Gene. 108 (1991) 81-89 0

1991 Elsevier

GENE

Science

Publishers

B.V. All rights reserved.

81

0378-l 119/91/$03.50

06158

Protein disulfide isomerase is essential for viability in Sacchavomyces cerevisiae (Recombinant

DNA;

yeast; disulfide

bond formation;

thioredoxin;

PDIl

Ronnie Farquhar”, Neville Honey”*, Susan J. Murant”, Peter Bossier Ronald W. Ellisb, Robert B. Freedman” and Mick F. Tuite”

gene)

a, Loren Schultz”, Donna Montgomery b,

” Biological Laboratory, Universityof Kent, Canterbury, Kent, CT2 7NJ (U.K.): and h Merck, Sharp and Dohme Research Laboratories, West Point, PA 19486 (U.S.A.) Tel. (215)661-6375 Received by J.R. Kinghorn: 27 June 1991 Revised/Accepted: 3 August/6 September 1991 Received at publishers: 10 September 1991

SUMMARY

Protein disulfide isomerase (PDI) is an enzyme involved in the catalysis of disulfide bond formation in secretory and cell-surface proteins. Using an oligodeoxyribonucleotide designed to detect the conserved ‘thioredoxin-like’ active site of vertebrate PDIs, we have isolated a gene encoding PDI from the lower eukaryote, Saccharolnyces cerevisiue. The nucleotide sequence and deduced open reading frame of the cloned gene predict a 530-amino-acid (aa) protein of M, 59082 and a p1 of 4.1, physical properties characteristic of mammalian PDIs. Furthermore, the aa sequence shows 30-32”b identity with mammalian and avian PDI sequences and has a very similar overall organisation, namely the presence of two approx. lOO-aa segments, each of which is repeated, with the most significant homologies to mammalian and avian PDIs being in the regions (a, a’) that contain the conserved ‘thioredoxin-like’ active site. The N-terminal region has the characteristics of a cleavable secretory signal sequence and the C-terminal four aa (-His-Asp-Glu-Leu) are consistent with the protein being a component of the S. cerevisiueendoplasmic reticulum. Transformants carrying multiple copies of this gene (designated PDIl ) have tenfold higher levels of PDI activity and overproduce a protein of the predicted M,.. The PDIl gene is unique in the yeast genome and encodes a single 1.8-kb transcript that is not found in stationary phase cells. Disruption of the PDIZ gene is haplo-lethal indicating that the product of this gene is essential for viability.

INTRODUCTION

Protein enzyme reactions, lumen in enzyme’s

disulfide-isomerase (PDI; EC .5.3.1.4.), an which catalyzes thiol : disulfide interchange is a major resident protein component of the ER secretory cells. A body of evidence on the cellular distribution, its subcellular location and

Correspondence to: Dr. M.F. Tuite, Biological Laboratory, Kent, Canterbury, Kent, CT2 7NJ (U.K.) Tel. (44)227-764000, ext. 3699; Fax (44)227-7639 12. * Permanent Massey

address:

University,

Tel. (64)063-69099.

Department

Palmerston

North

of Microbiology (New Zealand)

University

of

its developmental properties suggests that it plays a role in secretory protein biosynthesis (Freedman, 1984) and functions as a catalyst of native disulfide bond formation in the biosynthesis of secretory and cell surface proteins (Bulleid and Freedman, 1988; Freedman et al., 1989). Of the many protein factors currently implicated as mediators of protein folding, assembly and translocation in the cell

Abbreviations: aa, amino acid(s); bp, base pair(s); reticulum; kb, kilobase or 1000 bp; nt, nucleotide(s); ribonucleotide;

and

Genetics,

ORF,

open reading

frame;

PAGE,

electrophoresis; PDI, protein disulfide isomerase; PDI; S., Saccharomyces; SDS, sodium dodecyl NaCI/O.O15 M Na,

citrate

pH 7.6; UWGCG,

Genetics

Group

(Madison,

state.

Computer

WI);

ER, endoplasmic oligo, oligodeoxypolyacrylamide-gel

PDIl,

sulfate;

University

gene encoding SSC, 0.15 M of Wisconsin

[ 1,denotes plasmid-carrier

82 (Rothman, 1989), PDI catalytic activity. Mammalian PDI is characteristically acidic The enzyme has also et al., 1991), and from (Kaska et al., 1990)

is unusual

in having a well-defined

a homodimer (2 x 57 kDa) with a pI (4.0-4.5) (Hillson et al., 1984). been purified from wheat (Bulleid the alga Chlun~ydon~onns reinhardii and from the yeast S. cerevisiae

(Mizunaga et al., 1990). Recently, the complete aa sequences of a number of PDIs have been reported, largely derived from cloned cDNA sequences; these include the PDIs from rat (Edman et al., 1985), ox (Yamauchi et al., 1987), human (Pihlajaniemi et al., 1987) and chick (Parkkonen et al., 1988). The proteins from these vertebrate species show a high degree of sequence conservation and all show several overall features first noted in the rat PDI sequence (Edman et al., 1985). The most significant is the presence within the PDI sequence oftwo regions of approx. 100 aa (the a and a’ domains) strongly homologous to each other and closely related in sequence to thioredoxin, a small redox active-protein containing an active site disulfide: dithiol couple formed between vicinal Cys residues. In thioredoxin the active site sequence is WCGPCK, whereas the corresponding region, found twice in PDI, has the sequence WCGHCK. Sequences corresponding to, or closely related to PDI have been identified in work aimed at analysing functions

other than disulfide bond formation. For example, PDI acts as the fl subunits of the tetrameric a& enzyme prolyl-4hydroxylase, which catalyzes a major post-translational modification of nascent or newly-synthesized procollagen polypeptides within the ER (Pihlajaniemi et al., 1987; Koivu et al., 1987). There is also evidence suggesting that PDI participates in the system for cotranslational Asnglycosylation (Geetha-Habib et al., 1988) and recently the proposal has been made that the enzyme participates in the complex which transfers triglyceride to nascent secretory lipoproteins (Wetterau et al., 1990). Thus, PDI may be multifunctional in the co- and post-translational modification of secretory

proteins

(Freedman,

1989).

Our understanding of the structural basis of PDI activity would benefit from an analysis of sequence conservation based on PDI sequences from a wider range of organisms

Fig. 1. SDS-PAGE formant

carrying

plasmid.

Lanes:

[pMA3a];

analysis a putative

of a cell-free lysate of an S. cerevisioe transyeast

PDI-encoding

1, untransformed

strain

gene on a multicopy

MD40/4c;

2, strain

MD40/4c-

: C7] (pMA3a : C7 carried the puta-

3. strain MD40/4c[pMA3a

tive PDI-encoding gene). The large arrow indicates a polypeptide of 5X kDa overexpressed in the MD40;c[pMA3:C7] transformant. The small arrow indicates a ‘IO-kDa MD40/4c[pMA3:C7] transformant. markers, ments

M,. A yeast genomic

library,

containing

from the S. cerevisiae

of DNA

BamHI

polypeptide overexpressed Also shown are molecular

site of the high copy number

strain

partial

SKQ2n

LEU2-d,

Scru3A frag-

cloned

2p-based

in the weight into the

vector pMA3a

and Tuite, 1987) was used to screen for the PDII gene. A 30-mer

(Crouzet

oligo (5’-CTTACAGTGACCACACCATGGAGCGTAGAA-3’) synthesised

against

(FYAPWCGHCK), ng

kDa

205 116 97 66

4

2

+

+

of the

graphy

was

end-labelled

bias (Sharp

of Cohen

et al. (1972)

colonies

as follows: each nitrocellulose

16 h at 37’C

labelled

were screened

in 35”,,

formamide

pg per ml denatured

oligo (specific

activity

salmon

nt. Following approx.

on nitrocellulose

6 x SSC;‘l x Denhardt’s

sperm DNA/O.1 ‘