Gone. 95 (1990) 155-160 Elsevier
155
GENE 03731
C l o n i n g o f usp45, a g e n e encoding a s e c r e t e d p r o t e i n f r o m Lactococcus lactis subsp, lactis M G 1 3 6 3 (Recombinant DNA; lactic acid bacteria; chromosomally encoded gone; signal sequence; periplasm; secretion vectors; serine~rich protein; product of P54 of Enterococcus faecium)
Martien van Asseldonk, Ger Rutten, Marco Oteman, Roland J. Siezen, Willem M. de Vos and Guus Simons Molecular Genetics Group, Department of Biophysical Chemistry, Netherlands Institute for Da~y Research (NIZO), 6710 BA Ede (The
Netherlands) Received by J.-P. Lecocq: 21 February 1990 Revised: 27 May 1990 Accepted: 28 May 1990
SUMMARY
We have cloned usp45, agene encoding an extracellular secretory protein of Lactococcus lactis subsp, lactis strain MG 1363. Unidentified secreted 45-kDa protein (Usp45) is secreted by every mesophilic L. lactis strain we tested so far and it is chromosomally encoded. The nucleotide sequence of the usp45 gone revealed an open reading frame of 1383 bp encoding a protein of 461 amino acids (aa), composed of a 27-aa signal peptide and a mature protein initiated at Asp 2s. The gone contains a consensus promoter sequence and a weak ribosome-binding site; the latter is rather uncommon for Gram-positive bacteria. Expression studies in £scherichia coil showed efficient synthesis and secretion of the protein. Usp45 has an unusual aa composition and distribution, and it is predicted to be structurally homologous with P54 of Enterococcusfaecium. Up to now, no biological activity could be postulated for this secreted protein.
INTRODUCTION
Lactococci (L. lactis subsp, lactis and L. lactis subsp. cremoris) are used on a large scale as starter cultures in the manufacture of dairy products such as cheese, butter, and buttermilk. Recently, the complete nt sequence of the genes for two extracellular proteins, e.g., proteinase PrtP (Vos et al.,
Correspondenceto: Dr. G. Simons, NIZO, P.O. Box 20, 6710 BA Ede (The Netherlands) Tel. (31)8380-59557; Fax (31)8380-50400. Abbreviations: A., absorbance; aa, amino acid(s); bp, base pair(s); kb, kilobase(s) or 1000bp; L., Lactococcus; Lb., Lactobacillus; nt, nueleotide(s); oligo, oligodeoxyribonucleotide; ORF, open reading frame; PA, polyacrylamide; PAGE, PA-gel electrophoresis; RBS, ribosome-binding site; S., Streptococcus; SDS, sodium dodecyl sulfate; tsp, transcription start point(s); Usp45, unidentified secreted 45-kDa protein; usp45, gone encoding Usp45; [ ], denotes plasmid-carrier state. 0378-1119/90/$03.50© 1990Elsevier SciencePublishers B.V.(BiomedicalDivision)
1989) and the bacteriocin nisin (Buchman et al., 1988) have been determined. The proteinase is synthesized as a prepro-protein with a consensus Ala/Ala signal peptidase I cleavage site between aa 33 and 34. it has been demonstrated that this signal peptide directs the secretion of heterologous proteins in lactic acid bacteria (Simons etal., 1988). The gene for the bacteriocin nisin encodes a precursor 6.0-kDa polypeptide of 57 aa. A signal peptide of 23 aa has been deduced which deviates from the consensus signal peptide rules as postulated by yon Heijne and Abrahmsen (1989). To gain more insight in the properties and structure of lactococcal extracellular proteins, we screened lactococcal strains for extracellular proteins, and attempted to clone the gene encoding a 45-kDa protein which is secreted into the extracellular medium as an apparent 60-kDa protein and is produced by every lactococcal strain we tested so far. Based on the regulatory sequences of the usp45 gene, secretion vectors specific for lactococci will be constructed.
156 no antimicrobial activity against other Gram + bacteria, assayed by the method of Fowler (1975), could be assigned to the protein.
EXPERIMENTAL AND DISCUSSION
(a) Characterization of secreted proteins of laetococci L. lac~ strains were screened for secreted proteins by analyzing culture supernatant fractions on SDS-PAGE. As shown in Fig. IA, all screened strains secrete proteins ofwhich the most abundant one banded at 50-60 kDa. The amount of this secreted protein was estimated to be 2-4 ttg/ml/A6oonm. Western-blot analysis of the culture supernatant fractions showed a strong reaction between the 50-60-kDa proteins of all tested strains and antibodies against the 60-kDa protein of L. lactis MG1363 (Fig. 1B), isolated from the plasmid-free strain MG1363, indicating an immunological relationship between these proteins. No 50-60-kDa proteins could be detected in cell extracts of the tested strains, indicating that these proteins are tratzsported efficiently into the extracellular medium and are not associated with the cell envelope as is the case for components of the proteolytic system of lactococci (Vos et al., 1989). Extracellular proteins of Leuconostoc paramesenteroides, S. thermophilus, Lb. casei, Lb. acidophilus and Lb. bulgaricus strains showed no reaction with the antibodies (data not shown) suggesting that this 50-60-kDa protein is specific for L. lactis. Since every L. lactis strain we tested produces this 50-60-kDa protein and no L. lactis strains defective in synthesis of this protein were available, it was not possible to postulate a function for this protein. No proteolytic activity ofthe 60-kDa protein of MGl363 could be detected on casein, its degradation products or on synthetic substrates, indicating that the protein has presumably no function in the proteolytic system of L. lactis. In addition,
(b) Cloning and expression of the gone for the 60-kDa protein in EsckericMa eoli Since the plasmid-free strain MG1363 is able to produce the 60-kDa protein, chromosomal DNA of this strain was used to construct a genomic library in phage A47.1 (Loenen and Brammar, 1980). The genomic bank was screened with the isolated antibodies and positive phages were subsequently screened with probes directed against the N terminus of the mature protein. The sequence of the N-terminal 10 aa, as determined with a gas-phase sequenator (Applied Biosystems), was found to be D-T-N-S-D-IA-K-Q-D. Mixed oligo probes were derived against the last six aa of this sequence which had the least degenerate codon usage. Southern-blot analysis showed that the gene for the 60-kDa protein was found to be located on a KpnI-EcoRI fragment of + 3 kb. This fragment was cloned in pUCI9 generating pNZ 1011. Western-blot analysis of cell extracts of E. coli JM83[pNZ 1011], showed synthesis of the protein (Fig. 2, lane 2). Moreover, a substantial amount of the 60-kDa protein could be detected in the periplasmic fraction of the cells (lane 3), where only a small amount of 60-kDa protein was still present in the stripped cells (lane 4). However, the ratio of p-lactamase activity in the periplasmi¢ fraction compared to stripped cells (4: 1) was in complete agreement with the amounts of 60-kDa protein detected within these two fractions (results not shown). Hence, we conclude that the 60.kDa protein is transported efficiently into the periplasmic space of £. coil cells.
A kDa
1
2
3
4
5
6
7
8
B i
2
3
4
,5
6
7
8
9Z, 5 68 43
.... 29
.~,:~i~i )
Fig. 1. Gel analysis ofextracellular proteins ofL. lactbstralns. Supernatantsoflogarithmicogrowingeulturesweredialyzedagalnst water and concentrated by lyophilisation. Samples were analyzed by 0.1% SDS-10% PA gel (Laemmli, 1970) and Western blotting (Towbin et al., 1979). (Panel A) Coomassie brilliant blue staining of 1.0 ml ofsupernatant; (Panel B) Western blot of 0.2 m! ofsupcrnatanL To raise antibodies against the 60-kDa protein of MG1363, the 60-kDa protein was excised from SDS-PA gel and recovered by isota¢hophoresis (Ofverstedt et al., 1983), Lanes: 1, MG1363 (Gasson, 1983); 2, IL!403 (Chopin et aL, 1984); 3, RI*; 4, NZ22186"; 5, NCDO176; 6, SKI 128 (De Vos and Davies, 1984); 7, NCDOI200; 8, WG2* (*NIZO collection). The lanes are flanked on the lel~ side by Mr markers. The position of the $0-60-kDa prnteins is indicated by arrowheads.
157 l
2
3
4
5
Fig. 2. Western blot of proteins encoded by E. co/[ JM83[plqZ1011]. Samples were analyzed by 0.1% SDS-10% PA gel (Laemmli, 1970). Cells were fractionated by osmotic shock (Neu and Heppel, 1965) resulting in a periplasmic and a stripped-cell (cells without periplasmic space) fraction. Lanes: 1, JM83[pUCI9], unfractionated cells; 2, 3, and 4, JM83[pNZ1011]; 2, unfra¢tionated cells; 3, stripped cells; 4, periplasmic fraction; 5, MGI363, culture supernatant. The position of the 60-kDa protein is indicated by an arrowhead.
It has been demonstrated that regulatory sequences of Gram + organisms such as bacilli and staphylococci can function efficiently in E. coli (McLaughlin et al., 1981). Our data show that the regulatory sequences of this lactococcal gene and the signal sequence of its gene product are also recognized in E, cog. (c) Nucleotide sequence of the gem for the O0-kDa protein The nt sequence of the l~pnI.£coRl fragment from pNZI011 was determined and the region encoding the 60-kDa protein is shown in Fig. 3. The nt sequence encoding the first N-terminal 10 aa of the mature protein was found at nt position 97-113 and followed by an ORF encoding a mature protein of 434 aa. This protein has a predicted Mr of 44628 and it was designated U~p45. Upstream from the mature N-terminus, three in-frame start codons are located. The most likely start codon is ATG at nt position 1, since it precedes a precursor region which has the characteristics of a signal peptide: several basic aa followed by a hydrophobic core of 15-20 aa and a shorter uncharged region ending with an aa carrying a small side chain (yon Heijne and Abrahmsen, 1989). Although the highest cleavage probability by signal peptidase I is C-terminal of Ala 19 according to yon Heijne and Abrahmsen (1989), the observed cleavage after Ala 27 in the sequence Val-Xaa-Ala is also commonly found in prokaryotes, and particularly in Gram + bacI~ria. However, N-terminal processing cannot be excluded. The ATG start codon at nt position I is preceded by two
potential RBSs (Shine and Dalgarno, 1974), e.g., the sequence GGAGG at nt -26 to -22 and AAAG at -10 to -8. The calculated free energy of the latter one is only -4.6 kcal/mol (Tinoco et al., 1973). RBSs of genes from lactococci (De Vos, 1987) and other Gram + microorganisms like bacilli (McLaughlin et al., 1981) usually have free energy values between -10 and -15 kcal/mol. For the other possibility a free energy of -14.4 kcal/moi could be calculated. However, the spacing to the ATG codon is 21 bp. This length does not fit with the hypothesis proposed by Shine and Dalgarno (1974). Site-directed mutagenesis experiments need to be performed to determine whether the weak RBS is used and what the contribution is of the G G A G G sequence to the initiation of translation of the usp45 gene. Primer extension experiments (Debarbouille and Raibaud, 1983) were performed and showed that transcription of the usp45 gene was initiated with an A residue at nt position -44 (data not shown). A consensus -10-hexanucleotide sequence (TATAAT) was found at position -56 to -51. The most probable corresponding -35 sequence is TTAAGG at nt position -76 to -71, although the spacing between the two is smaller (14 bp) than that usually found in promoters in L. lactis (De Vos, 1987; Van der Vossen et al., 1987). Inspection of the nt sequence shows the presence of several repetitive sequences: the hexanucleotide sequence GCNCAA occurs 19 times in the gene, of which 16 times in the ORF, whereas the more degenerated sequence TC~AG c appears 16 times. Translation of this sequence would result in twelve Ser doublets of which seven are in tandem in one part of Usp45, resulting in an unusual aa sequence in the protein as discussed in section d, Downstream from the TAG stop codon of the usp4S gene an inverted repeat was found at nt position 1414-1452, which could form a hairpin-like structure with a free energy of -8.2 kcal/mol (Tinoco et al., 1973). The formed hairpin is flanked on both sides by A or T stretches and hence could function as a Rho-independent, bi-directional terminator of transcription. Northern-blot analysis (data not shown) revealed a messenger RNA of approx. 1500 nt which is in agreement with the deduced transcription termination. (d) The aa sequence, homology analysis and structural predictions of Usp45 The aa sequence of Usp45 as shown in Fig. 3 reveals a high content of Ser and Ala residues, together more than 32%. In one region of the protein, from aa 264-316, Ser residues constitute 33 out of 53 aa, including a continuous stretch of 16 aa. Database searches for homology were performed with the SWISSPROT (release 12.0) and the NBRF (release 22.0) aa sequence libraries. The N-terminal 280 aa of
158 X | &,i~ x
N
~.~,.'_]a,_.~.~&ak.C,L,r.kt%~-.~&"~,'kA, I L ' ~ & C l k ~ T & C ' ~ K
I(
r
Z
X
•
A
Z
L
M B
'~
V
C ' ~ CAGCC C 4 : ~ X r Z ~ G G " I L ' ~ & ~
Z
L
S
A
&
A
P
L
':
G
V
~ ¥
~
A
O
T T
&
~
M S
D
~ X
~
A
g
O
O
/k
T
Z
T C ~ _ _.~ ~ & i ~ T ~ A ~ C A C A A G C A ~ J ~ G C A C A A G ~ T A G C T ~ T C A ~ G T ~ C T T A C A A ~ G ~ A ~ ~ ~ C
4].
5
S
/4
Q
8
&
g
A
O
&
Q
&
U
V
O
S
241 G&&p.~qTtdk&GCAC~'TA&TE~N?.&&~T~G~ & IIIL
R
I,
Q
S
K
V
D
S
L
Q
O
K
Q
T
5
•
g
A
Q
X
A
Ig
1[
ig
S
ATC&&&G&&CGTACIC3t AG&~TTGG&J~CTC&&GC&~T>GCTC&P.GT~AP.XGCTC&GC&/~&&&~&TA~
~
g
~k L
M ~.
Q
Z
A
•
~
N
R
~=
X
[g
~
R
•
g
•
L
R
Jk Q
~.
R
S
.~
O
V
N
S
q
~
•
H
"If g
A
V
V
M
IS
K
'=
L
T
D
V
g
Q
K
V
T
A
Z
A
T
V
S
S
A
!1
I(
U
Z
L
R
Q
Q
E
IT
E
O
g
E
L
O
g
M
II
T
V
K
g
a
Y
H
q
P
V
S
L
S
~
S
L
O
S
U
A
q
E
1[, T
5
q
QI &
E
L
K
V
,It, T
L
E
A
0
5
T
36X 12"
O 48~
161
8
601 AACTATC~CA~TTI=CAACTG~GCA~TA~C~ 30!.
N
Y
O
A
T
Z
A
T
A
'0
D
K
K
"~. ~ A T T A G ~ T G / t ~ , & ~G C A G ~ C ~ C A c ~ r ~ A A G A A G C J ~ C ' T ~ C A A G C G G C T T A T I ~ Q
&
L
L
O
E
!(
A
A
&
E
g
k
A
U
E
A
*% g
g
Q
k
A
¥
731 ~tA.%AGA,'~,CAC~P.ACAAG~CAGC'['t~.AACAGCAC~AAC'I~CTAAAGCTel~AGAAGC A ~ C A A ~ A ~ A G ~ ~ ~ ~ A C ~ G C A ~ 341
Q
g
B
A
A
Q
A
U
A
A
S
?
A
A
•
A
g
A
V
E
A
A
T
841 GATAATACA~TCAAATGCTAG~'GC~.AAACAGTT~&~?J~PI~?ATC,~J~AC'TCAAGT TCA~ 36L
D
M T
T
5
M
A
S
A
6
M S
8
H
'=
S
A
N
S
5
S
S ~
8
8
S
A
S
& 8
A
~
8
5
~1 S
S
G -'= S
~ S
H
T
A
F
G
fl
961 AATACAAATTCAGGCACTAGTACT~T&OgGGAGGAAI~CTA~'~GTGGT&GC~CT ATA~TAG'I~CCAATTGC ~ T ~ & T G 321
U
8
U
V
~
9
9
T
~
B
~
S ~ 8
M
N
A
G
G
&~I'GTACTG~AT~'TATGG
H
S
G
T
8
T
G
N
T
G
q
T
T
T
G
G
9
~
g
H
8
-q
p
z
G
H
P
V
A
G
G
O
¢
T
D
Y
V
H
Y
P
A
A
Q
G
Z
Y
Z
R
N
|
M
IP
G
H
G
G
O
5/
A
S
N
G
P
A
Q
G
Y
L
II
V
V
G
A
A
P
G
V
A
|
B
P
B
/L
0
IF
V
G
¥
k
M
8
P
Y
G
II
V
AIr
V
K
8
V
g
8
D
O
T
Z
T
E
K
I
O
G
Y
G
t0OX
36}.
q 1.:101
4,1.
Z
12311, A C A A ~ q ~ C k T ~ C G T A C T ~ T ~ C G T ~ A ~ q ~ & T O C e A A ~ A ~ A A T A A A T A ~ . . ~ T A ~ T ~ T ~ 44~
T
i
T
M
O
H
e
M
T
V
M
A
Ii
G
V
T
V
L
N
P
M
1.441. TT~T~.~TTI~&~ATt~kTAATAATAAG~A~TGATA
Fig, 3, The nt sequence of the usp4$ gone and the deduced aa sequence. Seque,~ctng was performed following th~ Sequcnase protocol of' Tabor and Richardson (1987). The doable underlined sequences represent the -35 and -10 promoter boxes. The ~p is indicated by an upward arrowhead. The two putative RBSs as well as the determined N.terminal an (Asp 3B -Asp 3'/ ) ofthe mature protein ere underlined, Downstream from the gone an inverted repeat Is marked
with dashed arrows representing a Rho-independent
terminator
of transcription.
The OenBank
accession number
of this nt sequence
is M35374.
A ......
• I ....
I,.
.I..I
I .......
.....
I .I...
I, ,I,,.I
I • .. I.,,
,I.1.1,,I,
I .I ....
I,. IIl,.ll
QVg A| Wr~lk0OLd,TIg. UflPLMMJgBAQLIgQIg[ ledOLOgM! Iglgl4EA~ |
Q
~
..... II
H¥ |~10
GIJ)BQ&QIgLWl~llghgV~?lJ4YQATI ATe~MQgQ~I~IMglAAJ~IglgAAQIg&AR~y~ ~110
• I II
.I .-I1,1,1
, ,I.I.
II1,111,,I
,ll.il
I,I.
..11
A I ~ " I ~ U I J dM~QJ~I L~iViaTru Ii%AI~A?A k~UKI~I)LHIUU~I~K-~K~R ! I @ |
! .... I...
|, -I . I,..I ....
II- .,
.,
~g
2]U
I11,.
Fig. 4. Homology between Usp4$ and P54 determined according to Devet~ux et al, (1984), Depicted is our interpretation of translation of the
Usp45 show weak homology with myosins, keratins and streptococcal M proteins, which presumably reflects a common type of 0c-helical secondary structure (Cohen and Parry, 1986). Significant homology was detected with the recently published sequence of P54 protein of Enterococcusfaecium (Ftlrst et al., 1989), as shown in Fig. 4. This protein of unknown function is attached to the ceil-wall, and is generated from a 516-aa precursor. In the N-terminal 281 aa, including the signal peptide, 31 ~ ofthe residues are identical and another 53~o are conservative substitutions (Fig. 4A). published nt sequence of P54 (FOrst, 1989), which leads to an additional 9 aa residues at the N terminus compared to the published aa sequence. Identical residues arc connected by lines, conserved substitutions arc indicated by points. (A)aa 1-281; (B)aa 281-340.
159 TABLE I
Characteristics of the Usp45 and P54 domains" Domains
Residues Main aa Secondary structure Charge Hydropathy Flexibility
Signal peptide
Ab
Usp45
P54
Usp45
P54
1-27 A,S,I
1-28 A,S,I
28-281 A,Q,S,K
29-281 A,Q,K,E
~-helix low; positive hydrophobie low
,,-helix high; neutral amphipathic low
B¢
Usp45 1
--
Cd
D•
P54
Usp45
P54
Usp45
P54
282-336 S,T,F
282-340 S,N,T,G
337-395 N,T,S,P
341-461 G,A,V,S
396--516 G,V,A,S
turn/coil high; negative hydrophilic high
turn/coil none hydrophilic high
~-sheet low; neutral hydrophobic low
" Secondary structure, hydropathy and flexibilitypredictions are determined using the Pcgene (version 5.01) nucleic-acid and protein sequence analysis software system (Genofit). u Segment A, immediately alter the signal peptide, is predicted to be almost exclusively ae-helicalby the method of Gamier (Gamier et al., 1978). c Segment B is unique for P54. d Segment C consists mainly of the aa Ser, Ash, Gly and Thr (Pro instead of Gly in P54) which are commonly found in turns and random coil secondary structure. Characteristic is the long stretch of Ser and Asn aa in both proteins at the N-terminal end ofthis segment, as shown in Fig. 4, while Gly and Pro residues are concentrated at the C-terminal end. Segment C is also extremely hydrophih'c: 80~o of the aa in Usp45 contain either a hydroxyl or carboxyi group, but there are no charges in this segment. © Segment D is predicted to have largely ~-sheet secondary structure in short stretches, alternating with turns or bends. The aa compositions of segment D are very similar in Usp45 and P54, and contain much higher contents of aromatic and other hydrophobic as than segments A, B and C.
Besides the homology on primary aa sequence level, there appears to be an overall resemblance ofthe two proteins on the secondary and tertiary structure level. On the basis of structural predictions ~ d aa compositions, the two mature proteins can be divided into four segments with similar characteristics, as outlined in Table I. (f) Conclusions L. lactis strains produce extracellular secretory proteins with an apparent Mr of 50-60 × l0 s, which are strongly immunologically related. The gene for this protein of strain MG 1363, designated usp45, is chromosomally encoded and was cloned and expressed in E. coli under control of its own regulatory sequences. Moreover, the gene product was secreted efficiently into &e periplasmic space of this organism. The deduced aa sequence shows an unusual composition and distribution. Usp45 shows aa sequence homology with P54, an extracellular protein of Enterococcus faecium. Besides homology on the aa level the two proteins have similar structural characteristics, but until now their function remains unclear. The cloning ofthe usp4.~gene may be an introduction for further investigation of the secretory pathway of L. lactis and could lead to the construction of expression and secretion vectors in these food-grade dairy micro-organisms for the production of homologous and heterologous proteins.
ACKNOWLEDGEMENTS
We thank Joop Mondria, Simon van der Laan and Henk van Brakel for art work and photography. This work was financed by the Mesdag Foundation and the Cooperative Rennet and Dye Factory (CSKF).
REFERENCES
Buchmun, G.W., Banerjee, S. and Hansen, J.N.: Structure, expression and evolution of a gene encoding the precursor of Nisin, a small protein antibiotic. J. Biol. Chem. 263 (1988) 16260-16266. Chopin, A., Chopin, M.C., Moillo-Batt, A. and Langella, P.: Two plasmid-determined restriction and modification systems in 5, lac~s. Plasmid 11 (1984) 260-263. Cohen, C. and Parry, A.D.: ",-Helical coiled coil - a widespread motif in protein. Trends Biochem. Sci. I ! (1986) 245-248. Debarbouille, M. and Raibaud, O.: Expression of the £. co//malPO. operon remains unaffected after drastic alterations of its promoter. J. Bacteriol. 153 (1983) 1221-1227. Deverenx, J., Haeberli, P. and Smithies, O.: A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Reg. 12 (1984) 387-395. De Vos, W.M.: Gene cloning and expression in lactic streptococci. FEMS Mierobiol. Rev. 46 (1987) 281-295. De Vos, W.M. and Davies, F.L.: Plasmid DNA in lactic streptococci: bacteriophage resistance and proteinase plasmids in $. cremoris SKI !. 3rd European Congress on Biotechnology I!1. Vedag Chemic, Munich, 1984, pp. 201-206. Fowler, G.G., Jarvis, B. and Tramer, J.: The assay of nisin in foods. Sec. Appl. Bacteriol. Tech. Set. 8 (1975) 91-105.
160 Filrst, P.E., M6sch, H.U. and Solioz, M.: A protein of unusual composition from £ntemcoccusfaecium. Nucleic Acid~ Res. 17 (1989) 6724. Gamier, J., Osgnthorpe, DJ. and Robson, B.: Analysis of the accuracy and implications of simple methods for predicting the secondary structure of globalar proteins. J. Mol. Biol. 120 (1978) 97-120. Gasson, M J.: Plasmid complements of Streptococcus lactis NCDO712 and other lactic streptococci alter protoplast induced curing. J. Bacteriol. 154 (1983) 1-9, Laennnli, U.K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 (1970) 680-685. Loenen, W.A. and grammar, WJ.: A bacteriophage lambda vector for cloning large DNA fragments made with several restriction enzymes. Gone 10 (1980) 249-259. MeLaughlin, J.R., Murray, C.L. and Rabinowitz, J.C.: Unique features in the ribosome-binding site sequence ofthe Gram-positive Staphylococcus aureus ~-Iactamase gone. J. Biol. Chem. 256 (1981) 11283-11291. Neu, H.S. and Heppel, L.A.: The release of enzymes from E. coil by osmotic shock during the formation of spheroplasts. J. Biol. Chem. 240 (1965) 3685-3692. Ofverstedt, L.G., Sandelin, J. and Johansson, G.: Recovery of proteins on a milligram scale from polyacrylamide electrophoresis gels, exemplified by purification of a retinol-binding protein. Anal. Biochem. 134 (1983) 361-365. Shine, J. and Dalgarno, L.: The 3' terminal sequence of E. coil 16S ribosomal RNA: complementarity to nonsense triplets and ribosomebinding site. Proc. Natl. Acad. Sol. USA 71 (1974) 1342-1346,
Simons, G., Rutten, G., Homes, M. and De Vos, W.M.: Production of bovine prochymosin by lactic acid bacteria. In Breteler, H., Lelyveld, P.H. and Luyben, K.C.A.M. (Eds.), Proceedings 2nd Netherlands Biotechnology Congress, Netherlands Biotechnological Society, Amsterdam, The Netherlands, 1988, pp. 183-187. Tabor, S. and Richardson, C.C.: DNA sequencing analysis with a modified bacteriophage T7 polymerase. Proc. Natl. Acad. Sci. USA 84 (1987) 4767-4771. Tinoco Jr., I., Borer, P.N., Dengler, B., Levine, M.D., Uhlenbeck, O.C., Crothers, D.M. and Gralla, J.: Improved estimation of secondary structure in ribonucleic acids. Nature 246 0973) 40-41. Towbin, H., Staehelin, T. and Gordon, J.: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76 (1979) 4350-4354. Van der Vossen, J.M.B.M., Van der Lelie, D. and Venema, G.: Isolation and characterization of Streptococcus cremoris Wg2 specific promoters. Appl, Environ. Microbiol. 53 (1987) 2452-2457. yon Heijne, G. and Abrahmsen, L.: Species specific variation in signal peptide design. Implication for protein secretion in foreign hosts. FEBS Lett. 244 (1989) 439-446. Vos, P., Simons, G., Siezen, R.J. and De Vos, W.M.: Primary structure and organization of the gone for a prokaryotic cell envelope-located serine protease. J. Biol. Chem. 264 (1989) 13579-13585.
155
GENE 03731
C l o n i n g o f usp45, a g e n e encoding a s e c r e t e d p r o t e i n f r o m Lactococcus lactis subsp, lactis M G 1 3 6 3 (Recombinant DNA; lactic acid bacteria; chromosomally encoded gone; signal sequence; periplasm; secretion vectors; serine~rich protein; product of P54 of Enterococcus faecium)
Martien van Asseldonk, Ger Rutten, Marco Oteman, Roland J. Siezen, Willem M. de Vos and Guus Simons Molecular Genetics Group, Department of Biophysical Chemistry, Netherlands Institute for Da~y Research (NIZO), 6710 BA Ede (The
Netherlands) Received by J.-P. Lecocq: 21 February 1990 Revised: 27 May 1990 Accepted: 28 May 1990
SUMMARY
We have cloned usp45, agene encoding an extracellular secretory protein of Lactococcus lactis subsp, lactis strain MG 1363. Unidentified secreted 45-kDa protein (Usp45) is secreted by every mesophilic L. lactis strain we tested so far and it is chromosomally encoded. The nucleotide sequence of the usp45 gone revealed an open reading frame of 1383 bp encoding a protein of 461 amino acids (aa), composed of a 27-aa signal peptide and a mature protein initiated at Asp 2s. The gone contains a consensus promoter sequence and a weak ribosome-binding site; the latter is rather uncommon for Gram-positive bacteria. Expression studies in £scherichia coil showed efficient synthesis and secretion of the protein. Usp45 has an unusual aa composition and distribution, and it is predicted to be structurally homologous with P54 of Enterococcusfaecium. Up to now, no biological activity could be postulated for this secreted protein.
INTRODUCTION
Lactococci (L. lactis subsp, lactis and L. lactis subsp. cremoris) are used on a large scale as starter cultures in the manufacture of dairy products such as cheese, butter, and buttermilk. Recently, the complete nt sequence of the genes for two extracellular proteins, e.g., proteinase PrtP (Vos et al.,
Correspondenceto: Dr. G. Simons, NIZO, P.O. Box 20, 6710 BA Ede (The Netherlands) Tel. (31)8380-59557; Fax (31)8380-50400. Abbreviations: A., absorbance; aa, amino acid(s); bp, base pair(s); kb, kilobase(s) or 1000bp; L., Lactococcus; Lb., Lactobacillus; nt, nueleotide(s); oligo, oligodeoxyribonucleotide; ORF, open reading frame; PA, polyacrylamide; PAGE, PA-gel electrophoresis; RBS, ribosome-binding site; S., Streptococcus; SDS, sodium dodecyl sulfate; tsp, transcription start point(s); Usp45, unidentified secreted 45-kDa protein; usp45, gone encoding Usp45; [ ], denotes plasmid-carrier state. 0378-1119/90/$03.50© 1990Elsevier SciencePublishers B.V.(BiomedicalDivision)
1989) and the bacteriocin nisin (Buchman et al., 1988) have been determined. The proteinase is synthesized as a prepro-protein with a consensus Ala/Ala signal peptidase I cleavage site between aa 33 and 34. it has been demonstrated that this signal peptide directs the secretion of heterologous proteins in lactic acid bacteria (Simons etal., 1988). The gene for the bacteriocin nisin encodes a precursor 6.0-kDa polypeptide of 57 aa. A signal peptide of 23 aa has been deduced which deviates from the consensus signal peptide rules as postulated by yon Heijne and Abrahmsen (1989). To gain more insight in the properties and structure of lactococcal extracellular proteins, we screened lactococcal strains for extracellular proteins, and attempted to clone the gene encoding a 45-kDa protein which is secreted into the extracellular medium as an apparent 60-kDa protein and is produced by every lactococcal strain we tested so far. Based on the regulatory sequences of the usp45 gene, secretion vectors specific for lactococci will be constructed.
156 no antimicrobial activity against other Gram + bacteria, assayed by the method of Fowler (1975), could be assigned to the protein.
EXPERIMENTAL AND DISCUSSION
(a) Characterization of secreted proteins of laetococci L. lac~ strains were screened for secreted proteins by analyzing culture supernatant fractions on SDS-PAGE. As shown in Fig. IA, all screened strains secrete proteins ofwhich the most abundant one banded at 50-60 kDa. The amount of this secreted protein was estimated to be 2-4 ttg/ml/A6oonm. Western-blot analysis of the culture supernatant fractions showed a strong reaction between the 50-60-kDa proteins of all tested strains and antibodies against the 60-kDa protein of L. lactis MG1363 (Fig. 1B), isolated from the plasmid-free strain MG1363, indicating an immunological relationship between these proteins. No 50-60-kDa proteins could be detected in cell extracts of the tested strains, indicating that these proteins are tratzsported efficiently into the extracellular medium and are not associated with the cell envelope as is the case for components of the proteolytic system of lactococci (Vos et al., 1989). Extracellular proteins of Leuconostoc paramesenteroides, S. thermophilus, Lb. casei, Lb. acidophilus and Lb. bulgaricus strains showed no reaction with the antibodies (data not shown) suggesting that this 50-60-kDa protein is specific for L. lactis. Since every L. lactis strain we tested produces this 50-60-kDa protein and no L. lactis strains defective in synthesis of this protein were available, it was not possible to postulate a function for this protein. No proteolytic activity ofthe 60-kDa protein of MGl363 could be detected on casein, its degradation products or on synthetic substrates, indicating that the protein has presumably no function in the proteolytic system of L. lactis. In addition,
(b) Cloning and expression of the gone for the 60-kDa protein in EsckericMa eoli Since the plasmid-free strain MG1363 is able to produce the 60-kDa protein, chromosomal DNA of this strain was used to construct a genomic library in phage A47.1 (Loenen and Brammar, 1980). The genomic bank was screened with the isolated antibodies and positive phages were subsequently screened with probes directed against the N terminus of the mature protein. The sequence of the N-terminal 10 aa, as determined with a gas-phase sequenator (Applied Biosystems), was found to be D-T-N-S-D-IA-K-Q-D. Mixed oligo probes were derived against the last six aa of this sequence which had the least degenerate codon usage. Southern-blot analysis showed that the gene for the 60-kDa protein was found to be located on a KpnI-EcoRI fragment of + 3 kb. This fragment was cloned in pUCI9 generating pNZ 1011. Western-blot analysis of cell extracts of E. coli JM83[pNZ 1011], showed synthesis of the protein (Fig. 2, lane 2). Moreover, a substantial amount of the 60-kDa protein could be detected in the periplasmic fraction of the cells (lane 3), where only a small amount of 60-kDa protein was still present in the stripped cells (lane 4). However, the ratio of p-lactamase activity in the periplasmi¢ fraction compared to stripped cells (4: 1) was in complete agreement with the amounts of 60-kDa protein detected within these two fractions (results not shown). Hence, we conclude that the 60.kDa protein is transported efficiently into the periplasmic space of £. coil cells.
A kDa
1
2
3
4
5
6
7
8
B i
2
3
4
,5
6
7
8
9Z, 5 68 43
.... 29
.~,:~i~i )
Fig. 1. Gel analysis ofextracellular proteins ofL. lactbstralns. Supernatantsoflogarithmicogrowingeulturesweredialyzedagalnst water and concentrated by lyophilisation. Samples were analyzed by 0.1% SDS-10% PA gel (Laemmli, 1970) and Western blotting (Towbin et al., 1979). (Panel A) Coomassie brilliant blue staining of 1.0 ml ofsupernatant; (Panel B) Western blot of 0.2 m! ofsupcrnatanL To raise antibodies against the 60-kDa protein of MG1363, the 60-kDa protein was excised from SDS-PA gel and recovered by isota¢hophoresis (Ofverstedt et al., 1983), Lanes: 1, MG1363 (Gasson, 1983); 2, IL!403 (Chopin et aL, 1984); 3, RI*; 4, NZ22186"; 5, NCDO176; 6, SKI 128 (De Vos and Davies, 1984); 7, NCDOI200; 8, WG2* (*NIZO collection). The lanes are flanked on the lel~ side by Mr markers. The position of the $0-60-kDa prnteins is indicated by arrowheads.
157 l
2
3
4
5
Fig. 2. Western blot of proteins encoded by E. co/[ JM83[plqZ1011]. Samples were analyzed by 0.1% SDS-10% PA gel (Laemmli, 1970). Cells were fractionated by osmotic shock (Neu and Heppel, 1965) resulting in a periplasmic and a stripped-cell (cells without periplasmic space) fraction. Lanes: 1, JM83[pUCI9], unfractionated cells; 2, 3, and 4, JM83[pNZ1011]; 2, unfra¢tionated cells; 3, stripped cells; 4, periplasmic fraction; 5, MGI363, culture supernatant. The position of the 60-kDa protein is indicated by an arrowhead.
It has been demonstrated that regulatory sequences of Gram + organisms such as bacilli and staphylococci can function efficiently in E. coli (McLaughlin et al., 1981). Our data show that the regulatory sequences of this lactococcal gene and the signal sequence of its gene product are also recognized in E, cog. (c) Nucleotide sequence of the gem for the O0-kDa protein The nt sequence of the l~pnI.£coRl fragment from pNZI011 was determined and the region encoding the 60-kDa protein is shown in Fig. 3. The nt sequence encoding the first N-terminal 10 aa of the mature protein was found at nt position 97-113 and followed by an ORF encoding a mature protein of 434 aa. This protein has a predicted Mr of 44628 and it was designated U~p45. Upstream from the mature N-terminus, three in-frame start codons are located. The most likely start codon is ATG at nt position 1, since it precedes a precursor region which has the characteristics of a signal peptide: several basic aa followed by a hydrophobic core of 15-20 aa and a shorter uncharged region ending with an aa carrying a small side chain (yon Heijne and Abrahmsen, 1989). Although the highest cleavage probability by signal peptidase I is C-terminal of Ala 19 according to yon Heijne and Abrahmsen (1989), the observed cleavage after Ala 27 in the sequence Val-Xaa-Ala is also commonly found in prokaryotes, and particularly in Gram + bacI~ria. However, N-terminal processing cannot be excluded. The ATG start codon at nt position I is preceded by two
potential RBSs (Shine and Dalgarno, 1974), e.g., the sequence GGAGG at nt -26 to -22 and AAAG at -10 to -8. The calculated free energy of the latter one is only -4.6 kcal/mol (Tinoco et al., 1973). RBSs of genes from lactococci (De Vos, 1987) and other Gram + microorganisms like bacilli (McLaughlin et al., 1981) usually have free energy values between -10 and -15 kcal/mol. For the other possibility a free energy of -14.4 kcal/moi could be calculated. However, the spacing to the ATG codon is 21 bp. This length does not fit with the hypothesis proposed by Shine and Dalgarno (1974). Site-directed mutagenesis experiments need to be performed to determine whether the weak RBS is used and what the contribution is of the G G A G G sequence to the initiation of translation of the usp45 gene. Primer extension experiments (Debarbouille and Raibaud, 1983) were performed and showed that transcription of the usp45 gene was initiated with an A residue at nt position -44 (data not shown). A consensus -10-hexanucleotide sequence (TATAAT) was found at position -56 to -51. The most probable corresponding -35 sequence is TTAAGG at nt position -76 to -71, although the spacing between the two is smaller (14 bp) than that usually found in promoters in L. lactis (De Vos, 1987; Van der Vossen et al., 1987). Inspection of the nt sequence shows the presence of several repetitive sequences: the hexanucleotide sequence GCNCAA occurs 19 times in the gene, of which 16 times in the ORF, whereas the more degenerated sequence TC~AG c appears 16 times. Translation of this sequence would result in twelve Ser doublets of which seven are in tandem in one part of Usp45, resulting in an unusual aa sequence in the protein as discussed in section d, Downstream from the TAG stop codon of the usp4S gene an inverted repeat was found at nt position 1414-1452, which could form a hairpin-like structure with a free energy of -8.2 kcal/mol (Tinoco et al., 1973). The formed hairpin is flanked on both sides by A or T stretches and hence could function as a Rho-independent, bi-directional terminator of transcription. Northern-blot analysis (data not shown) revealed a messenger RNA of approx. 1500 nt which is in agreement with the deduced transcription termination. (d) The aa sequence, homology analysis and structural predictions of Usp45 The aa sequence of Usp45 as shown in Fig. 3 reveals a high content of Ser and Ala residues, together more than 32%. In one region of the protein, from aa 264-316, Ser residues constitute 33 out of 53 aa, including a continuous stretch of 16 aa. Database searches for homology were performed with the SWISSPROT (release 12.0) and the NBRF (release 22.0) aa sequence libraries. The N-terminal 280 aa of
158 X | &,i~ x
N
~.~,.'_]a,_.~.~&ak.C,L,r.kt%~-.~&"~,'kA, I L ' ~ & C l k ~ T & C ' ~ K
I(
r
Z
X
•
A
Z
L
M B
'~
V
C ' ~ CAGCC C 4 : ~ X r Z ~ G G " I L ' ~ & ~
Z
L
S
A
&
A
P
L
':
G
V
~ ¥
~
A
O
T T
&
~
M S
D
~ X
~
A
g
O
O
/k
T
Z
T C ~ _ _.~ ~ & i ~ T ~ A ~ C A C A A G C A ~ J ~ G C A C A A G ~ T A G C T ~ T C A ~ G T ~ C T T A C A A ~ G ~ A ~ ~ ~ C
4].
5
S
/4
Q
8
&
g
A
O
&
Q
&
U
V
O
S
241 G&&p.~qTtdk&GCAC~'TA&TE~N?.&&~T~G~ & IIIL
R
I,
Q
S
K
V
D
S
L
Q
O
K
Q
T
5
•
g
A
Q
X
A
Ig
1[
ig
S
ATC&&&G&&CGTACIC3t AG&~TTGG&J~CTC&&GC&~T>GCTC&P.GT~AP.XGCTC&GC&/~&&&~&TA~
~
g
~k L
M ~.
Q
Z
A
•
~
N
R
~=
X
[g
~
R
•
g
•
L
R
Jk Q
~.
R
S
.~
O
V
N
S
q
~
•
H
"If g
A
V
V
M
IS
K
'=
L
T
D
V
g
Q
K
V
T
A
Z
A
T
V
S
S
A
!1
I(
U
Z
L
R
Q
Q
E
IT
E
O
g
E
L
O
g
M
II
T
V
K
g
a
Y
H
q
P
V
S
L
S
~
S
L
O
S
U
A
q
E
1[, T
5
q
QI &
E
L
K
V
,It, T
L
E
A
0
5
T
36X 12"
O 48~
161
8
601 AACTATC~CA~TTI=CAACTG~GCA~TA~C~ 30!.
N
Y
O
A
T
Z
A
T
A
'0
D
K
K
"~. ~ A T T A G ~ T G / t ~ , & ~G C A G ~ C ~ C A c ~ r ~ A A G A A G C J ~ C ' T ~ C A A G C G G C T T A T I ~ Q
&
L
L
O
E
!(
A
A
&
E
g
k
A
U
E
A
*% g
g
Q
k
A
¥
731 ~tA.%AGA,'~,CAC~P.ACAAG~CAGC'['t~.AACAGCAC~AAC'I~CTAAAGCTel~AGAAGC A ~ C A A ~ A ~ A G ~ ~ ~ ~ A C ~ G C A ~ 341
Q
g
B
A
A
Q
A
U
A
A
S
?
A
A
•
A
g
A
V
E
A
A
T
841 GATAATACA~TCAAATGCTAG~'GC~.AAACAGTT~&~?J~PI~?ATC,~J~AC'TCAAGT TCA~ 36L
D
M T
T
5
M
A
S
A
6
M S
8
H
'=
S
A
N
S
5
S
S ~
8
8
S
A
S
& 8
A
~
8
5
~1 S
S
G -'= S
~ S
H
T
A
F
G
fl
961 AATACAAATTCAGGCACTAGTACT~T&OgGGAGGAAI~CTA~'~GTGGT&GC~CT ATA~TAG'I~CCAATTGC ~ T ~ & T G 321
U
8
U
V
~
9
9
T
~
B
~
S ~ 8
M
N
A
G
G
&~I'GTACTG~AT~'TATGG
H
S
G
T
8
T
G
N
T
G
q
T
T
T
G
G
9
~
g
H
8
-q
p
z
G
H
P
V
A
G
G
O
¢
T
D
Y
V
H
Y
P
A
A
Q
G
Z
Y
Z
R
N
|
M
IP
G
H
G
G
O
5/
A
S
N
G
P
A
Q
G
Y
L
II
V
V
G
A
A
P
G
V
A
|
B
P
B
/L
0
IF
V
G
¥
k
M
8
P
Y
G
II
V
AIr
V
K
8
V
g
8
D
O
T
Z
T
E
K
I
O
G
Y
G
t0OX
36}.
q 1.:101
4,1.
Z
12311, A C A A ~ q ~ C k T ~ C G T A C T ~ T ~ C G T ~ A ~ q ~ & T O C e A A ~ A ~ A A T A A A T A ~ . . ~ T A ~ T ~ T ~ 44~
T
i
T
M
O
H
e
M
T
V
M
A
Ii
G
V
T
V
L
N
P
M
1.441. TT~T~.~TTI~&~ATt~kTAATAATAAG~A~TGATA
Fig, 3, The nt sequence of the usp4$ gone and the deduced aa sequence. Seque,~ctng was performed following th~ Sequcnase protocol of' Tabor and Richardson (1987). The doable underlined sequences represent the -35 and -10 promoter boxes. The ~p is indicated by an upward arrowhead. The two putative RBSs as well as the determined N.terminal an (Asp 3B -Asp 3'/ ) ofthe mature protein ere underlined, Downstream from the gone an inverted repeat Is marked
with dashed arrows representing a Rho-independent
terminator
of transcription.
The OenBank
accession number
of this nt sequence
is M35374.
A ......
• I ....
I,.
.I..I
I .......
.....
I .I...
I, ,I,,.I
I • .. I.,,
,I.1.1,,I,
I .I ....
I,. IIl,.ll
QVg A| Wr~lk0OLd,TIg. UflPLMMJgBAQLIgQIg[ ledOLOgM! Iglgl4EA~ |
Q
~
..... II
H¥ |~10
GIJ)BQ&QIgLWl~llghgV~?lJ4YQATI ATe~MQgQ~I~IMglAAJ~IglgAAQIg&AR~y~ ~110
• I II
.I .-I1,1,1
, ,I.I.
II1,111,,I
,ll.il
I,I.
..11
A I ~ " I ~ U I J dM~QJ~I L~iViaTru Ii%AI~A?A k~UKI~I)LHIUU~I~K-~K~R ! I @ |
! .... I...
|, -I . I,..I ....
II- .,
.,
~g
2]U
I11,.
Fig. 4. Homology between Usp4$ and P54 determined according to Devet~ux et al, (1984), Depicted is our interpretation of translation of the
Usp45 show weak homology with myosins, keratins and streptococcal M proteins, which presumably reflects a common type of 0c-helical secondary structure (Cohen and Parry, 1986). Significant homology was detected with the recently published sequence of P54 protein of Enterococcusfaecium (Ftlrst et al., 1989), as shown in Fig. 4. This protein of unknown function is attached to the ceil-wall, and is generated from a 516-aa precursor. In the N-terminal 281 aa, including the signal peptide, 31 ~ ofthe residues are identical and another 53~o are conservative substitutions (Fig. 4A). published nt sequence of P54 (FOrst, 1989), which leads to an additional 9 aa residues at the N terminus compared to the published aa sequence. Identical residues arc connected by lines, conserved substitutions arc indicated by points. (A)aa 1-281; (B)aa 281-340.
159 TABLE I
Characteristics of the Usp45 and P54 domains" Domains
Residues Main aa Secondary structure Charge Hydropathy Flexibility
Signal peptide
Ab
Usp45
P54
Usp45
P54
1-27 A,S,I
1-28 A,S,I
28-281 A,Q,S,K
29-281 A,Q,K,E
~-helix low; positive hydrophobie low
,,-helix high; neutral amphipathic low
B¢
Usp45 1
--
Cd
D•
P54
Usp45
P54
Usp45
P54
282-336 S,T,F
282-340 S,N,T,G
337-395 N,T,S,P
341-461 G,A,V,S
396--516 G,V,A,S
turn/coil high; negative hydrophilic high
turn/coil none hydrophilic high
~-sheet low; neutral hydrophobic low
" Secondary structure, hydropathy and flexibilitypredictions are determined using the Pcgene (version 5.01) nucleic-acid and protein sequence analysis software system (Genofit). u Segment A, immediately alter the signal peptide, is predicted to be almost exclusively ae-helicalby the method of Gamier (Gamier et al., 1978). c Segment B is unique for P54. d Segment C consists mainly of the aa Ser, Ash, Gly and Thr (Pro instead of Gly in P54) which are commonly found in turns and random coil secondary structure. Characteristic is the long stretch of Ser and Asn aa in both proteins at the N-terminal end ofthis segment, as shown in Fig. 4, while Gly and Pro residues are concentrated at the C-terminal end. Segment C is also extremely hydrophih'c: 80~o of the aa in Usp45 contain either a hydroxyl or carboxyi group, but there are no charges in this segment. © Segment D is predicted to have largely ~-sheet secondary structure in short stretches, alternating with turns or bends. The aa compositions of segment D are very similar in Usp45 and P54, and contain much higher contents of aromatic and other hydrophobic as than segments A, B and C.
Besides the homology on primary aa sequence level, there appears to be an overall resemblance ofthe two proteins on the secondary and tertiary structure level. On the basis of structural predictions ~ d aa compositions, the two mature proteins can be divided into four segments with similar characteristics, as outlined in Table I. (f) Conclusions L. lactis strains produce extracellular secretory proteins with an apparent Mr of 50-60 × l0 s, which are strongly immunologically related. The gene for this protein of strain MG 1363, designated usp45, is chromosomally encoded and was cloned and expressed in E. coli under control of its own regulatory sequences. Moreover, the gene product was secreted efficiently into &e periplasmic space of this organism. The deduced aa sequence shows an unusual composition and distribution. Usp45 shows aa sequence homology with P54, an extracellular protein of Enterococcus faecium. Besides homology on the aa level the two proteins have similar structural characteristics, but until now their function remains unclear. The cloning ofthe usp4.~gene may be an introduction for further investigation of the secretory pathway of L. lactis and could lead to the construction of expression and secretion vectors in these food-grade dairy micro-organisms for the production of homologous and heterologous proteins.
ACKNOWLEDGEMENTS
We thank Joop Mondria, Simon van der Laan and Henk van Brakel for art work and photography. This work was financed by the Mesdag Foundation and the Cooperative Rennet and Dye Factory (CSKF).
REFERENCES
Buchmun, G.W., Banerjee, S. and Hansen, J.N.: Structure, expression and evolution of a gene encoding the precursor of Nisin, a small protein antibiotic. J. Biol. Chem. 263 (1988) 16260-16266. Chopin, A., Chopin, M.C., Moillo-Batt, A. and Langella, P.: Two plasmid-determined restriction and modification systems in 5, lac~s. Plasmid 11 (1984) 260-263. Cohen, C. and Parry, A.D.: ",-Helical coiled coil - a widespread motif in protein. Trends Biochem. Sci. I ! (1986) 245-248. Debarbouille, M. and Raibaud, O.: Expression of the £. co//malPO. operon remains unaffected after drastic alterations of its promoter. J. Bacteriol. 153 (1983) 1221-1227. Deverenx, J., Haeberli, P. and Smithies, O.: A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Reg. 12 (1984) 387-395. De Vos, W.M.: Gene cloning and expression in lactic streptococci. FEMS Mierobiol. Rev. 46 (1987) 281-295. De Vos, W.M. and Davies, F.L.: Plasmid DNA in lactic streptococci: bacteriophage resistance and proteinase plasmids in $. cremoris SKI !. 3rd European Congress on Biotechnology I!1. Vedag Chemic, Munich, 1984, pp. 201-206. Fowler, G.G., Jarvis, B. and Tramer, J.: The assay of nisin in foods. Sec. Appl. Bacteriol. Tech. Set. 8 (1975) 91-105.
160 Filrst, P.E., M6sch, H.U. and Solioz, M.: A protein of unusual composition from £ntemcoccusfaecium. Nucleic Acid~ Res. 17 (1989) 6724. Gamier, J., Osgnthorpe, DJ. and Robson, B.: Analysis of the accuracy and implications of simple methods for predicting the secondary structure of globalar proteins. J. Mol. Biol. 120 (1978) 97-120. Gasson, M J.: Plasmid complements of Streptococcus lactis NCDO712 and other lactic streptococci alter protoplast induced curing. J. Bacteriol. 154 (1983) 1-9, Laennnli, U.K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 (1970) 680-685. Loenen, W.A. and grammar, WJ.: A bacteriophage lambda vector for cloning large DNA fragments made with several restriction enzymes. Gone 10 (1980) 249-259. MeLaughlin, J.R., Murray, C.L. and Rabinowitz, J.C.: Unique features in the ribosome-binding site sequence ofthe Gram-positive Staphylococcus aureus ~-Iactamase gone. J. Biol. Chem. 256 (1981) 11283-11291. Neu, H.S. and Heppel, L.A.: The release of enzymes from E. coil by osmotic shock during the formation of spheroplasts. J. Biol. Chem. 240 (1965) 3685-3692. Ofverstedt, L.G., Sandelin, J. and Johansson, G.: Recovery of proteins on a milligram scale from polyacrylamide electrophoresis gels, exemplified by purification of a retinol-binding protein. Anal. Biochem. 134 (1983) 361-365. Shine, J. and Dalgarno, L.: The 3' terminal sequence of E. coil 16S ribosomal RNA: complementarity to nonsense triplets and ribosomebinding site. Proc. Natl. Acad. Sol. USA 71 (1974) 1342-1346,
Simons, G., Rutten, G., Homes, M. and De Vos, W.M.: Production of bovine prochymosin by lactic acid bacteria. In Breteler, H., Lelyveld, P.H. and Luyben, K.C.A.M. (Eds.), Proceedings 2nd Netherlands Biotechnology Congress, Netherlands Biotechnological Society, Amsterdam, The Netherlands, 1988, pp. 183-187. Tabor, S. and Richardson, C.C.: DNA sequencing analysis with a modified bacteriophage T7 polymerase. Proc. Natl. Acad. Sci. USA 84 (1987) 4767-4771. Tinoco Jr., I., Borer, P.N., Dengler, B., Levine, M.D., Uhlenbeck, O.C., Crothers, D.M. and Gralla, J.: Improved estimation of secondary structure in ribonucleic acids. Nature 246 0973) 40-41. Towbin, H., Staehelin, T. and Gordon, J.: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76 (1979) 4350-4354. Van der Vossen, J.M.B.M., Van der Lelie, D. and Venema, G.: Isolation and characterization of Streptococcus cremoris Wg2 specific promoters. Appl, Environ. Microbiol. 53 (1987) 2452-2457. yon Heijne, G. and Abrahmsen, L.: Species specific variation in signal peptide design. Implication for protein secretion in foreign hosts. FEBS Lett. 244 (1989) 439-446. Vos, P., Simons, G., Siezen, R.J. and De Vos, W.M.: Primary structure and organization of the gone for a prokaryotic cell envelope-located serine protease. J. Biol. Chem. 264 (1989) 13579-13585.
