Curr Genet (1997) 31: 361 369
O R I G I N A L PA P E R
Pradeep Kachroo Munish Ahuja Sally A. Leong
Bharat B. Chattoo
Organisation and molecular analysis of repeated DNA sequences
in the rice blast fungus Magnaporthe grisea
Received: 12 April / 12 November 1996
Abstract The distribution of a previously described re-
peated DNA sequence present as a 1.3-kb PstI fragment in
the genome of the rice blast fungus Magnaporthe grisea
was analysed by carrying out DNA fingerprint analysis of
36 isolates including rice, non-rice and laboratory strains.
The analysis of various higher-molecular-weight PstI frag-
ments with homology to the 1.3-kb repeat revealed that
these may arise predominantly from transposon insertions
or point mutations. Analysis of a 5.1-kb derivative revealed
both a point mutation at a PstI site and an insertion of a
putative transposable element which caused an increase in
molecular weight from 1.3 to 5.1 kb. Another repeat ele-
ment of 1.4 kb was identified and found to exist in associ-
ation with the 1.3-kb repeat. Both 1.3- and 1.4-kb elements
were found to be parts of MGR583 (Hamer et al. 1989), a
LINE-like element. These elements were present in a high
copy number in all the rice and a majority of non-rice patho-
gens indicating that MGR583 is not a host-specific se-
quence as reported earlier. Our results suggest that repeated
DNA elements in M. grisea have amplified independently
of one another and further indicate that different isolates
of M. grisea may have evolved from several distinct lines
of origin.
Key words Repeated DNA Rice blast
Fingerprinting Transposon
P. Kachroo M. Ahuja B. B. Chattoo () 1993).
Department of Microbiology and Biotechnology Centre,
M.S. University of Baroda, Baroda-390 002, India
P. Kachroo S. A. Leong
Department of Plant Pathology and
USDA-ARS Plant Disease Research Unit,
1630 Linden Drive, University of Wisconsin,
Madison, WI-53706, USA
Communicated by R. J. Rothstein
Springer-Verlag 1997
Introduction
Repeated DNA sequences are a common feature of eukar-
yotic genomes (Berg and Howe 1989). These DNA se-
quences may be dispersed, present in clusters, or arranged
in tandem arrays (Schmidt 1984; Jeffreys et al. 1985; Char-
lesworth et al. 1994) and are known to be involved in ge-
nomic rearrangements such as deletions, inversions
(Roeder and Fink 1980; Rothstein et al. 1987) or recombi-
nation (Petes 1980; Roeder and Fink 1980; Scherer and
Davis 1980). One unique group of repeated DNA se-
quences are transposable elements which are mobile and
capable of causing insertions at random locations in the
genome. The processes leading to changes in genomic ar-
chitecture are thought to be pre-requisites for evolution and
may facilitate adaptation of the organism to its environ-
ment (Dover 1982).
Several repetitive DNA sequences from the rice blast
fungus Magnaporthe grisea have been cloned and ana-
lysed. Among these, MGR586, MGR583 (Hamer et al.
1989; Dobinson and Hamer 1993) and pMG1065 (Sone
et al. 1993) are reported to be present at a copy number of
50, 100 and 40, respectively, in isolates of M. grisea which
infect rice but are present in 20 or less copies in isolates
infecting other grasses. Differences in the copy number and
polymorphisms seen at the DNA level have been exploited
by several workers to develop fingerprint profiles for
the rice-infecting isolates and as a means of analysing
phylogenetic relationships and distinguishing morpho-
logically similar subgroups of M. grisea at the molecular
level (Hamer et al. 1989; Levy et al. 1991; Borromeo et al.
Isolates of M. grisea are highly variable and generate
new pathogenic variants at a high frequency. Various mech-
anisms have been postulated to explain this (Suzuki 1965;
Giatgong and Frederiksen 1969; Genovesi and Magill
1976; Wyman and Blackburn 1991) and with the discov-
ery of transposable elements in this fungus (Kachroo et al.
1994, 1995; Farman et. al. 1996 a, b; Shull and Hamer
1996; M Lebrun, B Valent and F Chumley, unpublished),
362

it is plausible that these elements may be associated with Table 1 Geographic distribution of Magnaporthe grisea isolates
this pathogenic variability. used in this study
We have initiated a study of repetitive DNA sequences Isolate Host Origin
to determine if the genomic rearrangements caused by
these repeats could be associated with the high level of Isolate F Oryza sativa India
pathogenic variability seen among isolates of M. grisea. Isolate B O. sativa India
Isolate E O. sativa India
Earlier we reported the isolation of an inverted repeat trans- B101 O. sativa India
poson which was discovered as an insertion within a 1.3-kb B103 O. sativa India
PstI repeat element (Kachroo et al. 1994). Here we report B157 O. sativa India
the characterisation of a 5.1-kb derivative of the 1.3-kb JMB 840610a O. sativa Philippines
repeat, and demonstrate how complex repeat elements can P06-6a O. sativa Philippines
Isolate 101a O. sativa Philippines
be derived through insertions or point mutations. Isolate 102a O. sativa Philippines
Isolate 103a O. sativa Philippines
Isolate 104a O. sativa Philippines
V8601a O. sativa Philippines
4360-17-1 O. sativa USA
91-A-58 O. sativa USA
Materials and methods O-135 O. sativa China
Guy11 O. sativa French Guyana
Strains. Escherichia coli strains JM101 [F traD36 lacI q D (lacZ) 2539 Laboratory strain Laboratory strain
M15 pro A+ B/SupE thi D (lac-pro AB)] and DH5 [F/endA1 hsdR17 4360-4-3b Laboratory strain Laboratory strain
(rk mk+) supE44 thi-1 recA1 gyr A (Nalr) relA1 D (lacZYA-argF) 4091-5-8b Eragrostis curvula Laboratory strain
U169 deoR (f80 dlac D (lacZ) M15] were employed for bacterial Isolate A Setaria sp. India
transformation and plasmid propagation. M. grisea isolates used in Pg3395 Setaria sp. India
this study are listed in Table 1. Our analysis included 36 isolates, Pg 3393 Pennisetum sp. India
17 of which are pathogens of rice. All of the isolates were grown and Ei 476a Elusine indica Philippines
maintained on potato sucrose medium containing 1% sucrose as the Ec 522a Echinochloa colona Philippines
carbon source, as previously described (Kachroo et al. 1994). Pr 342a Panicum repens Philippines
Pr 886a Panicum repens Philippines
DNA isolation and Southern-hybridisation analysis. Isolation of ge- Re 012a Rottboellia exaltata Philippines
nomic DNA was performed as described previously (Kachroo et al. Lc 454 a Leptochloa chinensis Philippines
1994). Plasmid DNA was prepared by the Mini-maxi procedure Cb 334a Cyperus brevifolius Philippines
(Noguchi 1991). Genomic DNA blots were prepared by fractionat- Pd 8824c Paspalum distichum Philippines
ing DNA digested with restriction enzymes on a 0.8% gel (2530 cm, Cd 88215c Cynodon dactylon Philippines
35 V, 4C) run in 0.5 TBE and transferred onto nylon membranes Bd 8401c Brachiaria mutica Philippines
(MSI, Zetaprobe). DNA was fixed onto the membranes by baking at Lh 88490c Leersia hexandra Philippines
80C for 2 h followed by ultraviolet irradiation using a Fisher UV Dc 88428c Digitaris ciliaris Philippines
cross-linker (120 mJ/cm2). Probes for DNA hybridisation were la- Pr 8988c Panicum repens Philippines
belled by the random primer method (Feinberg and Vogelstein 1983)
or by nick translation (Sambrook et al. 1989). Hybridisations were a
DNA from these isolates was kindly provided by Dr. Naweed
carried out at 42C in 0.125 M Na2HPO4, pH 7.2; 0.01 M EDTA, 7% Naqvi, IRRI, Manilla, The Philippines
SDS; 50% formamide (Amasino 1986). Blots were washed once each b
Isolates kindly provided by Dr. F. Chumley, DuPont, USA
in 2 SSPE, 0.1% SDS and 0.1 SSPE, 0.1% SDS at 42C, with a c
Isolates kindly provided by Dr. R. Nelson, IRRI, Manilla, The
final wash in 0.1 SSPE, 0.1% SDS at 65C. Philippines
Nucleic-acid-analysis. The plasmids pBC157.3 (Kachroo et al. 1994)
and pBC101 carry 1.3-kb and 5.1-kb PstI inserts of genomic DNA,
respectively, from isolates B157 and B101, cloned into pUC19. The
5.1-kb fragment was isolated from a sucrose density gradient frac- Linkage analysis. The mapping population was derived from a cross
tionation (1040%, step gradient) of PstI-digested genomic DNA of between isolates Guy11 and 2539 (Skinner et al. 1993). Sixty one
isolate B101. The complete sequence of pBC101 was determined by random ascospore progeny were used to determine the segregation
constructing subclones in pUC19 and carrying out double-stranded of bands hybridising with the 1.3- and 1.4-kb probes and the segre-
sequencing with Sequenase (USB). The DNA sequence of the inter- gation data was analysed using MAPMAKER Version 2.0 (Lander
vening gaps was determined using specific oligonucleotide primers. 1987). Parameters for mapping were a minimum LOD of 4.0 and a
Partial sequence analysis of MGR583 was carried out by making spe- maximum recombination fraction of 0.2. The Kosambi mapping
cific subclones in pBluescript and by using oligonucleotide primers function was employed to compute recombination distances in cen-
based on the sequence of the 1.3-kb repeat. Sequence analysis was timorgans (cM).
carried out using DNA Inspector version 3.25 (Textco, W. Labanon,
N.H., USA) and Seqaid II version 3.70 (Rhoads and Roufa, Molec-
ular Genetics Laboratory, Kansas State University, USA) software
programs. DNA alignments were done using the University of Wis- Results
consin Genetics Computer group (Devereux et al. 1984) and Gen-
Bank software programs. The preparation of DNA samples for
pulsed-field gel electrophoresis was carried out as described by Skin- Analysis of the 1.3-kb repeat family
ner et al. (1993). Chromosome-sized DNA molecules were fraction-
ated in 0.8% agarose gels in 0.5 TBE, prepared using Fast-lane ag- Previously we reported the presence of a 1.3-kb multi-copy
arose (FMC). CHEF gel electrophoresis was carried out at 4C for
170 h at 35 V, with a switch time of 60 min in a Bio-Rad-DRII ap-
repeat element in the genomes of both rice and non-rice
paratus. DNA was transferred onto nylon membranes and hybridised M. grisea pathogens (Kachroo et al. 1994). The dispersion
with radiolabelled probes as described above. of the 1.3-kb repeat on different chromosomes was stud-

Fig. 1A, B Southern hybridisation analysis of electrophoretically
separated chromosomes of M. grisea probed with pBC157.3. Chro-
mosomal DNA from: isolate A, lane 1; B101, lane 2; and B157,
lane 3, was fractionated in a 0.8% agarose gel, transferred to a MSI-
nylon membrane and hybridised with the radiolabelled 1.3-kb PstI
fragment of pBC157.3. The molecular size of 4.6 Mbp corresponds
to chromosome 6 of isolate Guy11 (Skinner et al. 1993). A ethidium
bromide-stained gel. B autoradiogram of the Southern blot of the gel
shown in A
ied by CHEF analysis. Chromosomal DNA of three M. gri-
sea isolates, representing both rice and non-rice pathogens,
were separated on CHEF gels and hybridised to the 1.3-kb
repeat probe and showed hybridisation to all the chromo-
somes of M. grisea (Figure. 1). An intense hybridisation
signal was also observed for all the chromosomes, includ-
ing B-chromosomes, which are polymorphic in both num-
ber as well as in size among the isolates studied.
Southern hybridisation analysis of the genomic DNA of
isolates Guy11, 2539, B157 and B101 digested with PstI
and hybridised with pBC157.3 showed an intense band at
1.3 kb and several higher-molecular-weight bands (Figure.
2-II-A). Isolation and analysis of a 3.2-kb band from iso-
late B157 revealed the insertion of an inverted repeat trans-
poson within 1.3-kb repeat (Kachroo et al. 1994). This sug-
gests that other higher-molecular-weight bands may also
arise as a result of DNA rearrangements. The possible
mechanism by which other such high-molecular-weight
bands could arise was further investigated by isolating a
5.1-kb derivative (shown by arrowhead, Fig. 2-II-A) from
the isolate B101.
Analysis of the 5.1-kb derivative
Restriction analysis of the 5.1-kb PstI clone (pBC101) and
Southern hybridisation analysis with the 1.3-kb probe
(pBC157.3) showed that the region hybridising to the
1.3-kb repeat was present in the middle of the clone while
the flanking 2.7-kb BamHI and 1.4-kb HindIII (one of the
BamHI and HindIII sites each being contributed by the vec-
tor) fragments did not show any hybridisation (Fig. 2-I).
Since both pBC101 and pBC157.3 repeats are defined by
363
PstI sites at the ends, it was likely that the region homol-
ogous to the 1.3-kb repeat in pBC101 had lost both PstI
sites, thus generating a higher-molecular-weight fragment
of 5.1 kb. Alternatively, the 1.3-kb repeat could be part of
the transposable element which had lost the PstI sites dur-
ing insertion into a new genomic sequence (pBC101, in the
present context). These two possibilities were evaluated by
carrying out sequencing of pBC101 around the region ho-
mologous to the 1.3-kb repeat.
Sequence analysis of pBC101 revealed that the region
homologous to the 1.3-kb repeat contained only 966 bp of
the 5 end while the 395 bp at the 3 region were not present
in the clone (the orientation of the 1.3-kb repeat being ar-
bitrary, Fig. 2-I). The 5 PstI site of the 1.3-kb region of
pBC101 had undergone a transition mutation (A G,
CTGCAG CTGCGG) leading to loss of the PstI site.
Sequence analysis of the 966-bp region revealed identity
with that of pBC157.3 except for certain point mutations
which are common among members of the 1.3-kb repeat
family.
The 1.4-kb HindIII region upstream of the 5 mutated
PstI site of pBC101 was analysed by Southern hybridisa-
tion by using it to probe a blot containing digested genomic
DNA of M. grisea isolates Guy11, 2539, B157, B101.
The hybridisation pattern revealed it to be a repetitive
DNA element and, like the 1.3-kb repeat family, showed
an intense hybridising band of 1.4 kb (shown by arrow,
Fig. 2-II-B). Both the 1.3- and the 1.4-kb repeat probes also
hybridised to 410 higher-molecular-weight bands that
were polymorphic among different isolates (Fig. 2-II-B).
The similar intensity of hybridisation signal obtained
with the 1.3- and 1.4-kb repeats and their association in the
5.1-kb PstI clone prompted us to analyse if the presence
of the two elements in pBC101 was a chance occurrence
or whether they were parts of a larger repeat element. Anal-
ysis of a genomic DNA library prepared from isolate 2539
(Leong et al. 1994) using the 1.3-kb repeat probe and sub-
sequent hybridisation of positive clones to the 1.4-kb probe
revealed that all the clones hybridising to the 1.3-kb frag-
ment also hybridised to the 1.4-kb fragment. Southern hy-
bridisation carried out using the 1.3-kb and 1.4-kb repeats
as probes against MGR (Hamer et al. 1989) and other avail-
able repeats (Skinner et al. 1993; Kachroo et al. 1994; Le-
ong et. al. 1994; Kachroo et al. 1995; Kachroo and Chat-
too, unpublished results) revealed a strong hybridisation
of both probes to MGR583 (Hamer et al. 1989).
Further confirmation that the 1.3-kb and 1.4-kb repeats
are present as a contiguous fragment in the genome was
obtained by restriction mapping (Fig. 2-I) and sequence
analysis of MGR583. The sequence flanking the 1.3-kb re-
peat region in MGR583 was determined by carrying out
the sequencing reaction using the primers based on the
1.3-kb sequence. Sequence comparison between the region
upstream of the 1.3-kb repeat in MGR583 and the 1.4-kb
sequence revealed an exact match. Since the region up-
stream of the 1.4-kb region in MGR583 is just adjacent to
MGR586 in the lamda clone from which these sequences
were derived (Hamer et al. 1989), it is likely that arbitrary
designation of these clones may have yielded incomplete-
364
Fig. 2 (I) Restriction map of pBC101, MGR583 and pBC157.3.
Asterisks denote the cloning sites in vector pUC19. The black box
represents the 1.3-kb repeat while the stippled box represents the
1.4-kb repeat element. The triangle represents the absence of 151 bp
at the 5 end of Pot3 (Farman et al. 1996 a). The restriction enzyme
sites shown are PstI, P; BssHII, Bs; SalI, S; BamHI, B; HindIII, H;
EcoRI, E. Restriction sites shown by bold-type represent the same
sites in different clones, as determined by sequence analysis. The
DNA sequence shown in this figure is from the junction between the
1.3-kb and 2.7-kb repeats, which shows characteristic features of
LTRs. The TG dinucleotide and the 4-bp region present as an invert-
ed repeat is indicated by bold type. (II) Southern hybridisation anal-
ysis of the genomic DNA of M. grisea probed with (A) the 1.3-kb
PstI fragment of the plasmid pBC157.3 (B) the 1.4-kb HindIII frag-
ment of the plasmid pBC101 and (C) the 2.7-kb BamHI fragment of
length elements. This was recently confirmed by Farman
et al. (1996 a) who sequenced MGR586 and found it to en-
code the inverted repeat transposon Pot3 similar to Pot2
(Kachroo et al. 1994). DNA sequence comparison of Pot3
(Farman et al. 1996 a) with that of the region upstream of
the 1.4-kb repeat in MGR583 showed that, as expected,
this region contains 975 bp of the Pot3 element (EcoRI and
the plasmid pBC101. Genomic DNA (2 mg) was digested with re-
striction enzyme, electrophoresed through a 0.8% agarose gel, trans-
ferred onto a nylon membrane and hybridised with radiolabelled
probe. The lanes in A and B contain DNA from Guy11, lane 1; 2539,
lane 2; B157, lane 3; and B101, lane 4. The conserved 1.3- and
1.4-kb bands are marked by an arrow. The 5.1-kb derivative is
marked by an arrowhead. The band marked in B was present in
both the isolates Guy11 and 2539 as well as the progeny obtained
from a cross between them, while the band marked was absent in
three progeny out of 61 tested indicative of a possible genome re-
arrangement event involving the 1.4-kb repeat element (data not
shown). The lanes in C contain DNA from B157 digested with
EcoRV, lane 1; SalI, lane 2; PstI, lane 3; HindIII, lane 4; BamHI,
lane 5; EcoRI, lane 6; and BanI, lane 7. The conserved SalI and
BamHI sites are marked by arrowheads
SalI being internal sites within Pot3). The 151 bp at the 5
end of Pot3 was not seen in the MGR583 clone, indicating
that it may contain a possible insertion of the MGR583-
associated LINE-like sequence into Pot3. The LINE-like
element represents the major part of the length of MGR583
and both the 1.3- and 1.4-kb repeats are present within it
(Fig. 2-I).

Absence of the 3 end of the 1.3-kb element in pBC101
indicated disruption of the 1.3-kb repeat, possibly due to
the insertion of a putative transposable element. Since the
395 bp at the 3 end of the 1.3-kb repeat is replaced by a
2.7-kb sequence, it was likely that the insertion is at least
2.7 kb in size. Southern-hybridisation analysis of genomic
DNA from isolate B157 digested with several restriction
enzymes and probed with the 2.7-kb BamHI fragment from
pBC101 (other BamHI sites being contributed by the mul-
tiple cloning site of the vector), revealed it to be a multi-
copy repeat element and the minimum size of this element
was estimated to be 3.3 kb as judged by an intensely hy-
bridising band seen in the BamHI digest (Fig. 2-II C). The
repeat element was also present among both rice and non-
rice subgroups of M. grisea at roughly equal copy num-
bers (data not shown).
Sequence analysis of the 2.7-kb region revealed the
presence of a TG dinucleotide at the end (Fig. 2-I) A 4-bp
region at the terminus was present in an inverse orienta-
tion 314 bp from this end. These features are characteris-
tic of the Long Terminal Repeat (LTR) sequences of retro-
elements. Comparison of the translated sequence of the
2.7-kb BamHI fragment showed identity to the gag region
of several retroelements and the pol region of grh, a LTR-
retroelement present in the Eleusine pathogens of M. gri-
sea (Dobinson et al. 1993). A potential t-RNA binding site
located 3 to the 5 LTR and the primer binding site located
upstream of the 3 LTR of most retrotransposons (Bingham
and Zachar 1989) could not be identified in the 2.7-kb se-
quence. Southern-hybridisation analysis of various cos-
mids and genomic DNA digested with SalI, BamHI or
BssHII and hybridised to the putative LTR-like region (the
BamHI-SalI fragment of the 2.7-kb repeat) showed only
one intense hybridising band (data not shown). This indi-
cates that the element may not possess conserved LTR-like
ends characteristic of retroelements and may belong to the
LINE family of retroposons (Hutchison et al. 1989). Se-
quence comparison of the 2.7-kb region with retroelement
grh and Maggy (Dobinson et al. 1993; Farman et al. 1996
b), characterised recently from M. grisea, revealed no sim-
ilarities at the nucleic-acid level.
Genetic mapping of the 1.3- and 1.4-kb repeats
Genetic mapping of higher-molecular-weight derivatives
hybridising to the 1.3-kb and 1.4-kb repeat probes was car-
ried out to determine the incidence of point mutations at
the 3 PstI site. The experiment was based on the assump-
tion that mutation of a PstI site common to both the ele-
ments would result in higher molecular derivatives that
would hybridise to both elements and also map to the same
position. Additionally, such an analysis would also provide
an indication of the number of higher-molecular-weight
bands generated due to such a point mutation. Conse-
quently, the genomic DNA of parents Guy11 and 2539 and
61 random progenies was digested with PstI, fractionated
on a 0.8% gel and hybridised separately to the 1.3- and
1.4-kb repeat probes. The segregation of these bands
365
among the progeny is shown in Table 2. The higher-
molecular-weight bands, each representing a unique event
in the genome, were assigned a genetic linkage relative to
the previously assigned markers described by Skinner
et al. (1993) and Farman and Leong (1995). The deriva-
tives of both the 1.3- and 1.4-kb repeats mapped to differ-
ent locations of the genetic map, indicating that none of
these bands was derived by a point mutation at the PstI
site, common to the 1.3- and 1.4-kb repeats, in the genome
of isolate 2539 (Fig. 3).
DNA fingerprint analysis
DNA fingerprint analysis using the 1.3-kb repeat as a probe
against rice and non-rice pathogens collected from India
revealed that this repeat was present at a high copy num-
ber in both the subgroups of M. grisea. Since this 1.3 kb
is a part of the MGR583 repeat, it was expected MGR583
would also be present in a high copy number in the genome
of these isolates, in contrast to an earlier observation where
MGR583 was shown to be present at a copy number of 20,
or less, in non-rice pathogens (Hamer et al. 1989; Dobin-
son and Hamer 1993). To validate our results, we carried
out a fingerprint analysis of 36 M. grisea isolates (Table 1),
representing diverse geographical regions, using the 1.3-kb
and 1.4-kb elements as probes.
Southern hybridisation analysis of genomic DNA of
M. grisea isolates (data shown for 27) to the 1.3-kb repeat
probe (pBC157.3) revealed hybridisation to all the rice
pathogens (lanes am), which had almost an equal copy
number of the repeat as judged by the number and inten-
sity of the bands (Fig. 4 A). A majority of non-rice patho-
gens (lanes nw) also showed a strong hybridisation sig-
nal equivalent to that of the rice pathogens. Non-rice
isolates 4091-58, Bd8401 and Dc88428, representing
lanes xz, revealed a weak hybridisation signal and, in par-
ticular, showed poor hybridisation to the 1.3-kb band, in-
dicating that the copy number of this repeat in these iso-
lates is less than 10. Since the 1.3- and 1.4-kb repeats are
parts of a larger element, we expected to see a similar re-
sult upon hybridising the same blot with the 1.4-kb repeat
element. As shown in Fig. 4 B, isolates 4091-58, Bd8401
and Dc88428, as expected, revealed a very weak hybrid-
isation signal relative to other isolates.
The isolates used in this study were further analysed us-
ing Mg-SINE, a Short Interspersed Nuclear Element pres-
ent at a high copy number in both rice and non-rice patho-
gens (Kachroo et al. 1995), and Pot3 (Farman et al. 1996
a) which, with a few exceptions (Shull and Hamer 1996),
is present at a high copy number only in isolates of M. gri-
sea pathogenic on rice. Southern-hybridisation analysis re-
vealed a poor hybridisation signal in lanes xz to Mg-SINE,
indicating that these non-rice pathogens are highly diver-
gent (Kachroo et al. 1995). In addition, two other isolates,
Pd8824 (lane n) and Pr8988 (lane r), also revealed weak
hybridisation to Mg-SINE while possessing a high copy
number of the 1.3- and 1.4-kb repeats. In comparison, Pot3
showed a high copy number among all the rice pathogens,
366
Table 2 Segregation analysis of single copy derivatives of 1.3 kb
and 1.4 kb repeats. 61 progeny obtained from a single cross between
Guy11 and 2539 (Skinner et al. 1993) were analysed for the allele
inherited from Guy11 (A) or the allele inherited from 2539 (B). Blank
space indicates data not interpretable while ND indicates data not
done. Mapping of these markers was carried out using Mapmaker
version 2.0.

Progeny 1.31 1.32a 1.32b 1.33 1.34 1.35 1.41 1.43 1.44 1.45 1.46

1 5982 ND ND ND ND ND ND ND ND ND ND ND
2 6000 A B B B B B B B B B A
3 6001 B A A A A A A A A A A
4 6003 A B B B B B B B B B B
5 6004 A B B B B B B B B B B
6 6005 A A A A A B B A B B B
7 6007 B A A A A A A A A A A
8 6008 A B A B B A B B A B A
9 6029 B B B B B B A B B B B
10 6039 A A A A A B B A B B B
11 6047 A B B A B B B A B A B
12 6049 B A A B A B B B B B B
13 6050 B A B A A A B A A B A
14 6051 B B B A B B A A B A B
15 6052 A B B B B B B B B B B
16 6054 A B B A B B B A B B B
17 6055 A A A B A B B B B A B
18 6057 A B B A B B A A B A B
19 6058 A A A B A B A A B B B
20 6059 B B B A B A B B A B A
21 6061 A B B B B B
22 6062 B A A B A B B A B A B
23 6063 A A A A A B A B B B B
24 6064 B A A A A A B A A A A
25 6065 A B B B B A B B A B A
26 6066 B A A A A B A B B B B
27 6068 B B B A B A A A A B
28 6069 A B B A B A A A A B A
29 6070 B A A B A B B B B B A
30 6072 A A A A A A B A A B A
31 6073 B B B B B B A A B B B
32 6074 B A A A A A B A A A A
33 6075 A A A A A A B A A A A
34 6076 A B B A B A A B A B A
35 6077 B B B B B A A B A A A
36 6079 A B B A B A B A A A A
37 6080 A A A A A B B A B B B
38 6081 A B A B B A B B A B A
39 6082 A A A A A A A B A B A
40 6085 B A A B A B A A B B B
41 6087 B A A B A A A A A B A
42 6089 A B A B B A A B A B A
43 6090 A A A B A A B B A B A
44 6092 B A A B A B A A B B B
45 6093 B A A B A A A B A A A
46 6094 B B B B B B A B B A B
47 6095 A A A B A A A B A B A
48 6099 B B B B B A A A A A A
49 6100 A A A A A A A B A A A
50 6101 B A A B A A A A A A A
51 6102 B A A B A A A B A A A
52 6103 A A A A A A A B A A A
53 6105 B A A A A A B A A B A
54 6106 B A A A A B A A B A B
55 6108 B B B B B B A B B A B
56 6109 B A A B A A B A A B A
57 6111 A B B A A A A A
58 6112 B A A B B B A B
59 6115 B B B B B A A A
60 10058 B B B A B A B A
61 10077 A B A B

Fig. 3 Genetic mapping of higher-molecular-derivatives of the 1.3-
and 1.4-kb repeats (shown in bold type in boxes). The higher-
molecular-weight derivatives which showed proper segregation were
numbered 15 for the 1.3-kb repeat and 16 for the 1.4-kb repeat and
mapped using MAPMAKER. The linkage of these markers was de-
termined with respect to the M. grisea genetic map constructed by
Farman and Leong (1995) using 61 progeny obtained from a single
cross between isolates Guy11 and 2539
while a majority of non-rice pathogens showed a copy num-
ber of less than 10 (data not shown). However, four non-
rice pathogens, including Ec552 (lane o), Cb334 (lane s),
Lc454 (lane v) and Ei476 (lane w), showed a high copy
number of the Pot3 repeat. This indicates that the repeated
DNA sequences in M. grisea may have amplified indepen-
dently of one another or else that the various isolates may
have followed a divergent evolutionary course.
In order to rule out the possibility of uneven loading of
DNA samples, which may also influence the intensity of
bands, single-copy sequences were used to re-probe the
Southern blot shown in Fig. 4. A low-copy region which
mapped to chromosomes 1 (a 2.9-kb ApaI-NotI fragment
of Cosmid 32-7-H, M. Farman and S. Leong, unpublished
results) and 2 (pCRP, Kachroo et al. 1994) of isolate 2539
was used to re-probe the Southern blot shown in Fig. 4A.
As indicated in Fig. 4 C, Southern-hybridisation analysis
using the single-copy region from chromosome 1 as a probe
showed hybridisation to all the isolates of M. grisea ex-
cept for Pg3393 (lane t) and Dc88428 (lane z). Isolate
Dc88428 was the only one which showed a low copy num-
ber of all three repeat element probes which, taken together
with the absence of any detectable hybridisation to the
single-copy regions, raises doubts about its authenticity as
a M. grisea isolate. By contrast, the DNA of isolate Pg3393,
367
a Indian grass pathogen, was analysed with eight different
repeat elements of M. grisea and was found to hybridise
strongly to all of them (data not shown). Absence of a hy-
bridisation signal when probed with a single-copy region
could be explained by a DNA rearrangement event, result-
ing in deletion of the DNA which is homologous to the
probe.
Discussion
Our earlier studies on repeated DNA sequences from the
rice blast fungus along with those reported here have re-
vealed that insertion of one repeat element within an un-
related element may be of common occurrence in the
M. grisea genome (Kachroo et al. 1994, 1995; Farman
et al. 1996 a). These could be regarded as balancing events
which are required to regulate the copy number of actively
transposing elements and/or to prevent them from inacti-
vating essential genes. Alternatively, the presence of a
number of different elements may contribute to genome re-
arrangements involving the genes that control variability
in this fungus. The presence of various repeated DNA se-
quences having the structural characteristics of transpos-
able elements and our present data, which shows the ex-
tent of DNA rearrangement caused by them, makes these
a likely candidate for the generation of variability in the
M. grisea population. Since similar elements have been
shown to exist in other phytopathogenic fungi (Dobinson
and Hamer 1993) it is likely that such events involving re-
peated DNA may occur in these organisms and may also
contribute to variability. The transposon insertions in
M. grisea are not targeted to a specific site, although
AT-rich DNA is a preferred locus for most of these ele-
ments (Kachroo et al. 1994, 1995; Farman et al. 1996 a, b).
368
A
C this indicates that other DNA rearrangement events includ-
Fig. 4 Southern-hybridisation analysis of genomic DNA of rice and
non-rice infecting isolates of M. grisea probed with: (A) pBC157.3,
(B) the 1.4-kb repeat and (C) the 2.9-kb ApaI-NotI single-copy re-
gion of cosmid 32-7-H. Genomic DNA (2 mg) was digested with
PstI, electrophoresed through a 0.8% gel, transferred to a MSI-nylon
membrane and hybridised with the radiolabelled probe. The lanes
contain DNA from: F, lane a; B, lane b; B101, lane c; B157, lane d;
4360-17-1, lane e; P06-6, lane f; JMB840610, lane g; 101, lane h;
102, lane i, 103, lane j; V8601, lane k; 104, lane l; 91-A-58, lane m;
Pd 8824, lane n; Ec522, lane o; Pr886, lane p; Pr342, lane q; Pr8988,
lane r; Cb334, lane s; Pg3393, lane t; Cd88215, lane u; Lc454,
lane v; Ei476, lane w; 4091-58, lane x; Bd8401, lane y; Dc88428,
lane z; and Lh88490, lane z. The 3.2-kb band which is formed as a
result of a Pot2 insertion is marked by an arrowhead in (A)
This, along with the results obtained from other organisms
where transposable elements have been shown to partici-
pate in a variety of genomic rearrangements (Roeder and
Fink 1980; Rothstein et al. 1987; Thompson-Stewart 1994)
as well as altering local gene expression (McGinnis et al.
1983), makes their study essential to better understand their
role in genome organisation, variability and expression.
Our analysis has shown that the higher-molecular-
weight bands seen in the DNA fingerprint analysis of the
1.3-kb repeat could be formed as a result of either inser-
tion of transposable elements or by point mutations. Ge-
netic mapping studies carried out with all the higher-
B
molecular-weight bands hybridising to the 1.3- and 1.4-kb
probes revealed that point mutation of a PstI site may not
be a frequent event among the repeats as none of them
mapped to the same genetic location. In view of an earlier
result, where an insertion of an inverted repeat transposon
was identified within a 1.3-kb repeat element (Kachroo
et al. 1994), and the results obtained in the present study,
ing insertion and recombination may be largely responsible
for causing alterations within the 1.3-kb family of repeats.
Southern-hybridisation analysis using both the 1.3- and
1.4-kb repeats as probes demonstrated that both these re-
peats are also present in MGR583 which has been classi-
fied as a LINE-like sequence (Hamer et al. 1989; Dobin-
son and Hamer 1993). Sequence analysis of both the 1.3-
and 1.4-kb repeats and the database comparison did not re-
veal any homology to LINE-like sequences. However, it is
likely that the region conserved among LINE elements falls
outside the sequence of the 1.3- and 1.4-kb repeats. This
is supported by a database comparison of the 0.5-kb se-
quence, representing the right end of the MGR583 element,
which showed a high level of match to the Neurospora
crassa Tad LINE-like element. Our analysis, together with
those of Farman et al. (1996 a), shows that MGR583 is a
chimeric clone composed of a LINE-like element and an
inverted repeat transposon, Pot3 (Farman et al. 1996 a).
The LINE-like element appears to have inserted within the
Pot3 element since the 151 bp at the 5 end of Pot3 are not
present within the MGR583 clone. This is analogous to
results obtained with pBC101 where the 5 395 bp of the
1.3-kb PstI repeat are absent, possibly as a result of the in-
sertion of a putative transposable element.
Analysis of the fragments hybridising to the 1.3- and
1.4-kb repeats revealed a marked similarity in the banding
pattern of rice isolates derived from the same geographi-

cal region. This could be seen both among the Indian iso-
lates (lanes ad) and the isolates from the Philippines
(lanes fm). Many of these higher-molecular-weight bands
were common between the rice and non-rice isolates. The
3.2-kb band formed as a result of a Pot2 insertion within
the 1.3-kb repeat (Kachroo et al. 1994) was evident among
several rice and non-rice isolates, indicating that these iso-
lates have originated from a common stock (Fig. 4 A,
shown by arrowhead). Absence of the 3.2-kb band in the
rest of the isolates and the presence of other similar-sized
bands among rice and non-rice pathogens indicates that
these isolates may have originated from several distinct
lines of origin.
DNA fingerprint analysis using both the 1.3- and 1.4-kb
repeats as a probe revealed that they are present in both
rice and non-rice pathogens of M. grisea. Since both of the
repeats are present in the MGR583 clone, this observation
is contrary to an earlier result where MGR583 has been re-
ported to be present in a high-copy number only among
rice pathogens (Hamer et al. 1989; Dobinson and Hamer
1993; Shull and Hamer 1996). Only three non-rice patho-
gens (4091-5-8, Bd8401 and Dc88428) revealed a DNA
fingerprinting pattern which corresponds to that of an ear-
lier study where the MGR583 group of repeats was re-
ported to be present at a copy number of 1020 in non-rice
pathogens (Hamer et al. 1989). Isolate 4091-5-8 is the only
one n common between the present work and the earlier
study. This indicates that there may be a group of non-rice
pathogens which do not possess these repetitive elements
at a high copy number. This further suggests that different
M. grisea isolates may have followed different paths of ev-
olution. There may be at least two such paths, one which
involves co-evolution of two subgroups and the other
which involves evolution from a common ancestor and
subsequent divergence of subgroups. In the former event,
the isolates will share the same DNA fingerprinting pat-
terns whereas, in the latter event, the two subgroups may
amplify specific DNA sequences after their divergence
and, therefore, differ in their fingerprinting patterns. Since
neither of these mechanisms can explain the results ob-
tained both by us and by Hamer et al. (1989), we propose
that both mechanisms may have contributed to the evolu-
tion of divergent M. grisea isolates. Our present study
shows that the presence, or absence, of repetitive elements
among rice and non-rice pathogens of M. grisea cannot be
generalised and varying results may often be seen depend-
ing on the sample size and the origin of the isolates.
Acknowledgements We thank Drs. Naoto Nitta, (Department of
Plant Pathology, University of Wisconsin, Madison, USA) for pro-
viding the DNA samples of progeny obtained from a cross between
Guy11 and 2539, and Mark Farman (Department of Plant Patholo-
gy, University of Wisconsin, Madison, USA) for providing the DNA
samples of some of the non-rice pathogens, for the Pot3 clone and
also for sharing his unpublished data. Finally, we thank Dr. Autar
Mattoo (USDA, Beltsville, M.D., USA) for his critical reading of the
manuscript. This work was supported by grants from the Department
of Biotechnology, Government of India, and by the Rockefeller
Foundation to B.B.C., as well as by the United States Department of
Agriculture and the Rockefeller Foundation to S.A.L.
369
References
Amasino R (1986) Anal Biochem 152:304307
Berg DE, Howe MM (1989) In: Mobile DNA. American Society for
Microbiology, Washington, DC
Bingham PM, Zachar Z (1989) Retrotransposons and the FB trans-
poson from Drosophila melanogaster. In: Mobile DNA. Ameri-
can Society for Microbiology, Washington, D.C., pp 485502
Borromeo ES, Nelson RJ, Bonman JM, Leung H (1993) Phytopa-
thology 83:393399
Charlesworth B, Sniegowski P, Stephan W (1994) Nature 371:215
220
Devereux J, Haeberli P, Smithies O (1984) Nucleic Acids Res 12:
387389
Dobinson KF, Hamer JE (1993) Trends Microbiol 1:348352
Dobinson KF, Harris RE, Hamer JE (1993) Mol Plant-Microbe Inter-
act 6:114126
Dover G (1982) Nature 299:111117
Farman ML, Leong SA (1995) Genetics140:479492
Farman ML, Tosa Y, Nitta N, Leong SA (1996 a) Mol Gen Genet
251:665674
Farman ML, Taura S, Leong SA, (1996 b) Mol Gen Genet
251:675681
Feinberg AP, Vogelstein B (1983) Anal Biochem 132:613
Genovesi AD, Magill CW (1976) Can J Microbiol 22:531536
Giatgong P, Frederiksen RA (1969) Phytopathology 59:11521157
Hamer JE, Farrall L, Orbach MJ,Valent B, Chumley FG (1989) Proc
Natl Acad Sci USA 86:99819985
Hutchison CAIII, Hardies SC, Loeb DD, Shehee WR, Edgell MH
(1989) LINEs and related retroposons: long interspersed repeat-
ed sequences in the eucaryotic genome. In: Mobile DNA. Amer-
ican Society for Microbiology, Washington, D.C., pp 485502
Jeffreys AJ, Wilson V, Thein SL (1985) Nature 314:67
Kachroo P, Leong SA, Chattoo BB (1994) Mol Gen Genet 245:
339348
Kachroo P, Leong SA, Chattoo BB (1995) Proc Natl Acad Sci USA
92:1112511129
Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln SE,
Newburg L (1987) Genomica 1:174181
Leong SA, Farman ML, Budde AL, Smith RJ, Tosa Y, Nitta N (1994)
Molecular-genetic approach to the study of cultivar specificity in
the rice blast fungus. In: Rice blast disease. Zeigler RS, Leong
SA, Teng PS (eds). CAB, Wallingford, UK, pp 87110
Levy M, Romao J, Marchett MA, Hamer JE (1991) Plant cell
3:95102
McGinnis W, Shermoen AW, Beckendorf SK (1983) Cell 34:7584
Noguchi T (1991) Technique 3:3336
Petes TD (1980) Cell 19:765774
Roeder SG, Fink GR (1980) Cell 21:239249
Roeder GS, Farabaugh PJ, Chaleff DT, Fink GR (1980) Science
209:13751380
Rothstein R, Helms C, Rosenberg N (1987) Mol Cell Biol 7:11981207
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a la-
boratory manual, 2nd edn. Cold Spring Harbour Laboratory, Cold
Spring Harbour, New York
Scherer S, Davis RW (1980) Science 209:1380
Schmidt ER (1984) J Mol Biol 178:115
Shull V, Hamer JE (1996) Fungal Genet Biol 20:5669
Skinner DZ, Budde AD, Farman ML, Smith RJ, Leung H, Leong SA
(1993) Theor Appl Genet 87:545557
Sone T, Suto M, Tomita F (1993) Biosci Biotech Biochem
57:12281230
Suzuki H (1965) Origin of variation in Pyricularia oryzae In: The
rice blast disease. Johns Hopkins Press, Baltimore, Maryland,
pp 111149
Thompson-Stewart D, Karpen GH, Spradling AC (1994) Proc Natl
Acad Sci USA 91:90429046
Wyman C, Blackburn EH (1991) Genetics 128:5767