Content uploaded by George Khachatourians
Author content
All content in this area was uploaded by George Khachatourians on May 03, 2015
Content may be subject to copyright.
Relationships between fungal spore
morphologies and surface properties for
entomopathogenic members of the genera
Beauveria,Metarhizium,Paecilomyces,
Tolypocladium, and Verticillium
Lloyd B. Jeffs, Ilungo J. Xavier, Russell E. Matai,
and George G. Khachatourians
Abstract: The surface properties of aerial conidia (AC) from 24 strains of entomopathogenic fungi were studied and
compared using the salt-mediated aggregation and sedimentation (SAS) assay, electron microscopy, FITC-labelled
lectins, and spore dimensions. Spores with rugose surfaces were hydrophobic, whereas hydrophilic spores had smooth
surfaces. Correlation analysis found no link between spore dimensions and either hydrophobicity or surface
carbohydrates. However, there was a strong positive correlation between spore hydrophobicity and surface
carbohydrates. The three spore types of Beauveria bassiana were all shown to possess discrete surface
hydrophobicities, which were also strongly linked to surface carbohydrate profiles. Various chemical treatments had
pronounced effects on spore surface properties, with sodium dodecyl sulfate (SDS) and formic acid (FA) reducing both
lectin binding and surface hydrophobicity. When FA-protein extracts were separated and analysed using SDS-PAGE,
only the hydrophobic spores had low molecular weight hydrophobin-like peptides that were unglycosylated and
contained disulfide bonds. The strains with hydrophilic AC had much lower levels of FA-extractable protein per spore
dry weight compared to their more hydrophobic counterparts. Moreover, extracts of the more hydrophobic spores
tended to have greater protein:carbohydrate ratios.
Key words: fungi, spores, hydrophobicity, lectins, morphology, microbial insecticides, protein.
Résumé : A partir d’une collection de 24 souches de champignons entomopathogènes, nous avons étudié et comparé
les propriétés de surface des conidies aériennes (AC) par un essai de formation d’agrégats par le sel et de
sédimentation (SAS), par microscopie électronique, avec des lectines marquées au FITC et selon la dimension des
spores. Les spores qui avaient une surface ruqueuse étaient hydrophobes tandis que celles qui avaient une surface lisse
étaient hydrophiles. Une analyse de corrélation n’a pas établi de lien entre les dimensions de la spore et
l’hydrophobicité ou les hydrates de carbone de la surface. Il y avait par contre une forte corrélation positive entre
l’hydrophobicité de la spore et les hydrates de carbone de surface. Les trois types de spores de Beauveria bassiana
possédaient toutes une légère hydrophobicité de surface qui était fortement liée au profil des hydrates de carbone de la
surface. Différents traitements chimiques ont eu un effet prononcé sur les propriétés de surface de la spore et, entre
autres, le dodécyl sulfate de sodium (SDS) et l’acide formique (FA) ont diminué la capacité de liaison aux lectines et
l’hydrophobicité de surface. Lors de la séparation et de l’analyse d’extraits protéiques FA par SDS-PAGE, seules les
spores hydrophobes possédaient des peptides de faible poids moléculaire de type hydrophobine et ceux-ci étaient non-
glycosylés et contenaient des liens disulfures. Les souches porteuses de conidies AC hydrophiles possédaient de plus
faibles quantités de protéines pouvant être extraites à la FA par gramme de poids sec de spores comparativement à
leurs analogues plus hydrophobes. Le rapport hydrates de carbone: protéines avait tendance à être plus élevé dans les
extraits des spores hydrophobes.
Mots clés : champignons, spores, hydrophobicité, lectines, morphologie, insecticides microbiens, protéine.
[Traduit par la Rédaction] Jeffs et al. 948
Can. J. Microbiol. 45: 936–948 (1999) © 1999 NRC Canada
936
Received March 23, 1999. Revision received August 10, 1999. Accepted August 13, 1999.
L.B. Jeffs, I.J. Xavier, R.E. Matai, and G.G. Khachatourians.1Bioinsecticide Research Laboratory, Department of Applied
Microbiology and Food Science, University of Saskatchewan, 51 Campus Drive, Saskatoon, SK S7N 5A8, Canada.
1Author to whom all correspondence should be addressed (e-mail: khachatouria@sask.usask.ca)
Introduction
Entomopathogenic fungi (EPF) can provide safe and ef-
fective control of many important insect pests. Currently,
economic interest lies in focusing upon fungi from three
genera within the class Hyphomycetes;Beauveria,Verti-
cillium, and Metarhizium. Recent research has attempted to
meet the challenges of ensuring the stability (e.g. protection
from UV inactivation), efficacy (especially at low relative
humidities), and mass production of inexpensive yet high
quality fungal formulations (Feng et al. 1994). Naturally,
there is also continual interest in the discovery, isolation and
characterization of novel fungal isolates. It is estimated that
over 750 EPF species from over 100 genera have been de-
scribed to date (Samson et al. 1988; Carruthers and Hural
1990).
The driving force for studying EPF spores is that they ini-
tiate insect pathogenesis and are the active ingredient of
mycoinsecticide formulations. Surface properties determine
how these reproductive propagules interact with biotic and
abiotic factors in their particular habitats during dormancy,
dispersal, and their eventual association with a suitable sub-
strate or host. Before a fungal spore can even launch an at-
tack on the host cuticle, it must possess nonspecific
properties (e.g. surface hydrophobicity) and specific surface
antigens that allow for the specific binding to the insect cuti-
cle (Rath et al. 1995). The spore must also find sufficient
moisture and nutrients for germination, while resisting
fungistatic components before attempting to penetrate the
cuticle (Butt et al. 1995; Sitch and Jackson 1997). Infection
will then only be successful if the EPF can produce the nec-
essary combination of digestive enzymes and can overcome
the immune responses of the host (Khachatourians 1996).
Hyphomycetous entomopathogenic spores have tradition-
ally been divided into two general categories, based upon
the physical nature of their outer walls—wet and dry (re-
viewed by Boucias and Pendland 1991). The wet (hydro-
philic) conidia (e.g. those of Verticillium lecanii and
Aschersonia aleyrodis) characteristically lack a well orga-
nized outer rodlet layer as seen for dry spores but instead of-
ten possess a mucilaginous coat. This mucus coat facilitates
spore dispersal by rainfall and can act as a glue in adhesion
to insect hosts. The dry, hydrophobic conidia of Aspergillus
spp. and Penicillium spp. were found over 30 years ago to
possess an outer layer comprising bundles of rodlets ar-
ranged in a tight basket-weave arrangement (Hess et al.
1968; Hess and Stocks 1969). These proteinaceous rodlets
have now been found on the spores of many species of
fungi, and vary considerably in length, thickness and ar-
rangement in fascicles (Boucias and Pendland 1991; Wessels
1997).
Typically, the rodlet layer is very resilient and can only be
disrupted with a high powered sonicator and/or by harsh
chemical treatments involving very concentrated organic ac-
ids. Boucias et al. (1988) demonstrated that the resilience of
the rodlet layers on various EPF conidia can also vary, with
rodlets of Beauveria bassiana conidia being more prone to
removal by an incubation in 0.1 N NaOH than those of
Metarhizium anisopliae and Nomuraea rileyi conidia.
Hydrophobins are the proteins that are the main constituents
of the rodlet layer and were first characterized during a
search for genes expressed during the emergence of aerial
hyphae in Schizophyllum commune (Basidiomycota)
(Schuren and Wessels 1990; Wösten et al. 1993). Over 20
hydrophobins have now been identified in fungi from the
phyla Basidiomycota and Ascomycota, and some anamorphs
have also been shown to possess hydrophobin genes and/or
peptides (reviewed by Wessels 1997). For EPF species, a
putative hydrophobin gene has been identified for M. ani-
sopliae (St. Leger et al. 1992), while a hydrophobin peptide
(9.7 kDa) from the rodlet layer of B. bassiana conidia has
N-terminal similarities to other known hydrophobins
(Bidochka et al. 1995). Although other EPF are known to
produce conidia with rodlet layers (e.g. N. rileyi,Paecilo-
myces fumosoroseus), the peptides that make up these
rodlets remain to be isolated and characterized.
The purpose of this study was to characterize and com-
pare the surface properties of spores produced by entomo-
pathogenic hyphomycetes with the aim of gaining a further
understanding of how these attributes relate to spore attach-
ment. Light and scanning electron microscopy (SEM) were
employed to compare spore morphology and surface fea-
tures. The surface hydrophobicity of spores from various
strains as well as different spore types were estimated using
the salt-mediated aggregation and sedimentation (SAS) as-
say. Biochemical approaches were also used to identify,
quantify, and characterize surface proteins and carbohy-
drates.
Materials and methods
Preparation of spore suspensions
The EPF strains studied, with their functional abbreviations and
sources, are listed in Table 1. Aerial conidia were produced from
solid cultures grown on a yeast-peptone-dextrose (YPD) agar me-
dium, pH 5.9, comprising (per L): 2 g yeast extract (technical
grade), 10 g peptone, 20 g dextrose, and 15 g Bacto-agar; supplied
by Difco Laboratories (Detroit, Mich.). Beauveria bassiana
blastospores (BS) and submerged conidia (SC) were produced us-
ing liquid media. A peptone-dextrose (PD) medium, pH 6.1, was
used to produce BS and comprised (per L): 10 g peptone, 20 g dex-
trose. A broth medium possessing high carbon to inorganic nitro-
gen ratios was used to promote selective SC production, which was
originally developed by Thomas et al. (1987) and modified by
Bidochka et al. (1995). This broth was adjusted to pH 5.0 with
0.1 M NaOH and comprised (per L): 50.0 g glucose, 10.0 g KNO3,
5.0gKH
2PO4, 2.0 g MgSO4·7H2O, 50 mg CaCl2·2H2O, and 50 mg
yeast extract.
Agar plates were typically inoculated using approximately
0.1 mL of a 100:1 dilution of a stock spore suspension in sterile
distilled water (sdH2O). The inoculum was spread over the entire
surface of the plate with a bent glass rod to promote confluent
growth. Liquid media were inoculated using 0.1 mL of the stock
spore suspension per 100 mL of media. After inoculation, all me-
dia were incubated at 21–25°C for 14 d or other specified times,
with agar plates kept in closed plastic bags, whereas 100-mL
aliquots of liquid media in sealed 250-mL Erlenmeyer flasks were
constantly mixed using a rotary shaker set at 150 rpm. At specified
times, AC were harvested twice from plates by flooding with
10 mL of sdH2O, removing spores from the mycelial mat using a
bent glass rod and collecting the resulting spore suspension using a
Pasteur pipette. These spore suspensions were then filtered twice
through Pasteur pipettes containing 8-micron Pyrex fiber glass to
remove any mycelial fragments. Larger volumes (e.g. ≥50 mL)
were filtered using an autoclavable 60-mm diameter polypropylene
conical funnel lined with a fine mat of fiber glass, resting in a
© 1999 NRC Canada
Jeffs et al. 937
250 mL Erlenmeyer flask and kept together with aluminum foil.
Another piece of aluminum foil was placed over the funnel mouth
for a lid. After autoclaving, the tin foil lid was temporarily re-
moved and the spore suspension was added through a hole made in
the underlying foil. The filtered spores were then washed twice
with sdH2O water, and stored at 4°C before use. A similar method
was used for harvesting BS and SC, though the entire liquid cul-
ture was passed through the larger filters. The concentrations of
the various spore suspensions were estimated using a haemocyto-
meter, while optical densities were determined at 610 nm with a
spectrophotometer.
Microscopy
A Leitz phase contrast research microscope was used to enumer-
ate diluted spore suspensions with a haemocytometer, and to mea-
sure spore dimensions with an ocular micrometer, calibrated with a
stage micrometer. Spore volume and surface area were estimated
by measuring the long and short axes of the spores. A Jenamed 2
fluorescence microscope (Carl Zeiss, Jena, Germany) was used to
view the binding of FITC-labelled lectins, and various polystyrene
beads, to spores.
Fungal spores were prepared for viewing with scanning electron
microscopy by initially placing in a fixative solution, consisting of
3% (v/v) glutylaldehyde and 0.1 M cacodylic acid, for at least 24 h
(4°C). Samples were then dehydrated by immersion in a series of
increasingly concentrated ethanol solutions (10–95%) for ≥1h
each. Spore samples were placed in absolute ethanol for ≥1h,and
then this suspension was applied directly to pieces of solvent-
resistant Nucleopore polycarbonate 0.4 µm filters. The ethanol rap-
idly evaporated leaving a thin coat of spore material on the filter.
Samples were then mounted and viewed using scanning electron
microscopy (SEM) as described by Jeffs et al. (1997) in the facili-
ties of the Biology Department, University of Saskatchewan.
Hydrophobicity and lectin-binding assays
The salt-mediated aggregation and sedimentation (SAS) assay
was used to quantitatively estimate spore surface hydrophobicity
(Jeffs and Khachatourians 1997). The principle of the SAS assay is
that spores with greater surface hydrophobicity aggregate by salt-
ing out to a greater degree than more hydrophilic spores, and these
spore aggregates then sediment out of suspension faster than single
spores. The SAS assay was performed at 24°C and with spore
suspensions at 0.6 initial optical density (OD) at 610 nm using
6× 50 mm glass culture tubes with uniform wall thickness. All
buffers and stock salt solutions were passed through 0.4 µm filters
before use in these assays. Filtered and washed spores were first
resuspended in 2.0 mM sodium phosphate buffer (pH 6.8), then
20 µL of these spore suspensions were added 180 µLof2.0mM
sodium phosphate buffer (pH 6.8), to which was added 200 µLof
0.20 M ammonium sulfate in 2.0 mM sodium phosphate (pH 6.8)
© 1999 NRC Canada
938 Can. J. Microbiol. Vol. 45, 1999
Species Strain Abbreviation Source
Beauveria bassiana F6 Ba06 Costelytra zealandica, New Zealanda
GK 2016 Ba16 Lab isolateb
USSR 2274 Ba74 Not available (NA)c
GK 2051 Ba51 Lab isolated
F33 Ba33 Aphodius tasmaniae, New Zealanda
DAOM 210569 Ba69 Beetle, B.C., Canadae
MsGSM BaMs Melanoplus sanguinipes, Sask., Canadaf
ARSEF 2882 Ba82 Diuraphis noxia, Idahog
SRS Bb-86-5 Ba865 Melanoplus bivittatus, Sask., Canadah
ARSEF 2860 Ba860 Schizophis graminum, Idahog
DAOM 144746 Ba46 Soil, Alta., Canadae
ATCC 44860 Ba60 Soil, Georgiai
DAOM 195005 Ba05 Choristoneutra fumiferana, Que, Canadae
B. brongniartii 656 Br656 Nephotettix cincticepts, Chinaj
F156 Br156 Costelytra zealandica, New Zealanda
979 Br79 Melolontha melolontha, Francej
B. densa DAOM 57904 Bd04 Decaying tree, Canadae
Metarhizium anisopliae SL 297 Ma97 NAk
SL 549 Ma49 NAk
SL 2165 Ma65 NAk
Paecilomyces farinosus ATCC 1360 Pf60 NAi
Tolypocladium cylindropsorum ATCC 56519 Tc19
T. nivea ATCC 18981 Tn81 NAi
Verticillium lecanii ATCC 46578 Vl78 NAi
Superscripted letters indicate where strains were obtained.
aT. Glare, AgResearch, New Zealand Pastoral Agriculture Institute Ltd, Lincoln, New Zealand.
bR.G. Lidstone, Department of Applied Microbiology and Food Science, University of Saskatchewan, Saskatoon, Sask., Canada.
cas strain UFC 423 from Dr. Vitaley Sencenko, ALL Soviet Collection Center for Microorganisms, Moscowblast, Russia.
dMutant derived from GK2016 isolate (Kosir et al. 1991).
eCanadian Collection of Fungus Cultures, Ottawa, Ont., Canada.
fG.S. Miranpuri, Department of Applied Microbiology and Food Science, University of Saskatchewan, Saskatoon, Sask., Canada.
gUSDA-ARS Collection of Entomopathogenic Fungal Cultures.
hM. Erlandson, Agriculture Canada Research Station, Saskatoon, Sask., Canada.
iAmerican Type Culture Collection, Baltimore, Md.
jM.J. Bidochka, Boyce Thompson Institute, Ithaca, N.Y.
kR.J. St Leger, Boyce Thompson Institute, Ithaca, N.Y.
Table 1. Sources of entomopathogenic fungal strains.
for a total volume of 400 µL. These salt-spore suspensions were
vigorously mixed with a vortex mixer for 10 s, followed by an ini-
tial OD reading and a final reading at 2 h. The proportions of
spores remaining in suspension after specified times were cor-
rected using values from controls that possessed no ammonium
sulfate.
Various polystyrene latex microspheres were used to detect
electronegative, electropositive, and hydrophobic groups on the
surfaces of selected spore types and strains that had previously
been studied using the SAS and lectin binding assays. The spore-
bead attachment assay was adapted from an assay developed by
Hazen and Hazen (1987) to investigate hydrophobic sites on the
surfaces of Candida albicans yeast cells. Initially, spores and beads
(supplied by Sigma Chemical Co., St. Louis, Mo.) were washed,
filtered and suspended in 2 mM sodium phosphate buffer (pH 6.8).
The concentration of the spore suspension was adjusted to2×10
8
spores/mL, while the latex bead suspensions were adjusted to ap-
proximately2×10
9beads/mL. One hundred µL of spore suspen-
sion and an equal volume of bead suspension were added toa6×
50 mm glass culture tube and vigorously mixed for 10 s using a
vortex mixer. After 5 min, 3–5 µL aliquots were taken from the
10:1 bead-spore mixture and viewed using light and epi-
fluorescence microscopy under oil immersion, since some beads
were impregnated with fluorescent dye to aid their detection.
To determine the presence of spore surface carbohydrates, six
FITC-labelled lectins with specific binding properties (Sigma
Chemical Co., St. Louis, Mo.) were used: concanavalin A (ConA)
specific for α-D-mannose and α-D-glucose; phytohaemagglutinin P
(PHA-P) for oligosaccharides; Ricinus communis toxin (RCA) for
β-D-galactose and N-acetyl-D-galactosamine; soybean agglutinin
(SBA) for N-acetyl-D-galactosamine; wheat germ agglutinin
(WGA) for N-acetyl-D-glucosamine; winged pea agglutinin (WPA)
for α-L-fucose. The lectin binding assay was initially optimized to
ensure that minimal amounts of lectin were used, and binding
specificities of these lectins were confirmed by using their target
carbohydrates as blockers of lectin-spore binding (see Pendland
and Boucias 1984). The following solutions were initially pre-
pared: phosphate-buffered saline (PBS; 10 mM, pH 7.2) with 0.1%
sodium azide; 1 mg/mL FITC-conjugated lectin solutions in PBS
with 0.1% sodium azide; spore suspensions (1 × 108spores/mL) in
PBS. These solutions were then added to a 1.5-mL micro centri-
fuge tube in the following amounts: 5 µL of lectin solution, 45 µL
of PBS, and 50 µL of spore suspension. The contents of the tube
were gently mixed using a vortex mixer at a low setting and the
tube was placed in a dark cupboard for 30 min, with tubes mixed
again after the first 15 min. After this incubation, the tube was cen-
trifuged at 10 000 × gfor 2 min and 50µL of the supernatant was
removed without disturbing the small spore pellet at the bottom of
the tube. The spores were then washed twice by adding 500 µLof
fresh PBS, centrifuging (10 000 × g, 2 min) and then removing an
equal volume of supernatant. Labelled spores were stored at 4°C in
darkness until viewing to minimize the loss of FITC fluorescence.
Spore-lectin binding was assessed by viewing spores under oil
immersion with a fluorescence microscope at 1000×. The follow-
ing scheme was used to record and compare lectin binding to
spores, manifested as wall fluorescence: ++ strong to moderate, +
moderate to weak, ± weak, – no observed fluorescence. Relation-
ships between lectin binding properties, surface hydrophobicity,
and spore dimensions were further analysed by generating Pearson
correlation coefficients and probability values that were then used
to determine whether paired variables were significantly corre-
lated.
Spore chemical and physical treatments
Aerial conidia of Ba16 were also subjected to additional chemi-
cal and physical treatments in attempts to alter cell surface proper-
ties, which can be separated into the categories of non-destructive
and destructive treatments involving boiling and/or harsh chemical
treatments. Aliquots containing2×10
9filtered and washed spores
were subjected to these various treatments, followed by refiltering
and washing before changes in their surfaces were determined us-
ing the SAS and lectin binding assays.
The non-destructive treatments involved incubating spores with
the poly-cationic poly-lysine; laminarase, an enzyme shown to
break β-1–3 glucan linkages present in yeast cell walls (Montijn et
al. 1994); and BSA, a relatively hydrophobic protein shown to as-
sociate with the hydrophobic surface of A. fumigatus conidia (Thau
et al. 1994). Poly-lysine and BSA were prepared as 2 mg/mL solu-
tions in 2 mM sodium phosphate (pH 6.8) buffer, with these solu-
tions passed through 0.4 µm filters before mixing 0.5 mL with
spores. These mixtures were incubated for1hat4°Cbefore spores
were washed twice in 1 mL sdH2O, filtered through glass wool and
washed and suspended in 2 mM sodium phosphate (pH 6.8) buffer.
The laminarase solution (0.2 units/mL) was prepared in a 0.1 M
sodium acetate buffer (pH 5.0) containing 1 mM phenylmethyl-
sulfonylfluoride (PMSF). Again, this solution was passed through
0.4 µm pore size filters before adding 0.5 mL to each spore aliquot.
Spore-laminarase mixtures were then incubated at 37°C for various
times (0–4 h), after which supernatants were measured for total
protein and carbohydrate, while spores were prepared for SAS and
lectin binding assays as for BSA-treated spores.
The destructive treatments included boiling spores in water,
DTT, or SDS extraction buffers; incubating spores in 97% (v/v)
FA; and boiling in SDS followed by FA treatment. The DTT treat-
ment buffer comprised 50 mM DTT in 0.1 M Tris HCl buffer (pH
8.6) with 5 mM EDTA, and has been used to deglycosylate yeast
cell surfaces of C. albicans (Hazen et al. 1990). The SDS extrac-
tion buffer consisted of 1% (w/v) SDS in 0.05 M Tris HCl buffer
(pH 8.0) with 1 mM EDTA and 100 µg/mL PMSF. Spores were
mixed with 0.5 mL of buffer or sdH2O and then boiled at 100°C
for 10 min. The FAtreatment involved incubating spores in 0.5 mL
of 97% FA at 4°C, in a sonicating water bath for 2 h. The acidic
supernatant was removed after centrifugation (10 000 × g, 5 min)
and the spores were neutralized by resuspending in 1.5 M Tris HCl
(pH 8.8). The spores from all of these treatments were then pre-
pared for SAS and lectin binding assays as previously described
for the non-destructive treatments.
Spore proteins
The surface proteins of spores from various strains and types
were studied, with interest concentrated on isolating hot-SDS in-
soluble, FA-soluble proteins that may convey hydrophobic proper-
ties to spore surfaces. Spore extracts from each treatment were
estimated for protein and carbohydrate, and proteins in these ex-
tracts were separated and analysed by using the sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) discontin-
uous buffer system of Laemmli (1970). Protein separation with
SDS-PAGE was performed using the Mini-PROTEAN II electro-
phoresis system (Bio-Rad). Proteins were then detected by staining
and gel images were collected and studied using an image ana-
lyzer.
Aliquots of spores (1–2 × 109) from various strains were ex-
tracted using: hot SDS, FA, performic acid, or a series of these
treatments. Corresponding spore aliquots of known amounts were
dried in a 90°C oven for4htofind the dry weight of each spore
sample undergoing protein extraction. Spore aliquots in 4 mL of
sdH2O were disrupted with a Bronwill Biosonik sonicator
(Bronwill Scientific, Rochester, N.Y.) at a high power setting for
three 1-min bursts in an attempt to loosen the resilient outer spore
wall and facilitate subsequent protein extraction. Spore aliquots
were kept on ice during and between sonication periods to mini-
mize sample heating. After sonication, spores pellets were concen-
trated into 1.5 mL micro centrifuge tubes. The hot SDS extraction
consisted of incubating spores with 0.4 mL of 1% SDS extraction
© 1999 NRC Canada
Jeffs et al. 939
buffer, and boiling at 100°C for 10 min. After centrifuging at
10 000 × gfor 5 min, the supernatant was saved for further analy-
sis while the extraction was repeated once for the spores. Twice
SDS-extracted spores were then subjected to treatments with
0.2 mL of either 97% FA or performic acid to extract
hydrophobin-like surface proteins. The performic acid solution is
capable of cleaving peptide disulfide bonds and was prepared ac-
cording to Wessels et al. (1991). The organic acid treatments in-
volved resuspending spores in 0.2 mL of acid and a 2-h incubation
at 4°C in a sonicating water bath. These treatments were also re-
peated for spore samples, while collected supernatants were dried
down using a desiccator attached to a vacuum for 36–48 h. Once
the acid supernatants samples had completely dried, they were re-
suspended in 300 µL 0.5 M Tris HCl buffer (pH 8.0) to neutralize
any residual acidity, and samples were centrifuged at 10 000 × g
for 10 min to separate supernatant from any insoluble precipitates.
Sample supernatants were then prepared for the estimation of total
protein and carbohydrate using colorimetric assays. A Bio-Rad
protein assay, based on the dye-binding assay of Bradford (1976)
was used to estimate total protein, while total carbohydrate in
spore extracts was also examined using an anthrone assay de-
scribed by Herbert et al. (1971).
After protein estimation, protein extracts were diluted at least
twofold with the sample buffer. The Tris-glycine sample buffer
comprised of the following: 62.5 mM Tris HCl (pH 6.8), 10%
(v/w) glycerol, 2% (w/v) SDS, 125 mM DTT, and 0.005% (w/v)
bromophenol blue. The diluted protein samples were then heated
with the sample buffer for 3 min at 95°C. Samples were allowed to
cool and then aliquots containing 2–15 µg protein were loaded on
Tris-glycine gels with 4% stacking gels and 16% resolving gels
prepared according to Bio-Rad’s standard protocol. All gels were
run at a constant voltage of 200 V. After electrophoresis, protein
bands on these gels were detected using either silver staining
(based on the method of Merrill et al. 1981) or periodic acid with
Schiff’s reagent (PAS) staining for glycoproteins (Wan and van
Huystee 1993).
Results
Physical properties and morphologies of spores from
various strains
The physical properties and morphologies of AC from 24
fungal strains were studied using light and electron micros-
copy. These AC were found to be either spheroidal, elon-
gated, or cylindroidal in shape (Fig. 1). The lengths and
widths of these spores were measured using light micros-
copy and from these values spore volumes and surfaces ar-
eas were calculated (Table 2). Aerial conidia from the
B. bassiana strains showed considerable variability with
spore mean lengths and widths ranging from 1.9 to 3.3 µm
and 1.7 to 3.0 µm. Additionally, spheroidal AC were pro-
duced by Br156, Bd04, Pf60, and Tn81. Unlike AC of
Br156, the other B. brongniartii strains (i.e. Br656, Br79)
had a slightly elongated shape. All M. anisopliae strains (i.e.
Ma49, Ma65, and Ma97) had similar cylindroidal
morphologies, while T. cylindrosporum and V. lecanii spores
were also cylindroidal. Scanning electron microscopy re-
vealed that spore surfaces were either rugose or smooth, and
© 1999 NRC Canada
940 Can. J. Microbiol. Vol. 45, 1999
DimensionsaHydrophobicity Lectin Bindingc
Strain Spore length, width µm Volume µm3Surface area µm2SAS assay (%)bConA PhaP RCA SBA WGA WPA
Ba06 1.9 ± 0.1, 1.8 ± 0.1 3.6 ± 0.3 11.2 ± 0.6 45.7 ± 1.8 ++ ++ ++ ++ + +
Ba16 3.2 ± 0.1, 3.0 ± 0.1 16.2 ± 2.1 29.8 ± 2.7 50.6 ± 1.2 ++ ++ ++ ++ ++ +
Ba74 2.7 ± 0.1, 2.5 ± 0.1 9.5 ± 1.2 21.3 ± 1.7 51.4 ± 2.4 ++ ++ ++ ++ + +
Ba51 3.3 ± 0.1, 3.0 ± 0.1 16.8 ± 1.6 31.3 ± 1.9 55.7 ± 1.3 ++ ++ ++ ++ ++ +
Ba33 2.6 ± 0.1, 2.0 ± 0.1 6.1 ± 0.6 16.2 ± 1.1 58.2 ± 3.7 ++ ++ ++ ++ ++ ++
Ba69 2.2 ± 0.1, 1.7 ± 0.1 3.8 ± 0.4 11.8 ± 0.8 59.1 ± 2.8 ++ ++ ++ ++ ++ ++
BaMs 2.2 ± 0.1, 1.8 ± 0.1 4.3 ± 0.4 12.8 ± 0.8 61.4 ± 1.6 ++ ++ ++ ++ ++ ++
Ba82 2.6 ± 0.1, 2.0 ± 0.1 6.1 ± 0.5 16.3 ± 0.9 62.8 ± 1.0 ++ ++ ++ ++ ++ ++
Ba865 3.2 ± 0.1, 2.8 ± 0.1 14.5 ± 1.3 28.5 ± 1.7 66.4 ± 1.8 ++ ++ ++ ++ ++ ++
Ba860 2.6 ± 0.1, 2.4 ± 0.1 8.3 ± 0.7 19.5 ± 1.2 71.0 ± 1.5 ++ ++ ++ ++ + ++
Ba46 2.4 ± 0.1, 2.2 ± 0.1 6.7 ± 0.6 17.0 ± 1.0 71.3 ± 3.9 ++ ++ ++ ++ ++ ++
Ba60 2.7 ± 0.1, 2.0 ± 0.1 6.5 ± 0.6 17.1 ± 1.0 80.1 ± 3.1 ++ ++ ++ ++ ++ ++
Ba05 3.1 ± 0.1, 3.0 ± 0.1 16.2 ± 2.1 29.8 ± 2.7 86.6 ± 1.3 ++ ++ ++ + + ++
Br656 3.6 ± 0.1, 2.1 ± 0.1 9.6 ± 0.4 23.0 ± 0.7 60.2 ± 1.2 ++ ++ + ++ + ++
Br156 2.5 ± 0.1, 2.4 ± 0.1 8.2 ± 1.5 18.8 ± 2.0 74.3 ± 2.4 ++ ++ + + + +
Br79 3.6 ± 0.1, 2.6 ± 0.1 14.4 ± 1.1 29.1 ± 1.5 88.6 ± 2.9 ++ – + + + +
Bd04 2.3 ± 0.1, 2.0 ± 0.1 5.7 ± 0.9 15.0 ± 1.5 99.3 ± 1.8 ++ – ± – ++ ±
Ma97 6.4 ± 0.1, 2.8 ± 0.1 34.3 ± 1.7 56.8 ± 1.8 65.8 ± 0.6 + + ± ± + +
Ma49 6.1 ± 0.2, 2.5 ± 0.1 25.7 ± 1.7 47.5 ± 2.0 68.6 ± 0.8 ++ ± ± ± ++ ±
Ma65 5.3 ± 0.2, 2.5 ± 0.1 22.2 ± 1.9 41.5 ± 2.3 69.7 ± 0.8 ++ + + ± + ±
Pf60 3.3 ± 0.1, 2.9 ± 0.1 16.1 ± 1.8 30.3 ± 2.2 60.1 ± 2.9 ++ ++ ++ + ++ ++
Tc19 5.0 ± 0.2, 1.5 ± 0.1 8.5 ± 1.0 23.8 ± 1.8 100.3 ± 0.3 ++ – – – + –
Tn81 2.6 ± 0.3, 2.1 ± 0.1 7.1 ± 1.4 17.5 ± 2.3 101.7 ± 1.6 ++ – – – + –
Vl78 4.9 ± 0.2, 2.2 ± 0.1 17.0 ± 2.0 34.3 ± 2.4 101.9 ± 1.3 + – – – – –
aValues are means ± S.E., n= 20.
bValues are means ± S.E., n=4.
cFluorescent lectin binding was scored as follows: ++, strong to moderate; +, moderate to weak; ±, weak; –, none observed.
Table 2. Spore dimensions, surface hydrophobicity, and lectin binding.
both spheroidal and elongated spores could possess either
surface morphology (Fig. 1).
The surface hydrophobicities of conidia from 24 fungal
strains are also presented in Table 2. The AC from the 13
B. bassiana strains exhibited quite a broad range of SAS
values from 45.7% for Ba06, to 88.6% for Ba05. Aerial
conidia from the B. densa strain were very hydrophilic and
exhibited minimal sedimentation in the presence of salt,
while AC from the three B. brongniartii strains possessed
SAS values similar to the upper half of the range seen for
B. bassiana strains. The hydrophobicity values for Ba16 and
Ba51 varied slightly from the values previously reported us-
ing 0.1 M ammonium sulfate (Jeffs and Khachatourians
1997). Aerial conidia from P. farinosus and three M. ani-
sopliae strains were also found to be quite hydrophobic and
were also previously shown to possess rugose surfaces as
seen for AC from all but one of the Beauveria strains. This
exception was Bd04, since its AC were found to possess
smooth surfaces, apparently lacking a rodlet layer. Both the
AC from Tolypocladium strains and the single Verticillium
strains also were very hydrophilic and possessed smooth
spore surfaces when viewed with SEM.
Aerial conidia from all B. bassiana strains exhibited simi-
lar binding affinities, with some variation, with WGA and
WPA binding (Table 2). The AC of the B. brongniartii
strains each possessed distinct lectin profiles, with Br656
AC showing the greatest lectin binding followed by those of
Br156, and Br79, which had no PHA-P binding. This obser-
vation is interesting since all of the most hydrophilic spores
showed no PHA-P binding. The very hydrophilic B. densa
only bound the lectins ConA and WGA strongly, and WPA
and PHA-P weakly. Metarhizium anisopliae strains bound to
all lectins, though binding was weak to moderate for most,
except for ConA and WGA where binding was stronger.
Paecilomyces farinosus exhibited moderate to strong bind-
ing for all lectins. The lectin binding profiles of Tolypocla-
dium and Verticillium strains were particularly interesting.
As with all strains tested, they did bind to ConA but pos-
sessed no reactivity to any other lectin, with the exception of
WGA for the Tolypocladium strains.
The collection of spore measurements, spore hydropho-
bicity values and the inference of surface carbohydrates with
FITC-labelled lectins allowed for the further investigation of
the relationships among these spore surface properties.
© 1999 NRC Canada
Jeffs et al. 941
Fig. 1. Scanning electron micrographs of cylindroidal and spheroidal EPF spores. A) B. bassiana GK2016; B) M. anisopliae 2165;
C) T. cylindrosporum;D)T. nivea. Scale bars = 1 µm.
Pearson’s Correlation analysis was used to statistically de-
termine which spore properties were linked, with the find-
ings summarized in Fig. 2. To relate surface carbohydrates
with both hydrophobicity and spore surface area, a total
lectin binding score was used, where the degree of lectin
binding was scored on a scale of zero to three for each of the
six lectins, giving a maximum overall value of 18. Apart
from a small negative correlation (r= –0.56, P= 0.005) be-
tween spore length and total lectin binding, there was no sig-
nificant correlation between spore dimensions and either
surface hydrophobicity (greater SAS values indicate more
hydrophilic spores) or surface carbohydrates. Total lectin
binding and SAS values showed a strong correlation (r=
–0.77, P= 0.0001), and similar values were see for SAS val-
ues and individual lectins; PHA-P (r= –0.81, P= 0.0001),
SBA (r= –0.78, P= 0.0001), RCA (r= –0.74, P= 0.0001),
WPA (r= –0.61, P= 0.002). However, WGA binding and
hydrophobicity were only weakly correlated (r= –0.44,
P= 0.03) and there was no correlation with ConA binding
(r= –0.23, P= 0.28). This latter result is not surprising,
since even the most hydrophilic spores bound ConA, indi-
cating that this lectin may actually be binding to the exposed
glucan component of the walls of the hydrophilic spores,
rather than the surface glycoproteins of the hydrophobin-
containing spores. Notwithstanding that the AC from
B. bassiana strains possessed varied hydrophobicities but
very similar lectin profiles, overall there still appeared to be
a clear correlation between SAS values and total lectin
scores (Fig. 2). It would appear that the more hydrophilic
spores generally have less affinity for lectins, indicating
fewer kinds and amounts of surface carbohydrates. Total
lectin binding did not appear to be related to spore surface
area, indicating that the criteria used to measure lectin bind-
ing (i.e. intensity of fluorescence at cell wall) was not bi-
ased.
Depending on culture conditions, B. bassiana is known to
produce three distinct spore types: AC on solid media, BS in
nutrient rich liquid media, and SC in nutrient-deficient liq-
uid media (see Thomas et al. 1987). Aerial conidia were
slightly smaller than SC (3.5 ± 0.1 µm), with BS (8.8 ± 0.3
µm) being almost three times the length of AC (3.0 ±
0.1 µm; Table 3). Of the different spore types, AC were
clearly the most hydrophobic, followed by B-SC, with T-SC
being slightly less hydrophobic counterparts, whereas BS
were very hydrophilic (Table 3). Using the SAS assay, this is
the first demonstration that AC and SC have different sur-
face hydrophobicities since two previous studies that relied
on the phase exclusion assay both reported that these two
spore types possessed similar hydrophobicities (Hegedus et
al. 1992, Bidochka et al. 1995). Comparison of lectin bind-
ing affinities for spores types showed that there were some
differences in lectin binding between SC and AC, with SC
having weaker PHA-P and SBA affinities (Table 3).
Blastospores showed no binding of WPA or WGA, while ex-
hibiting weak binding to PHA-P and SBA. Again, spore sur-
face hydrophobicity appeared to be positively correlated to
the overall lectin binding affinity with WPA, SBA, and
PHA-P binding affinities varying most dramatically.
Another approach to studying spore attachment properties,
was to evaluate their relative affinities to three kinds poly-
styrene latex beads with the following surfaces: hydropho-
bic, unmodified; electronegative, carboxylate modified; and
electropositive, amine-modified. The surface hydrophobicities
of the beads were initially estimated using the SAS assay
and, not surprisingly, both the amine- and carboxylate-
© 1999 NRC Canada
942 Can. J. Microbiol. Vol. 45, 1999
Fig. 2. Results from Pearson’s Correlation analysis of spore
properties. Significantly correlated paired variables are marked
with asterisks; * P< 0.05, ** P< 0.01. Abbreviations: length
(Len), total lectin binding value (Tlec), SAS hydrophobicity
value (SAS), width (Wid). Please refer to Materials and methods
for definitions of lectin abbreviations.
Hydrophobicity Lectin bindingc
Spore type Mean lengtha(µm) SAS value (%)bConA PHA-P RCA SBA WGA WPA
AC 3.0 ± 0.1 48.8 ± 0.3 ++ ++ ++ ++ ++ +
SC 3.5 ± 0.1 85.1 ± 2.3 ++ + ++ + ++ +
BS 8.8 ± 0.3 103.7 ± 1.5 ++ ± ++ ± – –
aSpore lengths given as means ± S.E., n= 100, except for BS where n= 50.
bValues givens as means of quadruplicate assays ± S.E.
cFluorescent lectin binding was scored as follows: ++ strong to moderate; + moderate to weak; ± weak; – none observed.
Specific lectin binding similar to untreated aerial conidia highlighted in bold.
Table 3. Hydrophobicities and lectin profiles for different spore types of B. bassiana GK2016.
modified beads were very hydrophilic (Table 4). The 0.8 µm
unmodified beads gave a very reproducible SAS value of
73.0 ± 0.6%, indicating that they were much more hydro-
phobic than the charged beads. Naturally the relative
hydrophobicity of the unmodified polystyrene beads com-
pared to those of spores using the SAS assay since the beads
are different sizes and likely have densities quite different
from spores. Nonetheless, these latex beads could potentially
be used as standards for SAS assays since they are easily
available, more homogeneous, and less prone to surface
changes than spores that can be altered by numerous afore-
mentioned factors during production, preparation, and stor-
age.
The interactions between spores and beads are summa-
rized in Table 4. The unmodified beads bound to both hy-
drophilic and hydrophobic spores, with the more hydrophilic
spores appearing to have a greater affinity for these beads.
This finding indicates that even so-called hydrophilic spores
possess some hydrophobic sites. The amine-modified beads
again bound to more hydrophilic spores with greater affinity
than to more hydrophobic spores, indicating that there are
more electronegative groups on the surfaces of hydrophilic
spores. Amine beads also promoted varying degrees of spore
aggregation, since they can allow spores to bind to free
electronegative groups on other spores. This phenomenon is
similar to the results seen with poly-lysine treated Ba16 AC
spores, which were found to undergo spontaneous aggrega-
tion (Table 5). The carboxylate spores did not attach to any
of the spores tested, indicating the absence of electropositive
surface charges. Not surprisingly, poly-lysine treated Ba16
AC spores were able to attach to carboxylate beads to a
small extent.
Effects of chemical treatments on surface properties
The enzyme laminarase, a β1-3 glucanase, was used in an
attempt to remove glucan-linked surface proteins from the
hydrophobic Ba16 AC. Laminarase-treated spores became
very hydrophilic (97.1 ± 0.3%) within 5 min (Table 5). To
determine whether the dramatic decrease in surface
hydrophobicity exhibited by the laminarase-treated spores
was a specific result of the laminarase or a general phenom-
enon of protein coating, spores were also incubated in
2 mg/mL bovine serum albumin (BSA). These BSA-treated
spores also became more hydrophilic but not to the same ex-
tent as laminarase-treated spores; SAS value 81.3 ± 1.3%.
The laminarase treatment also altered spore lectin binding,
reducing the binding of PHA-P, SBA, and WPA. A similar
reduction in PHA-P binding was seen for BSA-treated
spores though SBA binding was not reduced to the same ex-
tent, and WPA binding was unaffected. It would appear that
laminarase bound to spore surface components, but that it
neither removed surface protein nor carbohydrate since the
protein and carbohydrate levels in the supernatant were un-
changed duringa4hincubation (data not shown).
An attempt to change the surface charge of Ba16 AC was
made by treatment with the poly cation poly-lysine, which is
known to promote cell adhesion to substrates. Attempts to
estimate the hydrophobicity of spores treated with 2 mg/mL
poly-lysine were unsuccessful since there was significant
control sedimentation in the absence of salt (SAS value 66%
compared to control lacking poly-lysine 96%), indicating
that poly-lysine had bound to spores and allowed the spore
to attach to the unmodified anionic sites of other spores.
Poly-lysine treated spores showed no differences in lectin
binding as compared to control spores (Table 5).
The effects of different boiling treatments on AC surface
hydrophobicities and lectin binding were next studied. Ae-
rial conidia were shown to be slightly altered after boiling in
1% SDS extraction buffer for 10 min (Table 5). After this
treatment, spores became slightly more hydrophilic (SAS
value 59.1%) and the binding intensities for the lectins
PHA-P and WPA were significantly weakened. When AC
were boiled in buffer containing the deglycosylating agent
dithiothreitol (DTT) for 10 min, spores again became less
hydrophobic and exhibited reduced affinity for PHA-P. The
control treatment, involving boiling spores in dH2O had the
greatest effect, resulting in spores with a SAS value of
83.4% and reduced affinity for the lectins PHA-P, RCA, and
WPA. It would seem that the buffers used in the 1% SDS
and DTT treatments may actually protect the spore surface
from the harsh effects of boiling. When spores were incu-
bated with 97% FAin a sonicating water bath for 2 h, spores
became much less hydrophobic (SAS value 81%, Table 5).
Moreover, lectin binding was absent for PHA-P and dimin-
ished for SBA, WGA, and WPA. Spores extracted first with
SDS followed by FAwere more hydrophobic than the spores
treated with FA alone (SAS value 71%), though all lectin
binding was further reduced with PHA-P, SBA, and WPA,
showing no binding. It would appear that pre-treating spores
by boiling in 1% SDS protects the spore surface, preventing
© 1999 NRC Canada
Jeffs et al. 943
Latex bead Spore strain SAS valuea(%) and degree of bead attachmentb
Modification SAS valuea(%) Ba16a
50.6 ± 1.2 Ma97
65.8 ± 0.6 Ba16 AC
85.1 ± 2.3 Tn81
101.7 ± 1.6 Ba16 BS
103.7 ± 1.5
Unmodified
0.80 µm, blue 73.0 ± 0.6 + + +++ +++ +++
Carboxylate
0.43 µm, yellow-green fluorescent 100.6 ± 0.1 – – – – –
Amine
0.49 µm, red fluorescent 102.0 ± 0.8 + + +++ +++ ++
aValues given as means ± S.E.; (n = 4).
bBead attachment to spores was scored as follows: +++ abundant large spore-bead aggregates (25–100 cells); ++ good bead attachment with some
aggregates; + moderate bead attachment with few aggregates (<10 cells); – no evidence of bead attachment.
Table 4. Attachment of various polystyrene latex beads to fungal spores.
the subsequent FA treatment from exerting its full effect.
For all chemical treatments, spores maintained their shapes
and exhibited no lysis and all spores were refiltered to
remove any large spore clumps, and washed before hydro-
phobicity and lectin binding were assessed. Scanning elec-
tron microscopy indicated that SDS boiled spores retained
the rough warty surface features of control spores though the
surfaces of FA-treated spores become significantly smoother,
as reported previously by Bidochka et al. (1995), suggesting
that some of the hydrophobic proteins making up the spore
coat had been solubilized.
Spore surface proteins
The protein work was only pursued for those fungal
strains that produced large numbers of spores. After trying
Tris glycine gels of differing acrylamide/bis percentages to
separate protein in spore extracts, 16% gels were chosen
since these gels adequately resolved the low MW proteins of
special interest as well as allowing for the identification of
higher MW proteins (Fig. 3). From this figure, prominent
low MW proteins bands were evident for the hydrophobic
spores of Ba16 AC, Ba60, Br156, and Br656, and clearly ab-
sent for the hydrophilic Ba16 BS, Bd04, and Tn81. None of
these proteins bands appeared to be glycosylated; when the
same protein extracts were separated on a similar gel and
stained with periodic acid – Schiff’s stain, none of the
hydrophobin-like protein bands were stained, while the
horseradish peroxidase control did show staining (data not
shown).
Aerial conidia from four of the Beauveria strains found to
possess these low MW FA-soluble protein bands (i.e., Ba16,
Ba860, Br156, Br656) were extracted with performic acid,
© 1999 NRC Canada
944 Can. J. Microbiol. Vol. 45, 1999
Hydrophobicity Lectin bindingb
Treatment SAS value (%)aConA PHA-P RCA SBA WGA WPA
Control 48.8 ± 0.3 ++ ++ ++ ++ ++ +
Laminarase 97.1 ± 0.3 ++ ± ++ ± ++ ±
Poly-lysine no value ++ ++ ++ ++ ++ +
BSA 81.3 ± 1.3 ++ ± ++ + ++ +
Control boiling 83.4 ± 2.5 ++ + + ++ ++ ±
DTT boiling 75.1 ± 2.6 ++ + ++ ++ ++ +
1% SDS boiling 59.1 ± 4.1 ++ ± ++ ++ ++ ±
97% formic acid 80.9 ± 0.6 ++ – ++ + + ±
SDS + formic acid 71.0 ± 1.4 + – + – ± –
aValues givens as means of quadruplicate assays ± S.E.
bFluorescent lectin binding was scored as follows: ++ strong to moderate; + moderate to weak; ± weak; – none observed.
Table 5. Effects of chemical treatments on hydrophobicity and and lectin binding of Ba16 AC.
Fig. 3. Separation of FA-soluble spore proteins using SDS–PAGE with a 16% Tris glycine gel. In each lane, 2.6 µg protein was loaded
and the gel was silver-stained to resolve protein bands. Various protein samples were loaded in the following lanes: 1) Ba16 BS; 2)
Ba16 AC; 3) Ba860; 4) Bd04; 5) Br156; 6) Br656; 7) Tn81; 8) molecular weight standards.
instead of FA, following two 1% SDS extractions in an at-
tempt to determine whether these small peptides possessed
disulfide linkages as do hydrophobins (Wessels 1997). For
each of the performic acid samples, two clear low MW
bands were seen, instead of a single diffuse band for the FA
extracts, with a generally more prominent band at 10–
11 kDa and the other band close to 6.5 kDa. The corre-
sponding FA extracts all possessed more diffuse single
bands between 6.5 and 11 kDa. These results indicate that
the FA-soluble peptides from hydrophobic Beauveria AC all
contain disulfide bonds, since the hydrolysis of these bonds
linearizes the peptide, giving it a higher apparent MW when
separated using SDS–PAGE. These results also corroborate
earlier work by Bidochka et al. (1995), which initially re-
ported the presence of a hydrophobin-like peptide in the
conidial wall of another B. bassiana strain.
The sequential spore extraction protocol was successful in
aiding the resolution of spore hydrophobin-like proteins, but
also yielded data relating to the amounts of protein and car-
bohydrate extracted after each step. For SDS treatments the
majority of the protein (62–82%, dependent upon strain) was
extracted from the first SDS treatment. Again, the first FA
treatment typically yielded more protein than the subsequent
treatment (64–76% of total), with the notable exceptions of
Bd04 and Tn81, for which similar but low levels of FA-
soluble protein were extracted for both treatments. It was of
little surprise to find that the four hydrophilic spore strains
possessed by far the lowest levels of FA-extractable protein
per spore dry weight. Carbohydrate was detected in each of
the initial SDS extracts for the hydrophilic spore strains and
the moderately hydrophobic Br156, while no carbohydrate
was seen in the second SDS extracts. For the FA extracts,
very low levels of carbohydrate were detected or the some
of the initial spore extracts, notably Br156 and Ma97,
though increased levels of carbohydrate were seen in the
second FA extracts. This may indicate that removal of FA-
soluble protein exposes glucan in the underlying wall matrix
to attack from FA. Presumably, the hydrophilic spore strains
possess less of the resilient FA-soluble protein and are there-
fore more prone to have carbohydrate extracted along with
the protein.
The total protein and carbohydrate extracted from all of
the treatments, for each strain, is shown in Fig. 4. The more
hydrophobic spores tended to have greater protein:carbohy-
drate ratios than hydrophilic spores. Of the strains analyzed,
Ba16 AC were the most hydrophobic, and possessed the
greatest protein:carbohydrate ratio, while the very hydro-
philic spores (Ba16 BS, Tn81, and Bd04) had very low
spore ratios. Again, Br156 AC did not fit the general trend
very well since they possessed more total extractable carbo-
hydrate than any of the eight spore strains studied. It may be
that although Br156 possesses hydrophobin-like spore coat
proteins, these proteins may not be as tightly packed on the
spore surface, thus facilitating their removal and the extrac-
tion of the underlying carbohydrate by both SDS and FA
treatments.
Discussion
In this study, the surface morphology and properties of
spore strains and types have been found to be quite diverse,
possessing varied relative hydrophobicity, and discrete sur-
face carbohydrates and morphologies. This information may
contribute to understanding how spores from certain
entomopathogenic fungi (EPF) are specialized for their host
insects and habitat. The finding that AC from various
B. bassiana strains had varied hydrophobicities (SAS values
47%–87%) may reflect the adaptation of their surface prop-
erties to suit the conditions of their particular habitats and
target insects. Additionally, spore surface hydrophobicity
was positively correlated with both spore surface roughness
and increased lectin binding by surface carbohydrates. The
former finding supports the proposed role of hydrophobic
spore rodlet layers in conveying surface hydrophobicity.
However, rodlet layers are not necessarily required to make
spores hydrophobic, since conidia from the plant-pathogenic
fungus B. cinerea have been shown to possess hydrophobic
properties, as measured by a phase exclusion assay and their
adherence to polystyrene, but lack a rodlet layer (Doss et al.
1993, 1997). Using a field-emission scanning electron mi-
croscopy (SEM), Doss et al. (1997) found that unhydrated
B. cinerea conidia have rough surfaces possessing numerous
short (200–250 nm) protuberances, which were lost upon
spore hydration and subsequent drying. In our study, a rela-
tively harsh fixation procedure was used to prepare spores
for viewing of their surface features by conventional SEM.
Alhough this procedure was suitable for detecting rodlet lay-
ers, the possibility exists that other more delicate spore sur-
face features may have been lost that could have been
viewed using sample cryofixation in combination with low-
temperature SEM (see Read and Jeffree 1991).
Using polysterene latex beads, it was found that all spores
possessed surface electronegative charges and apparently
lacked significant electropositive groups. An unexpected
© 1999 NRC Canada
Jeffs et al. 945
Fig. 4. Total amount of spore protein and carbohydrate extracted
after sequential treatments. Each spore type was subjected to two
SDS extracts followed by two treatments with FA. Ba16 AC and
Ba16 BS extracts designated Ba16a and Ba16b, respectively.
Values are averages of duplicate assays.
finding was that even very hydrophilic spores possessed hy-
drophobic surface groups that formed extensive aggregates
with unmodified polystyrene beads. It may be that the SAS
assay is unable to promote the aggregation of these hydro-
philic spores because their hydrophobic sites are masked by
adjacent electronegative groups. However, the much smaller
hydrophobic beads, which have volumes 60 and 320 times
smaller than 2016 AC and BS, can bind to these hydropho-
bic sites, thus promoting the formation of large bead-spore
complexes. The reason there is a greater degree of spore-
bead interaction among more hydrophilic spores could be
that hydrophobic beads in an aqueous environment contain-
ing high densities of electronegative spore-based charges are
strongly attracted to the relatively sparse hydrophobic sites
on these same spores.
This is the first report of the fact that B. bassiana sub-
merged conidia (SC) possess surface hydrophobicities
greater than blastospores (BS), but less than aerial conidia
(AC), which suggests that the hydrophobic surface proteins
of SC are arranged differently than those of AC. Submerged
conidia were also shown to possess somewhat similar sur-
face morphologies and dimensions to AC, though they ex-
hibited reduced affinities for the lectins PHA-P and SBA
which preferentially bind oligosaccharides and N-acetyl-D-
galactosamine. The differences between AC and SC surface
properties suggest that physical constraints (e.g. the lack of a
media-air interface) prevent the proper assembly of the SC
outer coat. The intermediate surface hydrophobicity of sub-
merged conidia of B. bassiana SC is likely a compromise
adaptation where a vegetative aqueous culture experiences
stress from reduced water and/or nutrient availability, and
produces conidia possessing a more resilient hydrophobin-
containing outer wall, which will allow a spore to remain vi-
able for extended periods of time in a drier, less hospitable
environment. These findings may be of significance to the
large-scale production of B. bassiana conidia using sub-
merged culture techniques because the distinct surface prop-
erties of the SC product differ from AC, which may have
unforeseen effects upon the stability and efficacy of the re-
sultant spore-based formulation when it is applied against
pest insect populations.
Also of interest was the finding that Bd04 AC possessed
less FA-extractable protein than AC from any other
Beauveria strain, and that its smooth surface did not pos-
sess the characteristic low molecular weight hydrophobic
peptides seen for other hydrophobic AC. These Bd04 AC
more closely resembled AC from T. nivea in both surface
morphology, hydrophobicity, and low concentrations of
SDS-insoluble protein, though Bd04 AC exhibited addi-
tional lectin binding to RCA, WGA, and WPA. Using
MtDNA probes, it has been shown that B. bassiana strains
are quite closely related to B. caledonica and Bd04, while
Tn81 and a B. brongniartii strain are more distantly related
(Hegedus and Khachatourians 1993). It was concluded
from these findings that Bd04 may be considered a subspe-
cies of B. bassiana. A question that emerges from these
combined results is why do Bd04 AC lack hydrophobic
surface proteins unlike all other Beauveria strains studied
to date? Either the Bd04 strain has lost the ability to pro-
duce these hydrophobic proteins, or it requires different en-
vironmental cues to initiate their production. In either case,
further study of this strain may give new insights into the
control of the expression of Beauveria surface hydrophobic
proteins as have the rodletless mutants of S. commune and
Aspergillus nidulans (Wessels et al. 1991; Stringer et al.
1991).
The finding that spore surface hydrophobicity was posi-
tively correlated to the presence of surface carbohydrates
suggests that the hydrophobic proteins making up these
spore coats may be glycosylated. However, the predominant
FA-soluble proteins identified in AC for various hydropho-
bic Beauveria strains did not appear to be glycosylated using
periodic acid – Schiff’s staining. These proteins were shown
to have disulfide linkages as demonstrated by the performic
acid treatments. The possibility exists that these low molecu-
lar weight hydrophobin-like peptides may be glycosylated in
their native form, but the harsh extraction treatments may
have removed these carbohydrate groups. A hydrophobic
protein, SC3, isolated from aerial hyphae of S. commune,
has been shown to be a glycoprotein possessing an O-linked
chain of 16–22 mannose residues attached to its N-terminus
(de Vocht et al. 1998).
The demonstration that Ba16 AC surface hydrophobicity
and lectin binding are significantly reduced after FA treat-
ment is strong evidence to support the theory that FA-
soluble proteins are glycosylated and necessary for convey-
ing hydrophobic properties. The superiority of the SAS
assay over the phase exclusion assay was again demon-
strated, because a previous study reported no difference in
B. bassiana AC hydrophobicity after either SDS or FA
treatments (Bidochka et al. 1995). The incubation of Ba16
AC with laminarase also had the effect of rendering spores
very hydrophilic, though it was found that this glucanase
did not release surface carbohydrate, and that incubation
with BSA could partially reproduce these effects. Presum-
ably, this enzyme interacted with carbohydrate polymers on
the spore surface, reducing carbohydrate-mediated binding
of the lectins PHA-P, SBA, and WPA while masking sur-
face proteins conveying hydrophobicity. It would also
appear that electronegative moieties on Ba16 AC surfaces
are not closely associated with sites possessing lectin-
binding surface carbohydrate since treatment with the poly-
cation poly-lysine did not diminish binding for any of the
lectins.
The predominance of hydrophobin-like proteins of the
rodlet layer of dry hydrophobic spores is well established,
but do these proteins convey additional properties involved
in the specific attachment of spores to target insects, and are
other wall components (e.g., glycoproteins and glycolipids)
involved? The structures of hydrophobins have been com-
pared to other proteins rich in cysteine and possessing
disulfide bridges, such as agglutinins and chitin-binding
lectins (see Wessels 1997). The apparent commonalties be-
tween hydrophobins and these other proteins may hold the
key to answering both this question and how hydrophobins
interact with other components in the cell wall. With this
more complete understanding of spore surface properties, a
more comprehensive picture of EPF spore interactions with
their hosts will emerge and lead to the development of more
efficacious mycoinsecticides.
© 1999 NRC Canada
946 Can. J. Microbiol. Vol. 45, 1999
Acknowledgements
We would like to thank Mr. Y. Yano for his technical as-
sistance with the scanning electron microscopy and Dr. L.
Xavier for her statistical help. We are also grateful to
Agrium Inc. for financial support provided in the form a
graduate research grant.
References
Bidochka, M.J., St. Leger, R.J., Joshi, L., and Roberts, D.W. 1995.
The rodlet layer from aerial and submerged conidia of the
entomopathogenic fungus Beauveria bassiana contains hydro-
phobin. Mycol. Res. 99: 403–406,
Boucias, D.G., Pendland, J.C., and Latge, J.P. 1988. Nonspecific
factors involved in attachment of entomopathogenic Deutero-
mycetes to host insect cuticle. Appl. Environ. Microbiol. 54:
1795–1805.
Boucias, D.G., and Pendland, J.C. 1991. Attachment of myco-
pathogens to cuticle: the initial event of mycoses in arthropod
hosts. In The Fungal Spore and Disease Initiation in Plants and
Animals. Edited by G.T. Cole and H.C. Hoch. Plenum, New
York. pp. 101–127.
Bradford, M. 1976. A rapid and sensitive method for the
quantitation of microgram quantities of protein utilising the
principle of protein-dye binding. Anal. Biochem. 72: 248–254.
Butt, T.M., Ibrahim, L., Clark, S.J., and Beckett, A. 1995. The ger-
mination behavior of Metarhizium anisopliae on the surface of
aphid and flea beetle cuticles. Mycol. Res. 99: 945–950.
Carruthers, R.I., and Hural, K. 1990. Fungi as naturally occuring
entomopathogens. In New Directions in Biological Control: Al-
ternatives for Suppressing Agricultural Pests and Diseases.
Edited by R.R. Baker and P.E. Dunn. Alan R. Liss, Inc., New
York. pp. 117–138.
de Vocht, M.L., Scholtmeijer, K., van der Vegte, E.W., de Vries,
O.M.H., Sonveaux, N., Wösten, H.A.B., Ruysschaert, J.-M.,
Hadziioannou, G., Wessels, J.G.H., and Robillard, G.T. 1998.
Structural characterization of the hydrophobin SC3, as a mono-
mer and after self assembly at hydrophobic/hydrophilic inter-
faces. Biophys. J. 74: 2059–2068.
Doss, R.P., Potter, S.W., Chastagner, G.A. and Christian, J.K.
1993. Adhesion of nongerminated Botrytis cinerea conidia to
several substrata. Appl. Environ. Microbiol. 59: 1786–1791.
Doss, R.P., Potter, S.W., Christian, J.K., Soeldner, A.H., and
Chastagner, G.A. 1997. The conidial surface of Botrytis cinerea
and several other Botrytis species. Can. J. Bot. 75: 612–617.
Feng, M.G., Poprawski, T.J., and Khachatourians, G.G. 1994. Pro-
duction, formulation and application of the entomopathogenic
fungus Beauveria bassiana for insect control: Current status.
Biocontrol Sci. Technol. 4: 3–34.
Hazen, K.C., and Hazen, B.W. 1987. A polystyrene microsphere
assay for setecting surface hydrophobicity variations within
Candida albicans populations. J. Microbiol. Methods, 6: 289–
299.
Hazen, K.C., Lay, J.-G., Hazen, B.W., Fu, R.C., and Murthy, S.
1990. Partial biochemical characterization of cell surface hydro-
phobicity and hydrophilicity of Candida albicans. Infect.
Immun. 58: 3469–3476.
Hegedus, D.D., and Khachatourians, G.G. 1993. Identification of
molecular variants in mitochondrial DNAs of members of the
genera Beauveria,Verticillium,Paecilomyces,Tolypocladium,
and Metarhizium. Appl. Environ. Microbiol. 59: 4283–4288.
Hegedus, D.D., Bidochka, M.J., and Khachatourians, G.G. 1992. A
comparison of the virulence, stability and cell-wall-surface char-
acteristics of three spore types produced by the entomopatho-
genic fungus Beauveria bassiana. Appl. Microbiol. Biotechnol.
36: 785–789.
Herbert, D., Phipps, P.J., and Strange, R.E. 1971. Chemical analy-
sis of microbial cells. Methods Microbiol. 5B: 209–338.
Hess, W.M., and Stocks, D.L. 1969. Surface characteristics of
Aspergillus conidia. Mycologia, 61: 560–571.
Hess, W.M., Sassen, M.M.A., and Remsen, C.C. 1968. Surface
characteristics of Penicillium conidia. Mycologia 60: 290–303.
Khachatourians, G.G. 1996. Biochemistry and molecular biology
of entomopathogenic fungi. In The Mycota, Vol VI: Human and
Animal Relationships. Edited by K. Esser and P.A. Lemke.
Springer Verlag, Berlin. pp. 331–363.
Jeffs, L.B. and Khachatourians, G.G. 1997. Estimation of spore
hydrophobicity for members of the genera Beauveria,
Metarhizium and Tolypocladium by salt-mediated aggregation
and sedimentation. Can. J. Microbiol. 43: 23–28.
Jeffs, L.B., Feng, M.G., Falkowsky, J.E. and Khachatourians, G.G.
1997. Infection of the Migratory Grasshopper (Orthoptera:
Acrididae) by ingestion of the entomopathogenic fungus
Beauveria bassiana. J. Econ. Entomol. 90: 383–390.
Laemmli, U.K. 1970. Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature, 277: 680–
685.
Merrill, C.R., Goldman, D., Sedman, S.A., and Ebert, M.H. 1981.
Ultrasensitive stain for proteins in polyacrylamide gels shows
regional variation in cerebrospinal fluid proteins. Science, 211:
1437–1438.
Montijin, R.C., van Rinsum, J., van Schagen, F.A., and Klis, F.M.
1994. Glucomannoproteins in the cell wall of Saccharomyces
cerevisiae contain a novel type of carbohydrate side chain. J.
Biol. Chem. 269: 19 338 – 19 342.
Pendland, J.C., and Boucias, D.G. 1984. Use of labeled lectins to
investigate cell wall surfaces of the entomogenous hyphomycete
Nomuraea rileyi. Mycopathologia, 87: 141–148.
Rath, A.C., Guy, P.L. and Webb, W.R. 1995. Metarhizium spore
surface antigens are correlated with pathogenicity. Mycol. Res.
100: 57–62.
Read, N.D., and Jeffree, C.E. 1991. Low-temperature scanning
electron microscopy in biology. J. Microsc. 161: 59–72.
Samson, R.A., Evans, H.C., and Latge, J.P. 1988. Atlas of Entomo-
pathogenic Fungi. Springer-Verlag, Berlin.
Schuren, F.H.J., and Wessels, J.G.H. 1990. Two genes specifically
expressed in fruiting dikaryons of Schizophyllum commune:
homologies with a gene not regulated by mating-type genes.
Gene, 90: 199–205.
Sitch, J.C., and Jackson, C.W. 1997. Pre-penetration events affect-
ing host specificity of Verticillium lecanii. Mycol. Res. 101:
535–541.
St. Leger, R.J., Staples, R.C., and Roberts, D.W. 1992. Cloning and
regulatory analysis of starvation-stress gene, ssgA, encoding a
hydrophobin-like protein from the entomopathogenic fungus,
Metarhizium anisopliae. Gene, 120: 119–124.
Stringer, M.A., Dean, R.A., Sewall, T.C., and Timberlake, W.E.
1991. Rodletless, a new Aspergillus developmental mutant in-
duced by directed gene inactivation. Genes and Devel. 5: 1161–
1171.
Thau, N., Monod, M., Crestani, B., Rolland, C., Tronchin, G.,
Latgé, J.P., and Paris, S. 1994. Rodletless mutants of Aspergillus
fumigatus. Inf. Immun. 62: 4380–4388.
© 1999 NRC Canada
Jeffs et al. 947
© 1999 NRC Canada
948 Can. J. Microbiol. Vol. 45, 1999
Thomas, K.C., Khachatourians, G.G., and Ingledew, W.M. 1987.
Production and properties of Beauveria bassiana conidia culti-
vated in submerged culture. Can. J. Microbiol. 33: 12–20.
Wan, L., and van Huystee, R.B. 1993. Rapid determination of
glycoproteins and glycopeptides by periodic acid Schiff reagent
dot-blotting assay on nitrocellulose membrane. J. Agric. Food
Chem. 41: 896–898.
Wessels, J.G.H. 1997. Hydrophobins: proteins that change the na-
ture of the fungal surface. Adv. Microbial Physiol. 38: 1–45.
Wessels, J.G.H., de Vries, O.M.H., Ásgeirsdóttir, S.A., and
Springer, J. 1991. The thn mutation of Schizophyllum commune,
which suppresses formation of aerial hyphae, affects expression
of the SC3 hydrophobin gene. J. Gen. Microbiol. 137: 2439–
2445.
Wösten, H.A.B., de Vries, O.M.H., and Wessels, J.G.H. 1993. In-
terfacial self-assembly of a fungal hydrophobin into a hydropho-
bic rodlet layer. Plant Cell, 5: 1567–1574.
