Two isolates of Fusarium proliferatum from different global locations and habitats mineralized several natural and synthetic lignins. MUCL 31970 was isolated from a forest soil whereas the second strain, NRRL 31071, was a wheat endophyte causing disease in stressed seedlings. Onset and the fastest rate of lignin mineralization occurred during logarithmic and early stationary-phase of culture. Reduction of glucose in the medium shortened log-growth phase and advanced the onset of mineralization for both isolates. Mineralization correlated with the detection of extracellular laccase and aryl alcohol oxidase activities. The carbon–nitrogen ratio in the medium influenced laccase isozyme production and secretion by both strains. These studies suggest that both F. proliferatum strains degrade lignin via comparable routes, despite their different habitats and saprophytic or endophytic strategies.
Fusarium proliferatum occupies many different habitats. One characterized isolate from forest soils in the Canary Islands (F. proliferatum MUCL 31970), termed the TF strain, is ligninolytic  and mineralizes lignin in media with a C/N ratio of 13, a ratio comparable to that of its habitat. This ligninolytic strain is taxonomically related to a second strain (F. proliferatum NRRL 31071 termed the UT strain) isolated as an endophyte from wheat kernels in the USA . This isolate causes necrosis of young developing wheat seedlings grown under stressed conditions [3,4]. Pathogenicity of other strains of F. proliferatum has been reported, causing corn-ear rot in the field and in post-harvest storage in maize kernels [5,6], black point and tombstone in wheat [7,8], and rots affecting asparagus , bananas  and gladiolus corms .
Laccases are produced by the UT strain during infection of wheat seedlings . These same isozymes are present in extracts from carbon-limited in vitro cultures of the fungus, while other isoforms are detected on low nitrogen or high nutrient media . Apoplastic sugar levels are reduced by 70% in wheat when it was grown under stressed conditions. It is possible that the low carbon conditions in the stressed plants permitted laccase gene expression by the fungus . Laccases, produced by the necrosis-causing fungus, Botrytis cinerea, during host infection, are proposed to be important in degradation of host toxic materials and in the limitation of lignification as a defense response during pathogenesis [13,14]. Laccases also are implicated in lignin degradation by anamorphic fungi isolated from soils, including the isolate of F. proliferatum MUCL 31970 [1,15].
In this paper, we address whether the wheat F. proliferatum isolate has a similar capacity to degrade lignin as the forest soil TF strain. To this end, both isolates were cultivated in media containing 14C-labeled natural lignin and three lignin model substrates to determine the extent of mineralization. The activities for several enzymes, laccase, aryl alcohol oxidase and peroxidases, proposed in other systems to be involved in mineralization were examined over time courses to compare their presence with mineralization activity. Complex regulation of laccase production exists for the UT strain [12,16]. Similarly, transcript patterns for genes encoding several enzymes involved in wood digestion by Phanerochaete chrysosporium are dramatically different when the fungus colonizes wood chips, as opposed to those obtained in defined media [17,18]. Thus, the effect of nutrition on laccase isozyme production in TF was also compared between the wheat and forest soil isolates.
2Materials and methods
2.1Microorganisms, media and culture conditions
F. proliferatum NRRL 31071 (UT strain) was obtained from within surface-sterilized wheat kernels [15,18]. The strain was stored at −70°C in 15% glycerol. F. proliferatum MUCL 31970 (TF strain; ) was isolated from pine forest soil by using a mineral medium solidified with agar containing the polymeric fraction (1 mg ml −1) of Kraft lignin (PKL) (PKL, Indulin AT; Westvaco Co., Charleston, S.C.), purified as described by Rodríguez et al.  and used as the sole carbon and energy source. The TF strain was maintained at 4°C on this medium. Inoculum for the studies in this paper were generated by transfer of 1 cm2 plugs of mycelia on agar plates to the PKL medium modified with addition of 55.5 mM glucose.
Media used for lignin degradation studies contained different 14C-labeled lignins as employed by Rodríguez et al. . The uniformly 14C-labeled milled wood lignin from wheat straw (14C-MWL) had a specific activity (sa) of 0.009 μCi mg−1. The synthetic lignins were dehydrogenative polymerizates of coniferyl alcohol (DHP), specifically labeled in the 14C-β-side chain (14C-β-DHP, sa: 0.126 μCi mg−1), uniformly 14C- labeled in the aromatic ring (14C-ring-DHP, sa: 0.19 μCi mg−1), or specifically labeled in the methoxy-groups (14C-methoxy-DHP, sa: 0.083 μCi mg−1). All these substrates were purchased from Forstbotanisches Institut der Universität Göttingen, Germany. These DHPs were added at 0.09 μCi per flask. Chromatography of both DHPs and natural lignin on Sephadex G-50 (45 × 2 cm) showed no low molecular weight materials. The products had a size range between 15.5 and 1.5 kDa. D-glucose was added as an easily metabolized carbon source at final concentrations of 5.5 or 55.5 mM together with nitrogen provided by 13.48 mM asparagine and 12.2 mM NH4NO3 to give C/N ratios of 1.287 and 12.87, respectively (referred to in this paper by the C/N ratios of 1.3 and 13). 14CO2 evolved from the 14C-labeled sources was trapped with 10% sodium hydroxide as used by Haider and Trojanowski , and radioactivity was measured by combining 1 ml of the NaOH sample with 1 ml of distilled water and 4 ml of scintillation liquid (Optiphase III; Kabi Pharmacia) in a liquid scintillation counter (LKB Wallace mod.1214). Lignin mineralization was expressed as the percentage of the labeled lignins recovered as 14CO2. To determine the patterns of enzyme production, unlabeled DHP (1 mg ml−1) was used.
All the assays described above were performed in triplicate in 250 ml Erlenmeyer flasks containing 25 ml of medium and incubated with shaking (50 strokes min−1) at 28°C for up to 30 days. Non-inoculated controls were incubated in duplicate but otherwise were treated identically.
2.2Determination of the dry weight of mycelia
Spore suspensions were prepared by stirring a 1-cm2 plug from a PKL-glucose agar plate into 5 ml of 0.01% Tween 80 in sterile water (vol/vol). One ml of this suspension (about 9 × 106 spores) was inoculated into 25 ml of media with C/N 1.3 or 13 supplemented with unlabeled DHP (1 mg ml−1) contained in 250 ml Erlenmeyer flasks. Inoculated flasks were incubated at 28°C with shaking for 10 days. Studies were run in triplicate and flasks were harvested daily. The cultures were filtered through pre-weighed Millipore filters (0.45 μm) and the retained mycelia were repeatedly washed with distilled water. Filters were maintained at 80°C to reach a constant weight and the dry weights of the mycelia determined.
Enzyme activities from Fusarium strains were assayed periodically in cultures supplemented with unlabeled DHP. Cultures were harvested by centrifuging (9000 × g), discarding the pellet, and the extracellular fluids were evaluated for enzymatic activities. The activities were determined spectrophotometrically, all in final reaction volumes of 1 ml, using a UV–Vis Shimadzu 160 as described below. A unit (U) was defined as the amount of enzyme capable of transforming 1 μmol min−1 of its target substrate.
Laccase-like activities were tested by monitoring oxidation of ABTS using a modified assay as described by Wolfenden and Wilson . The reaction mixture contained 2 mM ABTS, sodium succinate and lactate buffers, 0.1 M at pH 4.5, and 50–200 μl of culture filtrate. Oxidation of ABTS was followed by an absorbance increase at 420 nm (ε420= 3.6 × 104 M−1 cm−1). Activity for oxidation of 2,6-dimethoxy phenol (DMP; ε468= 1 × 104 M−1 cm−1) was monitored using 2 mM DMP in 200 mM sodium tartrate pH 4.5, as previously described by Shuttleworth et al. . Laccase isozymes composition was visualized after separation of the enzymes by non-denaturing polyacrylamide gel electrophoresis and staining for activity with ABTS by the method of Niku-Paavola et al. .
Aryl-alcohol oxidase (AAO) was assayed by monitoring the veratraldehyde (ε310= 9300 M−1 cm−1) formed on oxidation of veratryl alcohol following a method modified from Bourbonnais and Paice  to contain 2 mM veratryl alcohol, 0.05 M sodium tartrate buffer pH 4.5, and 500 μl aliquots of culture filtrate.
Several different potential substrates for oxidases that would generate hydrogen peroxide were used (methyl glyoxal, glyoxal, d-cellobiose, d-glucose, d-galactose and oxalic acid). Tests for hydrogen peroxide production used a peroxidase-coupled assay with phenol red as the substrate, as previously described by Kersten and Kirk . Reaction mixtures contained 20 mM sodium 2,2-dimethyl succinate (pH 4.5), 10 mM oxidase substrate, 0.01% phenol red (ε520= 2.2 × 104 M−1 cm−1), 10 μg of horse-radish peroxidase (Sigma Chemical Co), and up to 300 μl of culture fluid. Reactions were stopped by adding 50 μl of 2 N NaOH. Controls lacked addition of culture filtrate. Product formation was detected at 520 nm.
Lignin peroxidase (LiP) was assayed by the method of Tien and Kirk , monitoring oxidation of veratryl alcohol (ε310= 0.93 × 104 M−1 cm−1), after addition of hydrogen peroxide (0.4 mM). A second method described by Archibald  used azure B (ε651= 4.88 × 104 M−1 cm−1) as the substrate.
Manganese peroxidase (MnP) was determined according to the method previously described by Camarero et al. , following the oxidation of DMP in a reaction mixture with 100 mM Na tartrate (pH 5), 100 μM MnSO4, 50–200 l of culture filtrate and starting the reaction by the addition of 50 μM hydrogen peroxide.
The protein content of extracts was determined spectrophotometrically at 595 nm with the Coomassie Blue method . Bovine albumin fraction V was used as protein standard (Bio-Rad Protein Assay).
3.1The ligninolytic potential of F. proliferatum strain UT is comparable to that of TF
Both isolates caused mineralization of natural and synthetic lignins (14C-MWL from wheat straw, 14COH3-DHP, 14C-β-DHP and 14C-ring-DHP, Fig. 1A and B). After a 30-day incubation in C/N 13 medium, the UT isolate mineralized about 30% of the 14C-MWL (Fig. 1A), a level that was comparable to that achieved by TF (Fig. 1B). Cumulative mineralization rates also were similar between UT and TF for the synthetic lignins: between 9% and 11% mineralization with 14C-β-DHP, and up to 7% when 14CH3O-DHP was used as substrate (Fig. 1A and B). However, mineralization of the aromatic-ring-labeled DHP by UT was about 75% higher than for TF (14% and 4%, respectively) in the 30-day period (Fig. 1A and B). The products obtained after TF cultures amended with either 14C-MWL and 14C-DHPs had a lower molecular size when compared with the starting substrates as determined by chromatography on Sephadex G-50 (data not shown). The rate of mineralization of 14C-β-DHP by both strains in C/N medium 13 was higher during the initial time period of 3-to-7 days rather than 9–30 days (Fig. 2B and D). A time course in the C/N 13 amended with unlabeled DHP showed glucose was used by 72 h although maximum fungal biomass was not attained until four to five days (data not shown).
3.2Effect of the C/N ratio on mineralization and production of enzyme activities by Fusarium strains of synthetic lignin
Mineralization of 14C-β-side chain-labeled DHP occurred earlier in the C/N 1.3 medium (Fig. 2A and C) than the C/N 13 medium (Fig. 2B and D) for both strains. About 50% of total mineralization in C/N 1.3 medium occurred in the first three days of incubation, whereas five days were necessary in C/N 13 cultures (Fig. 2A–D). Mineralization occurred prior to the attainment of maximal mass. Mycelial mass was 4 and 12 mg (dry weight) for the UT and TF strains at 48 h, with a maximum of 20 mg for both strains at 72 h of incubation, in the C/N 1.3. The UT and TF strains, respectively, had masses of 4 and 20 mg at 48 h, both with a maximum of 80 mg, respectively, at 96 and 120 h of incubation, in the C/N 13. Glucose was catabolized by 48 h in C/N 1.3 and 72 h for C/N 13 (data not shown).
To attempt to understand this effect of nutrition on enzyme activities potentially associated with lignin degradation, culture filtrates of C/N 1.3 and 13 media supplemented with unlabeled DHP were examined. No H2O2-generating oxidases that used methyl glyoxal, glyoxal, d-cellobiose, d-glucose, d-galactose or oxalic acid as substrates were detected. No Lip activities were detected when either veratryl alcohol or the dye azure B were used as substrates nor was MnP activity observed with DMP. However, the extracellular medium from both cultures showed laccase and aryl alcohol oxidase (AAO) activities (Figs. 3 and 4). For both strains extracellular laccase detected with ABTS was present early in culture and reached its highest levels during growth on the C/N 1.3 medium but was negligible in C/N 13 medium during the experimental period. DMP-detected laccase activity peaked much higher and earlier in the C/N 13 medium than the C/N 1.3 medium for the UT strain. For the TF strain, DMP-laccase activity increased as the cultures aged, thus exhibiting a different temporal pattern than for the UT strain. AAO activity patterns were similar for both strains in the C/N 1.3 medium and resembled those for the ABTS-laccase acitivity. In the C/N 13 medium, the patterns for AAO activity were again different between strains, being lower for the UT strain than the TF strain. Low levels of both AAO and laccase were present prior to the peak of mineralization measured with 14C-β-side chain labeled DHP (Figs. 2–4).
3.3Laccase isozyme production is similar in TF and UT
Previous studies by Kwon and Anderson with UT  demonstrated that laccase activity as detected by ABTS oxidase activity resulted from different isozymes. Isolate TF also produced different laccase isozymes (Fig. 5) that were detected in both mycelial extracts and extracellular culture filtrates in a manner similar to that for the UT strain. Production of the laccases was affected by the C/N ratio, occurring earlier in the C/N 1.3 medium than the C/N 13 medium. In addition, the production of the laccase isozymes was repressed temporally in the C/N 13 medium, in agreement with the spectrophotometric analyses of ABTS laccase activities (Figs. 3 and 4). The faster moving intracellular isozymes for the TF isolate were expressed extracellularly to a lesser extent in comparison to the UT strain.
Two F. proliferatum isolates from different global locations and ecological sites, strain TF from a forest soil in the Canary Islands, and strain UT, an opportunistic endophytic wheat pathogen from the USA, displayed similar abilities to degrade natural lignin from wheat (14C-MWL) and synthetic polymers. The maximal rates of mineralization were observed early in growth. This finding agrees with reports that actinomycetes solubilized and degraded lignin during their primary growth  but differs from mineralization by the white-rot fungi, like P. chrysosporium, where it is associated with late stationary-phase growth [31,32].
Differences were detected between the strains in mineralization of aromatic-ring-labeled DHP with the UT strain having the higher activity. The UT and TF strains also differed in the relative production of AAO and DMP-laccase activities in C/N 13 medium, in which the UT strain had higher DMP-laccase and lower AAO activies than the TF strain. Whether this difference relates to the more extensive degradation of the aromatic-ring-labeled DHP requires additional investigation.
Only laccase and AAO activities were found in the supernatant of both strains after searching for other oxidases and peroxidases associated with lignin degradation. Cooperative action between laccase and AAO or other oxidases (e.g., glucose oxidase) has been described for basidiomycetes when degrading lignin derivatives and soluble lignosulphonate [33,34]. The peak activities of the ABTS and DMP laccases did not correspond to the timing of maximum mineralization. However, only low levels of laccase activity were required for the ligninolytic system of Petriellidium fusoideum, where the activity was correlated with production of .hydroxyl radicals .
Laccase activity in both isolates was derived from different isozymes. The lower glucose concentration in the C/N 1.3 medium promoted a much earlier production of both the mycelial and extracellular activities. These findings confirmed studies with the UT strain where carbon and nitrogen levels were shown to regulate laccase isozyme production . The regulation of mineralization and laccase activities by nutrient availability may be relevant to the natural habitats of these F. proliferatum strains. Similar nutritional regulation was demonstrated for P. chrysosporium where lignin peroxidase activity was enhanced by the low nitrogen conditions thought to exist in woody materials . In non-living woody material, carbon is sequestered in polymeric forms until ligninolysis has proceeded. Thus, an ability to initiate mineralization under conditions of low carbon could be an advantage for a forest soil isolate. Maintenance of mineralization as carbon level increases also could be advantageous for the TF isolate. For the opportunistic pathogen, the ability to degrade lignin may be exploited when the F. proliferatum is growing as a pathogen rather than an endophyte. As discussed by Kwon and Anderson , the switch to the pathogenic state may have occurred when the supply of low molecular weight sugars in the wheat apoplast was low, as conditioned by plant growth under stress. Degradation of phenolic modifications to the plant cell wall may then promote pathogenic growth of the F. proliferatum. Indeed, the full complement of anionic laccases and expression of transcripts corresponding to lac-1 was detected in wheat showing symptoms of necrosis upon infection with UT strain . Continued studies of the mechanisms of lignin mineralization and the regulation of specific enzymes related to phenolic modification may help reveal the subtleties in the genomes of these F. proliferatum strains that enable them to occupy different habitats in nature.
This research has been partially funded by the Projects PI 2002/064 (Gobierno Autónomo de Canarias) and REN 2002–02732/TECNO (Ministerio de Ciencia y Tecnología) and Utah Agricultural Experiment Station funding to AJA.