Mycologia, 104(6), 2012, pp. 1281–1290. DOI: 10.
3852/11-316
# 2012 by The Mycological Society of America, Lawrence, KS 66044-8897
Molecular evidence for Neotyphodium fungal endophyte
variation and specificity within host grass species
Somaye Karimi
Aghafakhr Mirlohi1
Mohammad R. Sabzalian
Department of Agronomy and Plant Breeding, College of
Agriculture, Isfahan University of Technology, Isfahan
84156–83111, Iran
Badraldin E. Sayed Tabatabaei
Department of Plant Biotechnology, College of
Agriculture, Isfahan University of Technology, Isfahan
84156–83111, Iran
Bahram Sharifnabi
Department of Plant Pathology, College of Agriculture,
Isfahan University of Technology, Isfahan 84156–
83111, Iran
Abstract: Host specificity of Neotyphodium species
symbiotic with three grass species, Festuca arundinacea,
Festuca pratensis and Lolium perenne, was studied based
on comparisons of amplified fragment length poly-
morphisms (AFLP) between hosts and their corre-
sponding endophytes. Endophytic fungi were isolated
from 24 accessions of host plants. Neotyphodium
identity was determined based on morphological
characteristics observed in cultures and polymerase
chain reaction analysis using specific primers. The
results of AFLP data analysis revealed high genetic
variation in plant and fungal endophyte species. Plant
AFLP genotypes from different species clustered in
three distinctive groups, congruent with species. A
cluster analysis of AFLP data grouped endophytic
isolates according to their host species and secondarily
according to their host geographic distribution. The
result of the AMOVA on AFLP data accounted for a
large and significant proportion of genetic variation
due to differences among plant and endophyte
species. Phylogenetic groups of isolates corresponded
to their respective host genotypes based on maximum
parsimony phylograms. Comparisons of the two
phylograms illustrated a significant congruence be-
tween nodes and branches of host and endophyte
clades. These results strongly suggest host specificity of
Neotyphodium fungal endophytes with their geograph-
ically distant host grasses within each species.
Key words: AFLP, co-evolution, endophyte, Lo-
lium, specificity, variation
Submitted 30 Sep 2011; accepted for publication 7 May 2012.
1
Corresponding author. E-mail: mirlohi@cc.iut.ac.ir
1281
INTRODUCTION
Many pasture grasses are infected with fungal
endophytes of the genus Neotyphodium. The species
Neotyphodium lolii, N. coenophialum and N. uncinatum
are commonly detected in perennial ryegrass (Lolium
perenne), tall fescue (Festuca arundinacea. 5 Lolium
arundinaceum), and meadow fescue (Festuca pratensis
5 Lolium pratense) respectively (Christensen et al.
1993). Plants and fungus have an interactive relation-
ship. The grass provides the endophyte with nutri-
ents, shelter and seeds as means of propagation, while
the fungus improves host survival through improved
adaptations to biotic (Bacon and Siegel 1988, Schardl
and Phillips 1997, Sabzalian et al. 2004) and abiotic
stresses (Arachevaleta et al. 1989, Buck et al. 1997,
Malinowski and Belesky 2000, Sabzalian and Mirlohi
2010). Although endophytes are beneficial to their
host grasses, they produce alkaloid toxins that may
harm livestock (Vazquez de Aldana et al. 2001). These
toxins have encouraged researchers to look for novel
associations that are less harmful to livestock while
providing beneficial aspects for the hosts. Some of
these new combinations may be fully compatible and
stable, whereas others may result in cellular incom-
patibility reactions causing host tissue and hyphae
death (Koga et al. 1993, Christensen 1995).
Researchers do not clearly understand why at-
tempts to form artificial associations between Neoty-
phodium spp. and grasses not closely related to their
natural hosts usually are unsuccessful. Cross inocula-
tions of symbionts of related host species often are
unstable and elicit host defense responses (Kearney
et al. 1991, Wille et al. 1999, Johnson-Cicalese et al.
2000). Establishment of novel symbioses may even
result in ultrastructural changes in the host that are
not characteristic of natural symbiont-host combina-
tions (Leuchtmann 1993). The coordination of host
and symbiont reproductive life cycles has led to
discussions about host specificity resulting from
ancient relationships and a cooperative evolution of
both symbionts.
Several molecular techniques have been employed
to estimate evolutionary relationships between species
of grass endophytes and the genetic structures of
these populations (Leuchtmann and Clay 1990, Tsai
et al. 1994, Groppe et al. 1995, Glenn et al. 1996,
White and Huff 1996). The use of amplified fragment
length polymorphisms (AFLP) has proven to be a
reliable and powerful method for measuring genetic
1282 MYCOLOGIA
variability between and among a large number of
locations in the genome (Vos et al. 1995). This
technique has been used successfully for assessment
of genetic diversity in F. arundinacea (Mian et al.
2002) and in a number of fungal species including
Neotyphodium and Epichloë (Tredway et al. 1999,
Arroyo-Garcia et al. 2002).
Schardl and Phillips (1997) compared the molec-
ular phylogeny of Epichloë species with the phylogeny
of their hosts. In most cases Epichloë species were
limited to hosts within the same genus and the
endophyte phylogeny closely mirrored the host
phylogeny in the tribe. Yet certain Epichloë species
exhibited tribe or genus co-evolution but the signif-
icance of this evolutionary trend at the species or
subspecies level of plant host has not been fully
investigated. Multiple genetic determinants of com-
patibility and incompatibility with the host species
might govern this specificity (Chung et al. 1997).
The existence of different taxonomic groups of tall
fescue endophytes originating from different hybrid-
ization events (Schardl 1996) seems to support and
reflect distributions of taxonomic groups in host
plants. This could be the result of synergistic plant-
fungus associations in tall fescue, unique maternal
transmission of the fungus by seeds and physical or
genetic isolation of the fungus within its host. Piano
et al. (2005) reported that the morphology of the host
plant and identification of its own endophyte could
be combined to provide useful information for
identifying the phylogeny and taxonomy of tall
fescue. They declared that a co-evolutionary pattern
had been postulated (Schardl 1996), but any com-
prehensive investigation attempting to establish a
relationship between endophyte and host plant
variation is lacking. Therefore, the objectives of this
paper were to examine the variation in endophytes
and their respective hosts and to test host specificity
of fungal endophytes within grass species by using
AFLP molecular data.
MATERIALS AND METHODS
Fungal isolation and DNA extraction.—Twenty-four strains
of Neotyphodium endophytes were isolated from surface-
sterilized leaf sheaths (Bacon and White 1994) of three
grass species: F. arundinacea (11 genotypes), F. pratensis
(seven genotypes) and L. perenne (six genotypes) collected
from wild populations in regions of Iran (TABLE I). Isolates
were grown on potato dextrose agar containing 50 mg/mL
antibiotics (chloramphenicol, erythromycin and streptomy-
cin) at 22 C in the dark. Endophyte cultures were identified
on the basis of macroscopic and microscopic characteristics
of individual colonies (Morgan-Jones and Gams 1982).
For DNA extraction, 200 mg fresh mycelium was frozen in
liquid nitrogen, ground to fine powder and transferred to a
microcentrifuge tube. Fungal genomic DNA was extracted
with the method of Murray and Thompson (1980). DNA
from host grasses was purified, as described by Dellaporta
et al. (1983). To avoid endophyte DNA contamination in
plant tissues, only the upper third part of plant leaves were
sampled for DNA extraction. Microscopic examination
(Saha et al. 1988) revealed no evidence of endophyte
colonization in this leaf section. In addition, DNA from
isolated endophytic fungi on nutrient media was used for
PCR by plant primers and no amplification was observed.
Molecular identification of Neotyphodium species.—The
DNA extracted from fungal isolates was used in polymer-
ase chain reactions (PCR), using IS-1 (59–GGTGTTGAG-
CCCCCCTGATTT–39) and IS-3 (59–GTCTCATCTCCGG-
GGCGGTAT–39) primer pairs specific to the nucleotide
sequence of introns 1 and 3 of the tubulin 2–4 (tubB) gene
(Tsai et al. 1994). Each DNA sample was used as a
template for PCR (Doss et al. 1998). For each amplifica-
tion, 25 mL containing template DNA, 10 mM Tris-HCl
(pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 1.0 mM of each
primer (IS-1 & IS-3), 200 mM of each dNTP, and 0.75 U
Taq DNA polymerase was prepared. Each reaction
mixture was sealed with a thin layer of mineral oil, and
DNA was amplified with a Biometra DNA thermo-cycler
(Biometra, Germany). Thermal-cycling parameters con-
sisted of one initial cycle of denaturation at 94 C for 1 min,
followed by 35 cycles with denaturation at 94 C for 1 min,
annealing at 62 C for 1 min and extension at 74 C for
1 min. A final cycle was performed with an extension
segment of 72 C for 10 min. Amplification products were
separated on 2% agarose gels by electrophoresis and
stained with ethidium bromide.
AFLP analysis of endophyte-host genomic DNA.—The AFLP
procedure was conducted according to the protocol of Vos
et al. (1995) with minor modifications. Template DNA
(200 ng/mL) was digested with 2.5 U PstI and 2.5 U Tru9I
overnight at 37 C; then the products were ligated to primers
complementary to the adapter sequences (59–GACTGCG-
TAGGTGCA–39 and 59–GATGAGTCCTGAGTAA–39) with
no selective nucleotides. For pre-amplification reactions,
ligated template DNA (2 mL) was added to 2.5 mL 103 AFLP–
PCR buffer (100 mMTris-HCl [pH 8.3], 15 mM MgCl2,
500 mM KCl), 1 U Taq DNA polymerase, 50 ng PstI primer,
50 ng Tru9I primer, 0.2 mM each dNTP and water to a final
volume of 25 mL. PCR conditions were 26 cycles of 94 C for
1 min, 56 C for 1 min and 72 C for 1 min. The amplification
product was diluted fivefold in TE buffer and stored at 220 C
until used for selective amplification. Among the primers,
selective amplification was carried out with 11 combinations
of PstI + 3 and Tru9I + 3 primer pairs for grass samples and
eight selective primer pairs of PstI + 3 and Tru9I + 2 for
fungus samples (TABLE II). Selective PCR (20 mL vol.)
contained 5 mL pre-amplified template, 60 ng PstI primer,
60 ng Tru9I primer, 2 mL 103 AFLP–PCR buffer and 0.2 mM
each dNTP. PCR conditions were an initial cycle of 30 s at
94 C, 30 s at 65 C and 60 s at 72 C, followed by a touchdown
phase where the annealing temperature was lowered 0.7 C
each cycle for 12 cycles and finally 23 cycles of 30 s at 94 C, 30 s
at 56 C, and 60 s at 72 C.
 KARIMI ET AL.: HOST SPECIFICITY OF FUNGAL ENDOPHYTES 1283

TABLE I. Plant genotypes, province and corresponding endophytic isolates collected from Iran

No. Province (region) Host Isolate Host code

1 Chaharmahal bakhtiari (Gandoman) Festuca arundinacea FaGn1 M211

2 Chaharmahal bakhtiari (Gandoman) Festuca arundinacea FaGn2 M213
3 Chaharmahal bakhtiari (Gandoman) Festuca arundinacea FaGn3 M243
4 Chaharmahal bakhtiari (Gandoman) Festuca arundinacea FaGn4 M263
5 Chaharmahal bakhtiari (Galoogerd) Festuca arundinacea FaGd1 N622
6 Chaharmahal bakhtiari (Galoogerd) Festuca arundinacea FaGd2 N632
7 Chaharmahal bakhtiari (Galoogerd) Festuca arundinacea FaGd3 N651
8 Chaharmahal bakhtiari (Galoogerd) Festuca arundinacea FaGd4 N692
9 Kordestan (Kamyaran) Festuca arundinacea FaKn1 75A
10 Kordestan (Kamyaran) Festuca arundinacea FaKn2 75B
11 Kordestan (Kamyaran) Festuca arundinacea FaKn3 75C
12 Chaharmahal bakhtiari (Borujen) Festuca pratensis FpBn1 M812
13 Chaharmahal bakhtiari (Borujen) Festuca pratensis FpBn2 M813
14 Chaharmahal bakhtiari (Borujen) Festuca pratensis FpBn3 M822
15 Chaharmahal bakhtiari (Gandoman) Festuca pratensis FpGna1 M412
16 Chaharmahal bakhtiari (Gandoman) Festuca pratensis FpGna2 M413
17 Chaharmahal bakhtiari (Gandoman) Festuca pratensis FpGnb1 60A
18 Chaharmahal bakhtiari (Gandoman) Festuca pratensis FpGnb2 60B
19 Lorestan (Borujerd) Lolium perenne LpBd1 B5
20 Lorestan (Borujerd) Lolium perenne LpBd2 B7
21 Lorestan (Borujerd) Lolium perenne LpBd3 B112
22 Lorestan (Borujerd) Lolium perenne LpBd4 B114
23 Azarbayejan gharbi (Naghadeh) Lolium perenne LpNh1 N210
24 Azarbayejan gharbi (Naghadeh) Lolium perenne LpNh2 N212

Fifteen microliters of selective amplification product was
mixed with 15 mL deionized formamide and denatured at 95 C
for 10 min. Then the reaction products were separated on a 33 3
43 cm, 6% polyacrylamide gel (7.5 M urea, 8.5 g acrylamide, 1.5 g
bisacrylamide, 100 mL TBE 53, 3 ppm APS, 1 ppm TEMED and
water to a final volume of 500 mL), and the samples were
electrophoresed on a Biometra (Biometra, Goettingen, Ger-
many) sequencing gel. The injection time was 6 s, running time
was 1 h (100 W, 50 C) and gel was stained with AgNO3 (AgNO3
1% + 10 mL 1.53 formaldehyde 37%) (Bassam et al. 1991).
host species, a three-level AMOVA was performed on AFLP
data separately for plant and endophyte with ARLEQUIN 3
(Excoffier 2000). The variation attributable to the differ-
ences among species, among origin (collection) sites within
species and among individuals within origin sites was
estimated. For each level, variance component was calcu-
lated and expressed as a percentage. The significance of
each variance component was determined by the random
permutation test.

Data analysis.—Unambiguous AFLP bands were manually RESULTS

scored as present (1) or absent (0) in individual lanes. Both
monomorphic and polymorphic bands were included in the
binary dataset to provide unbiased estimates of genetic
similarity. Estimates of similarity among all genotypes were
calculated according to the Dice (1945) definition of
similarity: Sij 5 2a/(2a + b + c), where Sij is the similarity
between two individuals i and j, a is the number of bands
present in both individuals, b is the number of bands
present in i and absent in j, and c is the number of bands
present in j and absent in i. Scores were recorded and
formatted for analysis by NTSYS (Rohlf 1992). The similarity
matrix used for construction of a dendrogram was analyzed
by the unweighted pair group mean average (UPGMA)
method. Maximum parsimony (MP) also was applied to
data of the host species and their endophytes to estimate
phylogeny from the distance (1- Sij) matrix to obtain an
unrooted tree using the MEGA4 program from the MEGA
software package (Tamura et al. 2007). To reveal the
geographical basis for diversity within each endophyte and
Detection of Neotyphodium species.—All endophytes
isolated from surface-sterilized plant tissues agreed
with the morphological description of Neotyphodium
species (Christensen et al. 1993) and had slow to
moderate growth on PDA. Cultures of N. coenophia-
lum usually were slow growing (2.8 mm/wk), white to
tan, and hyphal growth generally resulted in wrinkled
colonies with irregular edges after 6 wk. The lunate to
reniform conidia were 6.5–12.0 mm long 3 1.5–3.0 mm
wide. The typical morphological features associated
with Neotyphodium uncinatum were short conidio-
phores (9.1–15.6 mm), singly branched producing
single uncinate conidia, 10.0–13.0 3 1.0–1.3 mm. In
N. lolii the colonies were characterized by a cerebroid
and waxy morphology and yellowish colony. Long
conidiophores (10.4–84.5 mm) give rise to small
reniform conidia, which were 3.2–6.0 mm long 3
 1284

TABLE II. Selective primer combinations for AFLP analysis of plant genotypes (top) and endophyte isolates (bottom), their total amplification results and the results
based on each species

F. arundinacea F. pratensis L. prenne Total

Plant primers No. band Polymorphism% No. band Polymorphism% No. band Polymorphism% No. band Polymorphism%
a
P(AAA)-bM(cCCT) 21 76.2 25 84 21 80.9 40 93
P(AGC)-M(CAC) 37 59.4 34 82.4 26 84.6 57 96.6
P(AAA)-M(CCC) 31 74.2 30 90 32 96.9 59 100
P(AAA)-M(CTT) 10 40 19 57.9 23 95.7 38 90.5
P(AAA)-M(CGG) 12 58.3 11 90.9 9 100 26 89.6
P(AAA)-M(CTC) 21 85.7 20 90 18 83.3 47 95.9
P(AGC)-M(CCC) 16 81.3 19 78.9 14 71.4 39 95.1
P(AGC)-M(CGG) 15 53.3 16 56.3 12 100 31 91.2
P(AAA)-M(CGT) 6 50 4 75 12 66.7 20 90.9
P(AAA)-M(CGC) 11 54.5 18 94.4 11 63.6 33 97
P(AGC)-M(CGT) 5 71.4 8 72.7 9 90 22 95.6
Total bands/average
polymorphism 185 64 204 79.3 187 84.8 412 94.7
MYCOLOGIA

N. coenophialum N. uncinatum N. lolii Total

Fungal primers No. band Polymorphism% No. band Polymorphism% No. band Polymorphism% No. band Polymorphism%

P (AGC)- M (TC) 7 71.4 7 57.1 7 100 18 94.7

P(AAA)-M(CG) 12 58.3 15 93.3 13 92.3 30 93.7
P(AAA)-M(TA) 33 72.7 35 97.1 40 95 55 91.6
P(AAA)-M(TG) 11 27.3 9 88.9 5 20 22 95.6
P(AAA)-M(CCG) 8 87.5 5 60 13 100 24 85.7
P(AAA)-M(CT) 21 80.9 23 78.3 24 79.2 48 90.6
P(AGC)-M(AG) 10 90.9 12 85.7 15 93.7 25 92.6
P(AGC)-M(TG) 15 88.2 5 100 6 85.7 20 90.9
Total bands/average
polymorphism 117 72.1 111 82.5 123 83.2 242 91.3
a
Primers complementary to the adapter sequence of PstI, 59–GACTGCGTAGGTGCA–39.
b
Primers complementary to the adapter sequence of Tru9I, 59–GATGAGTCCTGAGTAA–39.
c
Selected nucleotides added to 39 end of primers.
 KARIMI ET AL.: HOST SPECIFICITY OF FUNGAL ENDOPHYTES
FIG. 1. Conidia (arrow 1) and conidiophores (arrow 2) in isolates of Neotyphodium coenophialum (A) N. lolii (B) and N.
uncinatum (C). Courtesy of S. DehghanpoorFarashah, Isfahan University of Technology.
1.0–1.7 mm wide (FIG. 1). Based on these morpholog-
ical characteristics, it was determined that F. arundi-
nacea plants harbored N. coenophialum, L. perenne
contained N. lolii and F. pratensis plants were infected
by N.uncinatum (Christensen et al. 1993).
PCR using IS-1 and IS-3 primers successfully
amplified the intended DNA 444 bp fragment from
genomic DNA of fungal endophytes isolated from F.
arundinacea, F. pratensis and L. perenne. This frag-
ment, which is derived from intervening sequence of
a b-tubulin gene (tubB-4) has been reported to be
specifically amplified from genomic DNA of Neoty-
phodium species (Doss et al. 1998). Therefore, the
DNA analysis using specific primers confirmed the
macroscopic and microscopic identification of Neoty-
phodium species isolated from the three grass species.
AFLP analysis.—AFLP fingerprinting of the 24 plant
genotypes using 11 selective primer combinations
resulted in the amplification of a total 435 restriction
fragments. Restriction fragment was 70–500 bp. Of
the 435 fragments detected, 23 were common to all
genotypes and 412 (94.7%) were polymorphic, with
an average of 37.45 bands per primer (20–59). AFLP
with the P-AAA and M-CCC primer combinations
resulted in the greatest number of polymorphic
fragments (100%), whereas use of the P-AAA and M-
CTT primers and P-AAA and M-CGT primers yielded
the lowest number of polymorphic fragments (90.5
and 90.9% respectively). On average there was greater
polymorphism in genotypes of L. perenne (84.8%)
1285
than F. pratensis (79.3%) and F. arundinacea (64.0%)
(TABLE II).
AFLP fingerprinting of the 24 corresponding
fungal endophytes using eight selective primer
combinations resulted in the amplification of 265
restriction fragments. Restriction fragments were 70–
500 bp. Of the 265 fragments detected, 23 were
common to all isolates and 242 fragments (91.3%)
were polymorphic. A range of 18–55 polymorphic
bands were detected for the primers. Use of the P-
AAA and M-TG primer combinations generated the
greatest number of polymorphic fragments (95.6%),
while the P-AAA and M-CCG primers yielded the
fewest polymorphic fragments (85.7%). On average,
there was higher polymorphism in isolates of N. lolii
(83.2%) than N. uncinatum (82.5%) and N.coeno-
phialum (72.1%) (TABLE II).
Cluster and phylogenetic analysis.— For the host
species, genetic distances between genotypes were
0.19–0.96 based on Dice similarity coefficients. Geno-
types N622 and N632, collected from Chaharmahal
bakhtiari province, had the smallest genetic distance.
These two genotypes were from nearby locations. The
results of AFLP data for 24 plant genotypes showed
grouping in four distinct clusters in the dendrogram
generated with the UPGMA method (FIG. 2a). Cluster
1 included two L. perenne genotypes (N210, N212)
from Azarbayejan Gharbi province. Cluster 2 included
the other four L. perenne genotypes (B5, B7, B112,
B114) collected from Lorestan province. Cluster 3
 1286
MYCOLOGIA

FIG. 2. Dendrogram and phylogram for 24 accessions of F. arundinacea, F. pratensis and L. perenne and their respective endophytes, reconstructed by (a) UPGMA and
(b) MP methods respectively. The accession codes are the same as those in TABLE I. Numbers on the branches indicate the bootstrap values for 100 replicates and arrows
indicate the position of unmatched branches.
 KARIMI ET AL.: HOST SPECIFICITY OF FUNGAL ENDOPHYTES
TABLE III. Results of analysis of molecular variance (AMOVA)
Source of variation df Sum of variation
Endophytic fungi
Among species 2 256.36
Among collection sites within species 4 130.84
Within collection sites 17 140.5
Total 23 527.71
Plant hosts
Among species 2 623.57
Among collection sites within species 4 279.29
Within collection sites 17 269.67
Total 23 1172.54
a
After 1023 random permutations.
involved all genotypes of F. pratensis and the last
cluster contained all genotypes of F. arundinacea
(FIG. 2a). Cluster analysis indicated greater similarity
between F. pratensis and F. arundinacea.
Genetic distances between endophyte isolates were
0.06–0.98 based on Dice similarity coefficients.
Isolates M211 and M213, which belonged to two tall
fescue genotypes from nearby locations, had the
minimum genetic distance. The results of UPGMA
analyses on AFLP data of corresponding fungal
isolates grouped endophytes into three distinct
clusters. Cluster 1 included all isolates of N. lolii
obtained from L. perenne plants as the host, Cluster 2
contained all N. uncinatum endophytes isolated from
F. pratensis and Cluster 3 included all N. coenophialum
isolates from F. arundinacea (FIG. 2a).
Maximum parsimony (MP) yielded results similar
to UPGMA on AFLP data. F. arundinacea, F. pratensis
and L. perenne plus their corresponding endophytes
were distinctly separated from each other (FIG. 2b).
With the MP analysis, the grouping order of
endophyte species almost exactly matched grouping
of their respective host species and supported the
classification performed by UPGMA (FIG. 2a). At the
clade level, the endophyte and hosts phylogenetic
trees showed a better match than did the UPGMA
dendrogram.
The result of the AMOVA analysis revealed that
49.21% and 43.94% of the total variation was
attributable to differences among plant species
and their corresponding endophytes, respectively.
The results also showed that 25.62% and 26.32% of
the total variation was accounted for by the differences
among collection sites within plant and endophyte
species, respectively. Similarly, 25.17% and 29.74%
variation was attributable to the differences among
individual within collection sites for plants and corres-
ponding endophytic isolates. A random permutation
1287
Variance Percentage of
components variation Pa
12.21 43.94 ,0.001
7.31 26.32 ,0.001
8.26 29.74 ,0.001
27.79
31 49.21 ,0.001
16.14 25.62 ,0.001
15.86 25.17 ,0.001
test also revealed that these variance components all
were significant (P , 0.001) (TABLE III).
DISCUSSION
We detected a high degree of genetic diversity within
and between populations of endophytic fungi,
N.coenophialum, N.uncinatum and N. lolii, isolated
respectively from F. arundinacea, F. pratensis and
Lolium prenne. Other researchers have reported a
small amount of genetic variation within these
species. van Zijll de Jong et al. (2003, 2008) found
a low level of polymorphism within N. coenophialum
and N. lolii, whereas high polymorphism was found
between Neotyphodium and Epichloë species based on
SSR markers. In contrast, Groppe et al. (1995)
analyzed genetic diversity within Epichloë typhina
fungal endophytes isolated from Bromus erectus,
using SSR markers, and reported high genetic
diversity. Using AFLP markers, Arroyo- Garcia et al.
(2002) also observed significant genetic variation
within populations of E. festucae and a low differen-
tiation between populations isolated from natural
grasslands in Spain. The results of the present study
revealed that the differentiation between popula-
tions of Neotyphodium species and within population
variation detected by AFLP markers were significant
(TABLE III). This probably can be attributed to
collection of host genotypes from diverse geogra-
phical regions.
Maximum parsimony analysis of AFLP fragments
from host plants revealed a strong similarity between
meadow fescue and tall fescue. The origin of tall
fescue is still under debate, but Gaut et al. (2000)
suggested that meadow fescue (F. pratensis) and
ryegrass (L. perenne) originated from the same genus,
Schedonorus, and later F. arundinacea was derived
from F. pratensis through hybridization with 43 L.
1288 MYCOLOGIA
arundinaceum. Viewed in this context, the hexaploid
F. arundinacea contains a genome from F. pratensis,
thereby making it genetically much closer to F.
pratensis than to L. perenne. Results of the AFLP
analysis are consistent with a close phylogenetic
relationship between F. arundinacea and F. pratensis.
For Neotyphodium species, maximum parsimony
analysis did not show less genetic distance between
isolates of N. uncinatum and N. coenophialum than N.
lolii based on their hosts’ similarity. This may be
explained by molecular evidence indicating the hybrid
nature of Neotyphodium species. At least two hybridiza-
tion events were involved in the evolution of N.
coenophialum, the symbiont of tall fescue (Tsai et al.
1994). Actually it is suggested that N. coenophialum is a
hybrid of E. festucae, E. typhina and an undescribed
(perhaps extinct) Epichloë species most closely related
to E. baconii (Tsai et al. 1994). On the other hand, N.
uncinatum, the endophyte of F. pratensis, is character-
ized by a one-way hybrid (E. bromicola 3 E. typhina)
based on the analysis of tubB and tef-1(Moon et al.
2000, Craven et al. 2001). Considering this, the
endophyte of F. arundinacea does not have exactly
the same ancestors as the endophyte of F. pratensis.
Therefore a genetic relationship similar to that
inferred for hosts is not possible among these three
Neotyphodium species given the known hybrid pedi-
grees of N. coenophialum and N. uncinatum and the
non-hybrid origin of N. lolii.
The grouping order of endophyte species almost
matched the grouping of their respective host species
in the MP phylogram. This matching suggests
specificity of host-endophyte combinations even
within geographically distant populations of each
single-host species, possibly due to asexual reproduc-
tion, vertical transmission and geographical isolation
of fungal strains. Chung et al. (1997) reported such a
co-evolutionary specificity between perennial ryegrass
and orchard grass genotypes and sexually reproduc-
ing, horizontally transmitted E. typhina isolates. Piano
et al. (2005) reported the same coincidence between
the native Sardinian fescue germplasm and its
associated asexual reproducing, vertically transmitted
Neotyphodium endophytes. However, more evidence
of specificity such as cross-inoculation experiments is
needed to confirm the findings of this study. Using a
novel approach of MRCA (most recent common
ancestor), Schardl et al. (2008) suggested that with
relatively few exceptions evolution of symbiotic
Epichloe fungi largely track evolution of their grass
hosts. This implies that this symbiotic relationship has
emerged coincidentally with the emergence of grass
subfamily, Pooideae.
The result of the AMOVA on AFLP data accounted
for a large and significant proportion of genetic
variation due to differences among plant and
endophyte species. The remaining variation was
partitioned between collection sites (within species)
and individuals (within collection sites) almost
equally. The variation of collection sites within plant
and endophyte species (geographical diversification)
and the variation among individual plants and fungal
isolates also were significant. It has been reported that
out-crossing plant species such as Lolium and Festuca
have a significant percentage of variation attributable
to differences among individuals within populations
(Martin et al. 1997). Finding the same amount of
variation among asexual endophytic isolates, as in the
present study, seems remarkable. In this study the
percent of variation attributed to geographical
diversification was also as high as the variation
accounted for differences among individuals indicat-
ing the geographical significance for diversity within
each endophyte and host species. Interestingly, there
also was correspondence between the percentage of
variation in plant and endophyte AMOVA at each
level (species, collection sites and individuals). This
might confirm the coincidence observed between
plants and endophyte phylograms.
In summary, the clustering and phylogeny in F.
arundinacea, F. pratensis and L. perenne and their
corresponding endophytes was reconstructed on
AFLP molecular data using UPGMA and MP. The
analyses yielded similar results and both dendrogram
and phylogram outlined the same set of species
groups, with the conflicts occurring in the arrange-
ment of plant genotypes or endophytic isolates
within, not between, clades. Both methods also
suggested a close genetic relationship between F.
pratensis and F. arundinacea; however, evolutionary
interpretation of endophyte species was precluded
because of the hybrid origin of two of the three
endophytes. Despite of the sexual nature of Epichloë
species and hybrid nature of some Neotyphodium
species, the results presented here suggest a co-
evolutionary specificity between host plant and its
associated endophyte based on AFLP agreement
between both organisms. It seems that evolution of
host genotype may parallel the evolution of Neotypho-
dium fungal endophyte within each species in the
Festuca-Lolium complex.
ACKNOWLEDGMENTS
The authors acknowledge Dr Christopher L. Schardl for
critical review and constructive suggestions provided for this
manuscript. The authors also greatly appreciate Timothy
Flanagan from Writing Resource, Portland, Oregon, who
provided English language editing.
 KARIMI ET AL.: HOST SPECIFICITY OF FUNGAL ENDOPHYTES
LITERATURE CITED
Arachevaleta M, Bacon C, Hoveland C, Radcliffe D. 1989.
Effect of the tall fescue endophyte on plant response to
environmental stress. Agron J 81:83–90, doi:10.2134/
agronj1989.00021962008100010015x
Arroyo-Garcia R, MartinezZapater J, GarciaCriado B, Zabal-
gogeazcoa I. 2002. Genetic structure of natural
populations of the grass endophyte Epichloë festucae
in semiarid grasslands. Mol Ecol 11:355–364, doi:10.
1046/j.0962-1083.2001.01456.x
Bacon C, Siegel M. 1988. Endophyte parasitism of tall
fescue. J Prod Agric 1:45–55.
———, White JJ. 1994. Stains, media, and procedures for
analyzing endophytes. In: Bacon CW, White JJ, eds.
Biotechnology of endophytic fungi of grasses. Boca
Raton, Florida: CRC Press. p 47–56.
Bassam JB, Gustavo CA, Peter MG. 1991. Fast and sensitive
silver staining of DNA in polyacrylamide gel. Anal
Biochem 196:80–83, doi:10.1016/0003-2697(91)90120-I
Buck GW, West CP, Elbersen HW. 1997. Endophyte effect
on drought tolerance in diverse Festuca species. In:
Bacon CW, Hill NS, eds. Neotyphodium/grass interac-
tions. New Yor, London: Plenum Press. p 141–144.
Christensen MJ. 1995. Variation in the ability of Acremonium
endophytes of perennial ryegrass (Lolium perenne), tall
fescue (Festuca arundinacea) and meadow fescue (F.
pratensis) to form compatible associations in the three
grasses. Mycol Res 99:466–470, doi:10.1016/S0953-
7562(09)80647-3
———, Leuchtmann A, Rowan DD, Tapper BA. 1993.
Taxonomy of Acremonium endophytes of tall fescue
(Festuca arundinacea), meadow fescue (F. pratensis)
and perennial ryegrass (Lolium perenne). Mycol Res 97:
1083–1092, doi:10.1016/S0953-7562(09)80509-1
Chung KR, Hollin W, Siegel MR, Schardl CL. 1997. Genetics
of host specificity in Epichloë typhina. Phytopathol 87:
599–605, doi:10.1094/PHYTO.1997.87.6.599
Craven KD, Blankenship JD, Leuchtmann A, Hignight K,
Schardl CL. 2001. Hybrid fungal endophytes sym-
biotic with the grass Lolium pratense. Sydowia 53:44–
73.
Dellaporta SL, Wood VP, Hicks JB. 1983. A plant DNA mini-
preparation: version II. Plant Mol Biol Rep 1:19–21,
doi:10.1007/BF02712670
Dice LR. 1945. Measures of the amount of ecologic
association between species. Ecology 26:297–302,
doi:10.2307/1932409
Doss RP, Clement SL, Kuy SR, Welty RE. 1998. PCR-based
technique for detection of Neotyphodium endophytes in
diverse accessions of tall fescue. Plant Dis 82:738–740,
doi:10.1094/PDIS.1998.82.7.738
Excoffier L. 2000. ARLEQUIN: a software for population
genetics data analysis. Program manual, http://anthro.
unige.ch/arlequin
Gaut BS, Tredway LP, Kubik C, Gaut RL, Meyer W. 2000.
Phylogenetic relationships and genetic diversity among
members of the Festuca-Lolium complex (Poaceae)
based on ITS sequence data. Plant Syst Evol 224:33–53,
doi:10.1007/BF00985265
1289
Glenn AE, Bacon CW, Price RP, Hanlin RT. 1996. Molecular
phylogeny of Acremonium and its taxonomic implica-
tions. Mycologia 88:369–383, doi:10.2307/3760878
Groppe K, Sanders I, Wiemken A, Boller T. 1995. A
microsatellite marker for studying the ecology and
diversity of fungal endophytes (Epichloë spp.) in grasses.
Appl Environ Microbiol 61:3943–3949.
Johnson-Cicalese JJ, Secks ME, Lam CK, Meyer WA, Murphy
JA, Belanger FC. 2000. Cross species inoculation of
chewing and strong creeping red fescue with fungal
endophytes. Crop Sci 40:1485–1489, doi:10.2135/
cropsci2000.4051485x
Kearney JF, Parrott WA, Hill NS. 1991. Infection of somatic
embryos of tall fescue with Acremonium coenophialum.
Crop Sci 31:979–984, doi:10.2135/cropsci1991.00111
83X003100040028x
Koga H, Christensen MJ, Bennett RJ. 1993. Incompatibility of
some grass-Acremonium endophyte associations. Mycol
Res 97:1237–1244, doi:10.1016/S0953-7562(09)81292-6
Leuchtmann A. 1993. Nonreciprocal compatibility between
Epichloë typhina and four host grasses. Mycologia 85:
157–163, doi:10.2307/3760451
———, Clay K. 1990. Isozyme variation in the Acremonium/
Epichloë fungal endophyte complex. Phytopathol 80:
1133–1139, doi:10.1094/Phyto-80-1133
Malinowski DP, Belesky DP. 2000. Adaptations of endo-
phyte-infected cool-season grasses to environmental
stresses: mechanisms of drought and mineral stress
tolerance. Crop Sci 40:923–940, doi:10.2135/cropsci2000.
404923x
Martin C, Gonzalez-benito ME, Iriondo JM. 1997. Genetic
diversity within and among populations of a threatened
species: Erodium paularense Fern. Gonz. & Izco. Mol Ecol
6:813–820, doi:10.1111/j.1365-294X.1997.tb00135.x
Mian MAR, Hopkins AA, Zwonitzer JC. 2002. Determination
of genetic diversity in tall fescue with AFLP markers.
Crop Sci 42:944–950, doi:10.2135/cropsci2002.0944
Moon CD, Scott B, Schardl CL, Christensen MJ. 2000.
Evolutionary origins of Epichloë endophytes from
annual ryegrasses. Mycologia 92:1103–1118, doi:10.
2307/3761478
Morgan-Jones G, Gams W. 1982. Notes on Hyphomycetes
XLI. An endophyte of Festuca arundinacea and the
anamorph of Epichloë typhina, new taxa in one of two
section of Acremonium. Mycotaxon 15:316–318.
Murray MG, Thompson WF. 1980. Rapid isolation of high
molecular weight plant DNA. Nucleic Acids Res 8:
4321–4325, doi:10.1093/nar/8.19.4321
Piano E, Bertoli FB, Romani M, Tava A, Riccioni L, Valvassori M,
Carroni AM, Pecetti L. 2005. Specificity of host-endophyte
association in tall fescue population from Sardinia, Italy.
Crop Sci 45:1456–1463, doi:10.2135/cropsci2004.0287
Rohlf FJ. 1992. NTSYS-Pc 2.02 k: numerical taxonomy and
multivariate analysis system. State University of New
York, Stony Brook.
Sabzalian MR, Hatami B, Mirlohi A. 2004. Mealybug,
Phenococcus solani, and barley aphid, Sipha maydis,
response to endophyte-infected tall and meadow
fescues. Entomol Exp Appl 113:205–209, doi:10.1111/
j.0013-8703.2004.00227.x
1290 MYCOLOGIA
———, Mirlohi A. 2010. Neotyphodium endophytes trigger
salt resistance in tall and meadow fescues. J Plant Nutr
Soil Sci 173:952–957, doi:10.1002/jpln.200900345
Saha DC, Jackson MA, Johnson-Cicalese JM. 1988. A rapid
staining method for detection of endophytic fungi
in turf and forage grasses. Phytopathol 78:237–239,
doi:10.1094/Phyto-78-237
Schardl CL. 1996. Epichloë species: fungal symbionts of
grasses. Annu Rev Phytopathol 34:109–130, doi:10.1146/
annurev.phyto.34.1.109
———, Craven KD, Speakman SAS, Lindstrom A, Yoshida R.
2008. A novel test for host-symbiont codivergence
indicates ancient origin of fungal endophytes in grasses.
Syst Biol 57:483–498, doi:10.1080/10635150802172184
———, Leuchtmann A, Chung KR, Penny D, Siegel MR. 1997.
Coevolution by common descent of fungal symbionts
(Epichloë spp.) and grass hosts. Mol Biol Evol 14:133–143,
doi:10.1093/oxfordjournals.molbev.a025746
———, Phillips TD. 1997. Protective grass endophytes:
Where are they from and where are they going? Plant
Dis 81:430–438, doi:10.1094/PDIS.1997.81.5.430
Tamura K, Dudley J, Nei M, Kumar S. 2007. MEGA4: molecular
evolutionary genetics analysis (MEGA) software. Mol
Biol Evol 24:1596–1599, doi:10.1093/molbev/msm092
Tredway LP, White JJ, Gaut BS, Reddy PV, Richardson MD,
Clarke BB. 1999. Phylogenetic relationship within and
between Epichloë and Neotyphodium endophytes as
estimated by AFLP markers and rDNA sequences. Mycol
Res 103:1593–1603, doi:10.1017/S0953756299008874
Tsai HF, Liu JS, Staben C, Christensen MJ, Latch G, Siegel
MR, Schardl CL. 1994. Evolutionary diversification of
fungal endophytes of tall fescue grass by hybridization
with Epichloë species. Proc Natl Acad Sci USA 91:2542–
2546, doi:10.1073/pnas.91.7.2542
van Zijll de Jong E, Dobrowolski MP, Bannan NR, Stewart AV,
Smith KF, Spangenberg GC, Forster JW. 2008. Global
genetic diversity of the perennial ryegrass fungal endophyte
Neotyphodium lolii. Crop Sci 48:1487–1501, doi:10.2135/
cropsci2007.11.0641
———, Guthridge KM, Spangenberg GC, Forster JW. 2003.
Development and characterization of EST-derived simple
sequence repeat (SSR) markers for pasture grass endo-
phytes. Genome 46:277–290, doi:10.1139/g03-001
Vazquez de Aldana BR, Garcia Ciudad A, Zabalgogeazcoa I,
Garcia Criado B. 2001. Ergovaline levels in cultivars of
Festuca arundinacea. Anim Feed Sci Technol 93:169–
176, doi:10.1016/S0377-8401(01)00285-1
Vos P, Hogers R, Bleeker M, Reijans M, Lee T, Hornes M,
Frijters A, Pot J, Peleman J, Kuiper M, Zabeau M. 1995.
AFLP: a new technique for DNA fingerprinting. Nucleic
Acids Res 23:4407–4414, doi:10.1093/nar/23.21.4407
White JJ, Huff DR. 1996. Endophyte-host associations in grasses
XXIV. Some evidence to support the occurrence of endo-
phyte micro-species complexes centered around sexually
reproducing species of Epichloë. Symbiosis 20:219–227.
Wille PA, Aeschbacher RA, Boller T. 1999. Distribution of fungal
endophyte genotypes in doubly infected host grasses. Plant
J 18:349–358, doi:10.1046/j.1365-313X.1999.00462.x