Aspergillus ibericus: a new species of section Nigri isolated from grapes

Aspergillus ibericus: a new species of section Nigri isolated from grapes

Rita SerraCentro de Engenharia Biologica, Universidade doMinho, Campus de Gualtar, Braga, 4710-057Portugal

F. Javier CabanesDepartment de Sanitat i d’Anatomia Animals,Universitat Autonoma de Barcelona, 08193 Bellaterra,Barcelona, Espana

Giancarlo PerroneInstitute of Sciences of Food Production, CNR, VialeEinaudi 5, 70125 Bari, Italia

Gemma CastellaDepartament de Sanitat i d’Anatomia Animals,Universitat Autonoma de Barcelona, 08193 Bellaterra,Espana

Armando Venancio1

Centro de Engenharia Biologica, Universidade doMinho, Campus de Gualtar, Braga, 4710-057Portugal

Giuseppina MuleInstitute of Sciences of Food Production, CNR, VialeEinaudi, 70125 Bari, Italia

Zofia KozakiewiczCABI Bioscience UK Centre, Bakeham Lane, Egham,Surrey TW20 9TY, United Kingdom

Abstract: As part of a study on the ochratoxinproducing mycoflora of grapes, several Aspergillusstrains were isolated and tested for their ochratoxin A(OTA) producing abilities. Aspergillus strains of thesection Nigri, which did not produce detectableamounts of OTA but which had a similar morphologyto A. carbonarius, were isolated from wine grapesand/or dried vine fruit in Portugal and Spain. Thesestrains, however, have characters that allow morpho-logical distinction from the other species in thesection, particularly the conidia size (5–7 mm), whichallows separation of the species from the two mostcommon biseriate species in section Nigri: A.carbonarius (7–9 mm) and A. niger and its aggregatespecies (3–5 mm). The strains are described here asbelonging to a new species, named A. ibericus. Thevalidation of this new taxon is supported further byanalysis of the ITS-5.8S rDNA and calmodulin genesequences and by analysis of the amplified fragmentlength polymorphism (AFLP) patterns, which were

consistent in separating these strains from otherspecies in the section. A. ibericus strains do notproduce OTA therefore they are interesting forbiotechnological exploration because many metabo-lites with commercial value are produced by otherspecies in the section.

Key words: A. carbonarius, A. niger, black asper-gilli, fungi, ochratoxin A, systematics

INTRODUCTION

Aspergillus species within section Nigri are importantin biotechnological processes as well as biodeteriora-tion. Species such as A. niger have a GRAS status fromthe FDA due to its extensive commercial and in-dustrial uses. Conversely species in the section haveproved to be particularly significant in the productionof ochratoxin A (OTA) (Abarca et al 1994, Horie1995) in several food commodities, particularly ingrapes and grape products (Abarca et al 2003;Battilani and Pietri 2002; Battilani et al 2003; Bau etal 2005a; Bellı et al 2004; Heenan et al 1998; Leong etal 2004; Magnoli et al 2003; Rosa et al 2002; Sage et al2002, 2004; Serra et al 2003, 2005b; Tjamos et al2004). A. carbonarius was highlighted as the mainspecies responsible for mycotoxin production ingrapes, with some researchers claiming that 100% ofthe strains have the ability to produce OTA underlaboratory conditions (Bau et al 2005a, b; Leong et al2004; Sage et al 2002, 2004; Serra et al 2003). OTAproduction also was found in other species, namely inA. niger but at a much lower rate (Abarca et al 2004).

Recent surveys on the mycoflora of distinct com-modities revealed new species of the section unknownto science, with some having the ability to produceOTA. The taxonomy of the section was revisedrecently (Abarca et al 2004), and 15 taxa wereaccepted in the section in the last critical revision(Samson et al 2004): A. aculeatus, A. brasiliensis ined.,A. carbonarius, A. costaricaensis, A. ellipticus, A.japonicus, A. foetidus, A. heteromorphus, A. homomor-phus, A. lacticoffeatus, A. niger, A. piperis, A.sclerotioniger, A. tubingensis and A. vadensis. Howeveronly four species in the section are relativelycommon: the uniseriate species A. aculeatus and A.japonicus and the biseriate species A. carbonarius andA. niger aggregate (with its two molecular types, A.niger and A. tubingensis). The remaining taxa arerarely reported or known only from type isolates.

In surveys in Europe to characterize OTA-pro-ducing mycoflora of wine grapes and dried vine fruit,

Accepted for publication 25 Feb 2006.1 Corresponding author. E-mail: [email protected]

Mycologia, 98(2), 2006, pp. 295–306.# 2006 by The Mycological Society of America, Lawrence, KS 66044-8897

295

Aspergillus strains were isolated, identified, tested forOTA production and characterized with three molec-ular tools: ITS-5.8S rDNA and calmodulin genesequencing and amplified fragment length polymor-phism (AFLP) analysis. Some strains initially identi-fied as A. carbonarius were unable to producedetectable OTA (Abarca et al 2003; Bau et al 2005b;Serra et al 2005a, b). These strains exhibited otherdifferences from A. carbonarius, both in morpholog-ical and molecular aspects.

In this report we describe a new black Aspergillusspecies, Aspergillus ibericus, and compare it withothers in the section.

MATERIALS AND METHODS

Fungal cultures.—Origin of the strains analyzed are asindicated (TABLE I). All strains except the referencestrains used for comparison were isolated from grapesor dried vine fruit in Europe. The strains are stored inthe CABI culture collection (IMI) and in other culture

collections (viz. Micoteca da Universidade do Minhoculture collection [MUM] and Culture Collection ofInstitute of Sciences of Food Production, Bari [ITEM]).

Media and growth conditions.—Czapek Dox agar (CZ)(Raper and Thom 1949), Czapek yeast extract (CYA),Blakeslee’s malt extract (MEA), and G25N agar (Pitt1979) were formulated as described. Czapek’s agar with20% sucrose (CZ20) and malt and yeast extract agarwith 40% sucrose (M40Y) were made as described(Raper and Fennell 1965). Colony transfers were madeaccording to the method of Pitt (1979). Growth rateswere studied and compared with reference strains asfollows: on CYA at 5 C, 25 C and 37 C, on G25N, CZ andMEA at 25 C in the dark, all for 7 d. Cultures also weregrown on CZ20 and M40Y for 7 d at 25 C in the dark toassess growth on media with reduced water activity.

SEM microscopy.—SEM micrographs of conidia wereobtained from CZ according to the method describedby Kozakiewicz for Aspergillus conidia (1989) (viz. byrubbing a standard SEM aluminium stub across a growingcolony, coating it with gold [SC502 Fisons Instruments

TABLE I. List of strains included in this study for ITS1-5.8S-ITS2 and partial calmodulin sequence analyses

Isolate numberGeographical

origin SourceITS-5.8S-ITS2

GenBank Calmodulin EMBL

A. ibericus IMI 391428 (5A-1082,ITEM 6600)

Spain Dried vine fruits AY656622 AJ971806

A-1229 Spain Dried vine fruits AY656623A-1604 Spain Wine grapes AY656624IMI 391429 (501UAs294,

MUM 03.49, ITEM 4776)Portugal Wine grapes AY656625 AJ971805

IMI 391430 (5MUM 03.50,ITEM 6601)

Portugal Wine grapes AY656626 a

IMI 391431 (5MUM 03.51,ITEM 6602)

Portugal Wine grapes AY656627 a

A. carbonarius IMI 016136 (T) unknown unknown AY656628 AJ964873IMI 041875 unknown unknown AY656629 bIMI 387223 Portugal Wine grapes bIMI 387242 Portugal Wine grapes AJ582714

A. niger IMI 050566 (T) USA Tannin-gallic acidfermentation

AY656630 AJ964872

A. niger IMI 091881 unknown unknown cA. ellipticus IMI 172283 (T) Costa Rica Soil AY656631 AM117809A. awamori IMI 211394 (T) Brazil AJ964874A. tubingensis IMI 172296 (T) unknown unknown AJ964876A. helicotrix IMI 278383 (T) The Netherlands Culture

contaminantAM117810

A. japonicus IMI 211387 Panama Soil AJ582717A. japonicus IMI 387343 Italy Wine grape AJ582716A. aculeatus IMI 211388 (T) unknown Tropical soil AJ964877A. foetidus IMI 15954 (T) unknown unknownA. phoenicis IMI 211395 (T) Japan Kuro-kojiAspergillus flavus NRRL 21882 unknown unknown AY974341

a identical to AJ971805.b identical to AJ964873.c identical to AJ964874.

296 MYCOLOGIA

Sputter Coater] and examining directly in a Leica Cam-bridge S360 microscope). For conidiophore observationthe strains were grown on CZ medium for 10 d, 25 C inthe dark. Agar blocks were cut and fixed on a 1% OsO4

solution in sodium cacodylate buffer (pH 7.3) overnight.After fixation the solution was removed by suction andsamples rinsed several times with distilled water. Sampleswere dehydrated through a graded series of ethanol (25–100%) and left to air dry. Mounted samples were coatedwith gold and examined immediately.

OTA assay.—Strains were evaluated with a previouslydescribed HPLC screening method (Bragulat et al2001). Briefly, the isolates were grown on CYA andincubated at 25 C, 30 C and 35 C for 5 d and 10 d. Fromeach isolate and at each sampling time three agar plugswere removed from different points of the colony andextracted with 0.5 mL methanol. The extracts werefiltered and injected into the HPLC. OTA detection,and quantification was made by a Waters LCM1chromatograph with a fluorescence detector Waters2475 (excitation wavelength: 330 nm/emission wave-length: 460 nm) and with a column C18 SpherisorbS5 ODS2, 250 3 4.6 mm. Twenty mL of each extractwere applied. The mobile phase, with a flow rate of1 mL/min, consisted of this linear gradient: acetoni-trile, 57%; water, 41%, and acetic acid, 2%. The extractswith the same retention time as OTA (ca 6.8 min)were considered positive. Confirmation was madethrough derivatization of OTA in its methyl-ester (Huntet al 1980). The detection limit of the extractionprocedure and the HPLC technique was 0.02 ng OTAand the quantification limit of HPLC technique withthe extraction procedure was 0.05 mg/g for thismycotoxin.

Molecular methods.—For ITS-5.8S rDNA sequencinganalysis fungal DNA was extracted as described byAccensi et al (1999). The strains were inoculated in1.5 mL Eppendorf tubes containing 500 mL of Sabour-aud broth (2% glucose, w/v; 1% peptone w/v)supplemented with chloramphenicol (1 mg/L) andincubated overnight in an orbital shaker at 300 rpmand 30 C. Mycelium was recovered after centrifugationand washed with NaCl 0.9% (w/v), frozen in liquidnitrogen and ground to a fine powder with a pipette tip.The powder was incubated 1 h at 65 C in 500 mLextraction buffer (Tris-HCl 50 mM, EDTA 50 mM, SDS3% and 2-mercaptoethanol 1%). The lysate was ex-tracted with phenol : chloroform (1 : 1, v/v), 3 MNaOAc and 1 M NaCl. DNA was recovered by isopro-panol precipitation. The pellet was washed with 70% (v/v) ethanol, dried under vacuum and resuspended in TEbuffer (Tris-HCl 10 mM, EDTA 1 mM, pH 8). DNA wascleaned with Geneclean kit II (BIO 101 Inc., La Jolla,California), according to the manufacturer’s instruc-tions. ITS rDNA and 5.8S rDNA were amplified asdescribed by Gene et al (1996) with a Perkin Elmer 2400thermal cycler. Primer pairs ITS5 and ITS4 weredescribed by White et al (1990). The amplificationprocess consisted of a predenaturation step at 94 C,

5 min, followed by 35 cycles of denaturation at 95 C/30 s, annealing at 50 C/1 min and extension at 72 C/1 min, plus a final extension of 7 min at 72 C. Themolecular masses of the amplified DNA were estimatedby comparison with the 100-bp DNA ladder (Bio-RadLaboratories S.A, Barcelona, Spain) standard lane. ThePCR product was purified with the GFX PCR DNA andgel band purification kit (Amersham Pharmacia Bio-tech, Uppsala, Sweden), following the supplier’s pro-tocol. Purified PCR products were used as a sequencingtemplate. The protocol BigDye Terminator v3.1 CycleSequencing kit (Applied Biosystems, Gouda, TheNetherlands) was used for sequencing. Primers ITS5and ITS4 described by White et al (1990) were used inthe sequencing reaction. An Applied Biosystems mod.3100 sequencer was used to obtain the DNA sequences.The sequences were aligned with Clustal X (1.81) of themultiple sequence alignment program (Thompson et al1994). Adjustments for improvement were made by eyewhere necessary. Cladistic analyses with the neighborjoining method (Saitou and Nei 1987) were performedwith the MEGA 2.1 computer program (Kumar et al2001) with Kimura 2-parameter model, includingtransitions and transversions and with pairwise deletionfor the treatment of the handling gaps/missing data.Confidence values for individual branches were de-termined by bootstrap analyses (1000 replications) andmaximum parsimony. The nucleotide sequences de-termined in this study have been deposited at theGenBank database and are identified by accessionnumbers (TABLE I).

For DNA amplification and sequencing of the partialcalmodulin gene and for fluorescent AFLP analysis, fungalstrains were grown in shake culture (150 rpm) in Wicker-ham’s medium (40 g glucose, 5 g peptone, 3 g yeast extract,3 g malt extract and water up to 1 L). About 40 mg offiltered, frozen and lyophilized mycelium from each strainwere used for total genomic DNA extraction with the EZNAFungal DNA Miniprep Kit (Omega Bio-tek, Doraville,Georgia). DNA was recovered and dissolved in sterile water.Concentrations of DNA were determined by gel electro-phoresis, by measuring the ultraviolet-induced fluorescenceemitted by ethidium bromide molecules intercalated intoDNA and comparing the fluorescent yield of the sampleswith a standard.

Amplifications of the partial calmodulin gene were set upwith 2.5 U of Taq Gold DNA polymerase (Applied Biosys-tems) in 100 mL reaction mixtures, containing 30 pmol ofeach outside primer, 12.5 mM of each deoxynucleosidetriphosphate (Applied Biosystems), and 1 mL (approximate-ly 10 ng) of fungal template DNA. All isolates wereamplified with primers CL1 and CL2A (O’Donnell et al2000). The reactions were performed under these PCRconditions: denaturation at 94 C for 10 min; 35 cycles ofdenaturation at 94 C for 50 s, annealing at 55 C for 50 s,extension at 72 C for 1 min; final extension at 72 C for7 min, followed by cooling at 4 C until recovery of thesamples. After PCR amplicons were purified with a centrif-ugal filter device (Millipore) and sequenced with AppliedBiosystems BigDye Terminator cycle sequencing kit in

SERRA ET AL: ASPERGILLUS IBERICUS 297

a 9700 GeneAmp PCR system. All sequencing reactionswere purified through Sephadex G-50 (Pharmacia) equili-brated in double-distilled water and analyzed on an ABIPRISM 310 Genetic Analyzer (Applied Biosystems). Theresulting regions sequenced of all the isolates were alignedby the Clustal method with the DNAMAN program(Lynnon BioSoft). The unique calmodulin sequences weredeposited at the EMBL nucleotide sequence database(TABLE I).

For preparation of AFLP template we used the AFLPMicrobial Fingerprinting Kit (Applied Biosystems-PerkinElmer Corp., Foster City, California) according to manu-facturer’s instructions. Approximately 10 ng of genomicDNA of each isolate was cut with EcoRI and MseI (NewEngland Biolabs, Hitchin, Hertfordshire, United Kingdom)and the DNA fragments were ligated to double-strandedrestriction site-specific adaptors from the kit. A preselectivePCR (72 C 2 min, 20 cycles of 94 C 20 s, 56 C 30 s, 72 C2 min and held at 4 C) was carried out in a 20 mL (finalvolume) mixture. For the selective PCR, 1.5 mL of a 1 : 20dilution of the first PCR reaction was amplified in a 10 mL(final volume) mixture with selective primers. Two separateprimer combinations were used: (i) EcoRI+AC andMseI+CC; (ii) EcoRI+AT and MseI+CG. EcoRI primerswere labeled with fluorescent dye (Applied Biosystems).The PCR program for selective AFLP amplification was: onecycle of 94 C for 2 min, one cycle of 94 C for 20 s, 66 C for30 s, and 72 C for 2 min; this cycle was followed by ninecycles in which the annealing temperature was lowered by 1C at each cycle from 65 C to 57 C., after which 20 cycles of94 C for 20 s, 56 C for 30 s, and 72 C for 2 min wereperformed, followed by a final extension step of 60 C for30 min, then held indefinitely with a model 9700 GeneAmpPCR system.

After amplification 1 mL of reaction product was mixedwith 20 mL formamide and 0.5 mL GeneScan-500 (ROX)size standard (Applied Biosystems), of 35–500 bp long. Themix was heated 2 min at 95 C and snap-cooled on ice. Theproduct was separated by capillary electrophoresis on anABI PRISM 310 Genetic Analyzer (Applied Biosystems).After electrophoresis the pattern was extracted withGeneScan collection version 3.1.2 software (Applied Biosys-tems) and the fingerprints were analyzed with Genotypersoftware (Applied Biosystems). DNA samples from the fiveA. ibericus strains were tested in triplicate, and DNA samplesfrom other strains were tested in duplicate. DNA from threereplicate cultures of five strains also was tested.

Peak height thresholds were set at 200. Genotypersoftware (Applied Biosystems) was set to medium smooth-ing. Bands of the same size in different individuals wereassumed to be homologous and to represent the sameallele. Bands of different sizes were treated with AFLPmanager database developed by ACGT BioInformatica S.r.1.(via Principe Amedeo 347- 70100 BARI) and were exportedin a binary format with ‘‘1’’ for the presence of a band/peakand ‘‘0’’ for its absence. For clustering fragments of 50–500 bp were analyzed with NTSYS software with the Dicesimilarity coefficient based on presence/absence of thebands and clustered by the unweighted pair group method(UPGMA) (Nei and Li 1979).

TAXONOMY

Examination of morphological characters combinedwith the analysis of OTA production and moleculardata revealed that the Aspergillus Section Nigri strainsisolated do not match other closely related species ofthe section. Therefore a new species, A. ibericus, isproposed.

Aspergillus ibericus Serra, Cabanes et Perrone sp. nov.Coloniae in agaro Czapekii celeriter crescentes, in

septem dies 25 C, 38–43 mm diam attigentes,granulosae, superficie nigra, facie inferiore primoalba, deinde obscure ochraceae; exudata non con-spicuosa. Conidiogenesis abundata. Capitula conidicaglobosa, in columnas aegre formatas fissantia, 500–600 mm diam. Stipites leves incolores, superne pallideochracei, 1200–2000 mm longissimi, parie plerumque14–20 mm crassis; vesiculae globosae 50–60 mm diam,per toto fertilibus. Aspergilla biseriata; metulis etphialidibus 30.0–40.0 3 5.0–7.5 mm, phialides 8.0–10.0 3 6.0–7.0 mm. Conidia nigra oculo nudo visa,globosa vel subglobosa, 5.0–7.0 mm, conspicue verru-culosa in maturitate et spinis 1.0 mm projecti.

Colonies on Czapek agar growing rapidly, attaining38–43 mm diam within 7 d at 25 C, granular, uppersurface black (FIG. 1); reverse white, wrinkled, be-coming dull yellow with age; exudates not conspicu-ous. Conidiogenesis abundant. Conidial heads glo-bose (FIG. 2) splitting into poorly defined columnswith age, 500–600 mm diam. Stipes smooth, thick-walled, 1200–2000 3 14–20 mm, uncolored, upperportion light yellowish-brown; vesicles globose, 50–60 mm, fertile over entire surface. Aspergilli, biseriate(FIG. 3), with lightly packed metulae and phialides;metulae 30–40 3 5.0–7.5 mm; phialides 8.0–10.0 3

6.0–7.0 mm. Conidia black when seen with the nakedeye, globose to subglobose, 5.0–7.0 mm, conspicuouslyverruculose at maturity, with spines projecting up to1.0 mm (FIGS. 4–6).

HOLOTYPE: Dried colonies of IMI 391429, after7 d growth on CZ, deposited in the Herbarium ofCABI Bioscience, Egham, UK, formerly InternationalMycological Institute. Ex-type culture: IMI 391429.

Etymology. ibericus, from the Iberian Peninsula.Known distribution. Iberian Peninsula, Europe.Origin of strains. HOLOTYPE. PORTUGAL.

ALENTEJO, Evora. Vineyard at 38u339380N,7u549300W, on healthy grapes, 4 Oct 2001, RitaSerra 01UAs294.

Additional specimens examined. IMI 391430 (MUM03.50), PORTUGAL. ALENTEJO, Evora, from healthygrapes, 25 Aug 2003, Rita Serra 03UAs89; IMI 391431(MUM 03.51), DOURO, Pinhao, from healthy grapes, 1 Oct2003, Rita Serra, 03UAs254.

298 MYCOLOGIA

IMI 391428 (CCFVB A1082. CCFVB: Collection ofthe Veterinary Faculty of Barcelona) SPAIN. Fromraisins collected in a Spanish market survey, Dec 2000,M.L. Abarca; CCFVB A1229. SPAIN. From raisinscollected in a Spanish market survey, Oct 2001, F.J.

Cabanes; CCFVB A1604. SPAIN. JEREZ de la Fronteravineyard, from healthy grapes, Aug 2003, F. J.Cabanes.

Colonies grown 7 d on CYA overgrew a 90 mmplate (TABLE II). Conidiogenesis is abundant. Co-

FIGS. 1–6. Aspergillus ibericus. 1. Colony grown in CZ (9 d). 2. Biseriate aspergilli of a 4 d old culture in CZ (bar 5 10 mm).3. Aspergilli at SEM (bar 5 200 mm). 4. Conidia seen at Nomarski microscope (bar 5 10 mm). 5. SEM picture of the conidiawith variable ornamentation at different maturation stages (bar 5 20 mm). 6. SEM picture of a mature conidia (bar 5 2 mm).

SERRA ET AL: ASPERGILLUS IBERICUS 299

nidial heads black, mycelium white, usually incon-spicuous; exudates inconspicuous; reverse pale or dullyellow. When grown at 37 C on CYA, colonies wereheavily overgrown and were identical to coloniesgrown at 25 C. When grown on G25N coloniesattained (12–)15–18(–23) mm diam but otherwisewere identical to those on CYA at 25 C. No growth orgermination was observed at 5 C. Growth in CZ20 was44–46 mm diam in 7 d, with abundant conidiogen-esis. Colonies were identical to those on CYA at 25 C,apart from reverse, which was yellow. On M40Ycolonies were overgrown (90 mm plates) in 7 d at25 C, otherwise identical to those on CZ20. Colonies

on MEA (36–)38–43(–51) mm in 7 d at 25 C; conidiablack, mycelium inconspicuous or as a white basal felt;reverse uncolored; otherwise as on CYA.

RESULTS

OTA assay.—It proved negative for all the A.ibericus strains studied under experimental condi-tions. Conversely all A. carbonarius strains in thisstudy produced the mycotoxin at detectable levels,from 18 mg/g to 42 mg/g of culture medium.

Molecular analyses.—A phylogenetic tree (FIG. 7)

FIG. 7. Neighbor joining tree based on phylogenetic analysis of the ITS1-5.8S rRNA gene-ITS2 sequences. The numbers atbranch points are the percentages of 1000 bootstrapped datasets that supported the specific internal branches. Species withGenBank numbers represent sequences obtained from GenBank.

TABLE II. Growth rates (mm) of black Aspergillus species at different culture media and temperatures (n.g. 5 no growth).The main differences in growth rates of A. ibericus and A. carbonarius were found in CYA at 37 C

Species Strain code

CYA

MEA G25N CZ CZ20 M40Y5 C 25 C 37 C

A. ibericus IMI 391428 n.g. 63–70+ 63–70+ 56–60 13–17 38–43 45–46 63–70+A. ibericus IMI 391429 n.g. 63–70+ 63–70+ 53–55 15–17 (31–)38–41 43–46(–52) 63–70+A. ibericus IMI 391430 n.g. 63–70+ 63–70+ 47–51 23–25 44–48 50–52 63–70+A. ibericus IMI 391431 n.g. 63–70+ 63–70+ 58–63 17–18 49–53 49–53 63–70+A. carbonarius IMI 390417 n.g. 63–70+ (46–)49–53 (37–)42–44 17–22 34–36(–40) (37–)42–45 63–70+A. carbonarius (T) IMI 016136 n.g. 63–70+ (34–)45–49 42–45 17–20 34–37 43–49 63–70+A. ellipticus (T) IMI 172283 n.g. 51–54 n.g. 63–70+ 13–14 19–21 30–35 63–70+

300 MYCOLOGIA

was generated with the neighbor joining method.The section of DNA sequenced from A. carbonariusincluded 612 base pairs. The ITS1 region occu-pied nucleotides 57–240, the 5.8S rDNA genefrom nucleotides 241–397 and the ITS2 fromnucleotides 398–566. The section of DNA se-quenced from A. ibericus included 614 base pairs.The ITS1 region occupied nucleotides 57–241, the5.8S rDNA gene from nucleotides 242–398 andthe ITS2 from nucleotides 399–567.

Strains MUM03.49 and MUM03.50 showed identi-cal ITS1-5.8S-ITS2 sequences and differed from thesequence of A. carbonarius IMI 016136 (type strain) atfive nucleotide positions, one at ITS1 region (aninsertion at position 106), another one at 5.8S region(a T to C transition at position 366) and three at ITS2region (a G to T transversion at position 442, a T to Ctransition at position 445 and an insertion at position562). The sequences of strains IMI 391428, A-1229and A-1604 were identical and differed from thesequence of A. carbonarius IMI 016136 at fournucleotide positions, one at ITS1 region (an insertionat position 106) and three at ITS2 region (a G to Ttransversion at position 442, a T to C transition atposition 445 and an insertion at position 562). Thesequence of MUM 03.51 strain differed from A.carbonarius IMI 016136 sequence at only threenucleotides, one at ITS1 region (an insertion at

position 106) and two at ITS2 region (a G to Ttransversion at position 442, a T to C transition atposition 445).

We identified a significant number of differenceswhen the ITS1-5.8S-ITS2 sequences of the A. ibericusstrains were compared with the corresponding se-quences of other black Aspergillus species. Dissimila-rities between A. ibericus strains and A. niger IMI050566 and A. ellipticus IMI 172283 in their ITS1-5.8S-ITS2 sequences were 3.7% (23 of 614 nucleotides)and 5.5% (34 of 614 nucleotides) respectively.

The PCR product of the partial calmodulin genewas a fragment of about 660 bps long for the analyzedstrains. The alignment of the sequences of the four A.ibericus strains showed a high similarity 99–100%, onlythe strain IMI 391428 differs in the calmodulinsequences from the other three A. ibericus strainsfor eight nucleotide positions (FIG. 8). The clado-gram indicates that there are four major clades insection Nigri. One of these clades includes twosubclades in which the A. carbonarius and A. ibericusstrains are clearly separated. In fact the A. ibericussubclade resulted in clear separation from the A.carbonarius reference strain (IMI 16136) with a simi-larity of 88% which led to a difference of 78–80 bps inthe partial calmodulin sequences alignment betweenthe A. carbonarius and A. ibericus strains. On theother hand the A. ibericus clade, although close to A.

FIG. 8. Homology tree obtained by comparison of partial calmodulin gene sequences. The dendrogram obtained clearlyseparated the four atypical strains (A. ibericus) from Aspergillus carbonarius strains and also from other closely related species.

SERRA ET AL: ASPERGILLUS IBERICUS 301

carbonarius, was clearly separated from the A. nigeraggregate (similarity .80%) and also from the otherspecies of Aspergillus Section Nigri (similarity of72%). Aspergillus flavus (NRRL 21882) was used asoutgroup and showed a similarity of 67% with theAspergillus section Nigri cluster.

The UPGMA dendrogram (FIG. 9) calculated fromthe AFLP fingerprints obtained from differentAspergillus strains clearly differentiated the four A.ibericus strains from the IMI referenced strains of A.carbonarius. Similarity among the four A. ibericusstrains is 52% in AFLP profile; strains MUM 03.49 andIMI 391428 showed a higher variability while twostrains, MUM 03.50 and 03.51, were more similar andconstitute a subclade at a similarity of 72%. On theother hand the low similarity of these strains incomparison with all other species in the section Nigri(with a range of 19–23% in the dendrogram) issignificant for differentiation at species level.

Members of this new biseriate species form a wellsupported clade using comparative analysis with thethree molecular techniques. This clade is close to A.carbonarius and clearly separated from the A. nigeraggregate and the rest of the species belonging toAspergillus section Nigri.

DISCUSSION

A. ibericus is most similar phenotypically to A.carbonarius. Both species have black biseriate asper-gilli with long stipes and relatively large conidia (.

5 mm) when compared to the species of the A. niger

aggregate (3–5 mm). Of the 15 species in the sectionNigri accepted by Samson et al (2004) in the lastrevision, only two taxa had conidia of more than 6 mmdiam, namely A. carbonarius and the rare species A.ellipticus. Nevertheless the size, shape and ornamen-tation of A. ibericus conidia are different (FIGS. 10–15). A. carbonarius (FIG. 10) has the largest echinu-late conidia of all the black aspergilli; A. ellipticus(FIG. 11) has elliptical conidia echinulate of 7–9(–10)mm; A. ibericus (FIGS. 12, 13) has smaller echinulateconidia of 5–7 mm diam.

Other characters may help in the distinction ofboth species (e.g. colony reverse) The colony reverseof A. carbonarius on CZ usually has a dark or olivegreen center, while the reverse of A. ibericus is palethroughout. Furthermore, on MEA and CZ, A.ibericus grows faster than A. carbonarius (TABLE II).

Regarding the distinction between A. niger aggre-gate and A. ibericus, apart from conidia size, theornamentation of the conidia is a useful characteris-tic. A. niger aggregate conidia are verrucose (FIGS. 14,15) (Kozakiewicz 1989, Varga et al 2000) andechinulate in A. ibericus (FIGS. 12, 13).

The taxonomy of this section has received muchattention from molecular analysis and from extroliteprofiles, particularly for mycotoxin production by thespecies (Samson et al 2004). A. ibericus strains isolatedto date are unable to produce detectable OTA inculture media, unlike A. carbonarius, in which allstrains are strong OTA producers (Bau et al 2005a, b;Cabanes et al 2002; Leong et al 2004; Sage et al 2002,2004; Serra et al 2003, 2005b). Some researchers

FIG. 9. UPGMA dendrogram assessed from the comparison of AFLP fingerprints generated with primer A. Fragments of50–500 bp are shown. The Dice similarity index was used for similarity calculation. Numbers on the tree indicate thepercentage of similarity according to Dice. The dendrogram clearly differentiate the four atypical A. carbonarius strains (A.ibericus) from the type-strain of A. carbonarius (IMI 16136-ITEM 4503).

302 MYCOLOGIA

claim that certain A. carbonarius strains do notproduce OTA (Battilani et al 2003, Rosa et al 2002),but this claim must be investigated further. Bau et al(2005a, b) did not confirm these findings, stating that100% of the A. carbonarius strains tested produceOTA, apart from those here now described as A.ibericus.

A summary of the characters of A. ibericus

compared with other species in section Nigri is listed(TABLE III). A. carbonarius and A. ibericus strains canbe distinguished based on two characteristics: (i)conidia size, with A. ibericus having smaller conidia,5–7 mm, compared with A. carbonarius, 7–9(–10) mm;(ii) secondary metabolite profile, with none of the A.ibericus strains tested producing OTA, while all A.carbonarius are recognized as good OTA producers. A.

TABLE III. Some characteristics of the main black Aspergillus species and A. ibericus

Species Conidiophore Conidia size (mm) OTA production

A. japonicus/A. aculeatus Uniseriate 4–5 NegativeA. niger aggregate Biseriate 3–5 Positive (low %)A. sclerotioniger Biseriate 5–6 PositiveA. carbonarius Biseriate 7–9 Positive (high %)A. ibericus Biseriate 5–7 Negative

FIGS. 10–15. SEM pictures of conidia of black Aspergillus species. 10. A. carbonarius (bar 5 4 mm). 11. A. ellipticus (bar 5

2 mm). 12, 13. A. ibericus (bar 5 2 mm). 14, 15. A. niger aggregate (bar 5 2 mm).

SERRA ET AL: ASPERGILLUS IBERICUS 303

sclerotioniger is a recently described species isolatedfrom coffee beans related to A. carbonarius (Samson etal 2004) but with relatively smaller conidia (5–6 mm).A. ibericus can be distinguished from A. sclerotionigerbased on cultural traits, with A. sclerotioniger producingabundant yellow to red-brown sclerotia in CYA andMEA and by the secondary metabolite profile, becauseA. sclerotioniger produces OTA.

A. ibericus isolates phylogenetically form a wellsupported clade with comparative ITS-5.8S rRNAsequencing analysis. This clade is close to A.carbonarius and clearly separated from the A. nigeraggregate and the rest of the species belonging toAspergillus section Nigri (Abarca et al 2004). Thesequenced region is conserved within the speciesbelonging to Aspergillus section Nigri. Accensi et al(1999) described nearly identical sequences for A.niger and A. tubingensis. Parenicova et al (2001) alsoanalyzed the ITS1-5.8S-ITS2 region of the closelyrelated species A. japonicus and A. aculeatus. Theyfound nucleotide exchanges at only three positions.High sequence similarity of this DNA region betweenrelated species also has been demonstrated by otherauthors. OTA producing penicillia have been reclas-sified recently into two species, Penicillium verrucosumand Penicillium nordicum, based on their differentchemotypes (Larsen et al 2001). The ribosomal ITS1-5.8S-ITS2 sequences of these species were similar,except for two single nucleotide exchanges in severalstrains (Castella et al 2002).

Calmodulin results give a similar confirmation ofthe uniqueness of A. ibericus relative to othermembers of section Nigri. Both genes analyzed(rDNA and calmodulin) for sequence identity re-vealed a higher homology of A. ibericus strains to A.carbonarius strains (99% and 88% respectively) thanto other black aspergilli (FIGS. 7, 8). In other studiesthe calmodulin gene has been shown to be highlyreliable for phylogenetic analysis on the Gibberellafujikuroi complex and Fusarium related species(O’Donnell et al 2000), a candidate gene forpopulation genetic analysis (Geiser et al 2000) anduseful to discriminate among the Aspergillus ofsection Nigri (Perrone et al 2004). The AFLP profilesevidenced that A. ibericus strains showed a similarity. 50% among them when analyzed by AFLP, whilethe homology to the other species was 20%; thismeans that they must belong to a separate new specieswithin the section Nigri. AFLP analysis results,supported also by DNA sequence analysis of di-agnostic genes, confirmed the validity of this tech-nique in analyzing genetic relatedness among fungalspecies (Zeller et al 2003).

The three molecular techniques clearly differenti-ate these strains from A. carbonarius and together

with the morphological and biochemical differencesunequivocally underline that these strains representa new species in section Nigri.

In grapes A. ibericus is not significant regardingOTA production because the strains do not producedetectable OTA in laboratory media and in grapejuice-based medium. The species responsible for OTAproduction in grapes is A. carbonarius. In Portuguesevineyards A. ibericus strains were less frequentlyisolated compared with A. niger aggregate and A.carbonarius isolates. They were isolated from five outof 17 vineyards studied, colonizing 2–8% of thehealthy berries in the samples. A. ibericus strains werenot isolated from rotten berries. Nevertheless itspathogenic capabilities must be evaluated.

In other European vineyards from France, Greece,Italy, Spain (and also from Israel), A. ibericus strainswere much less frequently isolated than A. nigeraggregate and A. carbonarius strains (Bau et al2005b). Furthermore in previous studies A. ibericus(IMI 391428) was not able to produce OTA at anytemperature (Esteban et al 2004) or pH value(Esteban et al 2005) tested.

Aspergillus species in section Nigri are widely usedin industrial processes and have several biotechno-logical applications (Roehr et al 1992). Because A.niger is a GRAS organism and recent studies haveshown that some isolates are capable of producingOTA in low concentrations, a more careful study intocommercially used strains is advocated. The fact thatstrains of A. ibericus do not produce detectable OTAin culture media is an advantage for further studieson its potential biotechnological applications.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the support of the EC,Quality of Life Programme (QoL), Key Action 1 (KA1) onFood, Nutrition and Health; contract number QLK1-CT-2001-01761, Wine-Ochra Risk. Rita Serra was supported bygrant SFRH/BD/1436/2000 from Fundacao para a Cienciae Technologia.

LITERATURE CITED

Abarca ML, Bragulat MR, Castella G, Cabanes FJ. 1994.Ochratoxin A production by strains of Aspergillus nigervar. niger. Appl Environ Microbiol 60:2650–2652.

———, Accensi F, Bragulat MR, Castella G, Cabanes FJ.2003. Aspergillus carbonarius as the main source ofochratoxin A contamination in dried vine fruits fromthe Spanish market. J Food Protect 66:504–506.

———, ———, Cano J, Cabanes FJ. 2004. Taxonomy andsignificance of black aspergilli. Antonie van Leeuwen-hoek Int J Gen Mol Microbiol 86:33–49.

Accensi F, Cano J, Figuera L, Abarca ML, Cabanes FJ. 1999.

304 MYCOLOGIA

New PCR method to differentiate species in theAspergillus niger aggregate. FEMS Microbiol Let 180:191–196.

Battilani P, Pietri A. 2002. Ochratoxin A in grapes and eine.Euro J Plant Pathol 108:639–643.

———, ———, Bertuzzi T, Languasco L, Giorni P,Kozakiewicz Z. 2003. Occurrence of ochratoxin A-producing fungi in grapes grown in Italy. J FoodProtect 66:633–636.

Bau M, Bragulat MR, Abarca ML, Minguez S, Cabanes FJ.2005a. Ochratoxigenic species from Spanish winegrapes. Int J Food Microbiol 98:125–130.

———, Castella G, Bragulat MR, Cabanes FJ. 2005b. DNA-based characterization of ochratoxin-A-producing andnon-producing Aspergillus carbonarius strains fromgrapes. Res Microbiol 156:375–381.

Bellı N, Pardo E, Marın S, Farre G, Ramos AJ, Sanchis V.2004. Occurrence of ochratoxin A and toxigenicpotential of fungal isolates from Spanish grapes. J SciFood Ag 84:541–546.

Bragulat MR, Abarca ML, Cabanes FJ. 2001. An easyscreening method for fungi producing ochratoxin Ain pure culture. Int J Food Microbiol 71:139–144.

Cabanes FJ, Accensi F, Bragulat MR, Abarca ML, Castella G,Minguez S, Pons A. 2002. What is the source ofochratoxin A in wine? Int J Food Microbiol 79:213–215.

Castella G, Larsen TO, Cabanes FJ, Schmidt H, Alboresi A,Niessen L, Farber P, Giessen R. 2002. Molecularcharacterization of ochratoxin A producing strains ofthe genus Penicillium. Syst App Microbiol 25:74–83.

Esteban A, Abarca ML, Bragulat MR, Cabanes FJ. 2004.Effects of temperature and incubation time on pro-duction of ochratoxin A by black aspergilli. ResMicrobiol 155:861–866.

———, Abarca ML, Bragulat MR, Cabanes FJ. 2005.Influence of pH and incubation time on ochratoxinA production by Aspergillus carbonarius in culturemedia. J Food Protect (In press).

Geiser DM, Harbinski FM, Taylor JW. 2000. Molecular andanalytical tools for characterising Aspergillus andPenicillium species at the intra- and interspecific levels.In: Samson RA, Pitt JI, eds. Integration of moderntaxonomic methods of Penicillium and Aspergillusclassification. Amsterdam, the Netherlands: HarwoodAcademic Publishers. p 381–394.

Gene J, Guillamon JM, Guarro J, Pujol I, Ulfig K. 1996.Molecular characterization, relatedness and antifungalsusceptibility of the basidiomycetous Hormographiellaspecies and Coprinus cinereus from clinical andenvironmental sources. Antonie van Leeuwenhoek IntJ Gen Mol Microbiol 70:49–57.

Heenan CN, Shaw KJ, Pitt JI. 1998. Ochratoxin A pro-duction by Aspergillus carbonarius and A. niger isolatesand detection using coconut cream agar. J Food Mycol1:67–72.

Horie Y. 1995. Productivity of ochratoxin A of Aspergilluscarbonarius in Aspergillus section Nigri. Nippon King-akukai Kaiho 36:73–76.

Hunt DC, McConnie BR, Crosby NT. 1980. Confirmation ofochratoxin A by chemical derivatisation and high-

performance liquid chromatography. Analyst 105:89–90.

Kozakiewicz Z. 1989. Aspergillus on stored products. Wall-ingford, UK: CABI Publishing: Mycol Paper vol 161:1–188.

Kumar S, Tamura K, Jakobsen IB, Nei N. 2001. Mega2:molecular evolutionary genetics analysis software. In-formatics 17:1244–1245.

Larsen TO, Svendsen A, Smedsgaard J. 2001. Biochemicalcharacterization of ochratoxin A producing strains ofthe genus Penicillium. App Environ Microbiol 67:3630–33635.

Leong S, Hocking AD, Pitt JI. 2004. Occurrence of fruit rotfungi (Aspergillus section Nigri) on some dryingvarieties of irrigated grapes. Australian J Grape WineRes 10:83–88.

Magnoli C, Violante M, Combina M, Palacio G, Dalcero A.2003. Mycoflora and ochratoxin-producing strains ofAspergillus section Nigri in wine grapes in Argentina.Let App Microbiol 37:179–184.

Nei M, Li WH. 1979. Mathematical model for studyinggenetic variation in terms of restriction endonucleases.Proceed Nat Acad Sci USA 76:5269–5273.

O’Donnell K, Niremberg HI, Aoki T, Cigelnik E. 2000. Amultigene phylogeny of the Gibberella fujikuroi speciescomplex: detection of additional phylogeneticallydistinct species. Mycoscience 41:61–78.

Parenicova L, Skouboe P, Frisvad JC, Samson RA, Rossen L,Hoor-Suykerbuyk MT, Visser J. 2001. Combined mo-lecular and biochemical approach identifies Aspergillusjaponicus and Aspergillus aculeatus. App EnvironMicrobiol 67:521–527.

Perrone G, Susca A, Stea G, Mule G. 2004. PCR assay foridentification of Aspergillus carbonarius and Aspergillusjaponicus. Euro J Plant Path 110:641–649.

Pitt JI. 1979. The genus Penicillium and its teleomorphicstates Eupenicillium and Talaromyces. London, UK:Academic Press.

Raper KB, Thom C. 1949. A manual of the Penicillia.Baltimore: Williams & Wilkins.

———, Fennell DI. 1965. The genus Aspergillus. Baltimore:Williams & Wilkins.

Roehr M, Kubicek CP, Kominek J. 1992. Industrial acids andother small molecules. In: Bennett JW, Klich MA, eds.Aspergillus: biology and industrial applications. Boston,USA: Butterworth-Heinemann. p 91–131.

Rosa R, Palacios V, Combina M, Fraga ME, Oliveira ReksonA, Magnoli CE, Dalcero AM. 2002. Potential ochratoxinA producers from wine grapes in Argentina and Brazil.Food Add Contamin 19:408–414.

Sage L, Krivobok S, Delbos E, Seigle-Murandi F, Creppy EE.2002. Fungal flora and ochratoxin A production ingrapes and musts from France. J Ag Food Chem50:1306–1311.

———, Garon D, Seigle-Murandi F. 2004. Fungal microfloraand ochratoxin A risk in French vineyards. J Ag FoodChem 52:5764–5768.

Saitou N, Nei M. 1987. The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Mol BiolE 4:406–425.

SERRA ET AL: ASPERGILLUS IBERICUS 305

Samson RA, Houbraken J, Kuijpers A, Frank JM, Frisvad JC.2004. New ochratoxin A or sclerotium producing speciesin Aspergillus section Nigri. Stud Mycol 50:45–61.

Serra R, Abrunhosa L, Kozakiewicz Z, Venancio A. 2003.Black Aspergillus species as ochratoxin A producers inPortuguese wine grapes. Int J Food Microbiol 88:63–68.

———. 2005a. Micoflora das uvas portuguesas e seupotencial para contaminacao das uvas com micotox-inas, com destaque para a ocratoxina A [Doctoraldissertation]. Braga: Univ. Minho. 399 p.

———, Braga A, Venancio A. 2005b. A. Mycotoxin pro-ducing and other fungi isolated from grapes for wineproduction with particular emphasis on ochratoxin A.Res Microbiol 156:515–521.

Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F,Higgins DG. 1994. The Clustal X windows interface:flexible strategies for multiple sequence alignmentaided by quality analysis tools. Nucl Acid Res 24:4876–4882.

Tjamos SE, Antoniou PP, Kazantzidou A, Antonopoulos DF,Papageorgiou I, Tjamos EC. 2004. Aspergillus niger and

Aspergillus carbonarius in Corinth raisin and wine-producing vineyards in Greece: population composi-tion, Ochratoxin A production and chemical control. JPhytopathol 152:250–255.

Varga J, Kevei F, Hamari Z, Toth B, Teren J, Croft JH,Kozakiewicz Z. 2000. Genotypic and phenotypic vari-ability among black aspergilli. In: Samson RA, Pitt JI,eds. Integration of modern taxonomic methods forPenicillium and Aspergillus classification. Amsterdam,the Netherlands: Harwood Academic Publishers.p 397–411.

White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification anddirect sequencing of fungal ribosomal RNA genes forphylogenetics. In: Innis MS, Gelfand DH, Sninsky JJ,White TJ, eds. PCR protocols, a guide to methods andapplications. San Diego, USA: Academic Press. p 315–322.

Zeller KA, Summerell BA, Bullock S, Leslie JF. 2003.Gibberella konza (Fusarium konzum) sp. nov. fromprairie grasses, a new species in the Gibberella fujikuroispecies complex. Mycologia 95:943–954.

306 MYCOLOGIA