m y c o l o g i c a l r e s e a r c h 1 1 3 ( 2 0 0 9 ) 1 1 3 7 – 1 1 4 5

journa l homepage : www.e l sev i er . com/ loca te /mycres

The ITS region as a taxonomic discriminator betweenFusarium verticillioides and Fusarium proliferatum

I. VISENTINa, G. TAMIETTIa, D. VALENTINOa, E. PORTISb, P. KARLOVSKYc,A. MORETTId, F. CARDINALEe,*aDiVaPRA, Plant Pathology, University of Turin, via L. da Vinci 44, 10095 Grugliasco (TO), ItalybDiVaPRA, Plant Genetics and Breeding, University of Turin, via L. da Vinci 44, 10095 Grugliasco (TO), ItalycMolecular Phytopathology and Mycotoxin Research Unit, University of Gottingen, Grisebachstrasse 6, D-37077 Gottingen, GermanydISPA, National Research Council, Via Amendola, 122/O – 70126 Bari (BA), ItalyeDept. of Arboriculture, University of Turin, via L. da Vinci 44, 10095 Grugliasco (TO), Italy

a r t i c l e i n f o

Article history:

Received 6 April 2009

Received in revised form 15 July 2009

Accepted 16 July 2009

Available online 23 July 2009

Corresponding Editor:

Stephen W. Peterson

Keywords:

Fusarium proliferatum

Fusarium verticillioides

ITS

Maize

PCR

* Corresponding author. Tel.: þ39 011 670 8E-mail address: francesca.cardinale@unit

0953-7562/$ – see front matter ª 2009 The Bdoi:10.1016/j.mycres.2009.07.011

a b s t r a c t

The maize pathogens Fusarium verticillioides (Fv) and Fusarium proliferatum (Fp) are morpho-

logically very similar to one another, so Fp isolates have been often mistaken as Fusarium

moniliforme (the former name of Fv). The only presently accepted morphological discrimina-

tor between these species is the presence/absence of polyphialides. Here, a collection of 100

Fusarium strains, isolated from infected maize kernels on plants grown in north-western

Italy, were assigned as Fv or Fp on the basis of the presence/absence of polyphialides.

This classification was tested on a subset of isolates by sexual crosses, ITS and calmodulin

sequencing and AFLP profiling. An ITS-RFLP assay was extended to the full collection and to

a number of Fv and Fp isolates of different geographical origin and hosts. The ITS region is

proposed as taxonomically informative for distinguishing between Fp and Fv.

ª 2009 The British Mycological Society. Published by Elsevier Ltd. All rights reserved.

Introduction by insects (Munkvold 2003). Ear rot not only reduces grain

Fusarium verticillioides (Fv) (sexual stage Gibberella moniliformis)

and Fusarium proliferatum (Fp) [sexual stage Gibberella interme-

dia] are both common as pathogens of maize worldwide, caus-

ing [in conjunction with Fusarium subglutinans] ear, stalk and

root rot (Bottalico 1998; Leslie et al. 1990; Nelson et al. 1981).

They can be particularly aggressive in temperate climates

(Munkvold 2003), and maize produced in northern Italy is of-

ten heavily contaminated. Infection can be spread by both

soil- and seed-borne inoculum, but usually invasion of the

growing plant occurs through the silk or the kernels damaged

875; fax: þ39 011 236 887o.itritish Mycological Society

yield, but also grain quality, since both Fv and Fp produce my-

cotoxins within the infected kernel. In particular, significant

quantities of the class B fumonisins (FB) are frequently

detected in maize grown in northern Italy (Bottalico et al.

1989; Moretti et al. 1995). Over 20 natural analogues of fumoni-

sins are known, but it is fumonisin B1 (FB1) which is the com-

monest and most dangerous one, and the one which was

associated with severe mycotoxicosis in both domesticate an-

imals and humans (Rheeder et al. 2002). Kernels can become

heavily contaminated long before harvest, so an understand-

ing of the epidemiology of ear, stalk and root rot is necessary

5.

. Published by Elsevier Ltd. All rights reserved.

1138 I. Visentin et al.

before control strategies can be elaborated. The availability of

a robust and reliable method to taxonomically identify Fv and

Fp (the major two fumonisin-producing species) is an essen-

tial requirement for this purpose.

Both Fv and Fp belong to the Gibberella fujikuroi species

complex, in which at least eleven different mating popula-

tions (MPs) (i.e. biological species) have been identified (Leslie

& Summerell 2006). Among the species in the G. fujikuroi

complex which are pathogenic on maize, Fv and Fp can be

distinguished by their microconidial chains (not formed by

F. subglutinans), and it is the presence of polyphialides which

differentiates Fp from Fv (Nirenberg & O’Donnell 1998). As

sexual crosses are both labour- and time-consuming to gen-

erate, species diagnosis relies currently on morphological

traits, supported by the DNA sequence of several genes,

most often, tubulin and calmodulin (O’Donnell et al. 1998). Al-

though a number of criteria show Fv and Fp to be true evolu-

tionary species (Taylor et al. 2000), the morphological

distinction between Fp and Fv is rather fine and requires

trained personnel to recognize, while DNA sequencing can

Table 1 – Primer sequences used for genotyping Fusarium spp

Primer pairs Target sequence Amplicon size

Heat shock protein

FUS1 50-cttggtcatgggccagtcaagac-30 1600

FUS2 50-cacagtcacatagcattgctagcc-30

IGS

VERT1 50-gtcagaatccatgccagaacg-30 800

VERT2 50-cacccgcagcaatccatcag-30

IGS

Fp3F 50-cggccaccagaggatgtg-30 230

Fp4R 50-caacacgaatcgcttcctgac-30

Calmodulin

VER1 50-cttcctgcgatgtttctcc-30 578

VER2 50-aattggccattggtattatatatcta-30

Calmodulin

PRO1 50-ctttccgccaagtttcttc-30 585

PRO2 50-tgtcagtaactcgacgttgttg-30

Calmodulin

CL1 50-gartwcaaggaggccttctc-30 670

CL2A 50-ttttgcatcatgagttggac-30

ITS

ITS1 50-tccgtaggtgaacctgcgg-30 620

ITS4 50-tcctccgcttattgatatgc-30

ITS

verITS-F 50-aaatcgcgttccccaaattga-30 172

ITS4 50-tcctccgcttattgatatgc-30

ITS

ITS1 50-tccgtaggtgaacctgcgg-30 390

proITS-R 50-gcttgccgcaagggctcgc-30

MAT1

fusALPHArev 50-ggartaracyttagcaatyagggc-30 200

fusALPHAfor 50-cgccctctkaaygscttcatg-30

MAT2

fusHMGfor 50-tgggcggtactggtartcrgg-30 260

fusHMGrev 50-cgacctcccaaygcytacat-30

be expensive if large numbers of strains require identifica-

tion. PCR-based genotyping offers a more cost-effective

means of molecular diagnosis. Primer pairs which should

specifically amplify DNA from Fv have been designed based

on sequence variation within either the intergenic spacer

(IGS) region of the ribosomal locus (VERT1–VERT2) (Patino

et al. 2004) or the calmodulin gene (VER1–VER2) (Mule et al.

2004). Similarly, it has been possible to design the Fp-specific

primers Fp3F–Fp4R (IGS) (Jurado et al. 2006) and PRO1–PRO2

(calmodulin) (Mule et al. 2004) (Table 1). An important issue

in the context of PCR-based species diagnostics is amplifica-

tion specificity, especially where the fungal strains to be

tested have been isolated from an environment which was

not sampled during the process of primer design. The inten-

tion of the present research was to classify Fusarium strains

in a collection of isolates from maize samples grown in Pied-

mont (north-western Italy) using morphological, biological

and molecular tools, and to develop robust PCR primers to

distinguish between Fv and Fp as a complement to the

primers already available.

.

Specificity Reference

Fusarium moniliforme Murillo et al. (1998)

Fusarium verticillioides Patino et al. (2004)

Fusarium proliferatum Jurado et al. (2006)

F. verticillioides Mule et al. (2004)

F. proliferatum Mule et al. (2004)

Fusarium spp. O0Donnell et al. (1998)

All fungal species White et al. (1990)

F. verticillioides This work, White et al. (1990)

F. proliferatum This work, White et al. (1990)

Fusaria species with Calonectria,

Gibberella and Nectria teleomorphs

Kerenyi et al. (2004)

Fusaria species with Calonectria,

Gibberella and Nectria teleomorphs

Kerenyi et al. (2004)

The ITS region as a taxonomic discriminator 1139

Materials and methods

Fungal isolates

A collection of 100 fungal strains (see Table 2 and Table S1)

was isolated from maize kernels of naturally infected plants

in 25 fields in Piedmont, Italy (north-western Italy, Fig S1).

Monoconidial cultures of all strains were obtained, and culti-

vated on Komada medium (Komada 1975) to allow a prelimi-

nary taxonomical assignment based on colony pigmentation

and the presence/absence of conidial chains (Summerell

et al. 2003). Only those isolates that produced conidial chains

were retained, and the morphology of their conidiophores

when cultured on SNA medium was noted (Nirenberg 1976).

Moreover a strain of Fusarium subglutinans (NOVb1) was iden-

tified and retained to be used as outgroup (negative control) in

the phylogenetic analyses. Isolates were stored on malt-agar

at 10 �C.

DNA extraction and PCR amplification

Whatman FTA cards were used to extract DNA from freshly

grown mycelium of isolates grown on malt-agar, following

the manufacturer’s instructions. Five primer pairs were tested

(Table 1): VERT1–VERT2 (Patino et al. 2004), Fp3F–Fp4R (Jurado

et al. 2006), VER1–VER2 and PRO1–PRO2 (Mule et al. 2004), and

FUS1–FUS2 (Murillo et al. 1998). PCRs were performed using

a T-Gradient thermal cycler (Biometra) replicating the

Table 2 – Fp and Fv strains isolated from maize grown in northpolyphialides. PCR amplification used extant species-specific p

Strains Morphology MATallele/MP

AFLPgroup

Calmodulingroup

Rpr

VP2 Fusarium

verticillioides

1/NT A A

CH2 F. verticillioides 2/NT A A

CHI1 F. verticillioides 1/NT A A

FR3 F. verticillioides 1/NT A A

SAL4 F. verticillioides 1/NT A A

GE1 F. verticillioides 1/NT A A

PI1 F. verticillioides 1/A A A

CA2 F. verticillioides 1/NT A A

CM3 F. verticillioides 2/A A A

SA3 F. verticillioides 1/A A A

PO1 Fusarium

proliferatum

1/not fertile B B

RT1 F. proliferatum 2/not fertile B B

BU1 F. proliferatum 2/D B B

CE1 F. proliferatum 2/not fertile B B

MATA-1 F. verticillioides 1/A NT A

MATA-2 F. verticillioides 2/A NT A

MATD-1 F. proliferatum 1/D NT B

MATD-2 F. proliferatum 2/D NT B

NT: not tested.

a Weak amplification.

b Quantitative analysis for FB1 and FB2.

c FB3 analysis also performed.

conditions reported by the originators of the primers. The

newly developed primers based on ITS sequence (verITS-F

and proITS-R, see Table 1) were used in combination with

the primer pair ITS1–ITS4. Multiplex reactions were carried

out in 25 ml final reaction volume. Each tube contained

0.4 mM ITS1 and 0.4 mM ITS4, 0.5 mM verITS-F and 0.5 mM

proITS-R, 0.625 U GoTaq polymerase (Promega, 5 U ml�1),

1� PCR buffer and 0.4 mM dNTP. The amplification conditions

consisted of a denaturation of 94 �C/2 min, followed by 35 cy-

cles of 96 �C/60 s, 60 �C/60 s and 72 �C/45 s, and a final exten-

sion of 72 �C/10 min. PCR products were separated by

agarose gel electrophoresis (1 % w/v in 0.5� TBE) and visual-

ized by ethidium bromide staining.

ITS-RFLP

An aliquot of the ITS DNA amplicon [generated by the ITS1–

ITS4 primer pair (White et al. 1990)] from each isolate was

digested overnight at 37 �C with one of the 4 bp cutters AluI,

MboI, HaeI or TaqI, or the 5 bp cutter HinfI (all enzymes pur-

chased from Fermentas). The restriction fragments were sep-

arated by electrophoresis through 2 % (w/v) agarose gels and

visualized by ethidium bromide staining.

ITS and calmodulin sequencing

DNA from all isolates was amplified with both ITS1–ITS4

(White et al. 1990), and the calmodulin gene primers CL1–

CL2A (O’Donnell et al. 2000) (Table 1). The PCRs consisted

-western Italy. Fp strains identified by the presence ofrimer pairs

FLPofile

Fus1 VERT1 VER1 PRO1 Fp3F FBproductionb

Fus2 VERT2 VER2 PRO2 Fp4R

A þ þ þ � � þþþ

A þ þ þ � � þþþA þ þ þ � � þþþA þ þ þ � � þþþA þ þ þ � � þþþA þ þ þ � � þþþA þ þ þ � � þþþA þ þ þ � � þþþA þ � þ � � þþc

A þ þ þ � � þþþB þ þ þa þ þ �c

B þ � þa þ � þþþB þ þ � þ þ þc

B þ þ þa þ þ þc

A þ þ þ � � NT

A þ þ þ � � NT

B þ þ þa þ � NT

B þ þ þa þ � NT

1140 I. Visentin et al.

of a denaturation of 94 �C/5 min, followed by 30 cycles of

94 �C/30 s, 55 �C/60 s and 72 �C/60 s, and a final extension of

72 �C/10 min. The amplicons were directly sequenced from

both ends by Genelab (ENEA, Rome). Phylogenetic analyses

were carried out using MEGA software (www.megasoftware.-

net) based on the sequences of 14 Fusarium strains plus mat-

ing-type tester strains for Fv and Fp teleomorphs (MATA-1/2

and MATD-1/2; see further on) and one Fusarium subglutinans

strain (NOVb1). Neighbour-joining trees were constructed us-

ing Kimura’s two-parameter model (Kimura 1980) with 1000

bootstrap replicates. The isolates listed in Table 2 were depos-

ited in the collection of the Institute of Sciences of Food Pro-

duction (ISPA-CNR, Bari, Italy; http://server.ispa.cnr.it/ITEM/

Collection) and the genomic sequences were deposited in

Genbank. Accession numbers for ISPA and Genbank (ITS and

calmodulin, in this order) for each isolate are as follows: VP2

(10676; EU151483; EU430618); CH2 (10677; EU151473;

EU430621); CHI1 (10678; EU151479; EU430604); FR3 (10679;

EU151474; EU430608); SAL4 (10680; EU151472; EU430614); GE1

(10681; EU151480; EU430612); PI1 (10682; EU151481;

EU430622); CA2 (10683; EU151478; EU430610); CM3 (10684;

EU151469; EU430616); SA3 (10685; EU151482; EU430613); PO1

(10686; EU151487; EU430607); RT1 (10687; EU151490;

EU430606); BU1 (10688; EU151489; EU430611); CE1 (10689;

EU151486; EU430620); NOVb1 (10690; EU151476; EU430603);

MATA-1 (7581; EU151467); MATA-2 (7583; EU151468); MATD-1

(7596; EU151484); MATD-2 (7595; EU151485). The four latter

mating-type tester strains were not sequenced on calmodulin

(available Genbank sequences were used and are reported in

Fig 2 along with their accession numbers). Additional strains

of Fv and Fp from different hosts and/or geographical origin

were also tested; they were obtained from public collections

and are listed in Table S1.

AFLP analysis

Strains were grown in Czapek-Dox mineral medium (Sigma)

in still culture for two weeks. Mycelium was harvested by fil-

tration and frozen in liquid nitrogen, and DNA extracted by

the CTAB method (Murray & Thompson 1980). The AFLP pro-

tocol followed Vos et al. (1995) with minor modifications

(Lanteri et al. 2004). Briefly, 5 ml (400–500 ng) DNA was dou-

ble-digested with EcoRI and MseI and ligated to adapters.

Digested and ligated DNA fragments were pre-amplified

with primers carrying one selective base (EcoRIþA and

MseIþC), and selectively amplified using primers carrying

two or three selective bases. Initially, 12 primer pairs (four

EcoRI and three MseI primers) were tested against four tem-

plates, and the outcome of this pilot resulted in the choice of

the four primer combinations EþAAT/MþCAA, EþAAT/

MþCAG, EþACT/MþCAA, EþACT/MþCAG. AFLP ampli-

cons were resolved through 5 % denaturing polyacrylamide

gels, and visualized by silver staining as described (Bassam

et al. 1991).

Data scoring and analysis

AFLP reactions were repeated at least twice in order to as-

sure consistency. Electrophoretic patterns were documented

using a commercial gel documentation system (Quantity

One Programme, BioRad). Each fragment was assumed to

represent a single locus and only reproducible polymorphic

bands were scored as present (1) or absent (0). The binary

data were used to calculate the polymorphism information

content (PIC) by applying the simplified formula 2f (1�f )

(Anderson et al. 1993) for expected heterozygosity, where f

represents the percentage of individuals in which the

marker is present. The binary matrix was imported into

NTSYS-pc v1.80 (Rohlf 1993) to perform a cluster analysis.

Genetic similarity among isolates was calculated according

to Jaccard’s similarity index (JSI) (Jaccard 1908), using the

SIMQUAL routine. The similarity coefficients were used to

construct a dendrogram by UPGMA with 1000 bootstraps, us-

ing PHYLIP package [http://evolution.genetics.washington.

edu/phylip.html (Felsenstein 1993)]. A co-phenetic matrix

was produced using the hierarchical clustering system, by

means of the COPH routine, and correlated with the original

distance matrices for AFLP data, in order to test for the asso-

ciation between the cluster in the dendrogram and the JSI

matrix. Mantel tests (Mantel 1967) were performed to check

the correlation between the similarity matrices generated by

each AFLP primer combination and the global similarity

matrix. A principal coordinate (PCO) analysis was also per-

formed, based on the triangular matrix of genetic similarity

estimates, and the first two axes were plotted graphically.

Fertility test and vegetative compatibility

Before crossing, each strain was PCR genotyped using the

fusALPHAfor/rev and fusHMGfor/rev primer pairs (Table 1)

to check which idiomorphic allele they carried at the MAT lo-

cus (Kerenyi et al. 2004; Yun et al. 2000). Strains tested for sex-

ual compatibility (PI1, CM3, SA3, BU1, PO1, RT1 and CE1) were

co-cultivated on carrot agar in combination with each of the

mating-type tester strains MATA-1, MATA-2, MATD-1 and

MATD-2 (obtained from the Institute of Sciences of Food Pro-

duction, Bari, Italy) to create a set of crosses following Klittich

and Leslie (1988). The female strain was grown on carrot agar

and sprinkled with a conidial suspension obtained from water

agar plates of the male. The cultures were taken as fertile

when cirri were observed as extruding from the perithecia.

The carrot agar plates were incubated for four weeks at 23/

24 �C under 12/12 h light/darkness. A vegetative compatibility

group (VCG) assignment for all strains was based on comple-

mentation between nitM and nit1 mutants grown on minimal

medium (Correl et al. 1987). Pairs of vegetatively compatible

isolates produced vigorous aerial growth at the contact site

of the mycelia of two nit mutants. Isolates that exhibited

this morphology were assigned the same VCG.

HPLC analysis of fumonisins

20 g samples of cracked maize kernel were combined with

2 ml water each, and autoclaved twice (120�C, 1 h). Each maize

aliquot was inoculated with 3 ml of sterile water containing

106 conidia of each isolate in turn, and maintained at room

temperature under 12/12 h light/darkness with daily manual

shaking. The quantity of FB1 and FB2 in the culture was deter-

mined by HPLC 20 d after inoculation, and cultures containing

little or no FB1 and FB2 were re-analysed for FB3 content.

Coordinate 1

-0.70 -0.41 -0.13 0.16 0.45

Coo

rdin

ate

2

-0.95

-0.65

-0.35

-0.05

0.25

MATA-1

MATA-2

MATD-1

MATD-2

NOVb1

PI1

CM3

CHI1SA3GE1VP2

PO1

RT1

CA2BU1

SAL4

CH2

CE1

FR3

A

B

54

91

55

52

65

79

74

65

65

52

77

7285

67

88

83

Jaccard's Similarity Coefficient

0.25 0.50 0.75 1.00

MATA-1

SA3

CHI1

VP2

GE1

MATA-2

PI1

CH2

FR3

CM3

CA2

SAL4

MATD-1

PO1

RT1

MATD-2

CE1

BU1

NOVb1

A

B

Fig 1 – UPGMA dendrogram (A) and principal coordinate

analysis based on Jaccard’s similarity coefficient (B) of 19

The ITS region as a taxonomic discriminator 1141

Fumonisins were detected by described methods (Adejumo

et al. 2007).

Results

Morphological identification and toxin production

A collection of 100 Fusarium strains producing microconidial

chains was assembled from 25 different fields covering the

Piedmont territory (Fig S1). Monoconidial cultures were

scanned for presence/absence of polyphialides to discriminate

between Fv and Fp. The results of this morphologically based

classification are given in Table S1. A sample of 14 isolates,

each originating from a different field and including some Fv

and Fp based on morphological scoring, was selected for fur-

ther analyses. All Fv isolates were highly toxigenic in vitro; FB

production was more variable among Fp isolates (Table 2).

AFLP analysis and genetic relatedness

The four primer combinations amplified 168 fragments, of

which 151 were polymorphic. The number of polymorphic

fragments per primer combination ranged from 28 to 51

(Table 3). All the genotypes were characterized by a unique

banding pattern. The EþACT/MþCAG combination pro-

duced the greatest number of polymorphic fragments, the

highest PIC and could discriminate, on its own, between

all 14 isolates and the testers. The lowest JSI value was

0.194 (Fusarium subglutinans NOVb1 and CM3), while the

highest (0.922) applied to the contrast MATA-1 and SA3

(JSI values of the other comparisons are available

on request from the authors). The resulting

UPGMA dendrogram is given as Fig 1. The co-phenetic cor-

relation coefficient between the data matrix and the co-

phenetic matrix for the AFLP data was 0.975, indicating

a close fit between the dendrogram clusters and the JSI

matrix from which they were derived. As expected, isolate

NOVb1 was the most highly differentiated from the other

isolates (mean JSI 0.250). Two major clusters (bootstrap

probability� 88 %) were resolved: clusters A and B corre-

spond to isolates identified by their morphology as Fv

and Fp, respectively. The principal coordinate scatter plot

confirmed the clustering of the genotypes. The first two

axes accounted for, respectively, >49 % and >19 % of the

Table 3 – AFLP profiling of a sample of Fv and Fp isolates.TNB: total number of bands, NPB: number of polymorphicbands, %: percentage of polymorphic bands, PIC:polymorphic information content, r: correlation betweenthe similarity matrices generated by each individualprimer combination and the global similarity matrix

PC TNB NPB % PIC r

EþAAT/MþCAA 33 30 90.9 0.318 0.879

EþAAT/MþCAG 32 28 87.5 0.324 0.912

EþACT/MþCAA 46 42 91.3 0.252 0.954

EþACT/MþCAG 57 51 89.5 0.352 0.964

Average 42 38 89.9 0.311 0.927

Total 168 151

Fusarium isolates. Genotypic data generated by AFLP pro-

filing with four primer combinations. Numbers associated

with each dendrogram node represent the proportion of

1000 bootstrap samples in which the particular clade was

found. Only percentages above 50 % are shown.

variation. The former distinguished the isolates belonging

to cluster A from those belonging to cluster B, with

NOVb1 in an intermediate position; while the latter sepa-

rated NOVb1 from the rest of the collection.

Calmodulin sequence analysis

A phylogeny analysis was carried out based on the partial cal-

modulin gene sequences derived from the 14 strains (Fig 2A).

A

B

CM3

SA3

GE1

CH2

SAL4

VP2

AF158315

CA2

PI1

FR3

CHI1

NOVb1

CE1

BU1

AF291057

RT1

PO1

100

53

100

57

0.0000.0050.0100.0150.0200.0250.030

A

B

MATA-2SAL4

SA3GE1CH2

NOVb1VP2PI1CM3MATA-1FR3

CHI1CA2

MATD-2BU1CE1PO1RT1MATD-1

66

100

67

0.002

A

B

Fig 2 – Phylogeny based on a multiple alignment of the

CL1–CL2A amplified fragment of the calmodulin gene (A)

and the ITS1–ITS4 amplified fragment of the ITS (B). Both

Neighbour-joining trees were constructed with Kimura’s

two-parameter model with 1000 bootstrap replicates.

Edge length is indicated in terms of substitution rates per

nucleotide, together with bootstrap % values above 50 %.

1142 I. Visentin et al.

This indicated the existence of two major groups. One group

was centred on AF291057, a known type-Fv sequence, and

the other on AF158315, a type-Fp sequence. The cluster pat-

tern derived from this analysis perfectly matched the one gen-

erated on the basis of AFLP genotypes.

Fertility and vegetative compatibility

As far as sexual fertility, PO1, RT1 and CE1 (among the isolates

tested) did not mate with the tester strains under the condi-

tions imposed. Isolate BU1 was attributed to MP D (Fp) and iso-

lates PI1, CM3 and SA3 to MP A (Fv). As far as vegetative

compatibility, only 2 isolates out of the 14 (SA3 and VP2,

both Fv) belonged to the same VCG, resulting in the identifica-

tion of 13 distinct VCGs for 14 isolates. Such a high level of in-

ter-strain incompatibility is consistent with previously

observed rates of differentiation between Fv isolates (Desjar-

dins et al. 1994).

PCR amplification using VERT1 and VERT2, Fp3F and Fp4R,VER1 and VER2, PRO1 and PRO2, FUS1 and FUS2 primerpairs

The results of the PCR genotyping are summarized in Table 2,

where expected and obtained results for each isolate and spe-

cies are easily visualized. The VERT1–VERT2 (Patino et al. 2004)

and VER1–VER2 (Mule et al. 2004) primer pairs generated

amplicons of the expected size from most of the strains. The

exceptions were, for VERT1–VERT2: CM3 (Fv, but negative),

PO1, BU1 and CE1 (Fp, but positive). VER1–VER2 amplified inap-

propriately, though only weakly, from Fp strains PO1, RT1, CE1

and the MP D testers (Table 2). The sample of 14 isolates plus

testers was tested with the Fp-specific PRO1–PRO2 (Mule et al.

2004) and Fp3F–Fp4R (Jurado et al. 2006) primer pairs. PRO1–

PRO2 amplified the expected w600 bp product from five Fp

strains out of six (PO1, RT1, BU1 and CE1, along with the

MATD-1 tester). Fp3F–Fp4R did not amplify from three (RT1,

MATD-1, MATD-2) out of the six Fp strains expected to be pos-

itive (Table 2). The FUS1–FUS2 primer pair (Murillo et al. 1998)

amplified from all isolates, so all Fp strains were false positive

(Table 2). Thus to summarize, VERT1–VERT2, which was

designed to specifically amplify only Fv DNA, amplified some

25 % of the Fp strains and NOVb1 (Fusarium subglutinans), while

VER1–VER2 (Fv-specific) amplified weakly from the DNA of all

but one of the Fp strains. The FUS1–FUS2 primers (specific for

Fusarium moniliforme, i.e. Fusarium verticillioides sensu Niren-

berg – 1976) amplified from the DNA of all Fp isolates as well.

PRO1–PRO2 primers, which target the calmodulin gene, effec-

tively identified all the Fp strains except for one tester isolate,

whereas Fp3F–Fp4R (IGS region) gave a positive result for just

50 % of the Fp isolates.

ITS-RFLP and ITS sequence analysis

The digestion of the w600 bp amplicon generated by primers

ITS1 and ITS4 (White et al. 1990) with AluI, MboI, TaqI, HaeIII

or HinfI produced two profiles for each endonuclease

(see Table 4). The MP testers A and D were associated with, re-

spectively, profiles A and B, so the inference was that the Fv

isolates would have ITS-RFLP profile A and Fp isolates would

have profile B. A comparison of ITS amplicon sequences

among the sample of 14 isolates plus testers also showed

two clear groups (Fig 2B). Fp isolates differed from Fv types

by a 6 bp insertion, together with a few single base substitu-

tions (Fig S3). Group A and B correspond to Type I and Type

II ITS2 (O’Donnell & Cigelnik 1997).

Discrimination between Fv and Fp by primers targeting theITS region

Two primers, proITS-R and verITS-F, were designed to anneal

to the polymorphic region of the ITS amplicon (Fig S3), and

were intended to be combined with the universal primers

Table 4 – Sizes of restriction fragments obtained from digestion of the ITS region from restriction profile A (associated to Fvtester strains) and B (associated to Fp tester strains). Values in brackets refer to bands too small to be detected on agarosegel but inferred from sequence analysis

Restrictionprofile

Enzyme

AluI MboI HinfI TaqI HaeIII

A 332, 114 (61) 305, 110 (60, 29, 3) 245, 156, 91 (8, 7) 192, 118, 85 (59, 53) 347, 93 (67)

B 404, 114 174, 143, 110 (60, 28, 4) 266, 245 (8) 215, 192 (59, 53) 281, 93 (77, 68)

The ITS region as a taxonomic discriminator 1143

ITS1 and ITS4 in simplex or multiplex reactions. Fig 3A shows

the primer positions and orientations, along with expected

amplicon sizes. All 14 isolates and the tester strains for MP A

and MP D were tested (Fig 3B). A w600 bp fragment corre-

sponding to the ITS1/ITS4 product was generated from all tem-

plates, providing a positive amplification control. Fp and Fv

isolates produced species-specific amplicons of about 390

and 170 bp, respectively. Both this PCR assay and the ITS-

RFLP test were extended to all 100 isolates (Table S1), and the

classifications based on PCR all matched those based on mor-

phological analysis. Similarly, proITS-R and verITS-F were

assayed with corresponding results on the MP A and D tester

strains (Table 2) and on 9 isolates of Fv and 7 Fp from different

countries of origin (USA, Italy, Spain, China) and/or hosts

(maize, ornamental palm tree, asparagus, wheat) (Table S1).

Discussion

One prerequisite for a detailed understanding of the ecological

behaviour of the fungi belonging to the Gibberella fujikuroi

complex is to have a robust and reliable means of

600 bp

M 1 2 3 4 5B

ITS1

ITS4

verITS-F

proITS-R

172bp

390bp

A

200 bp

400 bp

Fig 3 – A. Placement of the PCR primers ITS1–ITS4, verITS-F

and proITS-R in the ITS region. B. Representative multiplex

PCRs using primers ITS1–ITS4, verITS-F and proITS-R.

Lanes 1–2: Fv tester strains MATA-1 and MATA-2. Lanes

3–4: Fp tester strains MATD-1 and MATD-2.

discriminating between the various species. Although host

range and morphology are normally rather species-specific,

Fp and Fv are quite similar to one another at the morphological

level, and are often both present in infected maize. Distin-

guishing between them has been based mainly on their ge-

netic diversity and on their sexual incompatibility with one

another (O’Donnell et al. 1998). Here, we have characterized

monoconidial strains of fumonisin-producing Fusarium iso-

lated from pink rotted maize kernels. The majority of these

(83) were Fv and the rest (16) mainly Fp. One Fusarium subglu-

tinans was identified from the same sampling and retained in

the collection as a supplementary member of the G. fujikuroj

complex. The Fv and Fp strains could be distinguished from

one another by their AFLP and ITS-RFLP profiles, and by the

DNA sequence present in the ITS and the calmodulin gene.

Species classification based on any of the molecular analyses

was in agreement with conclusions based on sexual crosses

and vegetative morphology.

Because Fv and Fp are so similar to one another at the mor-

phological level, and because they are often simultaneously

present in diseased maize kernels, the availability of a simple,

cheap and reliable diagnostic method is clearly desirable, and

a variety of PCR-based analyses to be used in combination

might be the best compromise between reliability and cost ef-

fectiveness. A number of PCR-based diagnostic assays have

been published as able to discriminate between these two spe-

cies with a high level of accuracy (Jurado et al. 2006; Patino et al.

2004). The effectiveness of some of them was, however, lim-

ited when tested against the present collection of Italian Fv

and Fp isolates (with VER1–VER2 and PRO1–PRO2 performing

best). Similarly, conflicting results were obtained by others

with primer pairs VERT1–VERT2 and PRO1–PRO2 on Fp isolates

from Allium fistulosum in Japan (Dissanayake et al. 2009). A

probable explanation for the observed discrepancies is that

the primer sequences were designed on the basis of too lim-

ited a sample of fungal strains. Since the extant Fusarium

‘‘species-specific’’ primers proved not all sufficiently robust

in our conditions, we searched for additional amplification

targets, and the ITS region was identified as a likely candidate.

The ITS sequence of the sample of 14 isolates and four tester

strains showed that species-specific polymorphisms were in-

deed present. Most likely, they reflect the relative intrage-

nomic abundance of the nonorthologous type I and type II

ITS2 (O’Donnell & Cigelnik 1997). The resulting PCR assays

proved to be an effective, species-specific diagnostic for Fv

and Fp, regardless of the geographical and host origin.

Sequence variation within the ITS region on its own can

discriminate between Fv and Fp, but ITSs cannot be consid-

ered phylogenetically informative more widely within the ge-

nus Gibberella and particularly are not sufficient for species

1144 I. Visentin et al.

discrimination between Fv and F. subglutinans, among maize

pathogens (this work and Mule et al. 2004; O’Donnell & Cigel-

nik 1997). However these two species can be distinguished rel-

atively easily from one another at the morphological level.

Overall, we conclude that sequence variation within the ITS

region should be exploited as a taxonomic diagnostic, particu-

larly in the context of discriminating between Fv and Fp. The

ITS-RFLP and PCR assays described here provide a simple

and reliable means of discriminating between Fp and Fv iso-

lates, and should be added to the toolbox of mycologists

researching the fumonisin-producing pathogens of maize.

Acknowledgements

The authors are thankful to Dr. Mariangela Girlanda for dis-

cussing the results, to Marzia di Maio and Federica Mattio

for technical help, and to Dr. Ursula Hettwer for fumonisin

analysis. Work supported by Regione Piemonte – Progetti

CIPE ‘‘Tecniche di controllo delle micotossine del mais per

impieghi alimentari e zootecnici’’ and ‘‘Progetto pilota per la

produzione di cereali ad uso alimentare a basso contenuto

in micotossine’’.

Supplementary information

Supplementary information associated with this article

can be found in the online version at doi:10.1016/

j.mycres.2009.07.011.

r e f e r e n c e s

Adejumo T, Hettwer U, Karlovsky P, 2007. Survey of maize fromsouth-western Nigeria for zearalenone, alpha- and beta-zear-alenols, fumonisin B1 and enniatins produced by Fusariumspecies. Food Additives & Contaminants 24: 993–1000.

Anderson JA, Churchill GA, Autrique JE, Sorells ME, Tanksley SD,1993. Optimizing parental selection for genetic-linkage maps.Genome 36: 181–186.

Bassam BJ, Caetano-Anolles G, Gresshoff PM, 1991. Fast andsensitive silver staining of DNA in polyacrylamide gels.Analytical Biochemistry 196: 80–83.

Bottalico A, 1998. Fusarium disease of cereals: species complexand related mycotoxin profiles, in Europe. European Journal ofPlant Pathology 80: 85–103.

Bottalico A, Logrieco A, Visconti A, 1989. Fusarium species andtheir mycotoxins in infected corn in Italy. Mycopathologia 107:85–92.

Correl JC, Klittich CJR, Leslie JF, 1987. Nitrate non-utilizing mu-tants of Fusarium oxysporum and their use in vegetative com-patibility tests. Phytopathology 77: 1640–1646.

Desjardins AE, Plattner RD, Nelson PE, 1994. Fumonisin produc-tion and other traits of Fusarium moniliforme strains frommaize in northeast Mexico. Applied and Environmental Microbi-ology 60: 1695–1697.

Dissanayake MLMC, Tanaka S, Ito S, 2009. Fumonisin B1 produc-tion by Fusarium proliferatum strains isolated fromAllium fistulosum plants and seeds in Japan. Letters in AppliedMicrobiology 48: 598–604.

Felsenstein J, 1993. PHYLIP, Phylogenetic Inference Package version3.5.7. A Computer Program Distributed by the Author. De-partment of Genetics, University of Washington, Seattle. http://evolution.genetics.washington.edu/phylip.html.

Jaccard P, 1908. Nouvelles recherches sur la distribution florale.Bulletin de la Societe Vaudoise des Sciences Naturelles 44:223–270.

Jurado M, Vazquez C, Marin S, Sanchis V, Gonzalez-Jaen MT, 2006.PCR-based strategy to detect contamination with mycotoxi-genic Fusarium species in maize. Systematic andApplied Microbiology 29: 681–689.

Kerenyi Z, Moretti A, Waalwijk C, Olah B, Hornok L, 2004. Matingtype sequences in asexually reproducing Fusarium species.Applied and Environmental Microbiology 70: 4419–4423.

Kimura M, 1980. A simple method for estimating evolutionaryrates of base substitutions through comparative studies of nu-cleotide sequences. Journal of Molecular Evolution 16: 111–120.

Klittich C, Leslie JF, 1988. Nitrate reduction mutants of Fusariummoniliforme (Gibberella fujikuroi). Genetics 118: 417–423.

Komada H, 1975. Development of a selective medium for quan-titative isolation of Fusarium oxysporum from natural soil.Review of Plant Protection Research 8: 114–125.

Lanteri S, Saba E, Cadinu M, Mallica GM, Baghino L, Portis E, 2004.Amplified fragment length polymorphism for genetic diversityassessment in globe artichoke. Theoretical and Applied Genetics108: 1534–1544.

Leslie JF, Pearson CA, Nelson PA, Toussoun TA, 1990. Fusariumspp. from maize, sorghum, and soybean fields in the centraland eastern United States. Phytopathology 86: 343–350.

Leslie JF, Summerell BA, 2006. The Fusarium Laboratory Manual.Iowa, USA, Blackwell.

Mantel N, 1967. The detection of disease clustering and a gener-alized regression approach. Cancer Research 27: 209–220.

Moretti A, Bennett GA, Logrieco A, Bottalico A, Beremand MN,1995. Fertility of Fusarium moniliforme from maize and sor-ghum related to fumonisin production in Italy. Mycopathologia131: 25–29.

Mule G, Susca A, Stea G, Moretti A, 2004. A species-specific PCRassay based on the calmodulin partial gene for the identifi-cation of Fusarium verticillioides, F. proliferatum and F. subglu-tinans. European Journal of Plant Pathology 110: 495–502.

Munkvold G, 2003. Epidemiology of Fusarium diseases and theirmycotoxins in maize ears. European Journal of Plant Pathology109: 705–713.

Murillo I, Cavallarin L, San Segundo B, 1998. The development of arapid PCR assay for detection of Fusarium moniliforme. EuropeanJournal of Plant Pathology 104: 301–311.

Murray MG, Thompson WF, 1980. Rapid isolation of high molec-ular weight plant DNA. Nucleic Acids Research 8: 4321–4325.

Nelson PE, Toussoun TA, Cook RJ, 1981. Fusarium: disease, biologyand taxonomy. Pennsylvania State University Press, UniversityPark and London, USA.

Nirenberg HI, 1976. Untersuchungen uber die morphologischeund biologische Differenzierung in der Fusarium-Sektion Liseola. Mitteilungen aus der Biologische Bundesanstalt furLand- und Forstwirtschaft Berlin- Dahlem 169: 1–117.

Nirenberg HI, O’Donnell K, 1998. New Fusarium species andcombinations within the Gibberella fujikuroi complex. Mycologia90: 434–458.

O’Donnell K, Cigelnik E, 1997. Two divergent intragenomic rDNAITS2 types within a monophyletic lineage of the fungus Fusa-rium are nonorthologous. Molecular Phylogenetics and Evolution7: 103–116.

O’Donnell K, Cigelnik E, Nirenberg HI, 1998. Molecular systemat-ics and phylogeography of the Gibberella fujikuroi speciescomplex. Mycologia 90: 465–493.

O’Donnell K, Nirenberg HI, Aoki T, Cigelnik E, 2000. A multigenephylogeny of the Gibberella fujikuroi species complex: detection

The ITS region as a taxonomic discriminator 1145

of additional phylogenetically distinct species. Mycoscience 41:61–78.

Patino B, Mirete S, Gonzalez-Jaen MT, Mule G, Rodriguez MT,Vazquez C, 2004. PCR detection assay of fumonisin-producingFusarium verticillioides strains. Journal of Food Protection 67:1278–1283.

Rheeder JP, Marasas WF, Vismer HF, 2002. Production of fumo-nisin analogs by Fusarium species. Applied and EnvironmentalMicrobiology 68: 2101–2105.

Rohlf FJ, 1993. NTSYS-pc Numerical Taxonomy and Multivariate AnalysisSystem, version 1.80. Owners Manual. Exter Software, New York.

Summerell BA, Salleh B, Leslie JF, 2003. A utilitarian approach toFusarium identification. Plant Disease 87: 117–128.

Taylor JW, Jacobson DJ, Kroken S, Kasuga T, Geiser DM,Hibbett DS, Fisher MC, 2000. Phylogenetic species recognition

and species concepts in fungi. Fungal Genetics and Biology 31:21–32.

Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M,Frijters A, Pot J, Peleman J, Kuiper M, et al., 1995. AFLP: a newtechnique for DNA fingerprinting. Nucleic Acids Research 23:4407–4414.

White TJ, Burns T, Lee S, Taylor JW, 1990. Amplification and directsequencing of fungal ribosomal RNA genes for phylogenetics.In: Innis MA, Gelfald DH, Sninsky JJ, White TJ (eds), PCR Pro-tocol: a guide to methods and application. Academic Press, Inc.,New York, USA, pp. 315–322.

Yun SH, Arie T, Kaneko I, Yoder OC, Turgeon BG, 2000. Molecularorganization of mating type loci in heterothallic, homothallic,and asexual Gibberella/Fusarium species. Fungal Genetics andBiology 31: 7–20.