Gene. 120 (1992) 119-124 0 1992 Elsevier Science Publishers
GENE
B.V. All rights reserved.
119
0378-l 119/92/$05.00
0669 1
Cloning and regulatory analysis of starvation-stress gene, ssgA, encoding a hydrophobin-like protein from the entomopathogenic fungus, Met arhizium
anisopliae (Insect
pathology;
Raymond
appressorium
formation;
glycosylation
site; nutrient
deprivation;
gene expression)
J. St. Leger, Richard C. Staples and Donald W. Roberts
Boyce Thompson Institute for Plant Research. Inc., Ithaca, NY 14853-1801,
USA
Received
May 1992; Received
by J. Kinghorn:
20 March
1992; Revised/Accepted:
20 May/24
at publishers:
22 June 1992
SUMMARY
The nucleotide (nt) sequence of a starvation-stress gene (ssgA) of the entomopathogenic fungus, Metarhizium anisopliae, and its deduced amino acid (aa) sequence were determined. The primary structure of the SSGA (96 aa; deduced M, = 9925; p1 = 4.1) protein shares extensive similarities with fungal wall proteins of the ‘hydrophobin’ class, and the eight Cys residues and putative signal sequences are conserved. Secondary structure predictions suggest an additional resemblance to low-M, toxins and agglutinins. Northern (RNA) blot analysis and nuclear run-on assays demonstrated transcriptional control of expression of ssgA during nutrient deprivation and during formation of infection structures. Hybridizations of M. anisopliae genomic DNA indicate that there is only one form of ssgA in the genome.
INTRODUCTION
A number of low-molecular-weight, hydrophobic, extracellular polypeptides called hydrophobins have been identified recently in the basidiomycete Schizophyllum commune and the ascomycete Aspergillus nidulans. Each have eight cysteines conserved at identical positions (Schuren and Wessels, 1990; Stringer et al., 1991). Schuren and Wessels (1990) described one S. commune hydrophobin-encoding gene, designated Sc3, that is expressed abundantly during
Correspondence to: Dr. R.J. St. Leger, Boyce Thompson
Institute for Plant
Research,
USA.
Inc., Tower Road,
Tel. (607) 254-1252; Abbreviations:
Ithaca,
NY 14853-1801,
Fax (607) 254-1242.
aa, amino acid(s); bp, base pair(s); kb, kilobase
bp; M., Metarhizium; MM, minimal media (0.1 ?0 KH,PO,/O.OS
or 1000 “i, MgSO,
pH 6); nt, nucleotide(s); Prl, M. anisopliae protease; prl, gene (DNA, mRNA) encoding Prl; SDB, Sabouraud dextrose broth; SDS, sodium dodecyl
sulfate;
SSC, 0.15 M NaCl/0.015
starvation-stress gene; SSGA, NaCI/O.Ol M NaH2P0,/0.00125
M Na,.citrate
pH 7.6; ssgA,
protein product of ssgA; SSPE, M EDTA pH 7.4.
0.15 M
the formation of aerial hyphae, whereas two related genes (Scl and Sc4) are regulated by mating-type genes and are expressed during the formation of fruit bodies. Apparently, the gene-encoded polypeptides are secreted into the growth medium by submerged hyphae and accumulate in the cell walls of emerged hyphae (Wessels et al., 1991). Stringer et al. (1991) described a hydrophobin gene in A. nidulans which is expressed in conidiophore cells, the protein product of which accumulates in the rodlet layer of the conidium wall. It has been proposed that by forming a hydrophobic coating on their surfaces, hydrophobins are important in the formation and function of aerial structures such as conidiophores and mushrooms. Here we describe a new member of this hydrophobin gene family produced by the commercially important entomopathogenic deuteromycete, Metarhizium anisopliae. The gene, designated ssgA, was obtained from a cDNA library representing a nutrient-deprived culture of M. anisopliae. Nutrient deprivation is one of the interactive phenomena which correlate formation of infection structures (appressoria) and production of insect-cuticle-
-- 324
817
ME1
2951
23
>HH>WH,HH>#H>WH wIulwxwwIwwIwwIw
Fig. 2.
Fig. 1. Sequence
of ssgA cDNA
and the deduced
[35S]dATP
and a double-stranded
constructed
in the pBluescript
glycosylation
site (Se?‘-Th?“)
template
aa sequence.
procedure
vector (Stratagene), are underlined.
The nt sequencing
provided
in the Sequenasc
was performed
and the entire inserts were sequenced
The stop codon
is marked
by the dideoxy
kit (U.S. Biochemicals, on both strands.
by three dots. GenBank
chain-termination
Cleveland,
The eight Cys residues
accession
number
procedure
OH). The ssgA subclones
using wcrc
and a possible O-linked
is M8528 1.
Fig. 2. Southern blot analysis of restricted chromosomal DNA from five isolates of M. anisopliae. Genomic DNAs (5 pg) prepared from m)cclia of five isolates (MEl, 23, 2951, 817, and 324) (St. Leger et al., 1992) were digested to completion with EcoRI (EI), EcoRV (EV), or Hind111 (HIII) and scparated by electrophoresis random
in 0.1 x SSC/O.l% plasmid
on a 1.Oob agarose
primed ssgA cDNA
fragment
gel. The fragments
under high stringency
SDS at 68”C, and subjected
using [3ZP]CTP
as described
were transferred conditions
to autoradiography
by Feinberg
and Vogelstein
to Gene Screen PlusTM membranes
(507; formamidc/h
x SSPE/ 1 x Denhardt’s
(4 h). The probe was prepared
(NEN) and probed with a “P-labeled solution/ 1“,, SDS at 42” C), washed
by oligo labeling of an EcoRI fragment
from the ssg,4
(1984).
degrading enzymes (St. Leger et al., 1989a). Nuclear run-on experiments confirmed that the gene is expressed synchronously with the gent encoding the cuticle-degrading protease. Pr 1. the only pathogenicity determinant so far described from an entomopathogenic fungus (St. Leger et al.,
1989b; 1991; 1992). One aim of our research is the molecular analysis of infection processes in entomopathogenic fungi. In this connection, a gene with a simple genomic organization that responds rapidly and strongly during the formation of infection structures is of great interest.
Number of Amino Acids 23 26 z: 46
ZB 80 50 96
1:: 117 89 146 100 111 125 159:
Fig. 3. Alignment
of the deduced
and Sc4 (S. commune) (Schuren (Lipman and Pearson, 1985).
SSGA aa sequence and Wessels,
(see Fig. 1) with other hydrophobin
aa sequences,
shown in a single letter code. Proteins:
1990) and RodA (A. nidulans) (Stringer et al.. 1991). The aa sequences
were aligned using the FASTP
SC 1, Sc3, program
121 EXPERIMENTAL
p1 = 4.1). The protein contains several hydrophobic quences including a highly hydrophobic N-terminal quence. The protein contains eight Cys residues and a tential O-glycosylation site ( Sers3-Thrs4) but lacks sites N-glycosylation.
AND DISCUSSION
(a) Sequence of ssgA and its putative protein Clones were isolated from the cDNA library (cloned into the EcoRI site of the vector igtl0) representing poly(A)‘RNA from mycelia of M. anisopliae deprived of nutrients for 3 h (St. Leger et al., 1992). We screened a total of 15000 plaques at a high stringency (50% formamide/ 6 x SSPE/l x Denhardt’s solution/0.3% SDS; washes ranged from 2 x SSC to 0.1 x SSC) and identified 45 clones differentially expressed during nutrient deprivation (St. Leger et al., 1992). The clones were rescued as Bluescript plasmids and segregated into four nonoverlapping clone classes by cross-hybridization. The 26 clones in class 1 were identified as prl (St. Leger et al., 1992). The 11 cDNA clones in class 2 ranged from 580 to 630 nt in length. Sequence analysis (Fig. 1) on three of the longest clones revealed identical putative coding regions. The ORF contained between the first ATG of the longest cDNA (ssgA, Fig. 1) and the first termination codon (TAA) is 288 bp long and encodes a 96-aa protein (deduced M,. = 9925,
sesepofor
(b) Genomic organization of the ssgA gene To determine whether ssgA is a member of a gene family, we performed DNA blot hybridization with genomic DNA prepared from five strains of M. anisopliae isolated from the USA (MEl, 23, 2951), Europe (817) and Australia (324) (Fig. 2). After digestion of the genomic DNA with appropriate restriction enzymes, only a single fragment hybridized to the ssgA cDNA, indicating that there is only one form of the gene in the genome. The same size fragments were present when DNA from the three American isolates was digested with EcoRI, EcoRV, and HindII1. Different fragments were obtained from the European and Australian isolates. The Australian isolate was distinctive in that two DNA fragments hybridized to the ssgA
+1.7 Janeson-Wolf (Antlgenlc Index) -1.7 GOR
GOR
/
/
turns
GIIR OL hekes
-
fl sheets
+1.7 Janeson-Wolf (Antlgenlc Index) ' -1.7 GLIR turns
KO
GOR
a hellces
GOR
B
n
e
w
sheets
h~l--_-
hydrophlllclty
rJ\,,
---
-5.0 +1.7 Janeson-Wolf (Antlgenlc 'nd~lx; ' GOR
Fig. 4. Computer-aided proteins. Hydrophilicity
o( hellces -
GOR
B sheets
I
\
n Al--I
secondary structure analysis of the deduced aa sequence of the SSGA (M. anisopliae), Scl (S. commune), and RodA (A. nidulans) was determined by the method of Kyte and Doolittle (1982). Hydrophobic regions are represented by negative numbers and
hydrophilic
regions
by positive
secondary
structure
predictions
were obtained
n
turns
GOR
from Schuren
ones, along the ordinate, according
and Wessels
to the algorithm
The antigenic of Garnier
(1990) and Stringer
index was calculated
according
et al. (1978) are also shown.
et al. (1991)
respectively.
to a formula
Primary
sequence
of Jameson
and Wolf (1988). The
data for Scl and RodA proteins
122 cDNA which may represent a different restriction pattern of a single gene or, less likely, multiple gene copies. (c) Sequence similarities and secondary structure of SSGA The predicted aa sequence of the SSGA shows 48”/,, 642, 67 %, and 48 y0 identity to the S. commune extracellular hydrophobins - Scl, Sc3, and Sc4 - and the rodlet protein from A. ~~~~Z~~~,respectively (Fig. 3). All five proteins contained a hydrophobic N-terminal sequence of about 20 aa resembling signal sequences for excreted proteins (see Fig. 4 for hydropathy plots). This suggests that the M. anisopliae SSGA is also secreted. The aa sequences of these leaders were quite homologous (8 out of 20 aa are shared between SSGA and Sc3 or Sc4). The eight Cys residues that were present in each of the five proteins are arranged in the same pattern, including the conserved tripkb
‘1
2
3
4
5
6
7
8
7.46 -
2.37-
eptide CCN. No other significant sequence similarities were identified in computer-assisted searches of the GenBank data bases. Computer-generated predictions of the secondary structure and hydrophilicity of S SGA were compared with data derived from the primary sequences of Scl (Schuren and Wessels, 1990) and RodA (Stringer et al., 1991) (Fig. 4). All three proteins contain regions of p-sheet with few, if any, regions of a-helix outside the N-terminal signal sequence. These features are shared with Sc3 and Sc4 (R.J.St.L., unpublished data). Apart from the signal sequences, hydrophobic regions in various domains of these molecules probably represent sequences passing through the interior of the protein (Kyte and Doolittle, 1982). The antigenic index [a surface probability plot (Jameson and Wolf, 1988)] also indicates that the O-glycosylation site in SSGA is located on a potentially exposed surface peak of the protein, making glycosylation possible. If the eight Cys residues formed disuhide bonds within the SSGA molecule, then its shape and contours would be altered in a manner dependent on bridging patterns. These cysteines may also be involved in cross-linking of the polypeptides to form polymeric structures (Shuren and Wessels, 1990; Stringer et al., 1991). However, as they were small, secreted hydrophobic proteins composed primarily of a p-sheet secondary structure and containing eight Cys residues, the hydrophobins closely resemble a wide range oftoxins and agglutinins (Drenth et al., 1980). The disullide bridging patterns which vary between toxin molecules result in four loops cont~ning a toxin-aggIutinin fold. Further biochemical and structural studies are necessary to identify the polypeptide structure of the ssgA product and relate it to the function.
0.24Fig. 5. Detection Poly(A)
of the
.q@
mRNA
during
+RNA isolated from mycehum of M.
in SDB and transferred transferred
nutrient
deprivation.
nksopliaegrown for 32 h
to different culture media for 3 h. Cultures
to fresh
SDB
N-acetylglucosamine/0.5’!0 3); to MM plus 0.5’:; chitin (3 h) followed by a second transfer plus either lo0 glucose
(2 h) (lane 4) or
(lane 5). SDB cultures
transferred
lug ~~-acetyl~lucosamine
to unsupplement~d
MM supplemented
with starch (O.S’:,) (lane 7) or cellulose (l”,) were grown in SDB and transferred
previously,
and total RNA was extracted
to MM (2 h)
MM (lane 6). to
Methods: Cultures described
were
(lane I), to MM supplemented with I “,, chitin (lane 2); to MM plus chitin (0.5”“) (lane
(lane 8).
to fresh media as
from fungal samples
using guanidine HCI (St. Leger et al., 1991). The poly(A)‘RNA fraction was purified by oligo(dT)-cellulose chromatography (Sambrook et al., 19X9). RNAs (0.5 &lane) for Northern blots were denatured with giyoxal and dimethylsulfoxide (Sambrook et al., 1989). The I ‘.,, agarose gels were blotted to nylon and hybridized lowing autoradiography
as for Southern
(12 h), the membranes
blots (see Fig. 2). Folwere incubated
in 0.4 N
NaOH at 7O’C for 30 min and rehybridized with a radiolabeled r-tubulin gene fragment ofM. aniwpfiaeto confirm the even transfer of RNA during blotting.
The a-tub&in
gene was obtained
in the igtl I vector with (R.J.St.L., unpublished).
monoclonal
by screening anti-~-tubuiin
a cDNA (Sigma,
library T9026)
(d) Expression of ssgA The ssgA gene was selected in a screen for genes expressed during nutrient deprivation. The pattern of the ssgA transcript accumulation was examined by Northern hybridization analysis using ssg_4 cDNA as a probe (Fig. 5). The transcript was present at low or undetectable levels in mycelia growing rapidly in SDB medium (Difco) or when the mycelia were transferred to medium containing N-acetylglucosamine, a readily utilized carbon and nitrogen source. By contrast, a hybridizing band of approx. 600 nt accumulated when cells were transferred to MM. Production was enhanced when the MM was supplemented with polymers (cellulose, chitin) at levels too low to produce carbon catabolite repression. A similar phenomenon has been recorded for Pr 1 production (St. Leger et al., 1988; 1991). The sJ;rA mRNA ceased to accumulate when chitin medium was supplemented with glucose or IV-acetylglucosamine for 2 h, confimling that transcript synthesis was catabolite repressed. We performed nuclear run-on
123 experiments
to confirm that regulation
of synthesis
N
D
C
was at
the level of transcription, and we compared the transcription rates of ssgA with those for mRNA of prl which had
bp
been previously determined (St. Leger et al., 1992) (Fig. 6). Neither transcript (ssgA or prl) was produced by nuclei from nutrient-rich cultures, but they were produced rapidly (< 2 h) when the cells were deprived of nutrients, reaching a maximum concentration after 24 h. Although the pattern of production of prl and ssgA transcripts was coordinated, levels of the prl transcript always exceeded those of ssgA, consistent with Pr 1 being the major starvation-specific protein. Pr 1 is produced at high levels by M. anisopliae germlings induced to differentiate infection structures by nutrient deprivation (St. Leger et al., 1989b). Coordinated regulation of ssgA with prl implies that the .ssgA gene would be expressed during infection processes. Indeed, the transcript was present at low or undetectable levels in conidia or undifferentiated germlings but at high levels in germlings producing infection structures (appressoria) (Fig. 7).
Fig. 7. Northern germlings
phal growth duced
analysis
producing
of the ssgA mRNA
appressoria
(N). Methods:
by germinating
Infection
conidia
glass petri dishes as described
in conidia
(D) or producing structures
(appressoria)
in yeast extract previously
(C) or from
nondifferentiated
media (0.0125%
w/v) in
(St. Leger et al., 1989a). Yeast
extract media (0.08% w/v) was used to obtain polar hyphal growth. RNA
(3 pgilane)
1991), denatured
was extracted with glyoxal
with a labeled random
primed f&RI
in Figs. 2 and 5. Following
branes were incubated
in 0.4 N NaOH
Stock Center) to confirm
Time
(h)
of prl and ssgA transcription
during nutrient
tion. Methods: Nuclei were isolated from nutrient-rich from cultures
transferred
to 72 h. Transcription isolated and assayed 1992). Assay MnCl,/200
depriva-
(SDB) cultures and
from SDB to MM (see legend of Fig. 5) for up
of prl and ssgA was measured by previously
mixes contained
described
in vitro. Nuclei were
procedures
(St. Leger et al.,
10 mM Tris pH S/l0 pM MgCl,/2
PM KC1/0.5 mM dithiothreitol/lO”;
glycerol/O.05
PM
mM each
ATP, CTP, and GTP/200 PCi of [3ZP]UTP (3000 Ci/mmol)/S x 10’ nuclei. Incubations lasted 30 min at 30°C. Incorporation of [32P]UMP into a trichloracetic-acid-precipitable et al., 1989). The synthesized
product was determined (Sambrook RNA was extracted and ethanol precipi-
tated, and 5 x IO’ cpm was used for hybridization natured
DNA (ssgA cDNA
ters as described
(Sambrook
plasmid
inserts)
to 10 pg of alkali de-
bound
to nitrocellulose
et al., 1989). Hybridizations
fil-
were carried out
in 50”,, formamide/ x SSCjl x Denhardts solution/O. 1 Of, SDS. Washes ranged from 2 x SSC to 0.1 x SSC. Counting error (two standard deviations) was 5% or less. cc-Amanitin added to the transcription that hybridized system.
(an inhibitor
of RNA polymerase
mix (0.5 mg/ml) blocked
to the ssgA gene, demonstrating
the synthesis the specificity
II)
of RNA of the
rDNA
Total
(St. Leger et al.,
and dimethylsulfoxide (Sambrook et al., w/v agarose gel), and blotted and probed
(1.25,
ized with a radiolabeled
Fig. 6. Regulation
using guanidine,HCl
1989), electrophoresed described
hy-
were in-
fragment
from the ssgA plasmid
autoradiography
as
(24 h), the mem-
at 70°C for 30 min and rehybrid-
fragment
the even transfer
(clone
1:7D, Fungal
Genetics
of RNA during blotting.
(e) Conclusions (1) Although the SSGA protein is not an enzyme and therefore presumably incapable of directly providing nutrients, the ssgA gene expression is coordinately regulated with prl, and it is transcribed when fungal growth is limited by nutrient deprivation, as during the formation of infection structures (appressoria). (2) As a hydrophobin, the SSGA peptide is presumably involved in building the walls of these infection structures and could assist with hydrophobic attachments to the cutitular surface and preventing desiccation. The resemblance of SSGA to toxins also merits further investigation to identify function. (3) The role of nutrient levels in regulating hydrophobin production by other fungi has not been determined. Nutrient deprivation plays a role in stimulating conidiation in many fungal species (Griffin, 1981) and aerial hyphae represent a potential nutrient sink. The effect of nutrient levels on the gene expression is therefore consistent with the location of other hydrophobins on aerial hyphae (Wessels et al., 1991) or conidial walls (Stringer et al., 1991).
124 Schuren,
ACKNOWLEDGEMENTS
F.H.J.
in fruiting
We thank David C. Frank for technical assistance. This work was supported in part by a grant (89-37263-4463) from the United States Department of Agriculture Competitive Research Grants Office.
and Wessels, dikaryons
J.G.H.:
St. Leger, R.J., Butt, T.M., Goettel, Production
agglutinin Feinberg, them.
J.S. and Wright,
C.S.: The toxin-
fold. J. Biol. Chem. 255 (1980) 2652-2655.
A.P. and Vogelstein,
restriction Garnier
of proteins
endonuclease
B.: A technique
fragments
for radiolabclling
DNA
to high specific activity. Anal. Bio-
137 (1984) 266-267. J., Osguthorpe,
D.J. and Robson,
of simple methods
ture of globular
proteins.
Griffin, D.H.:
Fungal
Kyte, J. and Doolittle, pathic character Lipman,
anisopliae.
searches. Sambrook,
CABIOS
for
4 (1988) 181-186.
Exp. Mycol.
13 (1989a) 274-288.
R.C. and Roberts, fungus
R.C. and Roberts,
by the entomopathogenic
protease
D.W.: Synthesis during differentia-
Metarhizium
unisopliue.
Exp.
St. Leger, R.J., Staples, R.C. and Roberts, D.W.: Changes in translatable mRNA species associated with nutrient deprivation and protease synthesis in Metarhizium
anisopliae.
J. Gen. Microbial.
for displaying
the hydro-
J. Mol. Biol. 157 (1982) 105-132.
W.R.: Rapid and sensitive protein and Maniatis,
anisopliae.
Stringer,
Laboratory Manual, 2nd ed. Cold Spring Harbor Cold Spring Harbor, NY, 1989.
Cloning.
Laboratory
gene from the Eur. J. Biochem.
inactivation. Wessels, J.G.H.,
similarity A
Press,
analysis
137 (1991) 807-
D.W. and Staples,
entomopathogenic fungus 204 (1992) 991-1001.
Genes
developmental
The rhn mutation
of aerial hyphae,
gcnc. J. Gen. Microbial.
W.E.: Rodlet-
171.
Asgeirsdottir,
of Schizophyllum
protease Metarhizium
mutant induced by directed gene
Dev. 5 (1991) 1161-l
DeVries, O.M.H.,
R.C.: Molec-
of the cuticle-degrading
M.A., Dean, R.A., Sewall, T.C. and Timberlake,
mation T.: Molecular
D.C., Roberts,
ular cloning and regulatory
less, a new Aspergillus
Science 227 (1985) 1435-1441. E.F.
Staples,
a cuticle-degrading
tion of the entomopathogenic Mycol. 13 (1989b) 253-262.
structural
index: a novel algorithm
R.F.: A simple method
of a protein.
J., Fritsch,
including
St. Leger, R.J., Frank,
struc-
John Wiley & Sons, New York, 1981.
determinants.
D.J. and Pearson,
of the accuracy
the secondary
J. Mol. Biol. 120 (1978) 97-120.
Physiology.
antigenic
B.: Analysis
for predicting
B.A. and Wolf, H.: The antigenic
predicting
with a
815.
and implications
Jameson,
M.S.,
in vitro of appressoria
fungus Metarhizium
J., Low, B.W., Richardson,
expressed
homologies
ulation of production of proteolytic enzymes by the entomopathogenic fungus Meturhizium anisopliae. Arch. Microbial. 150 (1988) 413-416.
St. Leger, R.J., Butt, T.M., Staples,
Drenth,
commune:
gene not regulated by mating-type genes. Gene 90 (1990) 199-205. St. Leger, R.J., Durrands, P.K., Cooper, R.M. and Charnley, A.K.: Reg-
D.W.:
REFERENCES
Two genes specifically
of Schizophyllum
S.A. and Springer,
commune,
affects expression
which suppresses
J.: for-
of the Sc3 hydrophobin
137 (1991) 2439-2445.