1Faculty of Agriculture, and 2United Graduate School of Agriculture, Tottori University, Tottori 680-0945, Japan

Background and objectives
Certain fungal pathogens, especially in the genera Alternaria and Cochliobolus, are known to produce host-specific toxins or host-selective toxins (HSTs) as agents of virulence or pathogenicity. In the genus Alternaria, at least nine examples of HST-producing pathogens have been reported. The tomato pathotype of A. alternata [A. alternata (Fries) f. sp. lycopersici], which causes stem canker on certain tomato cultivars such as Earlypak 7 in California and First in Japan, produces an HST named AAL-toxin. Precise definition of the pathological role of these HSTs as disease determinants requires detection of the toxins in diseased plant tissues. Here we report the development of an HPLC method and an immunoassay for the detection and quantification of AAL-toxins which were used to study in planta distribution of the toxin in tomato infected with the pathogen.

Materials and methods
Precolumn fluorescence-derivatization with o-phthalaldehyde (OPA) plus mercaptoethanol, or 4-fluoro-7-nitrobenzofulazan (NBD-F) were employed for the detection and quantification of AAL-toxins. The fluorescent derivatives were eluted from a reverse-phase column with methanol/phosphate buffer (pH 3.3) (70:30, v/v). Although AAL-toxins themselves are too small to be antigenic, coupling them to a protein carrier resulted in a functional immunogen. Cholera toxin (CT) produced by Vibrio cholerae was used as both a hapten carrier protein and an adjuvant. AAL-toxin TA was conjugated through the amino group to CT with glutaraldehyde as the coupling reagent. BALB/c mice were immunized by intraperitoneal or subcutaneous injection (three times) with 10 or 15 mg immunogen in 0.2 ml PBS. For selective detection of the inoculated pathogen within plant tissue, a transformant of the pathogen carrying plasmid pAN7-1, conferring resistance to hygromycin B, was used [1].

Results and conclusions
Pre-column derivatization of the toxins with OPA or NBD-F yielded highly fluorescent products showing well-resolved peaks on reverse-phase HPLC. The detection limit was in the order of 1 ng. Immunization of mice with the AAL-toxin-CT conjugate without adjuvant resulted in the production of antibodies reactive with AAL-toxins. The antibodies did not cross-react with fumonisin B1. The detection limit for AAL-toxin TA in an indirect enzyme-linked immunosorbent assay was 30 ng/ml.

Toxin production by germinating spores is required for the toxin to play a role in initial pathogenesis. Therefore, the presence of the toxin in spore germination fluids was quantified by these methods. The amount of TA and TB was estimated at about 0.25 pg per spore at 24 h after germination of virulent spores. This amount may be enough to cause damage on several host cells adjacent to the spore. AAL-toxins released by germinating spores therefore appear to be important for successful establishment of the pathogen in susceptible tomato.

A hygromycin-resistant transformant of the tomato pathotype of A. alternata was inoculated on the basal part of the stems of susceptible tomatoes grown in pots. Seven and 20 days after inoculation, serial sections were prepared from the infected stems and the distribution of the pathogen and the toxin was determined. On a resistant cultivar, AAL-toxin and the pathogen were detected only in sections around the inoculation site. On the other hand, a widespread distribution of the toxin and the pathogen were detected in the susceptible cultivar. Thus in planta production and release of the toxin by the pathogen were confirmed. Determination of toxin production in the infected susceptible tissue is an important step to examine the role of the toxin in disease development.

1. Akamatsu H, Itoh Y, Kodama M et al., 1997. Phytopathology 87, 967-972.