated, and incubated for 5 h. The oxidized products were separated by thin-layer chromatography (TLC) with a hexane–isopropanol 9 : 1 (v : v) (a) or 24 : 1 (v : v) (b,c) solvent system. Cholesterol 5α-hydroperoxide (Rf 0.4 or 0.1 in solvent A or B) formed distinct bands after staining with 1% of N,N-dimethyl-p-phenylenediamine. Radiation of cholesterol to UV light for 24 h resulted in multiple products, but no cholesterol 5α-hydroperoxide (c). Untreated cholesterol (in chloroform) was not visible after staining with the chromogen.

Electrolyte leakage induced by elsinochromes

When incubated for 24 h, the extracted elsinochromes induced an increase in electrolyte leakage of rough lemon leaf discs after illumination (Fig. 8a). Leaf discs treated with water or solvent alone, or leaf discs incubated in the darkness showed relatively minor electrolyte leakage. Electrolyte leakage induced by elsinochromes was steadily increased over time, whereas leakage of the controls remained low (Fig. 8b). However, addition of bixin or DABCO into the reaction mixture failed to inhibit electrolyte leakage induced by elsinochromes (data not shown).

Figure 8.

Leakage of electrolytes from illuminated rough lemon leaf discs treated with elsinochromes (ESC). (a) Light increased electrolyte leakage of citrus leaf discs treated with ESC compared with those treated with water or acetone (S) alone at 24 h. (b) Increasing electrolyte leakage of rough lemon leaf discs over time after treatment with ESC in the light. Water, open circles; acetone, closed circles; ESC, triangles. Data represent the means of two different experiments with five replicates.


Elsinoë fawcettii isolates obtained in Florida citrus-growing areas produced red/orange pigments in culture similar to elsinochromes produced by many Elsinoë spp. (Weiss et al., 1957, 1987), and had characteristic absorption spectra in the visible region. In addition, the orange/red pigments had other typical characteristics of elsinochromes, such as forming a distinct yellow/greenish fluorescence leuco compound when reacting with zinc dust in alkaline conditions (Weiss et al., 1957, 1987). Thin-layer chromatography analysis of the acetone-extracted pigments revealed significant variations of the pigments produced in culture (data not shown). The results strongly suggest that these pigments were elsinochrome analogs containing a phenolic quinone chromophore.

Compared with studies on chemical characterization, little is known about the biological function of elsinochromes. The photodynamic action of elsinochromes leading to cellular toxicity has previously been predicted on the basis of their structural similarity to many photosensitizing compounds such as cercosporin, hypericin, or hypocrellin A (Daub et al., 2005), yet has never been demonstrated experimentally. In the present study we provide evidence to support the notion that elsinochromes exhibit cellular toxicity to plant cells by functioning as photodynamic compounds that generate reactive oxygen species.

Since citrus suspension cells aggregated to form massive clumps in culture, the toxicity assays were performed only with citrus protoplasts. Elsinochromes extracted from E. fawcettii rapidly killed citrus protoplasts in a dose-dependent manner and only when cells were exposed to the light, consistent with the mode of action for other photosensitizing compounds (Yamazaki et al., 1975; Daub, 1982a; Spikes, 1989). The cellular toxicity of elsinochromes was attenuated considerably when antioxidant compounds were added into the culture, implying the involvement of reactive oxygen species, particularly inline image and/or 1O2. Carotenoids are efficient 1O2 quenchers in the biological systems (Krinksy, 1979). Bixin, a carotenoid carboxylic acid, has lower molecular weight than β-carotene and is more polar owing mainly to the presence of the carboxylic acid group (Daub, 1982a). Both bixin and β-carotene have the same isoprenoid chain length but bixin is more soluble in aqueous solutions. Carotenoids quench 1O2 via energy transfer mechanisms (Foote & Denny, 1968; Foote et al., 1970) and are slowly consumed in the reaction. 1,4-Diazabicyclo octane (DABCO) also is known as an effective 1O2 quencher. Unlike bixin, DABCO quenches 1O2 by a chemical reaction, and is quickly consumed in the presence of 1O2 (Oannes & Wilson, 1968; Daub, 1982a). These features may account for the fact that bixin provided a more persistent protection of citrus cells than did DABCO (Figs 3 and 4). Both bixin and DABCO have been shown to decrease the toxicity of cercosporin to tobacco cells (Daub, 1982a). Elsinochromes were also toxic to tobacco, a nonhost plant of E. fawcettii. Similarly, addition of antioxidants also provided some levels of protection against the toxicity of elsinochromes, indicating a nonhost specificity of elsinochromes. The antioxidants bixin, DABCO and ascorbic acid but not α-tocopherol, which decreased the toxicity of elsinochromes in culture, provided some protection with detached rough lemon leaves. However, the concentrations required for protection in planta were far higher than those in culture, likely owing to the poor penetration of antioxidants through the cuticle.

Both ascorbic acid and the reduced form of glutathione were also effective in protecting citrus cells against elsinochromes. By contrast, l-cysteine and α-tocopherol had little or no effect on cellular protection against elsinochromes in culture and on citrus leaves. Application of α-tocopherol alone at a concentration of 500 mm onto young rough lemon leaves induced necrotic lesions. However, α-tocopherol provided some levels of cellular protection against the toxicity of elsinochromes in tobacco cells. α-Tocopherol, with a well-known ability to terminate radical chain oxidation by binding to cell membranes, has been often used in nutritional studies or cell membrane functions (Tinberg & Barber, 1970; Tappel, 1972). The requirement of α-tocopherol binding to membranes may contribute to its ineffectiveness in cellular protection against elsinochromes on citrus protoplasts and leaf tissues.

All photosensitizing compounds are able to generate reactive oxygen species upon activation by light (Yamazaki et al., 1975; Daub, 1982a; Dobrowolski & Foote, 1983; Spikes, 1989; Girotti, 1990). Results derived from this study on the toxic effect of elsinochromes in vitro and in vivo also suggested that elsinochromes could be damaging citrus cells by the production of both inline image and 1O2. We have shown that antioxidants such as bixin, DABCO, and others capable of quenching 1O2 provided substantial protection from the toxicity of elsinochromes in both citrus and tobacco suspension cultured cells and on rough lemon leaves. Production of 1O2 by elsinochromes was confirmed by successful detection of the cholesterol 5α-hydroperoxide that is solely produced through the oxidation of cholesterol by 1O2 (Kulig & Smith, 1973), which is one of the best indications of separating 1O2 production from the production of inline image and other free radicals (Smith et al., 1973). Oxidation of cholesterol by inline image, UV or other radicals often generates multiple products, mainly the 7α- and 7β-hydroperoxides, but never produce the 5α-hydroperoxide (Smith et al., 1973). By contrast, the oxidizing reaction of 1O2 with cholesterol primarily produces the 5α-hydroperoxide. Addition of β-carotene, an effective 1O2 quencher, completely prevented the production of the 5α-hydroperoxide by hematoporphyrin, cercosporin and elsinochromes. Therefore, positive detection of the cholesterol 5α-hydroperoxide indicates that elsinochromes were able to generate 1O2 after exposed to the light. Singlet oxygen is highly reactive and has been shown to be toxic to a wide range of cells (Foote & Denny, 1968; Foote et al., 1970; Daub & Ehrenshaft, 2000). The results strongly implicate the involvement of 1O2 in the toxicity of elsinochromes because 1O2 quenchers reduced cultured citrus cell mortality and prevented lesion formation on host plants.

Production of inline image by elsinochromes was shown by a superoxide production assay that was developed to determine SOD enzymatic activity using NTB as substrate (Beruchamp & Fridovich, 1971; Daub & Hangarter, 1983). It seemed that elsinochromes produced inline image more efficiently than cercosporin or riboflavin. Reduction of NTB induced by elsinochromes was inhibited by adding the inline image scavenging enzyme (SOD) but not the 1O2 quencher, providing evidence to support that photodynamic reaction of elsinochromes also generates inline image. The detection of inline image has important implication for involvement in elsinochrome toxicity since inline image also is toxic in biological systems.

To explore the potential toxic mechanisms of elsinochromes, electrolyte leakage of rough lemon leaf discs was measured after treatment with elsinochromes. The study revealed that elsinochromes induced higher electrolyte leakage than those of the water or solvent controls under light. The results suggested that elsinochromes exert their toxicity by damaging cell membranes, likely by inducing lipid peroxidation as demonstrated in cercosporin-treated tobacco tissues (Daub, 1982b). Unlike the rapidly (< 2 min) induced electrolyte leakage caused by cercosporin in tobacco, elsinochrome-induced ion leakage was not observed until several hours after irradiation, indicating that the toxic activity on membranes might not be a direct consequence. Alternatively, this slow response of electrolyte leakage might result from the complexity of cell membranes of citrus or the structure of its cuticle. Disruption of cell membranes followed by nutrient release might be beneficial to the invading pathogen. In this regard, elsinochromes may play a critical role in pathogenesis during fungal penetration and colonization. Continuing studies will further define the biological function of elsinochromes and the biosynthetic pathway leading to their production.


We thank J. K. Burns for initial editing and two anonymous reviewers for their helpful comments. This research was supported by the Florida Agricultural Experiment Station and grants from the Florida Citrus Production Research Advisory Council (FCPRAC) #012–04P and #033–03P.