Morphological, ontogenetic and molecular characterization
of Scutellospora reticulata (Glomeromycota)
Francisco Adriano DE SOUZA1*#, Ste ´ phane DECLERCK3, Eric SMIT4and George A. KOWALCHUK2
1Empresa Brasileira de Pesquisa Agropecua´ria, Embrapa Agrobiologia (CNPAB), Caixa Postal 74505, CEP 23851-970,
Rodovia BR 465-RJ, km 7, Serope´dica, Rio de Janeiro State, Brazil.
2Netherlands Institute of Ecology (NIOO-KNAW), Centre for Terrestrial Ecology, Boterhoeksestraat 48, 6666 GA Heteren,
3Universite´ Catholique de Louvain, Mycothe`que de l’Universite´ catholique de Louvain (MUCL*), Unite´ de microbiologie,
3 Place Croix du Sud, 1348 Louvain-La-Neuve, Belgium.
4National Institute for Public Health and the Environment (RIVM), Microbiological Laboratory for Health Protection,
P.O. Box 1, NL-3720 BA, Bilthoven, The Netherlands.
E-mail : firstname.lastname@example.org
Received 30 August 2004; accepted 20 January 2005.
The arbuscular mycorrhizal (AM) fungus Scutellospora reticulata (CNPAB11) was characterized using morphological,
ontogenetic and molecular approaches. Spore ontogenesis was studied using Ri T-DNA transformed carrot roots and
observations were compared with those published for eight other, pot-cultured, Scutellospora species. The sporogenesis
of S. reticulata exhibited an unreported pattern of outer spore wall differentiation. In addition, Denaturing Gradient Gel
Electrophoresis (DGGE), targeting the V9 region of the SSU nrDNA, was used to differentiate S. reticulata from
16 other Scutellospora species and results were confirmed by sequencing analysis. Phylogenetic analyses, using nearly
full length SSU nrDNA sequences, grouped S. reticulata in a cluster together with S. cerradensis and S. heterogama,
species that share similar spore wall organization and also possess ornamented external walls. PCR-DGGE and sequence
analysis revealed intragenomic SSU nrDNA polymorphisms in four out of six Scutellospora species tested, and
demonstrated that SSU nrDNA intragenomic polymorphism could be used as a marker to differentiate several closely
related Scutellospora species.
Species identification and phylogeny of arbuscular
mycorrhizal (AM) fungi have traditionally been based
on analysis of morphological characteristics of spores
and fungal mycelium (Morton & Benny 1990). How-
ever, the recent application of molecular phylogenetic
analysis, based on small subunit nuclear ribosomal
DNA (SSU nrDNA) sequences, has resulted in pro-
of a separate phylum (Glomeromycota), containing
new orders, families and genera (Schu ¨ ßler, Schwarzott
& Walker 2001, Walker & Schu ¨ ßler 2004).
Nevertheless, an integrated analysis combining de-
velopmental patterns during colony growth and spore
ontogenesis with molecular phylogenetic analysis has
never been performed with AM fungi. The comparative
studies of development patterns during morpho-
structure ontogenesis are a powerful way to establish
homology between morphological character states and
achieve a natural classification (Givnish & Sytsma
1997). Such an approach has been successfully used for
cladistic analyses of Scutellospora species (Franke &
Morton 1994, Morton 1995). However, these studies
involved soil-based systems that require destructive
sampling of the fungal material. One-way forward to
overcome this problem is to use a monoxenic culture
system to grow AM fungi (Fortin et al. 2002). The
monoxenic culture of AM fungi allows for cultivation,
real time observation and precise sampling of fungal
material throughout the life cycle. This technique has
already been used successfully to study spore ontogeny
(de Souza & Berbara 1999) and species characterization
using a polyphasic approach (Declerck et al. 2000).
Scutellospora belongs to the family Gigasporaceae,
order Diversisporales. This genus represents 17% of the
known AM fungi species (de Souza, 2000). Recently,
* Part of the Belgian Coordinated Collections of Micro-organisms
# Corresponding author.
Mycol. Res. 109 (6): 697–706 (June 2005).
f The British Mycological Society
doi:10.1017/S0953756205002546Printed in the United Kingdom.
a Scutellospora species was successful cultured in
monoxenic system with transformed carrot roots (de
Souza & Declerck 2003). The in vitro culture of S. reti-
culata (CNPAB11) was successfully used to investigate
extramatrical mycelium development, i.e. hyphal mor-
phology and branching (de Souza & Declerck 2003),
the dynamics of spore production, and the function of
auxiliary cells (Declerck et al. 2004).
The objective of the present study was to provide
a thorough characterization of S. reticulata and per-
form an integrated analysis combining developmental
patterns during spore ontogenesis with molecular
phylogenetic analysis. In addition, the application of
(DGGE) was tested as a rapid identification tool to
discriminate Scutellospora species.
MATERIAL AND METHODS
The Scutellospora species used in this study are listed
in Table 1. The accessions used were obtained from
collections of fungus cultures except for S. coralloidea,
which was collected from a trap culture (de Souza et al.
2004). Spores were extracted from soil using a wet
sieving technique and prepared for molecular analyses
according to de Souza et al. (2004). From the isolates
obtained, only S. reticulata was established in mono-
xenic culture, while the other strains were used for
Establishment of monoxenic culture
Scutellospora reticulata (CNPAB11) was established
under monoxenic culture as described by de Souza &
Declerck (2003). Briefly, surface-sterilized spores were
germinated in water-agar at pH 6. After germination,
single spores were used to inoculate explants of Ri
T-DNA transformed carrot roots. Ten cultures were
used as replicates (experimental units) for assessing
spore development and size. Each experimental unit
consisted of a transformed carrot root organ explant
inoculated with a single S. reticulata spore in a Petri-
plate containing 30 ml of the Modified Strullu-Romand
(MSR; Declerck, Strullu & Plenchette 1998, following
Strullu & Romand 1986) medium, and incubated in
inverted position at 27 xC for up to eight months.
Sporogenesis in monoxenic culture conditions
and data collection
In order to compare the data obtained with monoxenic
cultures with the data generated using pot culture
conditions, all development stages used here followed
the definitions and procedures established by Franke &
Morton (1994). To assess subcellular differentiation
during spore ontogeny, spores were sampled at differ-
ent developmental stages, which were differentiated by
changes in spore size and colour, ranging from white
opaque to dark brown (de Souza & Declerck 2003).
Juvenile and mature spores were differentiated by
colour, septa formation in the subtending hypha and
by absence of cytoplasmic activity in the sporogenous
subtending hypha. Spore dimensions were assessed
using 12 randomly chosen, mature spores in each
experimental unit. Sampled spores were mounted on
microscope slides with Polyvinyl-lactic acid-glycerol
medium (PVLG) (Omar, Bolland & Heather 1979) and
PVLG plus Melzer’s reagent (5:1 v/v). Observations
were made under a dissecting microscope and under
bright field through inverted and common compound
Intraradical structures assessment
Colonized roots were harvested from eight-month-old
cultures, cleared with 2.5% KOH overnight at room
temperature, washed with tap water and soaked in 1%
HCl solution for one hour. Roots were then stained
with 0.5% Quink Parker blue ink for 20 min at 60 x
(C. Walker, pers. comm.). After staining, the roots
were rinsed with tap water and preserved in 50%
glycerol solution with 1% HCl and stored at room
temperature (ca 20 x) until required.
DNA extraction and Denaturing Gradient Gel
Electrophoresis (DGGE) analysis
DNA was extracted from individual spores of each of
the fungi listed in Table 1, according to procedures
described previously (de Souza et al. 2004). PCR-
DGGE was used to providerapid fingerprint
Table 1. Species, code, contributor, origin and germplasm collection of the Scutellospora spores or isolates used in this study.
F. A. de Souza
F. A. de Souza
F. A. de Souza
J. O. Siqueira
F. A. de Souza
aBEG, European Bank of Glomeromycota, Dijon; CNPAB, Empresa Brasileira de Pesquisa Agropecuaria – Embrapa Agrobiologia, Rio de
Janeiro; MAFF, Ministry of Agriculture, Forest and Fisheries, Ibaraki; UFLA, Universidade Federal de Lavras, Minas Gerais, Brazil.
identification of Scutellospora reticulata CNPAB11,
which was compared to six other Scutellospora species.
The DGGE analysis was performed according to de
Souza et al. (2004), which developed a PCR-DGGE
system to assess the diversity of Gigasporaceae species,
targeting the V9 region of the SSU nrDNA using a
nested PCR approach. Briefly, in the first PCR round,
a set of specific Gigasporaceae primers (FM6 and
GIGA5.8R) was used. The resulting PCR product was
diluted 1:1000 and used as template for a second reac-
tion using the primer NS7 (White et al. 1990), with a
GC-clamp extension in its 5k end, in combination with
the fungal specific reverse F1Ra primer (de Souza et al.
2004). Spore-to-spore variation within accessions was
analysed using PCR-DGGE, five separate single-spore
DNA isolations for each fungus were compared.
To predict PCR-DGGE separation of sequences
deposited in the GenBank in relation to sequences
we obtained from S. reticulata and S. gregaria,
Scutellospora sequences that contain the fragment used
for PCR-DGGE analysis, were aligned and compared
to relate the sequence data with the migration of selec-
ted strains observed under DGGE. The PCR-DGGE
was also used to study the intraspecific SSU nrDNA
polymorphism of the species tested.
Cloning and sequencing
For cloning and sequencing purposes, only the DNA
extracted from spores of Scutellospora reticulata and
S. gregaria were used. The genomic DNA obtained
from individual spores was amplified (de Souza et al.
2004) with the forward primer NS1 in combination
with the reverse ITS4 (White et al. 1990). After ampli-
fication the PCR product was purified and cloned into
the pGEM-T easy vector, with Escherichia coli strain
JM109 used for transformation, according to pro-
cedures given by the manufacturer (Promega Benelux,
Leiden, The Netherlands). The clones obtained were
cultured and, after plasmid extraction used as template
for PCR-DGGE and for sequencing. Sequencing
reactions were performed for both DNA strands of
each clone using the Perkin Elmer Biosystems Big Dye
Terminator Sequence Reaction kit (Perkin Elmer,
Foster City, CA) and the reactions were analysed on
a Perkin Elmer 3700 capillary sequencer (RIVM;
Bilthoven, The Netherlands). The primers used for se-
quencing were NS1, NS2, NS6, NS7, ITS1, ITS4
(White et al. 1990), NS31 (Simon, Lalonde & Bruns
1992), AM1 (Helgason et al. 1998) and F1Ra (de Souza
et al. 2004).
Selecting polymorphic ribotypes
The PCR-DGGE targeting the V9 SSU nrDNA of
Scutellospora reticulata revealed the occurrence of
intraspecific polymorphism between thenrDNAcopies.
In order to analyse clones for different variants of the
ribosomal copies (ribotypes) occurring in S. reticulata
(CNPAB11) and S. gregaria (CNPAB7) plasmids con-
taining inserts obtained from those strains were used
as templates for PCR-DGGE analysis, as described
above. To help the selection of different ribotypes,
PCR products obtained from original isolates were
used as reference to select clones that matched to each
of the ribotypes detected in each isolate examined. We
analysed 46 clones of each strain. Plasmids containing
the desired insert were purified using Quiaquick purifi-
cation columns, and sequenced as described above.
Sequences were aligned with those obtained from Gen-
Bank (Benson et al. 2003) using Clustal-X (Thompson
et al. 1997), and the alignment was improved after-
wards by visual inspection. Phylogenetic trees were
constructed using distance, parsimony and maximum-
likelihood (ML) methods. The substitution model was
chosen after comparison of 56 different models using
the program ModelTest (Pousada & Crandall 1998)
version 3.5. The phylogenetic analyses were performed
using PAUP* version 4.0 Beta 10 (Swofford 2003).
Nucleotide sequence accession and alignment numbers
The sequences and alignment generated in this study
were deposited in EMBL-EBI nucleotide sequence
database (http://www.ebi.ac.uk) under the accession
numbers AJ871270 to AJ871275 and the alignment
Spore formation in Scutellospora reticulata takes
6–10 d to complete, and the spore development exhibits
changes in size, spore ornamentation, and colour
diameter of 379 mm (range 280–500 mm; CV=10.67%,
n=120). The main characteristics of the spore mor-
phology and ornamentation were in accordance with
the original description of this species (Koske, Miller &
Walker 1983). The spore ontogenesis of S. reticulata
consisted of six discrete stages defined by synthesis of
specific wall layers (Fig. 5A, murographic represen-
tation). During stage 1, two layers were synthesized
(Fig. 6). The external layer was hyaline (0.5–0.8 mm
thick) and the internal layer was pale yellow (1–2 mm
thick) and turned rust-red in Melzer’s reagent. At that
stage, some spores burst and released their contents
into the medium. The second stage began after the
spores reached their full expansion in size. The second
wall layer increased in thickness (6–20 mm thick in
PVLG; Fig. 7), with differentiation of external (reticu-
late; Fig. 8) and internal ornamentations (spines;
Fig. 9). In both stages, the spores were white to pale
white (Figs 1–2). The third stage was characterized by
R. A. de Souza and others699
a change in colour from pale white to greenish yellow
(Fig. 3), and later to dark red-brown (Fig. 4). At that
stage the typical ornamentation of the S. reticulata
outer wall layer could be seen (Figs 9–11). In the sub-
cellular structure, a laminar layer to 2–3 mm thick and
consisting of very thin adherent sublayers (the laminar
wall as defined by Walker 1983), was synthesized
(Fig. 9). In stages 4 and 5, two bi-layered inner wall
layers (IW) were synthesized (Fig. 12). These layers
were difficult to observe in CNPAB11, as they do not
detach easily from the spore wall. The IW1 was hyaline
(1–1.8 mm thick in PVLG), and the IW2 was light
yellow (1 mm thick in PVLG). No reaction in Melzer’s
reagent was observed in these layers. In stage 5, the
spores reached their mature dark red-brown colour.
Stage 6 was characterized by the synthesis of the
yellowish-brown germination shield (GS) between IW1
and IW2 (Fig. 13).
Intraradical mycelium structures
Scutellospora reticulata CNPAB11 formed typical
Gigasporaceae structures in excised carrot roots,
characterized by course mycelium and arbuscules, and
the absence of intraradical vesicles. The arbuscules
exhibited profuse hyphal coiling and were found in
Four of the seven Scutellospora species tested using
PCR-DGGE produced more than one band for the
V9 region of the SSU nrDNA (Fig. 14), demonstrating
the occurrence of intraspecific polymorphism in those
species, and no spore-to-spore variation was found.
Also, no difference was found between the DGGE
profiles of the two isolates of S. heterogama tested
The different species tested could be separated into
two major groups based on their DGGE profile. The
first group was composed of species with bands located
relatively high in the gel (S. calospora, S. castanea,
S. coralloidea and S. gregaria), while the second group
had lower bands (S. cerradensis, S. heterogama,
S. reticulata, see Fig. 14). In the first group, S. gregaria
and S. coralloidea bands had the same migratory
behaviour and could not be discriminated from each
other, but they could be separated from the two other
species in that group (S. calospora and S. castanea).
In the second group, all the species analysed contained
intraspecific polymorphism (i.e. multiple bands) and
the lower band of each species displayed the same
migratory behavior (position) in the gel (Fig. 14). These
species could be discriminated from each other on
the basis of the PCR-DGGE profiles.
Sequence comparison (Table 2) and phylogenetic
analysis (Fig. 15) confirmed the sequence similarity
within the two DGGE migration groups. The first
group was composed of sequences containing a higher
AT content than the sequences from the species of
the second group. In addition, sequences from SSU
nrDNA V9 region from S. aurigloba, S. nodosa,
S. projecturata, not analysed directly by DGGE, were
found to contain an intermediate AT/GC content
100 µm100 µm
100 µm 100 µm
100 µm100 µm
Figs 1–4. Changes in size and colour during the development of one spore of Scutellospora reticulata. As the spore
develops spore colour changes from white opaque (immature) to dark brown (mature). The subtending hypha (sh) moved
due to the growth of a root. Bold arrowheads show the sporogenous cell.
compared to the other two groups. Despite having a
more intermediate AT content in its V9 region, the
sequences of S. calospora migrated in the region typical
of high AT species.
Representative clones for each of the three S. reti-
culata DGGE bands were subjected to sequence
analysis: 8 (upper position), 9 and 18–2 (middle pos-
ition), and 10 (lower position): see Fig. 14 and Table 2.
Clones corresponding to the three S. reticulata bands
were recorded in a ratio of 11:15:20 (upper:middle:
lower; n=46) as determined by DGGE screening. The
two clones of S. gregaria sequenced were identical
within the region analysed by DGGE (Table 2), and
were in agreement with the predicted and observed
melting behavior (Table 2, Fig. 14).
Prior to phylogenetic analysis the Scutellospora origin
of the sequences was confirmed by BLAST search
(Altschul et al. 1990). The substitution model that best
fit the data, after removing the constant and the gapped
swswsw sw sw swsw
stage 1 stage 2
stage 3stage 4 stage 5 stage 6
spore cytoplasmspore cytoplasm
Fig. 5. Murographic representation of spore wall layers. (A) Scutellospora reticulata (CNPAB11) spore ontogenesis. Six
discrete stages of differentiation were observed. The first three stages comprised the spore wall (SW) layers and the last
three the Inner Wall (IW) layers and germination shield (GS) differentiation. Stage 1 is characterized by spore expansion,
stage 2A and 2B by differentiation of the second SW layer (wave lines) into a double ornamented layer (O), followed by
differentiation of a laminar layer (vertical dashes) in stage 3. Spore colour changes from pale white to dark brown during
the transition between stages 2 and 3. Stage 4 is characterized by formation of the first bilayered IW layer (horizontal lines).
In stage 5 the second bilayered IW layer is formed and in stage 6 the GS is differentiated. (B) Groups of Scutellospora
species based on comparative developmental sequences. *, designates species possessing a smooth outer layer. Underlined
species were grouped based on their description. Modified from Franke & Morton (1994), Morton (1995),
Stu ¨ rmer & Morton (1999) and INVAM.
R. A. de Souza and others701
sites, was HKY+G (Hasegawa, Kishino & Yano 1985,
Pousada & Crandall 1998), and the parameter were as
follows: proportion of invariable sites, I=0; gamma
distribution shape, G=9.5134; number of substitutions
type, NST=2; transition/transversion ratio=2.9585;
and base frequencies, A=0.2727; C=0.2009; G=
0.1971; T=0.3211. Usingnearly full-length SSU rDNA
sequences, Scutellospora sequences were grouped in
three separate clades, A–C in Fig. 15. The tree topo-
logy was consistent using distance, parsimony and
maximum likelihood methods. Clade A was composed
of S. aurigloba, S. calospora, S. nodosa and S. pro-
jecturata. The three sequence types of S. reticulata
clustered in clade B, together with sequences of S. cer-
radensis and S. heterogama. Sequences of S. dipapillosa
also cluster in that clade, however, the sequence avail-
able is only partial, and therefore it was not used in
the phylogenetic analysis. S. gregaria sequences clus-
tered in clade C together with S. castanea, S. fulgida, S.
gilmorei, S. pellucida, S. spinosissima and S. weresubiae.
50 µm50 µm50 µm
10 µm 10 µm10 µm
Figs 6–13. Spore wall differentiation of Scutellospora reticulata (CNPAB11). Fig. 6. Spore wall with two thin layers at
stage one. First layer is hyaline 0.5 mm thick (open arrowhead), the second one is pale yellow and 2.0 mm thick (bold
arrowhead). Fig. 7. Lateral view of layer two of the spore wall with an external reticulum (open arrowhead) and internal
spines (bold arrowhead). Fig. 8. Reticular ornamentation on the surface spore wall layer of a pale cream-coloured juvenile
spore at stage 2. Fig. 9. Lateral view of mature spore wall showing space between outer and double ornamented layers
(open arrowhead) and transition between the spine and laminate layers (bold arrowhead). Fig. 10. Lateral view of reticular
ornamentation of the surface spore wall layer of dark-brown-coloured mature spores at stage 6 and spore wall layer 1 sealing
the reticulum (open arrowhead). Fig. 11. Upper view of the reticular ornamentation of a mature spore where the layer 1 is
present obstructing the view of the internal spines. Fig. 12. Two bi-layered flexible inner walls (IW) layers. Fig. 13. Lateral view
of the germination shield (gs), bigger panel. Lower panel in the upper corner, magnification of the left upper side of the
bigger panel, separation of the two inner layers to form the GS. Small panel in the lower corner, spore wall (SW),
inner wall (IW) layers and the tip of the germination tube (arrowhead).
The monoxenic culture system proved ideal for the
study of spore development of Scutellospora reticulata,
as it facilitated sampling of spores at different devel-
opment stages and did not stimulate microbial activity
that can cause alterations in the outer wall layer.
Recently, mycelial characteristics and the role of auxil-
iary cells on spore production of this isolate were also
studied successfully using monoxenic culture system
(de Souza & Declerck 2003, Declerck et al. 2004).
In Scutellospora reticulata, the first spore wall layer is
thin and evanescent, adherent to the second layer, not
reported in the original description of the species
(Koske et al. 1983). However, it breaks easily during
spore manipulation and may be decomposed by micro-
bial activity in soil-based systems, probably explaining
why it has never been reported before. Layers similar to
the one we observed actually may occur in two highly
ornamented Scutellospora species from La Gran
Sabana, Venezuela: S. spinosissima (Walker, Cuenca
& Sa ´ nchez 1998) and S. crenulata (Herrera-Peraza,
Cuenca & Walker 2001).
The development pattern of the outer ornamented
layer in S. reticulata has not been reported before.
We found that this layer first expands and later
differentiates into thedouble
characteristic of that species. The discrimination of
spore wall (SW) layers during ontogeny is not easy,
and in some species they are more evident than in
others. For example, Spain & Miranda (1996) found
an additional inner wall layer in the SW of S. cer-
radensis. Later, a similar layer was found also in
S. heterogama, S. pellucida and S. rubra (INVAM;
http:\\invam.caf.wvu.edu). Such an additional layer
was not observed for S. reticulata. This does not
demonstrate the absence of that layer in S. reticulata,
as it may be too thin to be differentiated using light
Table 2. 23 DNA sequences from 15 different Scutellospora species, showing 43 variable positions in the SSU nrDNA V9 region and AT/GC
ratio of those positions.
Position in the alignmentb
Species/clone codea /GeneBank accession number
S. castanea AF038590
S. fulgida AJ306435
S. weresubiae AJ306444
S. gregaria F31; F37
S. nodosa AJ306437
S. gilmorei AJ276094
S. spinosissima AJ306436
S. calospora AJ306444
S. calospora AJ306443
S. calospora AJ306445
S. pellucida Z14012
S. aurigloba AJ276092
S. projecturata AJ242729
S. heterogama AJ306434
S. reticulata 8
S. aurigloba AJ276093
S. dipapillosa Z14013
S. cerradensis AB041344
S. cerradensis AB041345; S. reticulata 18-2; 9
S. reticulata 10
aDifferent clones in the same line are identical in DNA sequence for the fragment analysed by PCR-DGGE.
bThe 5k and the 3k end of the primers NS7 and F1Ra are located at positions 1420 and 1747 respectively, according to the alignment
provided at EMBL-EBI.
1 2 3 4 5 6 7 8 9 10 11 12 13
Fig. 14. PCR-DGGE analysis of the V9 region of the SSU
nrDNA sequences amplified from Scutellospora species
and run for 17 h at 95 v. Lane designations: 1, S. calospora
(BEG32); 2, S. gregaria (CNPAB7); 3, S. castanea (BEG1);
4, S. calospora (BEG32); 5, S. reticulata (CNPAB11); 6, S.
cerradensis (MAFF520056); 7, S. heterogama (CNPAB2); 8,
S. heterogama (UFLA); 9, S. gregaria (CNPAB7); 10, S.
coralloidea trap culture; (11), S. castanea (BEG1); 12, Blank;
and 13, S. reticulata (CNPAB11).
R. A. de Souza and others703
Comparison between sporogenesis in
We compared the sporogenesis of Scutellospora reti-
culata with other species of the genus studied using the
same criteria (S. coralloidea, S. fulgida, S. heterogama,
S. gregaria, S. pellucida, S. persica, S. rubra, S. verru-
cosa; Franke & Morton 1994, Morton 1995, Stu ¨ rmer &
Morton 1999). Based on Morton, Bentivenga & Bever
(1995), species-level definition is limited mainly to
characters of the SW, and the IW layers define higher
taxonomic groups. The latter appear to be linked with
germination events. Using these criteria it was possible
to cluster these species into three groups (Fig. 5B). The
first group is composed of S. reticulata, S. heterogama,
S. rubra, which all have two bilayered IW groups and
the GS positioned between them (Franke & Morton
1994, Stu ¨ rmer & Morton 1999). S. calospora, S. cerra-
densis and S. gilmorei (Spain & Miranda 1996;
INVAM website), not studied using developmental
patterns, also belong to this group. The second group
contains S. pellucida, and is discerned by characteristics
of the second layer of the IW2, which expand under
light pressure in mounting medium (Franke & Morton
1994). Species related to S. pellucida are S. dipapillosa,
S. erythropa, S. nodosa, and S. weresubiae (Blaszkowski
1991, Franke & Morton 1994), also possibly S. pro-
jecturata and S. spinosissima. The third is composed
of S. fulgida, S. verrucosa, S. persica, S. coralloidea, and
S. gregaria, which are species with a single bilayered IW
group and the GS located between the SW and the IW
(Fig. 5; Morton 1995). The last two species in this
group (S. coralloidea and S. gregaria) are most closely
related based on cladistic analyses, and S. castanea
is also known to belong to this group (Morton 1995).
S. aurigloba, was used in the molecular analysis, but
due to poor information on spore wall components
it was not included in any of developmental groups.
Phylogenetic analysis based on SSU nrDNA
The molecular phylogeny based on SSU nrDNA has
been proved to discriminate from genus and subgenus
level to above (Redecker, Morton & Bruns 2000,
Schwarzott, Walker & Schu ¨ ßler 2001, Schu ¨ ßler et al.
2001, Walker & Schu ¨ ßler 2004), but fails to discrimi-
nate between species because nrDNA is too conserved,
something confirmed by our analysis. In Scutellospora
the SSU nrDNA sequence produced three shallow
clades supported by strong bootstrap values. These
clades have poor resolution to discriminate between
species. However, they are indicative of subgroups
within Scutellospora. We have not tested for polyphyly
in our analysis; the tree topology suggests a different
ancestor for clade A in relation to the clades B and C,
and the latter clades were most closely related with the
sister genus Gigaspora. Polyphyly has to be confirmed
through the analysis of more taxa and genes.
Comparing molecular phylogenetic and
The three clades obtained by molecular phylogeny
clustered species that are closely related based on de-
velopmental patterns of the spores, but within each
clade more than one developmental group is present.
For instance, species on clade B are closely related
based on developmental patterns, except for Scutello-
spora dipapillosa (Fig. 5, not show in the tree; Fig. 15).
Using this criterion, species on clade C would be separ-
ated in three distinct groups: (1) S. castanea, S. fulgida,
and S. gregaria; (2) S. gilmorei; and (3) S. pellucida,
S. weresubiae, and S. spinosissima (Fig. 5). In clade A,
most of the species would cluster in the S. pellucida
group (Fig. 5). Besides, S. calospora (clade A) and
G. margarita F44
G. rosea X58726
G. albida Z14009
G. gigantea Z14010
S. gregaria F31
S. gregaria F37
S. gilmorei AJ276094
S. castanea AF038590
S. fulgida AJ306435
S. weresubiae AJ306444
S. cerradensis AB041344
S. cerradensis AB041345
S. heterogama AJ306434
S. reticulata 9
S. reticulata 18 2
S. reticulata 10
S. reticulata 8
S. calospora AJ306446
S. calospora AJ306443
S. calospora AJ306445
S. nodosa AJ306437
S. aurigloba AJ276092
S. aurigloba AJ276093
S. projecturata AJ242729
G. versiforme X86687
G. versiforme AJ276088
G. spurcum AJ276077
A. longula AJ306439
A. scrobiculata AJ306442
A. laevis Y17633
Acaulospora sp. AJ306441
Acaulospora sp. AJ306440
Fig. 15. Maximum likelihood SSU nrDNA phylogenetic tree
were clustered in three clades denominated A, B and C. The
tree was constructed using maximum parsimony, minimum
evolution (ME) and maximum likelihood (ML) methods. The
bootstrap supports (1000 repetitions) for each of these
methods are shown, respectively; thicker lines represent
clades with support higher than 80% for all three methods,
and clades supported by bootstrap values lower than 50%
were reduced to polytomies. Sequences from Glomus group A
(Schwarzott et al. 2001) were used as out group (Glomus
mosseae, AJ306438; and G. clarum, AJ276084).
S. gilmorei (clade C) should have clustered in clade B.
On one hand, these results could indicate the poor
resolution of SSU nrDNA to resolve species, or the
occurrence of recurrent mutation in these sequences
due to absence of selection (Givnish & Sytsma 1997).
On the other hand, if the molecular phylogeny is
correct then the developmental analysis might cluster
species with different ancestry based on convergence
of the IW layers. This might be possible if constraints
related to germination events take place during
species radiation. In such a scenario, nrDNA data must
give better resolution to discriminate sub-genus level
than developmental patterns based on IW layers
because nrDNA are probably under low selective
pressure during speciation in asexual organisms. The
comparison between molecular and morphological
characters is not easy, particularly in organisms were
genetics and evolutionary processes are poorly under-
stood. Besides, both data sets are subject to homoplasy.
Clearly, more research is needed to integrate morpho-
logical and molecular characters in Scutellospora
PCR-DGGE targeting the V9 region of the SSU
nrDNA could be used to discriminate Scutellospora
reticulata from all other Scutellospora species tested.
Due to the high level of sequence similarity some
closely related species (S. gregaria and S. coralloidea)
were not discriminated. Also, species in clades B and C
could be better discriminated from each other than
species from clade A (S. calospora). This poor separ-
ation within clade A may be due to the GC-rich nature
of the 3k end of amplicons in this clade, causing some
sequence differences to fall within a domain of rela-
tively high melting temperature, hampering discrimi-
nation (Table 2).
PCR-DGGE can only reveal changes that affect
melting behavior and thus some mutations (specifically
A$T and G$C changes) may go undetected. The
same sort of problems can occur during distinction of
sequences by RFLP. This technique overlooks mu-
tations that cause no changes in the restriction sites.
Also, the cloning procedure is known to affect the
integrity of the cloned DNA fragments, when using
mixed PCR products, as demonstrated by Speksnijder
et al. (2001). Here, prior to sequencing we screened
cloned products for correspondence to DGGE profiles
obtained from direct amplification of spores genomic
DNA. This step decreases the chances of selection of
clones containing PCR or cloning artifacts (Speksnijder
et al. 2001). Therefore we can be highly confident
of the sequence data. The SSU nrDNA sequences of
S. cerradencis deposited in the GeneBank were ob-
tained using the same accession used here. Based on
our DGGE analysis, the sequence deposited for S. cer-
radensis (AB041345) should have matched with S. reti-
culata clone 10 as opposed to clones 9 and 18-2
(Table 2, Fig. 14), which suggests a possible cloning or
PCR error in that sequence.
The great advantage of PCR-DGGE for analysis of
AM fungi is that it takes advantage of the intraspecific
polymorphism of the nrDNA copies to discriminate
closely related species. This was the case of S. reti-
culata, S. cerradensis, and S. heterogama. The same
approach was used to discriminate all known species
and even isolates in Gigaspora (de Souza et al. 2004).
The characterization of intraspecific nrDNA poly-
morphism using PCR-DGGE provides a detailed
molecular fingerprint suitable for species identification.
These characteristics make PCR-DGGE a good tech-
nique to control purity of AM fungal inoculum in
culture collections and to select clones before sequen-
cing, in addition, to its utility in assessing the diversity
of AM fungi species in environmental samples
(Kowalchuk, de Souza & van Veen 2002, Opik et al.
2003, de Souza et al. 2004).
revealed a novel pattern of outer spore wall differ-
(2) Phylogenetic analyses grouped S. reticulata in a
cluster together with S. cerradensis and S. hetero-
gama, which are species that share similar spore
(3) S. reticulata could be discriminated from all
16 known Scutellospora species by PCR-DGGE
analysis of the V9 region of the SSU nrDNA, a
result confirmed by sequence analysis.
FAdeS was supported by the Brazilian Council for Scientific
and Technological Development (CNPq) (grant #200850/98-9). SD
gratefully acknowledges the financial support received from the
Belgian Federal Science Policy Office (contract BCCM C3/10/003).
We would like to thank the contribution of the two anonymous
reviewers for their valuable comments. This represents publication
no. 3536 of the Netherlands Institute of Ecology (NIOO-KNAW).
Altschul, S. F., Gish, W., Miller, W., Meyers, E. W. & Lipman, D. J.
(1990) Basic Local Alignment Search Tool*1. Journal of Molecular
Biology 215: 403–410.
Benson, D. A., Karsch-Mizrachi, I., Lipman, D. J., Ostell, J. &
Wheeler, D. L. (2003) GenBank. Nucleic Acids Research 31: 23–27.
Blaszkowski, J. (1991) Polish Glomules VIII. Scutellospora nodosa, a
new species with knobby spores. Mycologia 83: 537–542.
Declerck, S., Cranenbrouck, S., Dalpe, Y., Seguin, S., Grandmougin,
F. A., Fontaine, J. & Sancholle, M. (2000) Glomus proliferum sp
nov.: a description based on morphological, biochemical, mol-
ecular and monoxenic cultivation data. Mycologia 92: 1178–1187.
Declerck, S., D’Or, D., Bivort, C. & de Souza, F. A. (2004)
Development of extraradical mycelium of Scutellospora reticulata
under root-organ culture: spore production and function of
auxiliary cells. Mycological Research 108: 84–92.
R. A. de Souza and others 705
Declerck, S., Strullu, D. G. & Plenchette, C. (1998) Monoxenic
culture of the intraradical forms of Glomus sp. isolated from
a tropical ecosystem: a proposed methodology for germplasm
collection. Mycologia 90: 579–585.
de Souza, F. A. (2000) Banco ativo de Glomales da Embrapa
Agrobiologia: catalogac ¸a ˜o e introduc ¸a ˜o de novos isolados desde
1995. Se ´ rie Documentos. Embrapa Agrobiologia, Brasilia.
de Souza, F. A. & Berbara, R. L. L. (1999) Ontogeny of Glomus
clarum in Ri T-DNA transformed roots. Mycologia 91: 343–350.
de Souza, F. A. & Declerck, S. (2003) Mycelium development and
architecture, and spore production of Scutellospora reticulata
in monoxenic culture with Ri T-DNA transformed carrot roots.
Mycologia 95: 1004–1012.
de Souza, F. A., Kowalchuk, G. A., Leeflang, P., van Veen, J. A. &
Smit, E. (2004) PCR-Denaturing Gradient Gel Electrophoresis
profiling of the inter- and intraspecies 18S rRNA gene sequence
heterogeneity is an accurate and sensitive method to assess species
diversity of arbuscular mycorrhizal fungi of the genus Gigaspora.
Applied and Environmental Microbiology 70: 1413–1424.
Fortin, J. A., Becard, G., Declerck, S., Dalpe, Y., St Arnaud, M.,
Coughlan, A. P. & Piche, Y. (2002) Arbuscular mycorrhiza on
root-organ cultures. Canadian Journal of Botany 80: 1–20.
Franke, M. & Morton, J. B. (1994) Ontogenetic comparisons of
arbuscular mycorrhizal fungi Scutellospora heterogama and
Scutellospora pellucida: revision of taxonomic character concepts,
species descriptions, and phylogenetic hypotheses. Canadian
Journal of Botany 72: 122–134.
Givnish, T. J. & Sytsma, K. J. (1997) Homoplasy in molecular vs.
morphological data: the likelihood of correct phylogenetic infer-
ence. In Molecular Evolution and Adaptive Radiation (T. J. Gvnish
& K. J. Systsma, eds): 55–101. Cambridge University Press,
Hasegawa, M., Kishino, H. & Yano, T.-A. (1985) Dating the human-
ape splitting by a molecular clock of mitochondrial DNA. Journal
of Molecular Evolution 22: 160–174.
Helgason, T., Daniell, T. J., Husband, R., Fitter, A. H. & Young,
J. P. W. (1998) Ploughing up the wood-wide web? Nature 394: 431.
Herrera-Peraza, R. A., Cuenca, G. & Walker, C. (2001) Scutellospora
crenulata, a new species of Glomales from La Gran Sabana,
Venezuela. Canadian Journal of Botany 79: 674–678.
Koske, R. E., Miller, D. D. & Walker, C. (1983) Gigaspora reticulata,
a newly described endomycorrhizal species from New England.
Mycotaxon 16: 429–435.
Kowalchuk, G. A., de Souza, F. A. & van Veen, J. A. (2002)
Community analysis of arbuscular mycorrhizal fungi associated
with Ammophila arenaria in Dutch coastal sand dunes. Molecular
Ecology 11: 571–581.
Morton, J. B. (1995) Taxonomic and phylogenetic divergence among
five Scutellospora species based on comparative developmental
sequences. Mycologia 87: 127–137.
Morton, J. B. & Benny, G. L. (1990) Revised classification of arbus-
cular mycorrhizal fungi (Zygomycetes): a new order, Glomales,
two new suborders, Glomineae and Gigasporineae, and two new
families, Acaulosporaceae and Gigasporaceae, with an emendation
of Glomaceae. Mycotaxon 37: 471–491.
Morton, J. B., Bentivenga, S. P. & Bever, J. D. (1995) Discovery,
measurement, and interpretation of diversity in arbuscular endo-
mycorrhizal fungi (Glomales, Zygomycetes). Canadian Journal of
Botany 73: 25–32.
Omar, M. B., Bolland, L. & Heather, W. A. (1979) A permanent
mounting medium for fungi. Bulletin British Mycological Society
Opik, M., Moora, M., Liira, J., Koljalg, U., Zobel, M. & Sen, R.
(2003) Divergent arbuscular mycorrhizal fungal communities
colonize roots of Pulsatilla spp. in boreal Scots pine forest and
grassland soils. New Phytologist 160: 581–593.
Pousada, D. & Crandall, K. A. (1998) Modeltest: testing the model of
DNA substitution. Bioinformatics 14: 817–818.
Redecker, D., Morton, J. & Bruns, T. D. (2000) Ancestral lineages of
arbuscular mycorrhizal fungi (Glomales). Molecular Phylogenetics
and Evolution 14: 276–284.
Schu ¨ ßler, A., Schwarzott, D. & Walker, C. (2001) A new fungal
phylum, the Glomeromycota: phylogeny and evolution. Myco-
logical Research 105: 1413–1421.
Schwarzott, D., Walker, C. & Schu ¨ ßler, A. (2001) Glomus, the
largest genus of the arbuscular mycorrhizal fungi (Glomales),
is nonmonophyletic. Molecular Phylogenetics and Evolution 21:
Simon, L., Lalonde, M. & Bruns, T. D. (1992) Specific amplification
of 18S fungal ribosomal genes from vesicular-arbuscular endo-
Microbiology 58: 291–295.
Spain, J. L. & Miranda, J. C. (1996) Scutellospora cerradensis:
an ornamented species in the Gigasporaceae (Glomales) from the
Cerrado region of Brazil. Mycotaxon 60: 129–136.
Speksnijder, A. G. C. L., Kowalchuk, G. A., de Jong, S., Kline, E.,
Stephen,J. R.& Laanbroek,
artifacts introduced by PCR and cloning of closely related 16S
rRNA gene sequences. Applied Environmental Microbiology 67:
Strullu, D. G. & Romand, C. (1986) Me ´ thod d’obtention d’en-
domycorrhizaes a ` ve ´ sicules et arbuscules en conditions axe ´ niques.
Comptes Rendus de l’Acadimie des Sciences, Paris se´rie 3 303:
Stu ¨ rmer, S. L. & Morton, J. B. (1999) Scutellospora rubra, a new
arbuscular mycorrhizal species from Brazil. Mycological Research
Swofford, D. L. (2003) PAUP*: phylogenetic analysis using parsi-
mony (*and other methods). Version 4. Sinauer Associates,
Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. &
Higgins, D. J. (1997) The CLUSTAL_X windows interface:
flexible strategies for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Research 25: 4876–4882.
Walker, C. (1983) Taxonomic concepts in the endogonaceae:
spore wall characteristics in species descriptions. Mycotaxon 18:
Walker, C., Cuenca, G. & Sa ´ nchez, F. (1998) Scutellospora spino-
sissima sp. nov., a newly described glomalean fungus from acidic,
low nutrient plant communities in Venezuela. Annals of Botany 82:
Walker, C. & Schu ¨ ßler, A. (2004) Nomenclatural clarifications
and new taxa in the Glomeromycota. Mycological Research 108:
White, T. J., Bruns, T., Lee, S. & Taylor, J. (1990) Amplification and
direct sequencing of fungal ribosomal RNA genes for phylo-
genetics. In PCR Protocols: a guide to methods and applications
(M. A. Innis, D. H. Gelfand, J. J. Sminski & T. J. White, eds):
315–322. Academic Press, San Diego.
Corresponding Editor: P. Bonfante