INTRODUCTION
Species of theFusarium genus are major plant pathogens, and are frequentlyisolated from soil and vegetable material (1). F. solani andother species are considered important agents in emergingopportunistic mycosis, since they are responsible for serioussystemic infections in immunodepressed patients (2, 3).Fusarium antigens can induce the development ofhypersensitive respiratory processes, although their frequency (4,5), epidemiological characteristics and clinical relevance arestill unknown partly due to the insufficient knowledge of theseantigens and the absence of any adequate standardization (6) fortheir use in diagnostic tests.
There are no establishedmethods for the preparation and standardization of fungal antigens,even though they are commonly cultivated in a synthetic, liquidmedia, without macromolecules and with a controlled pH. Incubationtime, temperature, humidity and aeration should be adequate foreach cultivated species; in some cases it is convenient to shakethe culture constantly to prevent sporulation, while other authors(7) prefer static cultures in order to promote sporulation by usingagar-containing media incubated for long periods oftime.
The requirements foroptimum growth and controlled sporulation differ from one speciesto another. In order to obtain optimized antigens, which willprovide reproducible results, it is crucial to apply a correctstandardization of the employed methodology. This includes theculture media composition, the incubation time, shaking andtemperature, the method for separating the mould from its culturemedia, as well as the dialysis, concentration and purificationmethods. All of these parameters influence the presence andrelative content of certain antigens which play a determinant rolein the diagnostic performance of the resulting fungalextract.
The main objective ofthis study was to obtain an appropriate standardization for thepreparation of the different antigen types of F. solani, aswell as to acquire more data about their biochemical andimmunological characteristics.
MATERIALS ANDMETHODS
F. solanistrains
Eight strains ofdifferent origin were used: two from type collections (CBS 49063and the Colección Española de Cultivos Tipo, CECT2199) and the other five from human patients; four were isolatedfrom systemically infected patients from Brazil, and the six, whichwas isolated in our laboratory, from an onichomycosis (MRL95650).
Strains were cultured inpotato-agar (Potato Dextrose agar, Merck) and cultivated for oneweek at 30 °C. Once the purity of the culture was assured, amixture of mycelium and conidia was inoculated into each of the twoculture media employed to prepare the antigens.
Culturemedia
A conventional Czapek-Dox(Difco) liquid medium containing 1% (wt:wt) yeast extract wasmodified by adding glucose and salts (39 g Cazpek-Dox, 20 gglucose, 1 g yeast extract, 0.01 g zinc sulphate and 0.005 g coppersulphate).
Antigenpreparation
Mould spore suspensionswere inoculated in one litre Erlenmeyer flasks containing 300 ml ofculture medium and were incubated for 21 days at room temperature(25-27 °C) with constant stirring. Mycelia were then separatedby vacuum-filtration through a Büchner porcelain funnel. Thisraw material was used to prepare three different antigens:metabolic, hydrosoluble and somatic (Figs. 1 and 2).
Figure1.--Schematic diagrams of antigen production.
Figure2.--Electrophoresis in SDS-PAGE of the different hydrosolubleantigens after silver staining. Lane number corresponds to thefollowing F. solani strains: 1: 4578; 2: JM (modifiedmedia); 3: CBS (modified media); 4: CECT 2199 (modified media); 5:F. solani JM (Czapek); 6: CBS (Czapek); 7: CECT(Czapek).
The metabolic antigen wasobtained by first filtering the culture media through a 0.7µm-pore filter (Millipore). It was thenconcentrated with polyethylene glycol 6000 (PEG) at 4 °C anddialyzed against running tap water followed by distilled water (20volumes) overnight. The resulting product was freeze-dried andstored.
The hydrosoluble antigenwas obtained by extracting the mycelia with ammonium bicarbonate(0.1 g/ml) overnight. The resulting solution was centrifuged andconcentrated with polyethylene glycol 6000, followed by dialysisand lyophilization.
The somatic antigen wasobtained after cool homogenization of the mycelia at 20.000 rpm for10 min. The homogenized mass was sonicated for 10 min, followed byconcentration, dialysis and lyophilization.
Biochemicalcharacterization
Protein and carbohydratecontents were determined for each of the extracts using theBradford (9) and Dubois (10) methods, respectively. Verticalelectrophoresis was carried out on each extract under denaturingconditions, in polyacrylamide gels (12.5% acrylamide) with sodiumdodecyl sulphate (11) staining with Coomassie or silver solution(12).
Immunologicalevaluation
Rabbit antisera wereobtained from the metabolic, hydrosoluble and somatic extracts.Weekly immunization was performed by intramuscular and subcutaneousroutes for 2 months (13). A mixture of the same type of antigenobtained was used for immunization.
The agarose geldouble-diffusion technique (14) was used to assay the differentantigens and to detect cross-reactions.
Western blotting (15) wascarried out by electrical protein transfer onto 0.45 µm Immobilon-P membranes (Millipore). Membranes were thenincubated two hours with a 1:1000 dilution of rabbit polyclonalantiserum. Anti-rabbit IgG conjugated to peroxidase (1:3000dilution) and diaminobenzidine were used for immunochemicalstaining according to the manufacturer's instructions (Sigma, StLouis, USA). The resulting immunoblotting band pattern was analyzedusing the software package (1998) Bio-Rad Diversity Database. Aphylogenetic tree was constructed based on UPGAMA with the Jaccarcoefficient to detect antigen similarities.
RESULTS
From the three F.solani collection strains, nine antigens (metabolic,hydrosoluble and somatic) were prepared by culturing the antigensin conventional Czapek media. The same antigens were also preparedin modified Czapek media. Another 15 antigens were prepared frompatient-isolated strains, cultivated exclusively in modified Czapekmedia. The antigen yield was quite variable, having no relation tothe composition of the culture media nor to the strain used in theproduction.
It was observed thatsomatic antigens presented a relatively higher protein content thanthe other antigens. Moreover, protein content in the somaticantigens was higher when prepared with the conventional Czapekmedia (table I).
| Table I Protein (Prot) carbohydrate (Carb) concentrations inantigens of Fusarium solani. In micrograms per milligram ofdry weight of antigen according to the culture media and strainused | ||||||
| Fusarium Strains | Hydrosoluble | Metabolic | Somatic | |||
| Prot | Carb | Prot | Carb | Prot | Carb | |
ConventionalCzapek | ||||||
| CECT | 40.38 | 162.66 | 61.825 | 217.66 | 112.90 | 129.57 |
| CBS | 93.93 | 70.66 | 32.525 | 122.14 | 59.23 | 133.42 |
| JM | 48.4 | 65.5 | 33.5 | 176.17 | 73.74 | 131.77 |
| Modified Czapek | ||||||
| CECT2199 | 57.53 | 111.61 | 38.06 | 122.9 | 66.23 | 123.27 |
| CBS | ND | ND | 33.11 | 186.34 | 48.08 | 37.113 |
| JM | 59.97 | 43.15 | 37.35 | 193.72 | 29.722 | 80.60 |
| 4578 | 23.83 | 27.18 | 21.64 | 26.23 | 106.29 | 50.87 |
| 5233 | 28.72 | 40 | 38.02 | 168 | 23.21 | 49.67 |
| 4558 | 11.55 | 116.1 | 49.27 | 219.86 | 81.60 | 137.47 |
| 5747 | 55.00 | 251.23 | 15.50 | 184.23 | 135.30 | 191.35 |
| 11020 | 25.10 | 38.32 | 36.30 | 237.36 | 51.71 | 91.82 |
ND: Notdone. | ||||||
Antigens prepared fromstrains belonging to CBS and CECT yielded higher protein contents,as well as a greater number of electrophoretic bands (Fig.2).
The hydrosoluble antigenspresented the highest number of protein bands, whilst the somaticantigens presented the lowest (table II).
| Table II Bands detected by vertical SDS-PAGE electrophoresis(IEF) and immunoblotting (IB) of the different pools of antigens ofFusarium solani incubated with the homologousantisera | ||||||
| Hydrosoluble | Metabolic | Somatic | ||||
| EF | IB | EF | IB | EF | IB | |
| 161 | * | |||||
| 99 | * | |||||
| 90 | * | |||||
| 83-80 | * | 81-79 | ||||
| 74 | * | |||||
| 69-66 | * | 67 | 69-71 | |||
| 61 | * | |||||
| 60-57 | * | 60 | ||||
| 53 | * | |||||
| 48 | * | * | ||||
| 44-43 | * | * | * | |||
| 43-42 | * | |||||
| 41 | * | * | ||||
| 39-36 | * | 36 | 36 | 38 | 39 | |
| 35 | * | |||||
| 34-32 | * | 33 | 32 | |||
| 29 | * | * | * | |||
| 28 | * | * | ||||
| 27 | * | |||||
| 25 | * | |||||
| 24-23 | * | 24 | * | |||
| 22 | * | * | ||||
| 17 | * | |||||
| 16-15 | * | * | * | |||
| 14-13 | * | 14 | ||||
| 11 | * | * | ||||
| 10 | * | |||||
| 8 | * | * | ||||
Common bands were foundin all extracts, and were distributed among 15; 16-17; 23-25;36-38; 44-46 and 73-77 kDa molecular weight. Antigens obtained frompatient-isolated strains presented the higher molecular weightbands in common (116 and 97 kDa). In some extracts a broad andintense protein band could be observed at 53-64 kDa, which couldconsist of several fractions. Somatic antigen electrophoresispresented two bands at 60 and 53 kDa, while the metabolic antigensexhibited just one band of 57-60 kDa, and the hydrosoluble antigensshowed a band of 61-62 kDa. The results obtained by immunoblottingof antigens and their corresponding homologous antisera are shownin table II. Somatic antigens presented the smallest number ofprotein bands both in electrophoresis and inimmunoblotting.
The combined analysis ofelectrophoretic results obtained with the three antigens pools(metabolic, hydrosoluble and somatic) confirmed the existence ofthe following common bands: 64, 61-60, 53, 43, 30-27, 24-22 and15-14 kDa. When evaluating the electrophoresis and immunoblottings,three immunogenic bands were common for the three types of antigenswere detected at 39-35 kDa, 29-32 kDa and 15-16 kDa,respectively.
The graphicrepresentation of these results as dendograms showed very littlesimilarity between the different antigens (Fig. 3), although themetabolic antigen presented the highest similarity independently ofthe serum employed. The highest similarity was found between thehydrosoluble and metabolic antigens.
Figure3.--Dendograms showing the relationship between the different poolsof antigens of Fusarium solani and their antibodies.Abbreviations used: Ac: Antibody. Ag: Antigen. Hydro: Hydrosoluble.Met: metabolic. Som: Somatic.
The double-diffusion andimmunoblotting confirmed several cross-reactions among theantigens. Metabolic antigens showed the highest degree ofcross-reactivity, whilst the hydrosoluble antigens showed thelowest cross-reactivity despite presenting the highest number ofbands in electrophoresis and immunoblotting.
DISCUSSION
F. solani isconsidered the most pathogenic species of the Fusariumgenus. The scarcity of reports existing on antigens fromFusarium refer mainly to this species, which is alsoconsidered the most allergenic and capable of causinghypersensitivity in atopic subjects exposed to thismould.
Allergies due toFusarium are not frequently reported in epidemiologicalstudies. This could be due to the lack of systematic studies thatinclude extracts from this genus, mainly because of the absence ofstandardized and chemically and immunologically characterizedantigens (16).
Qualitative differenceswere found according to the culture media used, as the antigensobtained after culturing in conventional Czapek showed a higherdiversity in protein composition. Thus, it is recommended thatconventional Czapek media be used for the culture of F.Solani. Modified Czapek media was found to be more appropriatefor other moulds such as Aspergillus (8).
It has been demonstratedhere that the collection strains, although stored for long periodsof time, were adequate for the preparation of F. solaniantigens. Such storage times however would not be adequate for thepreparation of A. fumigatus antigens (17).
In allergologicaldiagnoses, the extract considered as optimum is that which containsthe most potent and common allergens (18). In order to prove thisfact, IgE antibodies purified by affinity columns have been used(19, 20). The low affinity of the antibodies found for someantigens could be attributed to an insufficient amount of proteinand/or to an excessive carbohydrate content. Nevertheless, Baer(20) suggests that proteins bound to carbohydrates could also beallergenic; therefore, a high carbohydrate content, such as thatfound in Fusarium antigens, does not necessarily affecttheir quality.
Verna, et al (7, 21)produced metabolic antigens of F. solani with 11-daycultures, based on the maximum protein content and number of bandsresolved by isoelectric-focusing. The methodology for antigenpreparation used by these authors differs from that used in thepresent work, since the composition of the culture media wasdifferent. The hydrosoluble extract was obtained from myceliacultivated in solid media where a maximum sporulation is produced.In the present study, cultivation took place in liquid media, sothe extraction of the hydrosoluble antigen was obtained frommycelia with a small number of spores. Generally, the carbohydratecontent in extracts prepared from spores and mycelia predominantlyshowed a protein/carbohydrate ratio of around 1:2. For metabolicextracts a 1:4 ratio was obtained, which is much higher than thepublished mean ratio of 1:1 (21). These proportions remainedconstant, regardless of the culture medium used.
As previously described(22), the results obtained by electrophoresis were complemented bythe immunoblotting techniques, which allowed the detection of someproteins that probably, due to their low concentration, are onlyvisible by an antigen-antibody reaction. In order to simplify thepresent study, mixtures of each type of antigen in the sameproportions, were pooled. This method could, however, mask someprotein fractions, which may be present in individualextracts.
Á protein band of64 kDa common to the three F. solani extracts was described(23). In the present study, such a band was not detected in any ofthe different pools of antigens, but it was identified in some ofthe individual antigens. Somatic antigens presented a lowerdiversity of bands while at the same time had the highest totalprotein content. In contrast, the hydrosoluble antigens with a lowprotein content showed the highest electrophoretic diversity. Ahigher protein content does not necessarily assure an adequatediversity and better quality of antigens. This divergence betweenprotein content and diversity indicates the need for a systematicanalysis of extracts.
Three common bands of15-16, 24-25 and 27-30 kDa were detected in all antigens,independently of the method of preparation and the strain used. Twoof these proteins, those detected at 15-16 and 27-30 kDa, were alsodetected by others (7, 21, 23). The greatest diversity of proteinswith antigenic activity was provided by the hydrosoluble andmetabolic extracts, and considering their high yields and ease ofproduction, these latter antigens would be the most suitable fordiagnostic purposes.
ACKNOWLEDGEMENTS
The authors are indebtedto their colleagues in Brazil (Dr. B. Wanke, Fundaçao O.Cruz, Rio de Janeiro) for the isolation of Fusarium strains.Thanks also to Mrs. Fina Genè and colleagues from MycologyLaboratory, Universitat Rovira i Virgili from Reus, Tarragona,Spain for the identification of all the Fusarium strainsused here.
