Báo cáo khoa học: Characterization of recombinant forms of the yeast Gas1 protein and identification of residues essential for glucanosyltransferase activity and folding pot

Characterization of recombinant forms of the yeast Gas1 protein
and identification of residues essential for glucanosyltransferase
activity and folding
Cristina Carotti
1
, Enrico Ragni
1
, Oscar Palomares
2
, Thierry Fontaine
3
, Gabriella Tedeschi
4
,
Rosalı
´
a Rodrı
´
guez
2
, Jean Paul Latge
´
3
, Marina Vai
5
and Laura Popolo
1
1
Dipartimento di Scienze Biomolecolari e Biotecnologie, Universita
`
degli Studi di Milano, Milano, Italy;

2
Departamento de
Bioquimica y Biologia Molecular, Facultad de Ciencias Quimicas, Universidad Complutense, Madrid, Spain;
3
Institut Pasteur,
Laboratoire des Aspergillus, Paris, France;
4
Dipartimento di Patologia Animale, Igiene e Sanita
`
Pubblica Veterinaria,
Universita
`
degli Studi di Milano, Milano, Italy;
5
Dipartimento di Biotecnologie e Bioscienze, Universita
`
degli Studi
di Milano-Bicocca, Milano, Italy
Gas1p i s a glycosylphosphatidylinositol-anchored p lasma
membrane glycoprotein of Saccharomyces cerevisiae and is
a representative of Family GH72 of glycosidases/transgly-
cosidases, which also includes proteins from human fungal
pathogens. Gas1p, Phr1-2p from Candida albicans and
Gel1p f rom Aspergillus fumigatus have been shown to be
b-(1,3)-glucanosyltransferases required for proper cell wall
assembly and morphogenesis. G as1p is organized in to three
modules: a catalytic domain; a cys-rich domain; and a highly
O-glycosylated serine-rich region. In order to provide an
experimental system for the biochemical and st ructural
analysis of Gas1p, we expressed soluble forms in the m eth-

ylotrophic yeast Pichia past oris. Here we r eport that 48 h
after induction with methanol, soluble Gas1p was produced
at a y ield of % 10 mgÆL
)1
of medium, and this value w as
unaffected by the further removal of the serine-rich region or
by fusion to a 6 · His tag. Purified soluble Gas1 protein
showed b-(1,3)-glucanosyltransferase activity that was
abolished by replacement o f the putative catalytic residues,
E161 and E262, with glutamine. Spectral studies confirmed
that the recombinant soluble Gas1 protein assumed a stable
conformation in P. pastoris. Interestingly, thermal dena-
turation studies demonstrated that Gas1p is highly resistant
to heat denaturation, and a complete refolding of the protein
following heat treatment was observed. We also showed that
Gas1p contains five intrachain disulphide bonds. T he effects
of the C74S, C103S and C265S substitutions in the mem-
brane-bound Gas1p were analyzed in S. cerevisiae.The
Gas1-C74S protein was t otally unable to complement the
phenotype o f t he gas1 null mutant. We found that C74 is an
essential residue for the proper f olding and maturation of
Gas1p.
Keywords: b(1,3)-glucanosyltransferase; Gas1 protein;
Pichia pastoris; yeast cell wall.
The cell wall is an extracellular compartment that plays
several essential functions in yeast a nd fungal cells. It
determines the cell morphology and preserves o smotic
integrity. In fungal pathogens, the cell wall is involved in the
interaction with the host cells and in virulence. The
biogenesis of the extracellular matrix is a fascinating aspect

of yeast m orphogenesis. The elucidation of t he enzymatic
activities involved in its assembly could be relevant for the
development of new antifungal drugs [1].
The enzymes responsible for the architecture and remod-
elling of the cell wall are still largely unknown. Several lines
of evidence suggest that a class of recently identified
enzymes, endowed with b(1,3)-glucanosyltransferase activ -
ity, could play a role in the cross-linking of cell wall
components [2,3]. This activity was detected for the first
time in the Gel1 protein of Aspergillus fumigatus and
subsequentially in the homologous p roteins Gas1 of Sac-
charomyces cerevisiae and Phr1-2 of Candida albicans [3,4].
On the basis of the sequence similarity, these enzymes h ave
been grouped in a new family, called Family GH72, in the
classification of glycoside hydrolases (Class Glycosidases/
Transglycosidases; />Gas1p, Gel1p a nd Phr1-2p c atalyze t he splitting of an
internal b(1,3)-glycosidic bond of a donor laminarioligo-
saccharide followed by the transfer of the new reducing
end t o t he nonreducing end of an acceptor molecule,
with the formation of another b(1,3)-glycosidic bond [3].
As the anomeric configuration of the linkage was
conserved, these enzymes were also classified a s r etaining
enzymes.
Correspondence to L. Popolo, Dipartimento di Scienze Biomolecolari e
Biotecnologie, Universita
`
degli Studi di Milano, Via Celoria 26,
Milano, Italy. Fax: +39 02 50314912, Tel.: +39 02 50314919,
E-mail:
Abbreviations: DTNB, 5,5¢-dithiobis(2-nitrobenzoate); Endo H, endo-

b-N-acetylglucosaminidase H; GH, glycoside h ydrolases; GluTD,
b-(1,3)-glucan transferase domain; GPI, glycosylphosphatidyl-
inositol; H, 6 · His.
(Received 20 May 2004, revised 15 July 2004, accepted 21 July 2004)
Eur. J. Biochem. 271, 3635–3645 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04297.x
The reaction mechanism proposed for these enzymes is a
general acid/base catalysis [5]. Protonation of the glycosidic
oxygen by a catalytic acid residue is f ollowed by the release
of the cleaved product and stabilization of the carbon cation
by the catalytic nucleophile. The new reducing end is then
transferred to the hydroxyl group at the 3 -position of the
nonreducing end of another a cceptor molecule, yielding a
linear transfer product longer than the original substrate
[3,4]. At low c oncentrations of substrate, the reaction is
preferentially hydrolytic, the hydroxyl group of a water
molecule being the final acceptor. In glycoside hydrolases,
as for many cellulases, mannanases or glucanases, the
proton donor and the nucleophile residues are usually
aspartates or glutamates [6–8]. These residues are located in
different microenvironments that influence t he protonation
state of the carboxyl group of their side-chain [7].
The a im of the present study was to express Ga s1p a t
high levels for biochemical and structural characterization
of the protein as a r epresentative of the GH72 family.
Spectroscopic analyses o f t he purified proteins were
performed, and the behaviour of the purified protein
upon heat treatment was also monitored. By combining
heterologous expression and site-directed mutagenesis, the
role of two putative catalytic residues was investigated.
Moreover, the disulphide bonds present in Gas1p were

quantified and the intra- or intermolecular bonding was
determined. The effects of the rep lacement o f C 74, C105
and C265 with a serine residue on the expression and
complementation o f the mutant phenotype o f the gas1
null mutant of S. cerevisiae were examined. These data
provide a first insight into the biochemical features of
proteins of the GH72 family and d emonstrate that the
most N-terminal highly conserved cysteine is crucial for
the folding and maturation of Gas1p.
Materials and methods
Strains and growth conditions
Pichia pastoris strain GS115 (his4) (Invitrogen, Leek, the
Netherlands) was used for the heterologous expression of
Gas1p. To select His
+
transformants, regeneration dextrose
plates [2% (w/v) dextrose, 1
M
sorbitol, 1 .34% (w/v) Difco
yeast nitrogen b ase ( YNB), 4 · 10
)5
% (w/v) biotin, 2%
(w/v) agar] were used. F or Mut
+
or Mut
s
phenotype
screening, the minimal dextrose [2% (w/v) dextrose, 1.34%
(w/v) YNB, 4 · 10
)5

% (w/v) biotin, 2% (w/v) aga r] and
minimal methanol plates [0.5% (v/v) methanol, 1.34% (w/v)
YNB, 4 · 10
)5
% (w/v) biotin, 2% (w/v) agar] were used.
To induce the expression of recombinant proteins, the
His
+
Mut
s
colonies were shifted from a glycerol-complex
medium [1% (w/v) yeast extract, 2% ( w/v) peptone, 1%
(v/v) glycerol, 1.34% (w/v) YNB, 4 · 10
)5
% (w/v) biotin]
to a methanol-complex medium [1% (w/v) yeast extract,
2% (w/v) peptone, 0.5% (v/v) methanol, 1.34% (w/v) Y NB,
4 · 10
)5
% (w/v) biotin], according to the man ufacturer’s
instructions. Cells were grown in b atches at 30 °Cwith
strong agitation, and the growth was monitored th rough the
increase in attenuance at 600 nm.
The S. cerevisiae haploid strain, WB2d, carrying a n
inactivated GAS1 gene (ga s1::LEU2), was used for the
expression of the glycosylphosphatidylinositol (GPI)-
anchored forms Gas1-C74S, Gas1-C103S and Gas1-
C265S. S. cerevisiae cells were cultured in batches at 30 °C
in SC medium [0.67% (w/v) YNB, 2% (w/v) glucose and
the required supplements at 50 mgÆL

)1
foraminoacidsand
uracil and 100 mgÆL
)1
for adenine].
Construction of expression vectors
Recombinant plasmids for integrative recombination in
P. pastoris were generated by cloning BamHI/XhoI-digested
PCR fragments into the corresponding sites of the P. pas-
toris exp ression vector, p HIL-S1, to obtain in-frame fusion
with the secretion signal of the P. pastoris PHO1 gene. The
recombinant plasmids we re named as follows: pSC18
(sGas1
523
), pSC36 (sGas1
523
-H), pSC7 (sGas1
482
)and
pSC68 (sGas1
482
-H).
The plasmid pXH, carrying the full coding sequence of the
GAS1 gene, was used as a template for PCR amplifications
[9]. The s oluble forms – Gas1
523
p (lacking t he C-terminal
GPI-attachment signal) and Gas1
482
p (also lac king the Ser-

box region) – were obtained using the forward primer
XHup (5¢-GCATATTCGACTGA
CTCGAGACGATGT
TCCAGCGATTGAA-3¢) and the reverse primers XHdown
(5¢-ATCGTCGGGCTCA
GGATCCTTAAGATGAAGA
TGAAGCTGAAGA-3¢) or XH-Sdown (5 ¢-GTCGTCG
AGCTCA
GGATCCTTAATCAACACTACCTGATGC
AGA-3¢), respectively. XHup is complementary to nucleo-
tides +68 to +87 o f the coding region of GAS1 and has an
XhoI site incorporated (underlined). XHdown and XH-
Sdown are complementary to nucleotides +1549 to +1569
and to nucleotides +1426 to +1446, respectively, and have
an in-frame TAA stop codon (shown in bold) and a BamHI
site (underlined).
For each construct, a 6 · His ( H)-tagged s oluble form
was p repared using the same f orward primer, Xhup, and
the reverse primer His-XHdown (5¢-ATCGTCGGG
CTCA
GGATCCTTAGTGATGGTGATGGTGATGAGA
TGAAGATGAAGCTGAAGA-3¢), for Gas1
523
-H, or
His-XH-Sdown (5¢-GTCGTCGAGCTCA
GGATCCTTA
GTGATGGTGATGGTGATGATCAACACTACCTGAT
GCAGA-3¢), for sGas1
482
-H (the 6 h istidine coding

sequence is shown in italics).
Plasmid DNA was purified using plasmid purification
kits (Qiagen). DNA sequencing was routinely used to
confirm the correct fusions and the absence of undesired
mutations throughout the coding sequence (BMR-Servizio
sequenziamento; Universita
`
di Padova, Padova, Italy).
Mutagenesis of E161 and E262
The mutant soluble forms of sGas1
523
– sGas1E161Q-H
and s Gas1E262Q-H – were obtained by overlap extension
PCR [10]. In the first PCR step, two partially overlapping
fragments of GAS1 were amplified using two sets of
primers. For sGas1E161Q-H, a pairing of the forward
primer XHup and reverse primer RMGLN161
(5¢-TGTTAGTAACTTGATTACCGGCGAAG-3¢), and
a pairing of the forward primer FMGLN161 (5¢-CTT
CGCCGGTAATCAAGTTACTAACA-3¢) and reverse
primer His-XHdown were used. Similarly, for s Gas-
1E262Q-H, the amplification was carried out using a
pairing of the forward primer XHup and reverse primer
3636 C. Carotti et al.(Eur. J. Biochem. 271) Ó FEBS 2004
RMGLN262 (5¢-TACAACCGTATTGAGAGAAGA
AAAC-3¢), and a pairing o f the forward primer
FMGLN262 (5¢-GTTTTCTTCTCTCAATACGGTTG
TA-3¢) a nd reverse primer His-XHdown. RMGLN161
and FMGLN161 are complementary to nucleotides +467
to +493 of the coding region of GAS1 , while RMGLN262

and FMGLN262 are complementary to nucleotides +771
to +796. In these primers, a Gln codon (shown in bold),
instead of the Glu codon, was incorporated. For
Gas1E161Q-H, 20 cycles of a 45 s me lting step at 94 °C, a
1 m in annealing step at 50 °C and a 2.5 min extension step
at 72 °C were performed using t he Pfu Turbo DNA
polymerase (Stratagene). For Gas1E262Q-H, the tempera-
ture of the annealing step was 62 °C with p rimers XHup
and RMGLN262, and 57 °C with p rimers FMGLN262
and His-XHdown. The mutations of interest are located in
the region of overlap between the amplified fragments. The
pairing of overlapping fragments was u sed for a second
PCR step, using the forward primer XHup and the reverse
primer His-XHdown, to amplify the full-length mutated
sequence of GAS1. Twenty-five cycles o f a 45 s melting step
at 94 °C,a1 minannealingstepat50 °C for Gas1E161Q-H
and 5 3 °C for Gas1E262Q-H, and a 2 min extension step at
72 °C were performed and the Taq DNA polymerase was
used. The corresponding P. pastoris expression plasmids
derived from pHIL-S1 were named pE161Q and pE262Q.
Mutagenesis of C74, C103 and C265
The m utant GPI-anchored forms – Gas1-C74S, Gas1-
C103S and Gas1-C265S – were constructed by PCR-based
mutagenesis. For Gas1-C74S, two fragments of GAS1 were
amplified using two sets of primers: the primer pair OligoUP
(5¢-TACCATTTATCGATTACTGGCATACAATGGT-
3¢), complementary to nucleotides )830 to )800, and Oligo1
(5¢-TCTG
GAGCTCgaCTCATAATTGGCCAAAGG-3¢),
partially complementary to nucleotides +199 to +228; and

the primer p air Oligo2 (5 ¢-GAGtc
GAGCTCCAG
AGATATTCCATACCT-3¢), partially complementary to
nucleotides +214 to + 242, and OligoDOWN (5¢-ATAC
GCTCCATCTACATATGCTGACG-3¢) complementary
to nucleotides + 2408 to +2433. Oligo1 and Oligo2 have
a SacI site (underlined) c ontaining a serine codon (in bold)
instead of t he C74 c odon and two exchanged bases (lower
case) that allow the retention of residue S73. For Gas1-
C103S, the reverse primer Oligo3 (5¢-AGCCTTC
ATCGATTCGGAGTGATCTAGAGTG-3¢), partially
complementary to nucleotides +288 to +318, and Oligo4
(5¢-CTCCGA
ATCGATGAAGGCTTTGAATGATGC-3¢)
partially complementary to nucleotides +300 to +329
substituted for Oligo1 and Oligo2, respectively. Oligo 3
and Oligo4 have a ClaI site (underlined) containing a
serine codon (shown in bold) instead of the C103
codon. For Gas1-C265S, the reverse primer Oligo5
(5¢-TCGTTG
GATCCGTATTCAGAGAAGAAAAC-3¢),
partially complementary to nucleotides +772 to +800, and
the forward primer Oligo6 (5¢-AATAC
GGATCCAA
CGAAGTAACACCAAGAC-3¢), partially complement-
ary to nucleotides +784 to +814, were used instead of
Oligo1 and Oligo2. Oligo5 and O ligo6 have a BamHI site
(underlined) containing a serine codon (in bold) instead of
the C265 codon. Pfu Turbo DNA polymerase was used. All
mutant forms of the GA S1 gene were cloned into the

YCplac33 ARS-CEN shuttle vector and the resulting
plasmids were used to transform the WB2d ( gas1::LEU2)
strain. As a control, the same strain was transformed with
the wild-type GAS1 gene cloned in the same single copy
vector.
Transformation of
P. pastoris
and expression
of recombinant Gas1 proteins
Plasmids, linearized with BglII, were transformed into
P. pastoris cells using the ÔEasyCompÕ chemical transfor-
mation method (Invitrogen). His
+
Mut
s
mutants were
obtained by selecting His
+
transformants that grew well
on minimal dextrose, but poorly on minimal methanol
plates. To i nduce the expression of recombinant proteins,
the positive clones were cultured at 30 °C overnight in
10 mL of glycerol-complex medium with strong agitation
and the cells were spun down and resuspended in 20 mL of
methanol-complex medium to an attenuance o f 1.0 at
600 n m. Fresh m ethanol was added daily to 0.5% (v/v). The
culture medium was collected 48 h afte r the induction,
centrifuged and culture supernatants were quickly frozen
andstoredat)20 °C prior to purification.
Purification of His-tagged Gas1 proteins

A 10–20 mL sample of the culture supernatant was dialyzed
at 4 °C for 16 h against lysis buffer (50 m
M
sodium
phosphate buffer, pH 8.0, 200 m
M
NaCl). T wo m illilitres
of 50% (w/v) Ni-nitrilotriacetic acid agarose (Qiagen) was
added to the dialyzed supernatant, the mixture was
incubated at 4 °C for 1 h under gentle vertical rotation
andthenappliedtothecolumn(0.7· 10 cm or 1 · 10 cm;
Econo Column, Bio-Rad). The resin was washe d twice with
8mL of wash buffer (50m
M
sodium phosphate buffer,
pH 8.0, 200 m
M
NaCl, 3–5 m
M
imidazole). The His-tagged
protein was eluted with elution buffer (50 m
M
sodium
phosphate buffer, pH 8.0, 200 m
M
NaCl, 200 m
M
imidaz-
ole) and 1 mL fractions were collected. The position o f
Gas1 protein in the elution profile was determined. Protein

fractions, corresponding to the major peaks, were collected.
Protein concentration was determined by using the dye
reagent protein assay (Bio-Rad).
Endo-b-
N
-acetylglucosaminidase H treatment
Endo-b-N-acetylglucosaminidase H (Endo H) treatment
was performed on culture supernatant or on purified
proteins. For the treatment of culture supernatant, 80 lL
of a deglycosylation buffer [300 m
M
sodium citrate, pH 5.5,
0.5% (w/v) SDS, 2% (v/v) 2-mercaptoethanol] was a dded to
80 lL of culture supernatant and boiled for 2 min. After
repartition into two equal volumes, one aliquot was added to
100 mU of Endo H (Roche) and the other was used as a
control. After 18 h of incubation at 37 °C, an equal volume
of 2· Laemmli buffer was added and the samples were boiled
for 2 min prior to electrophoresis. For the treatment of
purified proteins, 2 lg of protein in 50 m
M
sodium acetate,
pH 5.5, was used. Samples were divided into two aliquots:
one was used as a control and the other was treated with 2 lL
of Endo H (10 mU). For treatment of the denatured protein,
Ó FEBS 2004 Production and characterization of Gas1p (Eur. J. Biochem. 271) 3637
SDS and 2-mercaptoethanol were added to final concentra-
tions of 0.02% (w/v) and 0.1
M
, resp ectively, prior to division

into two aliquots and the addition of t he enzyme. T o test the
effect of removal of N-linked chains on enzyme activity,
% 20 lg of native protein was treated with 10 mU Endo H in
0.1
M
sodium acetate, pH 5.5. After checking a small aliquot
of the sample for the complete removal of N-linked chains,
the remainder was used to assay the b(1,3)-endoglucanase
activity.
Electrophoresis and immunoblotting procedures
Total protein extracts from S. cerevisiae cells were obtained
as described previously [11]. Aliquots o f P. pastoris culture
supernatants, or fractions from the purification procedure,
were denaturated by b oiling for 3 min in SDS sample buffer
[0.0625
M
Tris/HCl, pH 6.8, 2.3% (w/v) SDS, 5% (v/v)
2-mercaptoethanol, 10% (v/v) glycerol and 0 .01% (v/v)
Bromophenol blue]. Proteins were separated by SDS/
PAGE in 7 or 8 % polyacrylamide gels. For analysis of
the proteins under nonreducing conditions, 2-mercaptoeth-
anol was omitted from the SDS sample buffer, and samples
were processed as previously described [12]. Samples were
boiled for 3 min and then divided into two samples of equal
volume. Dithiothreitol (20 m
M
final concentration) was
added to one sample, which was then reheated for 3 min.
When loaded side-by-side, all samples received 100 m
M

N-ethylmaleimide a fter cooling.
After electrophoresis, proteins were either stained with
Coomassie Blue R-250 or using a silver nitrate kit (Amer-
sham Pharmacia Biotech, Bucks., UK). For detection b y
Western blotting, proteins were transferred to nitrocellulose
membranes and processed as described previously [13].
Rabbit anti-Gas1p immunoglobulin, diluted 1 : 3000, was
used to detect Gas1p. A monoclonal anti-(polyHistidine)
immunoglobulin, diluted 1 : 3000 (Sigma), was used to
detect the 6 · His tag. Horseradish peroxidase-conjugated
anti-rabbit or anti-mouse secondary immunoglobulins were
used. Bound antibodies were revealed using the ECL
Western blotting detection reagents (Amersham Pharmacia
Biotech). To check the equivalence of protein loading,
primary antibodies were stripped and filters were treated
with anti-phosphofructokinase 1 ( Pfk1p) im munoglobulin
(kindly provided by J. J. Heinisch, Universitat H ohenheim,
Stuttgart, Germany), diluted 1 : 30 000.
Pulse–chase experiment and immunoprecipitation
A total of 2.5 · 10
8
logarithmically growing cells (equival-
ent to a value of 12 at an attenuance of 450 nm) were
resuspended in 4 mL of SC medium, incubated at 30 °Cfor
20 min, then pulse-labelled for 7 min with 350 lCi of
[
35
S]methionine. Pulse labelling was terminated upon the
addition of 40 lL of chase solution containing 0.3% (w/v)
methionine and 0.3

M
(NH
4
)
2
SO
4
. Immediately f ollowing
the addition of chase mixture, or a fter 10, 30 and 60 min
chase periods, 1 mL of culture was withdrawn and further
reactions were stopped by the additio n of NaF and NaN
3
to
final concentrations of 10 m
M
. Cells were rapidly collected
by centrifugation i n a microfuge at 4 °C and resuspended in
50 lL of TBS/SDS buffer [(50 m
M
Tris/HCl, pH 7.2,
150 m
M
NaCl, 1% (w/v) SDS] with protease inhibitors:
2m
M
phenylmethanesulfonyl fluoride, 1 lgÆmL
)1
pepsta-
tin, 50 lgÆmL
)1

aprotinin, and 1 0 lgÆmL
)1
leupeptin. Cells
were broken by vortexing with glass beads (0.45 mm
diameter) for four, 1 min periods, and then lysates were
denaturated f or 5 min at 100 °C. This treatment fully
solubilized Gas1p [14]. Then, 450 lL o f RIPA-minus buffer
[10 m
M
Tris/HCl, pH 7.2, 150 m
M
NaCl, 1% (v/v) Triton
X-100, 1% (w/v) sodium deoxycholate plus protease
inhibitors], was added and, i n this way, the p ercentage of
SDS was lowered to 0.1. A fter a 15 min incubation at 4 °C,
beads and cellular debris were sedimented by a 2 min
centrifugation i n a microfuge, followed by a further
centrifugation of the surpernatant for 15 min at 4 °C. Eight
microlitres of preimmune serum was added to the cleared
supernatant, and the tubes were gently mixed for 1 h
at 4 °C. Fifty microlitres of a 30% (v/v) suspension of
Protein A–Sepharose was added and, after incubation for
1 h, immune complexes were s edimented a t low speed at
4 °C. The supernatant was transferred to a new Eppendorf
tube and 8 lL of anti-Gas1p IgG were added. Incubation
was continued overnight at 4 °C. Then, 50 lL of the Protein
A–Sepharose suspension was added, incubation co ntinued
for 1 h and, after sedimentation, the Protein A–Sepharose
immune complexes w ere w ashed fi ve tim es w ith 1 mL of
RIPA buffer [the same composition of RIPA-minus buffer

but containing 0.1% (v/v) SDS], containing protease
inhibitors. Immunoprecipitated G as1p was t hen solubilized
by heating the pellet in 50 lL of SDS sample buffer (2·),
and Protein A–Sepharose beads were r emoved by centrif-
ugation. Supernatants were analysed by SDS/PAGE and
gels were stained with Coomassie Brilliant blue, fluoro-
graphed with Amplify (Amersham) and exposed to X-ray
films.
Enzyme assays
To test for b(1,3)-glucanosyltransferase activity, the puri-
fied proteins were incubated at concentrations of
0.09–0.32 mgÆmL
)1
with 3 m
M
reduced laminarioligosac-
charide G
13
in 50 m
M
sodium acetate buffer, pH 5.5, at
37 °C. Aliquots o f 2.5 lL were w ithdrawn at different time-
points, supplemented with 45 lLof50 m
M
NaOH and then
analysed by high-performance anion-exchange chromato-
graphy (HPAEC) though a CarboPAC-PA1 column (Dio-
nex 4.6 mm · 250 mm), as d escribed by Hartland et al.[4].
Spectroscopical analyses
CD spectra were obtained at different temperatures in the

far-UV range (200–250 nm) on a Jasco J-715 spectropola-
rimeter (Japan Spectroscopic Co., Tokyo, Japan), as
described previously [15]. The protein concentration was
0.20–0.28 mgÆmL
)1
in 50 m
M
sodium acetate buffer,
pH 5.5. Mean residue mass ellipticity was calculated based
on 10 7.98 as the average molecula r mass per residue,
obtained from the amino acid composition, and expressed
in terms of [h]
MWR
(degree · cm
2
· dmol
)1
). Thermal
unfolding of sGas1
523
-H was monitored b y recording the
ellipticity at 220 nm while heating from 20 °Cto80°Cand
cooling again at 1 °CÆmin
)1
using a computer-controlled
circulation waterbath. Fluorescence emission spectra were
obtained on an S LM Aminco 8000 s pectrofluorimeter at
3638 C. Carotti et al.(Eur. J. Biochem. 271) Ó FEBS 2004
25 °C and in 0.2-cm optical path-cells, using 4 nm s lits for
both e xcitation a nd emi ssion beams. The sample concen-

tration was 0.20 mgÆmL
)1
in 50 m
M
sodium acetate buffer,
pH 5.5.
Quantification of disulphide bonds and free sulphydryl
groups
The disulphide bonds were quantified using 5,5¢-dithio-
bis(2-nitrobenzoate) (DTNB), as described previously [16].
Purified sGas1 protein was dialyzed overnight against
50 m
M
phosphate buffer, pH 8.0. When the quantification
was performed using an Endo H-treated protein, a second
round of affinity purification was carried out in order to
remove Endo H. The reaction with DTNB was carried out
in phosphate buffer, pH 8.0, at 25 °Cusinga2l
M
protein
solution. The final vo lume of the reaction was 1 mL. The
formation of the product was monitored at 4 12 nm using
E ¼ 13 113
M
)1
Æcm
)1
. In order to expose thiol groups,
which may be buried in the interior of the protein, the
sample was denaturated by boiling f or 3 min before reaction

with DTNB. To analyze disulphide thiols, freshly prepared
100 m
M
1,4-dithioerythritol solution was a dded to the
denaturated sample and the reduction was carried out for
2 h at 25 °C. Excess dithioerythritol w as removed by gel
filtration on PD10 before incubation with DTNB.
Results and discussion
Production of recombinant soluble forms of Gas1p
in
P. pastoris
Gas1 is a plasma membrane GPI-anchored glycoprotein of
% 125–130 kDa. It contains a large N-terminal catalytic
domain of abou t 310 residues (D23–P332), known a s the
b-( 1,3)-glucan transferase dom ain (GluTD), a cystine-rich
region containing a motif of eight cysteines (C370–C462)
and a serine-rich region in which 28 serines are clustered in a
region between residues S485 and S525 (Fig. 1A). The
serine-rich region is a target for O-glycosylation and is
dispensable for activity [2,13]. A secretory signal peptide
(M1–A22) and a signal sequence for GPI attachment are
present at the N- and C-terminal ends, respectively. In order
to undertake a biochemical characterization of Gas1p, we
attempted to express it in P. pastoris. DNA sequences,
encoding different soluble forms of Gas1p, were subcloned
in the pHIL-S1 expression vector in-frame with the DNA
sequence e ncoding the P. pastoris Pho1p s ignal sequence.
The expression of the proteins was driven by the methanol-
inducible AOX1 promoter. Constructs encoding forms of
Gas1p lacking the GPI-attachment signal (sGas1

523
)orthe
GPI-attachment signal and t he serine-rich region ( sGas
482
)
and their His-tagged (H) versions are shown in Fig. 1B.
Fig. 1. Scheme of the mutant Gas1 proteins. (A) M odular organization of Gas1p. (B) Soluble recombinant proteins expressed in Pichia pastoris.The
different constructs were plac ed under the control of the AOX1 methanol-inducible promoter and expressed in s train GS115. (C) Glycosylphos-
phatidylinositol (GPI)-anchored proteins expressed in Saccharomyces cerevisiae. Con structs were placed under the c ontrol of the natural GAS1
promoter and cloned in the centromeric YCp Lac plasmid. Recomb inant plasmids were used for transformation of gas1 null mutant cells. The
position of the cysteine replacemen t is shown.
Ó FEBS 2004 Production and characterization of Gas1p (Eur. J. Biochem. 271) 3639
A major ban d of % 110 kDa was detected in the medium of
cells transformed with t he construct encoding sGas1
523
after
24 h of i nduction. Forty-eight hours after induction
(Fig. 2 A, lanes 4–6), levels reached a concentration equiv-
alent to 10 lgÆmL
)1
, determined using an ovalbumin
standard (data not shown). At 72 h the protein level was
equivalent to that observed at 48 h, and no degradation
products were detected, indicating that the secreted protein
was stable (data not shown). The 110 kDa protein was
identified as Gas1p because i t w as absent in the m edium
obtained from cells transformed with t he vector alone
(Fig. 2 A,B, lanes 1–3) a nd was recognized by anti-Gas1p
immunoglobulin (Fig. 2B, lanes 5 and 6). Removal of the
serine-rich region gave origin to a protein of % 90 kDa

(sGas1
482
), which was produced at levels equivalent to
sGas1
523
(Fig. 2 A, lanes 7–9). The presence of the His-tag
(sGas1-H) did not appreciably modify the expression level
of sGas1
523
and sGas1
482
, suggesting no deleterious effects
of the additional amino acids on Gas1p expression and
secretion ( Fig. 2A, lanes 10–12 and 13–15). Where the tag
was p resent the proteins were also recognized by a PolyHis
monoclonal antibody ( data not shown). T he difference in
molecular mass between sGas1
523
and sGas1
482
is % 20 kDa
and exceeds the 4 kDa predicted by the length of the
segment removed. This i s consistent with previous evidence
that the serine-rich region is a highly O-glycosylated
segment in the Gas1 protein [13].
The glycosylation profile of the proteins was examined.
Aliquots of medium containing the recombinant proteins
were treated with Endo H and analysed by Western blotting
using anti-Gas1p immunoglobulin (Fig. 2C). The apparent
molecular mass of both sGas1

523
-H and sGas1
482
-H
decreased by % 18–20 kDa, indicating th at the proteins
are N-glycosylated in P. pastoris. Interestingly, the contri-
bution of the N-linked glycans was % 10 kDa lower than in
sGas1p expressed in S. cerevisiae (data not shown). This is
in agreement with the evidence that N-linked oligosaccha-
ride chains are shorter in P. pastoris tha n in S. cerevisiae
[17].
Spectroscopical characterization of Gas1p
Because no direct structural information of Family GH72
proteins are available, we determined the signature spectra
of recombinant Gas1p. Gas1p was purified by affinity
chromatography on Ni-NTA agarose ( Fig. 5B below). The
CD spectru m of sGas1
523
-H in the f ar-UV range (peptide
bond region) that provides information about its secondary
structure, is shown in Fig. 3A. Deconvolution of the
spectrum of the protein at 20 °C by the convex-constraint
analysis method [18] gave the following composition: 15%
a-h elix, 29% b-sheet, 28% b-turn and 28% nonregular
conformation. Thermal unfolding of the protein was
followed by CD analysis in order to study the stability of
sGas1
523
-H. Changes in ellipticity were recorded at 220 nm
upon heating from 20 °Cto80 °C. The unfolding transition

of sGas1
523
-H was monophasic, with a melting point at
56.5 °C, and also h ighly cooperative (Fig. 3B). As shown in
the CD spectrum of Fig. 3A, the structural changes upon
heating to 80 °C resulted in a decrease of regular secondary
structure content ( 8% a-helix and 17% b-sheet). Interest-
ingly, these structural changes were totally reversible
because sGas1
523
-H recovered the initial structure at
20 °C after c ooling from 80 °C (Fig. 3A). Figure 3C shows
the fluorescence emission spectrum obtained for sGas1
523
-H
(spectrum 1). The emission of the protein was dominated by
the tryptophan contribution (spectrum 2, excitation at
295 n m) with a maximum at 320 nm and a shoulder at
332 nm. The intrinsic tryptophan fluorescence of a protein
is a sensitive indicator of the local environment of its
tryptophan residues. The mature G as1 protein contains five
tryptophan residues. The fact that the tryptophan emission
in sGas1
523
-H was s hifted to a lower wavelength than
expected for solvent-exposed tryptop hans strongly suggests
that these residues a re located in a hydrophobic environ-
ment in the protein. The low tyrosine contribution,
Fig. 2. Analysis of cultu re su pernata nts from Pichia pastoris-trans-
formed cells. (A) Coomassie Blue staining of 100 lLofculture

supernatant at 24 and 48 h from the shift (time zero) of cells containing
the pHIL-S1 vector (lanes 1–3), the cassette expressing sGas1
523
(lanes
4–6), sGas1
482
(lanes 7–9), sGas1
523
-H (lanes 10–12) and sGas1
482
-H
(lanes 13–15). (B) Immunoblot analysis of culture supernatants at 0, 24
and 48 h time-points from induction. A 4 0 lL sample was analysed by
SDS/PAGE and blotted proteins were incubated with anti-Gas1p IgG.
As representative samples, culture supernatants from cells containing
the pHIL-S1 vector (lanes 1–3), or the cassette expressing sGas1
523
(lanes 4–6), are s hown. (C) Endo-b-N-acetylglu cosaminidase H
(Endo H) t reatment of culture supernatants containing the indicated
recombinant proteins. A 40 lL sample of culture supernatant
obtained 48 h after induction w as incubated w ith (+) o r without (–)
Endo H and analysed by SDS/PAGE. Proteins were detected by
immunoblot using anti-Gas1p immunoglobulin.
3640 C. Carotti et al.(Eur. J. Biochem. 271) Ó FEBS 2004
30 residues in Gas1p, is also displayed in this figure
(spectrum 3). All these results support t he notion t hat
sGas1
523
-H produced in P. pastoris assumes a stable
conformation consistin g of a structure that contains both

a-he lices and b-sheets, and that the latter ones predominate.
Quantification of disulphides and free sulphydryl groups
present in Gas1p
Proteins of Family GH72 are rich in cysteines but no
determination of disulfide bonds has yet been reported. The
yeast Gas1 protein contains 14 cysteines. We used DTNB
to quantify t he number of disulphides plus free sulphydryl
groups in Gas1p. The recombinant protein lacking the
O-glycosylated region, which is dispensable for activity and
is devoid of cysteines, was analysed. As shown in Table 1,
sGas1
482
-H contai ns four thiols – one of which is readily
accessible to the reagent in native conditions – and five
disulphide bridges. The same results were obtained after
treatment with Endo H (Table 1), which completely
removes the N-linked c hains (data not shown). This
indicates that accessibility to t he reagent is not influenced
by the N-linked chains. Equivalent results w ere obtained for
sGas1
523
-H.
In order t o determine whether the disulphide bridges were
intra- or intermolecular, the electrophoretic mobility of
Gas1p was analysed under reducing and nonreducing
conditions. When s amples are d enaturated in th e absence
of a reducing agent a decrease in electrophoretic mo bility of
the nonreduced sample, with respect to the corresponding
reduced sample, in dicates the presen ce of interchain disul-
phide bridges, whereas an increase in mobility is related to

the presence of intrachain bonds. As shown in Fig. 4,
nonreduced Gas1p has a h igher m obility than reduced
Gas1p, indicating that the disulphide bo nds are intrachain.
Both the GPI-anchored form of Gas1p present in the yeast
cellular extract, and the purified sGas1
523
-H and sGas1
482
-H
proteins, gave similar results (Fig. 4A,B). In conclusion,
Gas1p has five intrachain disulphide bonds.
Conserved features of the catalytic domain of Family
GH72 and identification of the catalytic residues
in Gas1p
The sequences span ning the catalytic domain (GluTD) of 40
proteins have been aligned. Two glutamic acid residues
(shown in bold) are conserved in the A/S-G-N -E-V/I and
S-E-Y/F-G-C subsequences. Similarly to other protein
families of the GH-A clan, to which Family GH72 belongs,
it has been proposed that these r esidues might be important
for catalysis. A role of proton donor and nucleophile has
been proposed for these residues [2,19]. Moreover, six
glycine residues (159-197-243-264-290-304), five tyrosine
residues ( 92-113-198-231-294) and five cysteine residues
(74-103-216-234-265), are strictly conserved in all GluTD
sequences. Clan GH-A comprises many proteins that share
a low degree of conservation of the primary structure, but
contain a conserved fold consisting of a (b/a)
8
with the

putative catalytic residues located at the end of strands b-4
and b-7. This fold has also been predicted for Gel1p [19,20].
Fig. 3. Spectroscopical characterization of Gas1p. (A) Far-UV (200–250 nm) CD spectra of purified sGas1
523
-H at 20 °C(j), aft er heating at 80 °C
(d) and cooling again at 20 °C(s). [h]
MRW
, mean residue mass ellipticity. (B) Thermal unfolding of sGas1
523
-H in 50 m
M
sodium acetate buffer,
pH 5.5; changes in ellipticity at 220 nm were continuously monitored upon heating from 20 °Cto80°C. (C) Fluorescence emission spectra of
sGas1
523
-H. Spectrum 1 was obtained for excitation at 275 nm. Spectrum 2 ( tryptophan contribution) was obtained for excitation at 295 nm and
normalized at wavelengths above 380 nm. Spectrum 3 (tyrosine c ontribution ) was calculated as the difference spectrum (spectrum 1 minus spectrum
2). Fluorescence is expressed in arbitrary units. All the spectra were recorded at 25 °C and pH 5.5.
Table 1. Quantification o f disulphides and suphydryl groups in Gas1p.
The number of DTNB molecules per protein molecule as a mean of
three different determinations (SD < 10%) are shown in parenthe sis.
Endo H, endo-b-N-acetylglucosamin idase H.
Protein and condition Number of cysteines
sGas1
482
-H
Native conditions 1 (0.72)
After denaturation 4 (4.30)
After denaturation and reduction 14 (13.6)
Endo H-treated sGas1

482
-H
Native conditions 1 (0.71)
After denaturation 4 (3.92)
After denaturation and reduction 14 (14.4)
Ó FEBS 2004 Production and characterization of Gas1p (Eur. J. Biochem. 271) 3641
The c onserved glutamates correspond to residues E161
and E262 in Gas1p. In order to test the involvement of E161
and E262 in the active site of G as1p, we constructed two
mutant His-tagged forms of Gas1 – s Gas1E161Q-H and
sGas1E262Q-H (Fig. 1B). The mutant proteins were secre-
tedintheP. pastoris culture medium at the same level as
sGas1
523
-H, suggesting t hat they a re properly f olded
(Fig. 5A). To confirm t his, recombinant Gas1 proteins
were purified and their s ignature CD spectra were deter-
mined (Fig. 5B,C). No significant differences were detected
among the sp ectra of sGas1
523
, sGas1E161Q-H o r sGas-
1E262Q-H, demonstrating that sGas1
523
and the mutant
proteins assume the same folding (Fig. 5C). Next, we tested
whether the mutant proteins were enzymatically active.
Endo-b-(1,3)-glucanase activity was assayed by a reducing
sugar assay [4] and b-(1,3)-glucanosyltransferase activity by
an assay described in the Materials and methods. No endo-
b(1,3)-glucanase activity was observed for the two mutant

proteins (data not shown). Analysis of the b(1,3)-glucano-
syltransferase activity is shown i n F ig. 6. Using G
13
laminarioligosaccharide as a substrate, sGas1
523
-H pro-
duced smaller and larger oligosaccharides (Fig. 6A) that
corresponded to t he released and t ransferred products, a s
previously characterized [4], and confirmed the following
two-step enzyme activity:
E þ G
13
! E-G
x
þ G
13Àx
ð1Þ
E-G
x
þ G
13
! E þ G
13þx
ð2Þ
In contrast, with sGas1E161Q-H a nd sGas1E262Q-H, no
transferase activity was observed, indicating that both
glutamic acid residues are essential for th e two-step activity
(Fig. 6 B,C). I n a ddition, the sGas1
482
-H protein was

analyzed for activity before a nd after removal of the
N-linked chains by Endo H. The complete deglycosylation
occurred using both th e native and t he SDS-denaturated
Gas1 protein, suggesting that N-linked chains are com-
pletely accessible t o Endo H. The native deglycosylated
Fig. 5. Expression and purification of sGas1
523
and the mutant forms
sGas1E161Q and sGas1E262Q from Pichia pastoris culture super-
natants. (A) I mmunoblot with anti-Gas1p IgG of culture supernatants
(48 h from inductio n) from P. pastoris -transformed strains expressing
the His-tagged form s sGas1
523
, sGas1E161Q and sGas1E262Q. (B)
Representative purification of sGas1
523
-H. Silver staining of the dif-
ferent steps of the p urificatio n are shown. Lane 1, culture s upernatant;
lane 2, dialyzed fraction; lanes 3–5, three eluted fractions (E2, E3 and
E4). (C) CD spectra of His-tagged sGas1
523
(d), sGas1E161Q (n)and
sGas1E262Q (m).
Fig. 4. SDS/PAGE mobility of Gas1p under reducing and nonreducing
conditions. Non-reduced (nr) and reduced (r) samples of t otal extracts,
corresponding to % 60 lg of p rotein from W303-1B c ells (A) and
0.5 lg of the indic ated purified reco mbinant proteins (B), were loaded
on adjacent lanes and subjected to SDS/PAGE in a 7% polyacryl-
amide gel. Immunoblots obtained with anti-G as1p immunoglobulin
are shown. T he film in (A) was deliberately overexposed.

3642 C. Carotti et al.(Eur. J. Biochem. 271) Ó FEBS 2004
Gas1 protein was as active as the fully glycosylated form,
indicating that N-linked chains are not required for activity
(data not shown).
Essential role of C74 in the folding and stability
of the GPI-anchored Gas1p
To gain insight into the putative role played by the cysteine
residues in t he original GPI-anchored form of Gas1p, we
performed experiments based on site-dire cted mutagenesis.
We chose C74, the most N-terminal cysteine, the one next to
it (C103), and C265, which is c lose to the E262 catalytic
residue in the primary sequence. Each cysteine residue was
replaced with a serine. A gas1D mutant strain of S. cere-
visiae was transformed with a centromeric plasmid har-
bouring the wild-type GAS1 gene or the mutant constructs
(C74S, C103S and C265S, Fig. 1C). We reasoned that if
folded properly and localized correctly, the mutant proteins
should be able to fully complement the in vivo defects typical
of the mutant lacking Gas1p [21]. Upon microscopic
analysis, Gas1-C74S-expressing cells displayed the typical
gas1D abnormal morphology, whereas gas1D ce lls
expressing the mutant p roteins (Gas1-C103S or Gas1-
C265S) showed partial r eversal of the morphological
defects, and cells expressing the wild-type GAS1 exhibited
the normal morphology ( data not shown). Consistently with
these observations, the Gas1-C74S protein did not comple-
ment th e slow growth phenotype of gas1D cells, the
duplication time (T
d
)inSCmediumat30°Cbeing3.5h

either for cells expressing Gas1-C74S or for gas1 cells
Fig. 6. High-performance anion-exchange
chromatography (HPAEC) analysis of products
from the incubation of the recombinant
sGas1
523
p, sGas1E161Q-H and sGas1E262Q-
H mutant proteins with reduced laminari-
oligosaccharides. The recombinant purified
proteins were incubated wit h 3 m
M
reduced
laminarioligosaccharide of size G
13
,and
HPAEC analysis from samples taken at the
indicated time-points are shown. The size of
some of the major products is indica ted.
Fig. 7. C74 is essential for maturation of the glycosylphosphatidylinos-
itol (GPI)-anchored form of Gas1p in Saccharomyces cerevisiae.
(A) Immunoblot of tot al pro tein extracts from the gas1D strain of
S. cerevisiae transformed with the emp ty vector (YCplac33) or with
the same vector carrying the GAS1 ge ne or its mutant forms en coding
Gas1-C74S, Ga s1-C103S or Gas1-C265S. The filter was treated first
with anti-Gas1 I gG and, after stripping of the antibodies, was treated
with an ti-Pfk1p immunoglobulin to verify the l oading. T he arrow
indicates the position of the a-subunit. (B) Pulse-labelling and immu-
noprecipitation of cells harbouring the YCplac33-GAS1 or the YC-
plac33-GAS1-C74S plasmid. Cells were labelled for 7 min with
[

35
S]methionine and c hased fo r the indicated times before being
processed for immunoprecipitation w ith anti-Gas1p IgG .
Ó FEBS 2004 Production and characterization of Gas1p (Eur. J. Biochem. 271) 3643
transformed with t he empty v ector. Control gas1 cells
carrying the wild-type GAS1 gene showed a T
d
of 1.8 h. The
T
d
of cells expressing Gas1-C103S or Gas1-265S was 2.6 h,
indicating a partial complementation of the slow-growth
phenotype.
The total inability of Gas1-C74S and the partial ability
Gas1-C103S or GasC265-S proteins to rescue the gas1
phenotype could be primarily a result of alterations in
expression or folding of the mutant proteins. T o exclude
effects at t he mR NA level, a Northern blot analysis was
performed. Cells containing either the wild-type GAS1 gene
or the mutated alleles had comparable levels o f GAS 1
mRNA (data not shown). Therefore, the mutations do not
affect transcript accumulation.
The endoplasmic reticulum (ER) is the folding compart-
ment for proteins, such as Gas1p, destined for the plasma
membrane. F ailure to acquire a proper native conformation
leads to recognition b y the quality control machinery in the
ER, provoking retention and eventually degradation
[22,23]. Total protein from gas1D-transformed cells was
analysed by immunoblotting using anti-Gas1p immuno-
globulin. As shown in F ig. 7A, mature Gas1p (130 kDa)

was present in cells expressing the wild-type GAS1 and the
mutant Gas1-C103S or Gas1-C265S forms, while no
immunoreactive band corresponding to the mature protein
was detected in Gas1-C74S-expressing cells. H owever, a
faint band with the typical mobility of the ER 105 kDa
form was detectable in cells expressing Gas1-C74S. Cells
expressing Gas1-C103S or Gas1-C265S also showed an
accumulation of the E R form that w as not detectable in cells
expressing the wild-type Gas1p . These results suggest that
Gas1-C74S does not fold properly and is retained in the ER,
whereas th e processing of the Gas1-C103S a nd Gas1-C265S
precursors is slower than for wild-type Gas1p. To test, more
accurately, the effect of C74S replacement on Gas1p
maturation, a pulse–chase experiment was performed. Cells
expressing wild-type Gas1p or the Gas1-C74S mutant were
labelled for 7 min with [
35
S]methionine and the behaviour of
the l abelled forms was monitored at different time-points
after the chase (Fig. 7B). The E R form of Gas1p migrated
at 105 kDa, and t he mature f orm, after p rocessing in the
Golgi, migrated at 130 kDa (as determined by SDS/
PAGE), in agreement with the well-known maturation
process of Gas1p [24–26]. After a 7 min pulse, both the ER
form of Gas1p and of the mutant (C74S) w ere detectable at
equivalent l evels. During the chase, the maturation of
Gas1p proceeded and the mature form was already
detectable at 10 min. At 30 min all the p recursor had been
converted i nto t he mature protein. No maturation of the
mutant form occurred and the protein level progressively

decreased for 60 min when the immature form was almost
undetectable. This i ndicates that the G PI-anchored Gas1-
C74S is synthesized at the l evel of ER, but is not further
processed and undergoes degradation.
Conclusions
Proteins of Family GH72 play an active role in yeast and
fungal morphogenesis but, to date, no detailed biochemical
characterization has been reported. Previously, Gas1p has
proven to b e useful for studies on protein transport along
the secretory pathway in yeast. In this w ork, Gas1p w as
used as a model protein to undertake a first bioch emical
characterization of an enzyme of Family GH72. Here we
have shown that P. pastoris is a suitable host for the high-
level expression and secretion of Gas1p. The protein yield
was estimated to be about 100 times higher than for an
equivalent construct expressed in S. cerevisiae under the
control of the natural GAS1 promoter in a multicopy vector
(L. Popolo & M. Vai, unpu blished data). The purified
His-tagged sGas1 is active and could be used for future
structural characterization o f the protein. Our results
suggest that the high-level expression of secreted forms of
proteins in P. pa storis could constitute a valuable tool in the
study of fungal and p lant GPI proteins involved in c ell wall
biogenesis.
Multiple sequence alignment performed on t he 40
members o f F amily GH72 suggested t hat glutamates 161
and 262 in Gas1p could play the role of catalytic residues.
These predictions have been supported experimentally
because the replacement of these residues with glutamine
abolished the activity of Gas1p, whereas no significant

changes in the protein conformations were detected by CD
spectroscopy analysis. These results extend those obtained
for Gel1p of A. fumigatus, in which the replacement of E160
and E 261 with a Leu and a Phe, respectively, totally
abolished the in vitro glucanosyltransferase activity, and for
C. albicans where a glutamine substitution at position E169
or E270 yielded a Phr1 protein that was not able to
complement in vivo the morphological defects o f a phr1D
mutant [19,27]. These results could be valuable in the design
of inhibitors for b(1,3)-glucanosyltransferase activity, and
Gas1p could be considered a potential interesting molecular
target for the development of new antifungal agents.
Purified sGas1p was found to be quite resistant to heat
treatment. The structural changes induced in the protein
after h eating to 80 °C were totally reversible upon cooling
to 20 °C. This is a structural feature typical of proteins with
disulphide bridges. He re we have demonstrated that 10 out
of the 14 cysteines present in Gas1p are engaged in
intrachain disulphide bridges. The presence of disulphide
bonds in Gas1p is also consistent with a study on the effects
of reducing conditions on the transport of Gas1p from the
ER to the G olgi, w hich had indirectly revealed the
requirement of at least one disulphide bond for the exit of
Gas1p from the ER [25].
Multiple alignment indic ated that five cysteines are
conserved in the catalytic domain of members of Family
GH72. No data are available, to date, on the role of these
residues. Here we have shown that C74 of the catalytic
domain is crucial for the proper folding of Gas1p. C74 is the
most N-terminal cysteine in Gas1p and could be required

for non-native disulphide bond formation during folding
pathway, as has been suggested to occur for some
mammalian multidomain proteins [28], or for the sequential
and independent fold ing of the single domains – assuming
a vectorial mode of folding for Gas1p. In any case, the
replacement of C74 with a serine c ould e ntrap the protein in
a folding intermediate that is no longer processed and is
degraded. F rand & Kaiser reported that the ER form of
Gas1p accumulates in cells treated with dithiothreitol or in
cells defective in Ero1p, and is stable [25]. Our finding, that
the Gas1-C74S protein is unstable, is not necessarily in
contrast with their results. In the presence of the reducing
3644 C. Carotti et al.(Eur. J. Biochem. 271) Ó FEBS 2004
agent, or in an ero1 mutant, all disulphide bonds are
affected and probably present in a reduced state. Moreover,
both conditions are rather harsh and could a lso affect the
degradation pathway.
Given this prominent role of C74 in directing the folding
of Gas1p, it is premature to assess whether it is involved in
the formation of a structurally important disulphide bond,
and more direct analysis, such as mapping of disulphide
bridges, will be necessary to address t his point.
Acknowledgements
We wish to thank Michel Monod and Maria Antonietta Van oni for
plasmids and helpful suggestions, Carmela Gissi for the multiple
alignment, Prof. Gavilanes for helpful comments on spectral analysis,
Serena Crotti for technical assistance, David Horner for English
revision and Antonio Grippo for preparing the figures. This work was
partially supported by EU proje ct No. QLK3-CT-20009-01537
ÔEUROCE LLWALLÕ to L.P., and FIR ST 2001 and 20 02 grant to

L.P. C. Carotti was the recipient of a fellowship financed by the same
EU project.
References
1. Klis, F.M. (1994) Review: cell wall assembly in yeast. Yeast 10,
851–869.
2. Popolo, L. & Vai, M. (1999) The Gas1 glycoprotein, a putative
wall polymer cross-linker. Biochim. Biophys. Acta 1426, 385–400.
3. Mouyna, I., Fontaine, T., Vai, M., Monod, M., Fonzi, W .A.,
Diaquin, M., Popolo, L ., Hartland, R.P. & Latge, J.P. (2000)
Glycosylphosphatidylinositol-a nchored glucanosyltransferases
play an active role in the biosynthesis of the fungal cell wall.
J. Biol. Chem. 275, 14882–14889.
4. Hartland, R.P., Fontaine, T., Debeaupuis, J.P., Simenel, C.,
Delepierre, M. & Latge, J.P. (1996) A novel beta-(1-3)-glucano-
syltransferase from the c ell wall of Aspergillus fumigatus. J. Biol.
Chem. 271, 26843–26849.
5. Zechel, D .L. & Withers, S.G. (2000) Glycosidase me chanisms:
anatomy of a finely tuned c atalyst. Acc. Ch em. Res. 33, 11–18.
6. Chen, L., Fincher, G.B. & Hoj, P.B. (1993) Evolution of poly-
saccharide hydrolase subst rate specificity. Catalytic amino ac ids
are conserved in barley 1 ,3–1,4- and 1,3-beta-glucanases. J. Biol.
Chem. 268, 13318–13326.
7. Varghese, J.N., Garrett, T.P., Colman, P.M., Che n, L ., Ho j, P.B.
& Fincher, G.B. (1994) Three-dimensional structures of two plant
beta-glucan endo hydrolases with distinct substrate specificities.
Proc.NatlAcad.Sci.USA91, 2785–2789.
8. Davies, G.J., Mackenzie, L., Varrot, A., Dauter, M., Brzozowski,
A.M., Schulein, M. & Withers, S.G. (1998) Snapshots along an
enzymatic reaction co ordinat e: analysis of a r etain ing beta-gly co-
side hydrolase. Biochemistry 37 , 11707–11713.

9. Vai,M.,Gatti,E.,Lacana,E.,Popolo,L.&Alberghina,L.(1991)
Isolation and deduced amino acid sequence of the gene encoding
gp115, a yeast glycophosph olipid-anchored pro tein containing a
serine-rich region. J. Biol. Chem. 266, 12242–12248.
10. Horton, R.M., Hunt, H.D., Ho, S.N., Pullen, J.K. & Pease, L.R.
(1989) E ngineer ing hybrid g enes without t he use o f restriction
enzymes: gene splicing by overlap extension. Gene 77, 61–68.
11. Valdivieso, M.H., Ferrario, L., Vai, M., Duran, A. & Popolo, L.
(2000) Chitin synthesis in a gas1 mutant of Saccharomyces cere-
visiae. J. Bacteriol. 182, 4752–4757.
12. Braakman, I., Hoover-Litty, H., Wagner, K.R. & Helenius, A.
(1991) Folding of influenza hemagglutinin in the endoplasmic
reticulum. J. Cell. Biol. 114, 401–411.
13. Gatti, E ., Popolo, L., Vai, M., Rota, N. & Alberghina, L. (1994)
O-linked oligosaccharides in yeast glycosyl phosphatidylinositol-
anchored protein g p115 are clustered in a serine-rich region not
essential for its function. J. Biol. Chem. 269, 19695–19700.
14. Watanabe, R., Funato, K., Venkataraman, K., Futerman, A.H. &
Riezman, H. (2002) Sphingolipids are required for the stable
membrane association of g lycosylphosphatidylinositol-anchored
proteins in yeast. J. Biol. Chem. 277, 49538–49544.
15. Batanero, E., Villalba, M., Ledesma, A., Puente, X.S. & Rodriguez,
R. (1996) Ole e 3, an olive-tree allergen, belongs to a widespread
family of pollen proteins. Eur. J. Biochem. 241 , 772–778.
16. Tedeschi, G., Oungre, E., Mortarino, M., Negri, A., Maffeo, G. &
Ronchi, S. (2000) Purific atio n and primary structure of a n ew
bovine spermadhesin . Eur. J. Biochem. 26 7, 6175–6179.
17. Trimble, R.B., Atkinson, P.H., Tsc hopp, J.F., Townsend, R.R. &
Maley, F. (1991) Structure of oligosaccharides on Saccharomyces
SUC2 invertase secreted by the methylotrophic yeast Pichia pas-

toris. J. Biol . Chem . 266, 22807–22817.
18. Perczel, A., Hollosi, M., Tusnady, G. & Fasman, G.D. (1991)
Convex constraint analysis: a natural deconvolution o f circular
dichroism curves of proteins. Protein Eng. 4, 669–679.
19. Mouyna, I., Monod, M., F ontaine, T., Henrissat, B., Lechenne, B.
& Latge, J .P. (2000) Identification of the catalytic residues of the
first family of beta (1-3) glucanosyltransferases identified in fungi.
Biochem. J. 347, 741–747.
20. Henrissat, B., Callebaut, I., Fabrega, S., Lehn, P., Mornon, J.P. &
Davies, G. (1995) Conserved catalytic machinery and the predic-
tion of a common fold f or seve ral f amilies of g lycosyl hydrolases.
Proc.NatlAcad.Sci.USA92, 7090–7094.
21. Popolo, L., Vai, M., Gatti, E., Porello, S., Bonfante, P., Balestrini,
R. & Alberghina, L. (1 993) Physiological analysis of m utants
indicates involvement of the Saccharomyces cerevisiae GPI-
anchored prote i n gp115 in m orph ogenesis and cell separation.
J. Bacteriol. 175, 1879–1885.
22. Kostova, Z. & Wolf, D.H. (2003) New EMBO member’s review:
for whom the bell tolls: p rotein quality control of the endoplasmic
reticulum and the ubiquitin-proteasome connection . EMBO J. 22,
2309–2317.
23. Liu, C.Y. & Kaufman, J.R. (2003) The unfolded protein response.
J. Cell Sci. 116, 1861–1862.
24. Popolo, L., Vai, M. & Alberghina, L. (1986) Identification of a
glycoprotein involved in cell cycle progression in yeast. J. Biol.
Chem. 261, 3479–3482.
25. Frand, A.R. & Kaiser, C.A. (1998) The ERO1 gene of yeast is
required for oxidation of protein dithiols in the endoplasmic
reticulum. Mol. Cell 1, 161–170.
26. Doering, T.L. & Schekman, R. (1996) GP I anchor attachment is

required for Gas1p transport from the endoplasmic reticulum in
COP II vesicles. EMBO J. 15, 182–191.
27. Fonzi, W.A. (1999) PHR1 and PHR2 o f Candida albicans encode
putative glycosidases required for proper cross-linking of b eta-
1,3- and beta-1,6-glucans. J. Bacteriol. 181, 7070–7079.
28. Jansens, A., van Duijn, E. & Braakman, I. (2002) Coordinated
nonvectorial folding in a newly synthesized multidomain protein.
Science 298, 2401–2403.
Supplementary material
The following material is available from http://www.
blackwellpublishing.com/products/journals/suppmat/EJB/
EJB4297/EJB4297sm.htm
Fig. S1. Multiple sequence alignment among the catalytic
domains of Family GH 72 members.
Ó FEBS 2004 Production and characterization of Gas1p (Eur. J. Biochem. 271) 3645