Gene, 119 (1992) 183-190 0 1992 Elsevier Science Publishers

GENE

B.V. All rights reserved.

183

0378-l 119/92/$05.00

06664

Isolation

and

Agaricus

bisporus

(Mushroom;

characterization

cellulase;

basidiomycete;

of a cellulose-growth-specific

recombinant

DNA;

codon bias; introns;

multifunctional

gene

domains;

from

cellulose-binding

domain)

S. Raguz a*, E. Yagiie a, D.A. Woodb

and C.F. Thurston”

a Division of Biosphere Sciences, King’s College London, London W8 7AH. UK: and ’ Horticulture Research International, Worthing Rd., Littlehampton, West Sussex BNI 7 6LP, UK. Tel. (44-903) 716-123 Received by J.K.C

Knowles:

11 March

1992; Revised/Accepted:

2 June/3

June 1992; Received

at publishers:

11 June 1992

SUMMARY

The edible basidiomycete, Agaricus bisporus, produces extracellular endoglucanase. Endoglucanase production is induced by cellulose and repressed by fructose in A. bisporus grown on minimal medium, and is regulated in activity during fruiting body development. An anti-endoglucanase antibody was used to isolate cellulase-related genes. Three main polypeptides of 38, 58, and 60 kDa were immunoprecipitated by the antibody from products of in vitro cell-free translation of mRNAs isolated from cellulose-grown mycelium. No cross-reaction was detected with the translated products from fructose-grown mycelium. This antibody was used to immunoscreen a AZAPII-cDNA expression library made from mRNA isolated from cellulose-grown mycelium. Two cDNA cross-reacting clones, pSRcll0 and pSRc200, were isolated. Clones pSRc 110 and pSRc200 cross-hybridized and had the same restriction map. Clone pSRc200 hybrid selected an mRNA that on cell-free translation produced a 3%kDa polypeptide. The cDNA fragment from pSRc200 hybridized to a 1.3-kb mRNA from cellulose-grown mycelium. No hybridization was observed when using fructose-grown mycelium mRNA. Thus, the gene (cell) expressing the 1.3-kb mRNA, is differentially regulated by the carbon source of the culture medium. The cell gene was isolated in a 8.9-kb EcoRI genomic fragment after hybridization to pSRc200. Sequences similar to those in the egll and cbh2 genes from Trichoderma reesei were found upstream from the ATG start codon in cell. Nine short intervening sequences disrupt the cell coding sequence, and a strong bias against codons ending with G and A was observed. CELl (protein encoded by cell) showed a primary structure in which four putative functional domains were recognized: a predicted 29-amino-acid (aa) signal peptide, a core of 233 aa, a Pro-Ser-Thr-rich domain of 22 aa, and a C-terminal 36-aa cellulose-binding domain similar to those found in other fungal cellulolytic enzymes. No homology was observed between the CELI core and any /?-glycanase sequences described to date.

Correspondence to: Dr. E. Yagtte, Division College London,

Campden

Tel. (44-71)3334358; *Present

address:

of Biosphere

Hill Rd., London

Sciences,

King’s

W8 7AH, UK.

Fax(44-71)9375396. Gene Structure

tional Institute of Medical UK. Tel. (44-81)9593666,

and Expression

Research, The Ridgeway, ext. 2111.

hardt’s

(100

x ),

dNTP,

2% bovine serum albumin/2”i, deoxyribonucleoside

London

The NaNW7 IAA,

1000 bp; N, any nucleoside oligodeoxyribonucleotide; PCR,

A., Agaricus; aa, amino acid(s); bp, base pair(s); Bq, Bec-

querel; CAMP, cyclic AMP; cDNA, DNA complementary to RNA; ccl, cellulose-growth-specific gene; CEL, protein encoded by a ccl gene; Den-

polymerase acid];

chain poly(A)’

acid;

(in the sequence); PAGE,

reaction; RNA,

Ficoll/2%

triphosphate;

hydroxyethylpiperazine-N’-2-ethanesulfonic Laboratory,

sulfonic Abbreviations:

solution

vinylpyrrolidone;

kb,

kilobase

or

nt, nucleotide(s);

oligo,

polyacrylamide-gel Pipes,

electrophoresis;

piperazine-N,N’-bis[2-ethane-

polyadenylated

dodecyl sulfate; SSC, 0.15 M NaCl/lS

poly-

Hepes, N-2-

mRNA;

mM Na,.citrate

(w/v) tri-isopropyl naphthalene sulfonate/6% 10 mM Tris.HCl pH 7.4; u, unit(s).

SDS,

sodium

pH 7.6; TNS, 1%

(w/v) p-amino

salicylate/

184 INTRODUCTION

RESULTS

The basidiomycete Agaricus bisporus (the commercial button mushroom) is able to grow on microcrystalline cellulose as the sole carbon source (Manning and Wood, 1983). The complete enzymology of the conversion of cellulose into soluble sugars is not yet known, but the requirement for at least endo- and exo-glucanases acting synergistically on the cellulose microfibril is generally agreed

(a) Immunoprecipitation of cell-free translation products An endoglucanase fraction, showing five activity bands after Congo Red staining of non-denaturing PAGE zymograms, was isolated from A. bisporus D649 culture filtrates and used to raise a polyclonal antibody (Wood and Thurston, 1991). The anti-endoglucanase antibody reacted with several A. bisporus extracellular polypeptides of 75, 72, 54, 42 and 20 kDa from cellulose cultures but not from fructose cultures (results not shown). When the endocellulase antibody was used to immunoprecipitate cell-free translation products from cellulose-grown mycelium poly(A)’

(Coughlan and Ljungdahl, 1988; Beguin, 1990). Endoglucanase, /I-glucosidase, cellobiose oxidase, and filter paper hydrolyzing activities have been detected in A. bisporus cellulose-grown culture filtrates (Wood et al., 1988). Endoglucanase production is regulated by the carbon source in the culture medium. Cellulose, either amorphous or crystalline, acts as the best inducer. Moreover, endoglucanase activity is not found in cultures with fructose or lactose as the sole carbon source, and added glucose and cellobiose repress the increase in activity in carboxymethylcellulosegrown cultures (Manning and Wood, 1983). Therefore, the regulation of the A. bisporus cellulase system is very similar to those found in other fungi (Merivuori et al., 1985; Messner and Kubicek, 1991). Studies on cellulase production by A. bisporus on its commercial growth substrate, composted straw, have revealed that endoglucanase activity increases in parallel with the enlargement stage of sporophore development (Claydon et al., 1988). Other intracellular carbohydrate-metabolizing enzymes in the fruit bodies are also regulated both in-phase and out-of-phase with sporophore development. A model to explain the pattern of regulation of carbonmetabolizing enzymes has been proposed (Hammond, 1985; Wells et al., 1987; Wood et al., 1988). Despite the economic importance of mushroom cultivation (Spencer, 1985) the genetics of this organism is poorly understood (Elliott, 1985). This lack of knowledge and a complicated life cycle provide little opportunity for improvement by classical genetic and selection approaches. The use of recombinant DNA technology has been proposed for the development of strain improvement programmes but, to date, all attempts to transform A. bisporus have been unsuccessful. The isolation and characterization of functional genes are prerequisites for use in transformation strategies involving homologous recombination (Fincham, 1989). A glyceraldehyde-3-phosphate dehydrogenase gene recently isolated from A. bisporus (Harmsen et al., 1991) could be such a candidate, but no further characterization has yet been reported. In order to gain knowledge on the genetic regulation of cellulose utilization, and to pave the way for a transformation system, we decided to isolate cellulose-growth-specific genes from A. bisporus. In this paper we describe the isolation, regulation, and characterization of the cell gene.

AND DISCUSSION

RNA, three main polypeptides of 38, 58 and 60 kDa were observed (Fig. 1). None of these polypeptides were observed when using uninduced (fructose-grown) RNA or the preimmune serum. Furthermore, these polypeptides were present in all ages of culture assayed (from 7- to 22-dayold mycelium; results not shown); this could be due to the stability of these particular messengers or to continued de novo transcription. The additional band of 47 kDa was an artifact due to the labelling of proteins by an mRNAindependent process (Pelham and Jackson, 1976). The fact that the anti-endoglucanase antibody immunoprecipitated cell-free translation products from cellulose- and not from fructose-grown mycelium RNA suggested that the antibody was suitable for immunoscreening cDNA expression libraries. (b) Construction of a cDNA library A cDNA library of 1.4 x lo6 recombinants, made from poly(A)+ RNA isolated from 15-day-old cellulose-grown mycelium, was constructed in the vector AZAP II and immunoscreened with the Escherichia coli pre-adsorbed antiendoglucanase antibody. Detection of cross-reacting plaques on primary screening was performed with [ ‘251]Protein A, whereas for the secondary and tertiary screenings, an alkaline phosphatase-conjugated second antibody was used. Two positive phage 2 plaques were isolated and rescued as pBluescript recombinants named pSRcll0 and pSRc200. Restriction mapping and cross-hybridization under high stringency conditions showed that pSRcll0 and pSRc200 contained the same l.l-kb cDNA fragment. (c) Hybrid selection and in vitro translation of pSRcZOOselected mRNAs To confirm the identity of the ccl cDNA, mRNAs were selected from the total mRNA population by hybridization and translated in vitro (Fig. 2). The pSRc200-selected mRNA translated a single 38-kDa polypeptide that was not present in the control experiments. As in Fig. 1 the artifactual band of 47 kDa, and additional bands of 2027 kDa were seen in control experiments in which no DNA

185 EN

P

-69

69

6C

.46

50

38,46

38

.30

,30

.21

21

Fig. 2. Hybrid Fig.

1. Immunoprecipitation

of

cell-free

translation

poly(A)+ RNA from A. bisporus. Fungal mycelium minimal medium (Manning CC4 1 microcrystalline hum was harvested by the method

products

was grown on defined

and Wood, 1983) with either 0.05 y0 Whatman

cellulose or 0.1% D-fructose. and frozen in liquid nitrogen.

of Leonard

After 15 days myce-

Total RNA was isolated

et al. (198 1) modified as follows: frozen myce-

hum was blended with solid CO, in a coffee blender.

Powdered

was mixed

and

with TNS

buffer

phenol-cresokchloroform, Na,acetate

(2 ml/g wet weight)

ethanol-precipitated,

10 min at 4°C. RNA was dissolved modifications

NaCI/lO

SDS (instead

spectively)].

in a rabbit

mM

(lanes C) or fructose

to the manufacturer’s

and SDS, re-

cell-free system (AmerRNA from cultures grown reticulocyte

After 90 min at 30°C

incubated

SDS

at 37°C for 10 min. After

(v/v) Triton X-100) and 5 pl of crude rabbit antiserum

anti-endoglucanase; with gentle shaking. Protein

A-agarose

P, pre-immune) After removing (Sigma):binding

were added and incubated

(EN,

overnight

aggregates

by centrifugation

15 ~1 of

buffer (l:l,

v/v) were added

and in-

cubated for 3 h at 37°C with gentle rocking. Agarose beads were washed three times with 50 mM Tris~HCI/lSO mM NaC1/5 mM EDTA, pH 7.4/

1mM benzamidine/0.10/, Triton

Triton

and SDS were omitted).

10 min in 25 ~1 of SDS-gel 10% PAGE (Laemmli, in methanol:acetic

X-100/0.02% Polypeptides

loading

SDS (in the last wash were eluted by boiling for

buffer, and analyzed

1970). After electrophoresis,

acid:water

bated in 1 M Na,salicylate to x-ray film. Molecular

(40:10:50,

by 0.1%

SDS-

the gel was incubated

v/v) for 15 min and then incu-

(Chamberlain, 1979), dehydrated and exposed size (kDa) of “C-1abelled Rainbow markers

(Amersham) and immunoprecipitated right and left margins, respectively.

incubated

filters

were

Pipes.NaOH/SO ten-day-old washed

mM EDTA

cellulose-grown

sequentially

et al. (1982). For selecting mRNAs,

in 65%

formamide/0.4

mycelium

for 3 h at 50°C.

in 1 x SSC/O.5%

in 1 mM EDTA pH 7.5 at 60°C. and analyzed

Fig. 1. Controls with pBluescript Molecular

show in vitro translation (lane B) and no DNA

size (kDa) of “C-labelled products

and left margins,

respectively.

NaCI/lO mM from

Filters were then

SDS, in 2 mM EDTA pH 7.5, and

Selected mRNAs

on 0.1 y0 SDS-lo%

in vitro translation

M

pH 6.4 with 5 pg of poly(A)‘RNA

translated

diluting the reaction to 2% SDS, four volumes of antibody binding buffer (50 mM Tris,HC1/190 mM NaC1/6 mM EDTA pH 7.4/l mM benzamidine/2.5%

following Maniatis

the

Lin-

were eluted from filwith ethanol,

PAGE,

in vitro

as described

in

products

from hybrid selections

bound

to the filters (lane C).

Rainbow

of selected mRNAs

markers

(Amersham)

are indicated

and

on the right

mation that the recombinant isolated from the cDNA brary was derived from ccl mRNA from A. bisporus.

li-

in a total volume of 50 nl ac-

instructions.

was added to 4% and the reaction

filters (Amersham)

of selected mRNAs.

and spotted onto nitrocellulose

3M

Hepes,NaOH/0.5%

(lanes F), 25 ~‘1 of rabbit

lysate and 1.8 x 1O’Bq t_-[‘5S]methionine

(lane A) was denatured

ters by boiling in water for 1 min and precipitated

pH 7.4, and

of LiCI, Tris.HCI,

reticulocyte

sham) was carried out using 2.5 pg of poly(A)’

cording

in

selection and in vitro translation

earized pSRc200

with

at 5000 x g for

in 10 mM Hepes.NaOH

N-lauroylsarcosine/O.5”/,

on cellulose

extracted

isolated following Aviv and Leder (1972) with [0.4 M

Translation

mycelium

resuspended

pH 5.0 (0.5 ml/g wet weight), and centrifuged

the poly(A) + RNA fraction slight

from

polypeptides

are indicated

on the

was bound to the nitrocellulose filters (see section a). The presence of the same 38-kDa polypeptide in both immunoprecipitation and hybrid selection was taken as confir-

(d) Expression of the cell gene To determine whether the cDNA was derived from cellulose-grown mycelium mRNA and whether the size to generate the polypeptides seen on immunoprecipitation experiments was correct, Northern analysis was carried out (Fig. 3). The l.l-kb recombinant fragment from pSRc200 hybridized to a 1.3-kb mRNA isolated from cellulose- but not from fructose- or malt extract-grown mycelium. The gene expressing this mRNA was hereafter named cell. Expression of genes coding for enzymes involved in cellulose degradation occurs at the level of transcription (El-Gogary et al., 1989; Beguin, 1990). The results presented here suggest that this is also the case for cell. Because the gene is switched on or off by defined controlled conditions in the culture medium its regulatory regions are potentially useful in future transformation strategies. (e) Southern analysis and construction of Agaricus bisporus partial genomic DNA library Southern analysis of the l.l-kb fragment from pSRc200 showed hybridization to a single 8.9-kb EcoRI-digested

186 A C

genomic DNA from A. bisporus (results not shown). Therefore, a partial genomic DNA library was constructed (see

0

FM

C

F

M

legend to Fig. 4) and a clone, pSRg100, carrying the 8.9-kb EcoRI fragment was isolated. The probability of containing the whole of the cell gene within this clone was high because the l.l-kb fragment from pSRc200 mapped to the middle of pSRglO0 (data not shown). (f) Structure of the cell gene and architecture of the putative encoded protein Sequence analysis showed that both pSRcll0 and pSRc200 were truncated (Fig. 4). In order to complete the sequence data, clone pEYclO0 was constructed from

Fig. 3. Regulation

of the cell gene expression

RNA from A. bispoms mycelium ent carbon

sources

by the carbon

grown on minimal

(C, cellulose;

F, fructose)

EcoRI

fragment

belled (Boehringer)

membranes

from pSRc200

using [ a-32P]dCTP

with differ-

RNAs were analyzed

in 1% agarose gels as described

Gels were blotted onto Hybond-N The l.l-kb

source. Total

medium

and on 2% malt extract

broth (M) was isolated as in Fig. 1. Glyoxylated electrophoresis

cDNA by using the PCR primed with the oligos shown in Fig. 4 following the approach of Saloheimo et al. (1991). The ATG start codon was deduced because it was preceded by an in-frame stop codon. The G+C content of the fragment shown in Fig. 4 was 46.6% which agrees with that calculated for nuclear DNA from A. bispoms, 43.5% (Arthur et al., 1982). The G+C content of the cell coding exons was 50.9%, whereas the intervening sequences have an average of 36.5 %. The nine introns interrupting the cell coding sequence (Fig. 4) were 49-80 bp long, and a consensus GT,NNGz for the 5’and FAG for the 3’-splicing sites was observed. An internal consensus sequence CTNA close to the 3’-splice junction was found in eight out of the nine introns. The similarity with other fungal exon/intron splice junctions (Ballance, 1986) and dissimilarity with that of higher eukaryotes (Mount, 1982) raises the question whether the fungal splicing mechanism may differ from that of other

by

in Sambrook

et al. (1989).

(Amersham)

in 20 x SSC.

(A) was random-primer

and hybridized

la-

to the membranes

(hybridization

solution: formamide/ x SSCj5 x Denhardt’s 50% solution/O.1 % SDS) at 42°C for 12 h. Membranes were washed sequen-

tially in 1 x SSC/O.l%

SDS at room temperature

0.1% SDS at 42”C, dried and autoradiographed shows hybridization cDNA

library

to a noncellulose

(Fig. 4). The

(HindHI-digested

fragments

and then in 0.2 x SSC/ at -70°C.

Control

specific EcoRI fragment

positions

of glyoxylated

of 1. DNA)

are indicated

size

(B)

from the markers

in kb on the right

margin.

Fig. 4. Nucleotide show similarity

sequence

(identical

and TATA sequences sequences.

of the cell gene and deduced

nt underlined)

are double underlined.

The extent of cDNA

aa sequence

with regions in the promoter The presence

clones is indicated

of introns

with arrows.

of the CELl

protein.

Boxes covering regions between nt 218-233

of cbh2 and egll genes from T. reesei, respectively (lower-case

Deduced

letters) was deduced

aa sequence

of the CELl

from the comparison protein

is indicated

and nt 370-390

(see section f). Putative of the cDNA

CAAT

and genomic

below their respective

codons.

Numbers

indicate either nt from the beginning of the sequence (last digits are aligned with the corresponding nt) or aa from the start Met. The most important region is underrestriction enzyme recognition sites are marked. The arrow after Ala 29 shows the putative signal peptide cleavage. A Pro-Ser-Thr-rich lined. The putative cellulose-binding domain is indicated by upper-case letters. Boxed aa are conserved in all the fungal cellulose-binding section g). The sequence data reported in this paper will appear in the EMBL/GenBank/DDBJ nt sequence databases under accession Methods:

Sequence

analysis

was carried

out with IBI Pustell and the sequencing

‘GCG’

software

(Devereux

et al., 1984). The mRNA

domains (see No. M86356.

from 15-day-old

cellulose-grown mycelium was used to prepare a cDNA library in lZAP II (Stratagene, La Jolla, CA) using a Pharmacia cDNA synthesis kit and EcoRI-Not1 adaptors. The library was immunoscreened with E. coli pre-absorbed anti-endoglucanase antibody (diluted 1:200) and either [ ‘2SI]Protein A (Amersham) obtained

or an alkaline phosphatase-labelled

from immunocrossreacting

by the PCR as follows: cDNA kinased cDNA/2

second

antibody

(Sigma) (Sambrook

et al., 1989). Clones pSRcll0

and pSRc200

was made as above but using an Amersham

cDNA

synthesis

kit. Reaction

cocktail

contained

were

was constructed

0.24 mM dNTP/0.25

PM

chemically synthesized oligos (underlined nt)/50 mM KCljlO mM Tris.HCl pH 9.0/1.5 mM MgCl,/O.Ol% gelatin/O.1 % Triton X-100/7 ng u of Taq DNA polymerase (Promega, Madison, WI) in a total volume of 25 ~1. After 35 cycles (one cycle: 95 “C for 1 min, 45’ C for 1 min, 70’ C

for 1 min) 4 u of T4 DNA polymerase with glass beads (Geneclean, Broda,

goat anti-rabbit

plaques by in vivo excision and rescued using R408 helper phage (Short et al., 1988). Clone pEYclO0

BiolOl)

were added

and incubated

at 37°C for 15 min. A 200-bp band was isolated

and cloned into the SmaI site of pBluescript

1985) was depleted in mitochondrial

DNA by a bisbenzimide

KS + (Stratagene).

(0.12 mg/ml) density gradient

from an agarose

Mycelial DNA (isolated

centrifugation

(50000

x g

gel, DNA purified

according

to Raeder

and

for 24 h at 15°C). A genomic

DNA library of EcoRI-digested fragments ranging from 9 to 1 kb was constructed in the %ZAP II vector and screened with pSRc200 as a probe at high stringency (hybridization in 5 x SSCj5 x Denhardt’s solution/0.5% SDS, 65°C; washings with 1 x and 0.1 x SSC/O.l% SDS, 65°C). Plasmid rescue was as above. DNA fragments were subcloned into either pBluescript KS + or M13mp18 (Norrander et al., 1983). The nt sequence was obtained using the method of Sanger et al. (1977) on either single- or double-stranded Chemically

synthesized

oligos were used when necessary.

templates

(Mead et al., 1986; Mierendorf

and Pfeffer, 1987; Saunders

and Burke, 1990).

187 eukaryotic

organisms.

interrupted and Wessels,

Other basidiomycete

by a great number 1990; Saloheimo

genes are also

of short introns

(Schuren

et al., 1991) while the gap

gene from Ustilago maydis is interrupted by only one of 407 bp (Smith and Leong, 1990). As with the GPD genes from A. bisporus (M.C. Harmsen, pers. commun.), the first

nlJl.Gu.sa11 50 GTCGACTCTA CCCCTAACCT ACTAATAGAC GTCCCGGTGT GACCACCGAT CTTCATTGTC 150 TACCTAATCT CGGATEG CCTGGTAAAC TTTCAGCAGT AACCCCATGG TTTGATCTAC 250 300 GGAATGAACA AGCGCCATAC CCAGCTCAAA AGGCATGCCA CTGAACAGAC ACCATGTAGG 400 TTTGACTGAA GTTCCAGAGT TTCCGCGCCA 500 ATTCCAGGCA CACAGATGCT TCTCCGGGGC CGATTCCCTG TTATTTCACC GAGACCGTCC

intron in cell is positioned

within the three first bp of cod-

ing sequence. A fragment of 21 bp at the 5 ‘-noncoding region of cell had a 76% homology to one found at the same region in the egll gene (coding for endoglucanase I) of T. reesei (Van Arsdell et al., 1987). Another different fragment of 16 bp

100 GTGGTTAGCT CGTTCCTCTT TTGTTTCGGA GCGAAATTGT TCCGACGTAG 200 CACATGACTT CCCCACTGAC GTGCCGCTGG AATTTA @=%FATTCT AGCGATCAGC CGTTCACGCA TACTTaG GGGAATACGC GACTGGTGCT 4.50 TCTAGACTT TGT TGTCAAAATC ACAGAGGTGC 550 GCCACGATGA TCGCTC AT gtatgttttt ataggctatg atattttcat

GTTGGCCTCG

ATACTCTCTT CGCTTTGATG atcqctttct

650 pSRcll0 700 D_wr Me ATG -~/ZZ%~"~TC gtcggtctat ttaaagccaa agctqaagag attttatcag G CGT CTT CCC TCT CGA CAA CAA GTA rCTC AAA TCA CTA m TTG GGA a t Arg Le" Pro Ser Arg Gin Gin Val Le" Lys Met Leu Ala Thr Phe Ser Leu Ala Leu Gly Le" 10 20 pEYclO0 750 800 U GCA AAA GTC CAR GCT CAC GGA GGC GTC ATC GGA TAT TCC TGG GAT GGT ACT TGG TAT GAG GGG TGG CAC CCA TAC AAC ACT CCC GTA GGC CAG Phe Ala Ala Lys Val Gln Ala His Gly Gly Val Ile Gly Tyr Ser Trp Asp Gly Thr Trp Tyr Glu Gly Trp His Pro Tyr Asn Thr Pro Val Gly Cln 50 40 t 30 850 900 ACC TCC ATC GAG CGG CCT TGG GCG ACT TT gttagtcgat tgcaacttct aacggaaatt aatgcttatt agtcgcttat gacag T GAT CCT ATC ATG GAT GCA ACC Thr Ser Ile Glu Arg Pro Trp Ala Thr Ph e Asp Pro Ile Net Asp Ala Thr 60 70 950 1000 GCA TCC ACA GTC GGC TGC AAT AAC GAT GGC AAT CCT GGC CCG AAC CM TTG ACC GCT ACA GTT GCT GCT GGA ACG GCT ATC ACT GCT TAC TGG AAC CAG Ala Ser Thr Val Gly Cys Asn Asn Asp Gly Asn Pro Gly Pro Asn Gin Leu Thr Ala Thr Val Ala Ala Gly Thr Ala Ile Thr Ala Tyr Trp Asn Gln 80 100 90 1050 1100 GTT TGG CCA CAT CCC TAC GGG CCA ATG gtatgcactg gacatttccc ttcatgcgaa cgagagctga tqggagcaca g ACT ACC TAC TTG GGT AAA TGT CCT GGA AGC Val Trp Pro His Pro Tyr Gly Pro Met Thr Thr Tyr Leu Gly Lys Cys Pro Gly Ser 110 120 1150 1200 TCT TGC GAC GGC GTT AAC ACG AAC TCG CTC AAA TGG gtaggctata ttctaaatca tataaaadag gtgtcctgtc taatadttct aacag TTC RAG ATT GAC GAA GCC Ser Cys Asp Gly Val Asn Thr Asn Ser Le” Lys Trp Phe Lys Ile ASD Glu Ala 130 14b 1250 1300 GGC CTC TTG TCC GGC ACT GTT GGA AAA GGT GTA TGG GGT TCT GGT AAA ATG ATT GAC CAA AAT AAT AGT TGG ACC ACA ACT ATT CCT ACT ACT GTC CCG Gly Leu Leu Ser Gly Thr Val Gly Lys Gly Val Trp Gly Ser Gly Lys Met Ile Asp Gln Asn Asn Ser Trp Thr Thr Thr Ile Pro Ser Thr Val Pro 150 160 170 1350 1400 AGT GGG GCT TAC ATG ATT CGC TTC GAG ACC ATT GCG CTG CAC TCC CTT CCT GCT gtaagtcgat ccgtgattct taaaagccgc cdaactctta ccaaagttgc gatag Ser Gly Ala Tyr Met Ile Arg Phe Glu Thr Ile Ala Leu His Ser Leu Pro Ala 180 190 1450 1500 CAA ATC TAC CCC GAA TGT GCT CAA CTT ACG ATT ACT GGT GGT GGA AAC CGG GCA CCT ACA TCC ACT GAA TTA GTT TCC TTC CCC GGT GGA TAT TCT AAC Gln Ile Tyr Pro Glu Cys Ala Gln Leu Thr Ile Thr Gly Gly Gly Asn Arg Ala Pro Thr Ser Ser Glu Leu Val Ser Phe Pro Gly Gly Tyr Ser Asn 200 220 210 1550 1600 AGT GAT CCT GGT C gtaagttatg cagcgcttat gttgaagctg gtatttttct aagtcgcgca ttcgttag TG ACC GTC ART CTC TAC ACT CAA GAG GCT ATG ACC GA Ser Asp Pro Gly L e" Thr Val Asn Leu Tyr Thr Gln Glu Ala Met Thr As 230 240 1650 1700 gtgagttgaa atcacaagga gcggtgccat taatccaatc tattacattt tag T ACC ACA TAC ATT GTT CCC GGA CCT CCG CTT TAT GGA TCT GGT GGG AAC GGT GGT p Thr Thr Tyr Ile Val Pro Gly Pro Pro Leu Tyr Gly Ser Gly Gly Asn Gly Gly 250 260 1750 1800 1850 SDnI gtaagtcaat aatgaacaat gtattgaact cactcacatt tattcatag TCA CCA ACT ACT ACT CCC CAT ACT ACT ACT CCC ATT ACT ACT AGT CCC CCG CCT ACC AGC Thr Pro Tie mr Thr SLr EreErp Pro T_ 270 280 1900 1950 ACT CCA GGA ACT ATT CCT CAA TAC GGT CAA TGT GGT GGT ATT GGC TGG ACT GGA GGC ACC GGC TGT GTC GCC CCT TAT CAA TGT AAA GTC ATC AAT G gca Thr Gly THR ILE PRO GLN TYR GLY GLN CYS GLY GLY ILE GLY TRP THRmGLY THR GLYmVAL ALA PRO TYR GLNHLYS VAL ILEIASN(A 300 290 310 2050 2000 agtatat agctcggaac tgttgattct gcaactaaat ccttatcag AT TAC TAC TCT CAG TGC CTC TAG CAGTGGTACA TATCGCTTGA TAAGAAAACG GTGTACAAGA GTATAAA SD TYRmSER[GLN=+; End

7

2100 TTT CGCGATACAC TGTAATGTAG AAGATTTTAT TATGTCCTGC CCTCAAATTT ACGAGAATAT pSRc200 TGTTTCTA

188 was similar (81%) to a region in the cbh2 gene (coding for cellobiohydrolase II) from T. reesei common to the complementary strand of the putative eukaryotic CAMPdependent transcriptional signal sequence (Chen et al., 1987). These three genes are all regulated by an induction/ repression mechanism, and CAMP has been implicated in the regulation of cellulase production (Wood et al., 1984) but in the absence of experimental data it remains uncertain as to whether or not these regions of homology represent transcription factor recognition sequences. Two sequences resembling TATA and CAAT promoter elements were found 231 nt and 420 nt, respectively, upstream from the start codon

(Fig. 4).

The codons ATA (Ile), GTG (Val), AGA (Arg), and AGG (Arg) are not used in cell. There is a strong bias against

codons

finishing

in A (17’1;) and G (15;;)

(Ta-

TABLE Codon usage of the cell gene from Aguricus hisporus aa

Codon

Phc LCLI

LtXl

Quantity

aa

Codon

uuu uuc

Tyr

UAU UAC

UUA

End

UAA

End

UAG

I

cuu

His

CAU

2

CAC

CUA

Gin

CUG AUU

10

AUC

6

AUA

0

Met

AUG

1

Val

GUU

6 7

ASP

GUC GUA

3

GlU

GUG

0

Ser

ucu

Asn LyS

Thr

Ala

CAA CAG

3 II 3

AAU

6

AAC

9

AAA

7

AAG

I

GAU GAC GAA GAG

cys

5

UGU UGC

ucc

Pro

5 II

UUG

cut

Ile

Quantity

3

UCA

End

UGA

-

UCG

Trp

UGG

IO

ecu

Arg

CGU

ccc

CCC

CCA

CGA

CCG

CGG

ACU

22

ACC A CA

11

Ser

AGU

6

AGC AGA

2 0

AGG

0

Arg

ACG

5 3

GCU

12

G1)

GGU

15

CCC

CCC CGA

11

GCA

3 4

GCG

2

GGG

12 4

ble I). In other temporally controlled, and highly expressed basidiomycete genes the codon utilization is strongly biased towards using codons that end with C (M.C. Harmsen, pers. commun.). However, in cell more codons end with T (37.5%) than with C (30.39,;). The cell gene codes for a protein of 320 aa with a calculated A4,. of 33752. The mass of the protein was lower than that estimated from SDS-PAGE after hybrid selection (Fig. 2). This might be due to the abnormal shape of the SDS-polypeptide complex caused by the high Pro content in the C-terminal part of CELl (See and Jackowski, 1989) as has also been found in cellobiohydrolase I from T. reesei (Shoemaker et al., 1983) and in the polypeptides coded by the SC genes from Schizophyllurn commune (Schuren and Wessels, 1990). A classical signal peptide, very similar to that of the lactase gene of Phlehin rcrdirrtrr (Saloheimo et al., 1991). was found at the N terminus of CELl. and cleavage after Ala29 was predicted (Von Heijne, 1986). A Pro, Ser and Thr-rich region of 22 aa preceded the putative fungal cellulose-binding domain at the C terminus, similar to those found in the endoglucanases and cellobiohydrolases from T. reesei (Teeri et al., 1987), and cellobiohydrolases from Phanerochaete chr,:wsporium (Sims ct al.. 1988) and Humicolu grisea (Azevedo et al.. 1990). No sites for N-glycosylation (Asn-Xaa-Ser/Thr) were found in the CELl deduced aa sequence. No homology was found between the CELl core (the 233-aa region between the signal peptide and the Pro-Ser-Thr-rich region) and any of the members of the families into which the B-1,4-glycanases have been recently classified (Gilkes et al., 1991). This kind of architecture in which the putative catalytic core and the binding domain are separated by a linker region is very common among the cellulascs, xylanases, and, in general, in the multifunctional domain enzymes (Gilkes et al., 1991). The enzymatic activity of CELl is not yet known. The conserved terminal region (comprising Pro,Ser and Thrrich spacer and cellulose-binding domain) of T. reexei cellulases forms a strong antigenic epitope for polyclonal antibodies. As a result, anti-endoglucanase antibodies recognize cellobiohydrolases and vice versa (Aho and Paloheimo, 1990). Although the translation products from pSRcll0 and pSRc200 cross-reacted with the antiendoglucanase antibody, the possibility that CELl is a cellobiohydrolase cannot be ruled out. Cloning and expression of T. reesei cellulase genes in S~rcchorom~~ces cereviskie have been used to produce enzymatically active endoglucanases and cellobiohydrolases (Teeri et al., 1987: Van Arsdell et al., 1987) and a similar approach is currently being used to determine the enzymatic activity of CEL 1. Nevertheless, the regulation of cell gene expression and the presence of the fungal cellulose-binding domain in CELl indicate that this protein is involved in cellulose degradation by A hisporus.

189 (g) Conclusions (I) Two cDNA

Chamberlain,

clones

derived

from the same mRNA

were isolated from an A. bisporus expression library by cross-reaction with an anti-endoglucanase antibody. (2) This cDNA hybridized to 1.3-kb mRNA expressed only in cellulose- but not in fructose- or malt extract-grown mycelium. (3) The gene cell, expressing the 1.3-kb mRNA, was isolated from a partial A. bisporus genomic library. Two regions similar to those in the promoters of egll and chh2 genes from T. reesei, have been found upstream from the ATG start codon. The coding region is interrupted by nine short (48-80 bp) introns. A high bias against codons ending with G and A has been observed. (4) The architecture of CELl comprises several domains: a predicted 29-aa signal peptide, a core of 233 aa without any homology to the polypeptides in current data bases, a Pro, Ser and Thr-rich region, and a cellulosebinding domain similar to those found in other cellulolytic fungi.

J.P.: Fluorographic

detection

amide gels with the water-soluble them.

of radioactivity

fluor. sodium

in polyacryl-

salicylate.

Anal. Bio-

98 (1979) 132-135.

Chen, CM.,

Gritzali,

deduced

M. and Stafford,

primary

structure

reesei. Biotechnology Claydon,

of extracellular

Aguricus bisporus. Trans. Coughlan,

D.A.: Fruit body biomass

endocellulase Br. Mycol.

M.P. and Ljungdahl,

gal and bacterial

during

Degradation.

cellulolytic

Academic

J., Haeberli,

quence analysis

and

enzyme

fruiting by

biochemistry

systems.

In: Aubert,

and Genetics

Press, London,

P. and Smithies,

programs

regulated

periodic

Sot. 90 (1988) 85-90.

L.G.: Comparative

guin, P. and Millet, J. (Eds.), Biochemistry Devereux,

sequence

II from Trichoderma

5 (1987) 274-278.

N.. Allan, M. and Wood,

production

D.W.: Nucleotide

of cellobiohydrolase

of funJ.-P.,

Be-

of Cellulose

1988. pp. 1 I-30.

0.: A comprehensive

set of se-

for the VAX. Nucleic Acids Res. 12 (1984)

387-395. El-Gogary,

S., Leite, A., Crivellaro,

Mechanism

O., Eveleigh, D.E. and El-Dorry,

by which cellulose

triggers

cellobiohydrolase

H.:

I gene ex-

in Trichoderma reesei. Proc. Nat]. Acad. Sci. USA 86 (1989)

pression

6138-6141. Elliott, T.J.: The genetics P.B., Spencer, nology

of species of Agaricus. In: Flegg,

and breeding

D.M. and Wood. D.A. (Eds.), The Biology and Tech-

of the

Cultivated

Mushroom.

Wiley,

Chichester,

1985,

pp. 11 I-129. Fincham,

J.R.S.:

Transformation

in fungi.

Microbial.

Rev. 53 (1989)

148- 170. ACKNOWLEDGEMENTS

Gilkes, N.R., Henrissat, R.A.J.:

This work was supported by a grant from the Agricultural and Food Research Council (UK). We thank N. Claydon and M. Allan for the purification of endoglucanase and production of antibodies. We are grateful to K. Dudley, C. Joannou, P. Marsh, J. Mason, and J. Stirling for valuable discussions, B. Henrissat for hydrophobic cluster analysis, P. Cunningham for computer analysis, M.C. Harmsen for communicating unpublished results, and H. Arst and N. Mazarakis for critical reading of the manuscript. S.R. acknowledges the award of the studentship from the Science and Engineering Research Council (UK).

Domains

tion, function,

B., Kilburn,

in microbial

and enzyme families.

M.C.,

isolation

Scheer,

J., Schuurs,

of two tandemly

drogenase

M.: The conserved

terminal region of Trichoderma anti-

Arthur, R., Herr, F., Straus, N., Anderson, J. and Horgen, P.: Characterization of the genome of the cultivated mushroom, Agaricus brunnescens. Exp. Mycol. 7 (1982) 127-132. Aviv, H. and Leder, P.: Purification ger RNA by chromatography Nat]. Acad. Azevedo,

of biologically

on oligothymidylic

acid-cellulose.

Proc.

Sci. USA 69 (1972) 1408-1412.

M. de 0.. Felipe, M.S.S.,

Astolfi-Filho,

Hammond,

S. and Radford,

A.:

2569-2576. for gene expression

Beguin. P.: Molecular biology of cellulose crobiol. 44 (1990) 219-248.

degradation.

in filamentous Annu.

Rev. Mi-

J.G.H.:

The dehy-

L.J.L.D.

Wageningen,

ofAgaricus fructification.

J.B.W.: The biochemistry

D.. Casselton, opmental

L.A., Wood, D.A. and Frankland,

Biology

Cambridge,

1991,

of Higher

Fungi.

Laemmli, U.K.: Cleavage of structural head of bacteriophage Leonard,

Cambridge

In: Moore,

J.C. (Eds.), DevelUniversity

Press,

1985, pp. 389-401. proteins

T4. Nature

J.C., Dunham,

during the assembly of the

227 (1970) 680-685.

S.M. and Thurston,

C.F.: Isolation

ofpolysomes

RNA from Chlorella,fisca var. vacuolata. New

87 (1981) 39-51. T., Fritsch,

E.F.

Laboratory Manual. Harbor, NY, 1982.

and

Sambrook,

Cold Spring

J.: Molecular

Harbor

Laboratory,

Cloning.

A

Cold Spring

Manning, K. and Wood, D.A.: Production and regulation of extracellular endocellulase by Agaricus bisporus. J. Gen. Microbial. 129 (1983) 1839-1847. Mead,

D.A.,

Szczesna-Skorupa,

tem for cloning Merivuori,

E. and Kemper,

plasmids:

and protein

Trans. Messner,

B.: Single-stranded

a versatile tandem

engineering.

promoter

sys-

Prot. Eng. 1 (1986) 67-74.

H., Siegler, K.M., Sands, J.A. and Montenecourt,

ulation of cellulase biosynthesis

Cloning, sequencing and homologies of the cbh-2 (exoglucanase) gene of Humicola grisea var. thermoidea. J. Gen. Microbial. 136 (1990) Ballance, D.J.: Sequences important fungi. Yeast 2 (1986) 229-236.

of Aguricus. Pudoc,

and Breeding

DNA ‘blue’ T7 promoter active globin messen-

T.A. and Wessels,

linked glyceraldehyde-3-phosphate

pp. 140-144.

Phytol.

reesei cellulases forms a strong antigenic epitope for polyclonal bodies. Biochim. Biophys. Acta 1087 (1990) 137-141.

conserva-

Rev. 55 (1991) 303-

genes from Agaricus bisporus. In: Van Griensven,

(Ed.), Genetics

Maniatis,

Aho, S. and Paloheimo,

sequence

Microbial.

315. Harmsen,

and poly(A)-containing

REFERENCES

D.G., Miller Jr., R.C. and Warren,

b-1,4-glycanases:

and secretion

B.S.: Reg-

in fungi. Biochem.

Sot.

13 (1985) 411-414. R. and Kubicek,

C.P.: Carbon

source control of cellobiohydro-

lase I and II formation by Trichoderma reesei. Appl. Environ. biol. 57 (1991) 630-635. Mierendorf, R.C. and Pfeffer, D.: Direct sequencing of denatured DNA. Methods Enzymol. 152 (1987) 556-562. Mount. S.: A catalogue 10 (1982) 459-472.

of splice junction

sequences.

Microplasmid

Nucleic Acids Res.

190 Norrander,

J., Kempe, T. and Messing, J.: Construction

vectors

using

oligodeoxynucleotide-directed

(1983) 101-106. Pelham, H.R.B. and Jackson,

P.: Rapid preparation

M., Niku-Paavola,

structural

analysis

Eur. J. Biochem.

J., Fritsch,

Laboratory

E.F.

Manual,

J.K.C.:

Isolation

T.: Molecular

2nd ed. Cold Spring Harbor

5467. Saunders,

inhibitors.

Proc. Natl. Acad. Rapid

pp. l-8. Teeri, T.T., Lehtovaara,

A

of miniprep

in fruiting

DNA

of Schizophylium commune: homologies

dikaryons

gene not regulated

by mating-type

See, Y.P. and Jackowski,

D.H.

tides by SDS gel electrophoresis. Structure. Shoemaker, Myambo,

IRL Press, Oxford, S., Schweickart,

for

and expression

lase I derived

ZAP: a bacteriophage properties. Sims, P.F.G.,

In: Creighton,

T.E. (Ed.), Protein

lambda

Sorge, J.A.

S.,

L27. Biotechnology

expression

and Huse,

W.D.:

and

C. and Broda,

1

Lambda

vector with in vivo excision

characterization

P.: The identification, of

a

gene

from

I. and Knowles,

J.:

II. Gene 5 1 (1987) 43-

V.L., Ladner,

characterization,

M.B., Gelfand,

and expression

in

I from Trichodermu reesei. signal cleavage sites. Nucleic

J.B.W.

D.A. and Thurston,

Breeding

and Dickerson,

phosphorylase

molecular

Phanerochaete

A.G.:

and trehalase

Variation

during periodic

Agaricus bispotus (Lange)

Imbach.

in acfruitNew

105 (1987) 273-280.

Aguricus enzymes. D., Kwok,

S., Schweickart,

C.F.: Progress

In: Van Griensven,

of Aguricus. Pudoc,

in the molecular L.J.L.D.

Wageningen,

analysis

(Ed.), Genetics

of and

1991, pp. 81-86.

Wood, W.E., Neubauer, D.G. and Stutzenberger, F.J.: Cyclic AMP levels during induction and repression of cellulase biosynthesis in Thermomonosporu curvuru. J. Bacterial.

J.M.,

and 1985.

5 (1987) 60-64.

tivities of glycogen

Wood,

Nucleic Acids Res. 16 (1988) 7583-7600. James,

S., Salovuori,

von Heijne, G.: A new method for predicting

Phytol.

cloning of exo-cellobiohydro-

from Trichoderma reesei strain

(1983) 691-696. Short, J.M., Fernandez,

Wiley, Chichester,

of cellobiohydrolase

and Innis, M.A.: Cloning,

Wells, T.K., Hammond,

1989, pp. l-21. M., Gelfand,

Mushroom.

In: Flegg,

The Biology

in Trichodermu reesei cellulolytic enzymes: gene

ing of the edible mushroom,

weights of polypep-

V., Ladner,

(Eds.),

Acids Res. 14 (1986) 4683-4690.

with a

molecular

K. and Innis, M.: Molecular

and importance.

D.A.

Succhuromyces cerevisiae of endoglucanase

genes. Gene 90 (1990) 199-205.

G.: Estimating

its history Wood,

P., Kauppinen,

domains

Van Arsdell, J.N., Kwok, with chain-

double strand sequencing. Nucleic Acids Res. 18 (1990) 4948. Schuren, F.H.J. and Wessels, J.G.H.: Two genes specifically expressed

cloning

sequence

and

Press,

Sci. USA 74 (1977) 5463-

isolation

D.M.

of the Cultivated

Biotechnology S.E. and Burke, J.F.:

The mushroom,

Spencer,

Homologous

Cloning.

Laboratory

Cold Spring Harbor, NY, 1989. Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing terminating

and fun-

137 (1991) 1537-1544.

and Maniatis,

D.M.:

P.B.,

Technology

gene from the lignin-degrading

and characterization of a Ustilugo dehydrogenase-encoding gene.

Gene 93 (1990) 111-117.

67 (1976)

of DNA from tilamentous

M.-L. and Knowles,

of the lactase

to the exo-cellobiohydro-

lase I gene from Trichodermu reesei. Gene 74 (1988) 41 l-422. Smith, T.L. and Leong, S.A.: Isolation muydis glyceraldehyde-3-phosphate

trans-

1 (1985) 17-20.

gus Phlebiu rudiuta. J. Gen. Microbial. Sambrook,

26

Spencer,

fungi. Lett. Appl. Microbial. Saloheimo,

lysates.

chrysosporium that shows strong homology

Ml3

Gene

R.J.: An efficient mRNA-dependent

lation system from reticulocyte 247-256. Raeder, U. and Broda,

of improved

mutagenesis.

160 (1984) 1047-1054.

Wood, D.A., Claydon, N., Dudley, K.J., Stephens, S.K. and Allan, M.: Cellulase production in the life cycle of the cultivated musroom Agaricus bisporus. In: Aubert, J.-P., Beguin, P. and Millet, J. (Eds.), Biochemistry and Genetics of Cellulose Degradation. Academic Press. London,

1988, pp. 53-70.