Appl Microbiol Biotechnol (1991) 36:61-64 0175759891002410

App//ed Microbiologr Biotechnology © Springer-Verlag 1991

Production of heterologous proteins in Bacillus subtilis: the effect of the joint between signal sequence and mature protein on yield Harri Hemilii 1 and Mervi Sibakov 2

1 Institute of Biotechnology, Valimotie 7, 00380 Helsinki, Finland ~ Valio, Finnish Co-operative Dairies' Association, P. O. Box 176, 00181 Helsinki, Finland Received 8 April 1991/Accepted 17 June 1991

Summary. We have previously made a set of D N A constructs by fusing the mature part of Bacillus licheniformis a - a m y l a s e with the signal sequence o f B. amyloliquefaciens a - a m y l a s e at various distances from the signal sequence cleavage site. We observed that the level of a - a m y l a s e production in B. subtilis depended strongly on the distance of the junction from the signal sequence cleavage site, with quite a sharp o p t i m u m distance. To test whether the effect is limited to the pair of a - a m y l a s e signal sequence and mature protein, we analysed the protein production in a set of constructs in which an Escherichia coli fl-lactamase was similarly joined at different distances from the a - a m y l a s e signal sequence. Also in this case the distance seemed to be an important factor in affecting the level of production in B. subtilis. T h e observed effect might depend on the modulation of pre-protein folding, which in turn could affect the secretion level.

Introduction

Considerable efforts have been m a d e to produce heterologous proteins in Gram-positive bacteria with the help of p r o m o t e r s and signal sequences derived from secretory proteins (Palva 1989; U d a k a et al. 1989). We have attempted the production of several secretory proteins in Bacillus subtilis with the aid of expression vectors b a s e d on B. arnyloliquefaciens a - a m y l a s e p r o m o t e r and signal sequence (Palva et al. 1982; Sibakov 1986; Hemil~i et al. 1989; Heikinheimo et al. 1991). The general features required for the proper functioning of signal sequences have been extensively studied (Oliver 1985; Gierasch 1989). However, little attention has been focused on the type o f fusion between a heterologous pair of signal sequence and mature protein, and the characteristics required for maximal production capacity of the foreign protein. In attempting

Offprint requests to: M, Sibakov

to produce thermostable a - a m y l a s e in B. subtilis we m a d e several types of signal sequence to mature fusions to optimize enzyme production (Sibakov 1986). Although a fourfold variation in the production level was observed between the constructs, we could not correlate the differences with the amino acid patterns at the junctions. In connection with further studies related to the production of heterologous proteins in B. subtilis, we have made a closer analysis of the original results.

Materials and methods

Bacterial strains and plasmids. B. subtilis 168 derivative BRB1 (sacA321 metB5) was used as a host strain for the expression of fl-lactamase. Plasmids pKTH78, pKTH83 and pKTH84 have been previously described (Palva et al. 1982). Plasmid pKTH10 contains the B. amyloliquefaciens a-amylase gene in pUBll0 (Takkinen et al. 1983). Bacteria were propagated on Luria plates (Ausubel et al. 1987) supplemented with kanamycin (10 ~tg/ml). DNA methods. All the basic methods for DNA manipulation were based on Ausubel et al. (1987), Plasmid pKTH2044 was obtained as follows: pKTH10 was digested with EcoRI (site located 190 bp downstream from the cleavage site of a-amylase signal sequence) and treated with Bal31 to obtain a series with different lengths of mature a-amylase. The gene for E. coli TEM-fl-laetamase was cut off from pKTH78 by HindIII digestion, treated with Klenow enzyme to blunt the sticky ends, and ligated to the Bal31-series made of pKTH10. B. subtilis BRB1 was transformed with the ligation mixture according to Gryczan et al. (1978), and the signal sequence-mature joint of selected clones was sequenced directly from several plasmids as previously described (Hemil~i et al. 1989). Most of the resulting constructs had a long stretch of the a-amylase mature protein (over 20 amino acids) and only one of those characterized, pKTH2044, fell into the desired region. Flask cultivation ofB. subtilis. Cells were grown in Plasmid Broth (PB) medium (Hemil~i et al. 1989) supplemented with 10 ~tg kanamycin/ml, and 5% glucose, in a rotary shaker at 37° C. Samples were taken 3 h after the culture had reached an absorbance at 660 n m (A660) of 1.0. fl-Lactamase assay and gel electrophoresis, fl-Lactamase activity was determined by a spectrophotometer using nitrocefin (BBL Microbiology Systems, Cockeysville, Md., USA) as chromogenic

62 substrate (O'Callaghan et al. 1972). The rate of enzymatic reaction was calculated from the change in absorbance at 482 nm (25° C) one unit hydrolyses 1 ~tmol nitrocefin in 1 min. Proteins of the culture samples were precipitated with 10% trichloroacetic acid, followed by suspension into sample buffer and separation by 15% sodium dodecyl sulphate-polyacrylamide gel electrophoresis according to Laemmli (1970). The gels were stained with Coomassie Brilliant Blue R-250 (Serva, Heidelberg, FRG) according to Ausubel et al. (1987) and the fl-lactamase protein band was scanned with an Ultroscan XL densitometer (Pharmacia, Uppsala, Sweden). a-Amylase (Sigma, St. Louis, Mo., USA) was used as a standard for protein quantification.

•

Protein (%)

[ ~

Activity (%)

4O

-20

7FI

30

Results and discussion Previously we made a set of constructs for the production of thermostable B. licheniformis a-amylase in B. subtilis. The mature part of B. licheniformis a-amylase was fused to the signal sequence of B. amyloliquefaciens a-amylase (Sibakov 1986). The joints were located closely after the authentic cleavage site of the signal sequence (Fig. 1). Eight constructs were made that differed at the joint region. In four o f the constructs the proteins were fused directly to each other at different positions and in another four a HindlII-linker was located at the joint (Fig. 1B). There was over fourfold variation in the extracellular production level of a-amylase

signal sequence

+

mature

A B. amyloliquefaciens

I

I

20

10

10

213

I

I

I

0

I

I

I

I

I

5

I

I

I

I

I

10 D(aa)

Fig. 2. The extracellular production level of B. licheniformis aamylase as a function of the distance D (see Fig. 1). The relative protein amount ( • ) and the relative amylase activity ([]) are plotted. For both parameters 100% corresponds to the production level of authentic B. amyloliquefaeiens a-amylase in B. subtilis. The plot is based on the data in Fig. 4 in Sibakov (1986) and the numbering of the constructs corresponds to the same reference: aa, amino acids

B. lichenfformis

D ~ B e-amylase (B. amylofiquef.) (r-amylase (B. licheniformis)

mature

.oTSA V N G T L M Q Y F E WYTP.. ..AAA A N L N G T L M Q Y FEWY..

Construct

~ 2 3 4 5 e Z 8

(O=7) (O=7) (O=9) (O=6) (O=o) (O=3) (D=O) (D=-2)

..TSA ..TSA ..TSA ..TSA ..TSA ..TSA ..TSA ..TSA

VNGTAKLANL VNGTLQACNL VNGTLQACi%R ASLPAAI%NLN VNLNGTLMQY VNGANLNGTL VNGNGTLMQY VNGTLMQYFE

NGTL.. NGTL.. NLNG.. GTLM.. FEW~.. MQYF.. FEWY.. WYMP..

Fig. 1. A The structure of the a-amylase constructions. The middle region consists of mature Bacillus amyloliquefaciens a-amylase, HindlII-linker a n d / o r 1-2 C-terminal residues from the signal sequence of the B. licheniformis a-amylase (Sibakov 1986), Distance D is the number of amino acids between the signal sequence cleavage site (arrow) and the N-terminus of mature B. lio cheniformis a-amylase, which is extrapolated by comparison to the authentic sequence when the native N-terminal residue is absent. A negative value for D results when a truncated mature aamylase is fused to the native signal sequence. B The joint region of the a-amylase constructions. Signal sequence cleavage site for B. amyloliquefaciens a-amylase (arrow), HindlII-linker (underlinin~?) and the B. licheniformis a-amylase (bold) are indicated

among the constructs, but the variation could not be attributed to differences in the amino acid patterns in the joint region (Sibakov 1986). No accumulation of aamylase was observed in the cells in any construct (Sibakov 1986). In a closer analysis of the data we found that the distance (D) between the signal sequence and the mature protein is a significant factor in determining the production level. Figure 2 indicates that four to five amino acids would be the optimal distance in this series of constructs with respect to the protein production. Accordingly, the fusion of the authentic mature B. licheniformis a-amylase to the signal sequence ( D = 0 ) would not result in maximal protein production. However, maximal enzymatic activity is obtained with joint distances o f 0-3 amino acids. The discrepancy between protein production and enzymatic activity might be due to changes in the specific activity caused by variation in the N-terminus of the mature amylase. The N-terminal regions of the mature a-amylase o f B. licheniformis and B. amyloliquefaciens are nearly identical except that the latter is two residues shorter than the former (Stephens et al. 1984). Therefore, with respect to homology, the most native-like joint to the signal sequence would correspond to the distance

63 ~ e-amylase B-lactamase (mature)

..TSA

m~u~ VNGTLMQYFE

WYTPNDGQHW

..

HPETLVKVKD

~EDQLGI%RVG

..

1

(D=~

..TSA

QACPPETLVK

VKDAEDQLGA

..

2 3 4

(D=~ (D=~ (D=I ~

.. T S A

VNGTQACPPE

TLVKVKDAED

..

..TSA

VNGTLACPPE

TLVKVKDRED

..

plasmid pKTH78 pKTH~ pKTH2044

..TSA

VNGTLMQYFE

WYTPQACPPE

TL.

pKTH84

Con~ru~

Protein (%)

0 100 -

Activity (%)

[] -100

1

0

2

50

50 4

[]

I

0

I

I

I

I

10

I

I

I

I

I

20

I

I

O (aa)

Fig. 4. fl-Lactamaseprotein (0) and enzymaticactivity(rq) in the culture medium of B. subtilis BRB1 carryingthe fl-lactamase constructs: 100% corresponds to 28 mg/1 offl-lactamaseprotein and 31 units/ml of enzymaticactivityin the culture medium. For the numbering of the constructs see Fig. 3

D = - 2 amino acids, but only low activity was obtained with such a joint (construct 8). Constructs 2 and 3 contain 5 amino acids, and construct 8 contains 12 amino acids of the mature B. amyloliquefaciens a-amylase. Accordingly, these constructs have native-like surroundings for the signal sequence cleavage site of the B. arnyloliquefaciens a-amylase. However, the production level is much lower when compared to the best constructs and therefore the native-type cleavage site does not seem to be of primary importance. Thus, our results suggest that the determination of optimal joint distance is an empirical task, since none of the obvious alternatives: "exact joint" ( D = 0 amino acids), homologous joint (D = - 2 amino acids) or the joints preserving the surrounding of the cleavage site, represents the optimum in this series of constructs. To test whether the observed phenomenon is limited to the pair of signal sequence and mature protein originating from closely related a-amylases, we analysed an-

Fig. 3. The joint region of the fl-lactamase constructions. Intact signal sequence of B. amyloliquefaciens aamylase was joined to mature E. coli fl-lactamase at different distances. Signal sequence cleavage site (arrow), HindlII-linker (underlining) and the mature fl-lactamase (bold) are indicated

other set of fusions between the B. amyloliquefaciens aamylase and Escheriehia coli TEM-fl-lactamase (Figs. 3 and 4). This set of four constructs also showed a marked distance effect on the level of protein production. Ohmura et al. (1984) made three constructs to join the E. coli fl-lactamase to the signal sequence of B. subtilis a-amylase at different distances. The construct with the longest stretch of mature a-amylase before the fl-lactamase (D = 13) produced half the enzyme activity of that compared to the best constructs ( D = 3 and D = 5). Thus, according to the available data, the phenomenon seems not to be limited to the pair of homologous amylases. There are only few studies on the role of the N-terminal part of the mature protein on protein secretion, and in most of these a homologous pair of signal sequence and mature protein has been used. Li et al. (1988) found that the charge of the N-terminal region of the mature phosphatase affects its secretion in E. coli. Secretion was efficient with a neutral N-terminus but markedly reduced with two positive charges in the N-terminus. In all of our a-amylase constructs (Fig. 1B) the first five amino acids of the mature protein were neutral and accordingly this did not seem to affect the results. Pl~ckthun and Knowles (1987) found that deletion of several residues from the N-terminus of mature E. coli fl-lactamase led to a progressive decrease in secreted enzymatic activity. However, if only the first amino acid was deleted, there was no secretion at all, suggesting some other effect. Duffaud and Inouye (1988) expressed the OmpA (signal) - Staphylococcal nuclease A (mature) hybrid in E. coli. They found that deletion of one and two residues close to the N-terminus of the modified nuclease caused a progressive reduction in the rate of signal sequence cleavage. This was attributed to a decrease in the turn probability at the cleavage site. It has been found that the signal sequence, in addition to facilitating the binding of nascent protein to membrane, retards the folding of the secretory protein and thereby may affect the commitment of the protein to the secretion pathway (Park et al. 1988, Laminet and Pl~Jckthun 1989, Liu et al. 1989, Randall and Hardy 1989). Secretion is also dependent on the mature part of the protein even though no specific secretion-promoting signals have been found (MacIntyre and Henning 1990). A mutation has been found in the mature part of maltose-binding protein that affects secretion, and this has been attributed to a reduced rate of folding of the mutated protein (Liu et al. 1988, 1989). It appears reasonable to suppose that manipulation of the N-terminal

64 region o f the m a t u r e part o f the protein m a y affect the folding o f the p r e c u r s o r and thereby also the level o f secretion. There m a y be an optimal distance b e t w e e n the signal sequence a n d the mature part to allow efficient interactions regarding the rate o f pre-protein folding. This w o u l d explain our observations a n d those o f Pliickthun a n d K n o w l e s (1987). A n o t h e r e x p l a n a t i o n o f our results w o u l d be a reduction in the rate o f signal sequence cleavage due to local effects at the cleavage site ( D u f f a u d and I n o u y e 1988). H o w e v e r , the a - a m y l a s e and fl-lactamase constructs with the longest native s u r r o u n d i n g for the cleavage sites (constructs 3 a n d 4, respectively) were not the best b u t the p o o r e s t in the series, which suggests that this is not an a p p r o p r i a t e explanation for our resuits. Furthermore, the r e d u c t i o n in the rate o f signal sequence cleavage observed b y D u f f a u d a n d I n o u y e (1988) does not directly indicate that the cleavage reaction is involved, but an earlier step m a y be affected as well, Acknowledgements. The authors are grateful to Drs. M. Simonen

and T. Koivula for critical reading of the manuscript. This work was supported by a grant from the Foundation for Biotechnical and Industrial Fermentation Research.

References

Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K (1987) Current protocols in molecular biology. Wiley Interscience, New York Duffaud G, Inouye M (1988) Signal peptidases recognize a structural feature at the cleavage site of secretory proteins. J Biol Chem 263:10224-10228 Gierasch LM (1989) Signal sequences. Biochemistry 28:923-930 Gryczan TJ, Contente S, Dubnau D (1978) Characterization of Staphylococcus aureus plasmids introduced by transformation into Bacillus subtilis. J Bacteriol 134:318-329 Heikinheimo R, Hemil~ H, Pakkanen R, Palva I (1991) Production of pectin methylesterase from Erwinia chrysanthemi B374 in Bacillus subtilis. Appl Microbiol Biotechnol 35:51-55 Hemil~ H, Glode LM, Palva I (1989) Production of diphtheria toxin CRM228 in B. subtilis. FEMS Microbiol Lett 65:193198 Laemmli UK (1970) Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 227:680-685 Laminet AA, Pliickthun A (1989) The precursor of fl-lactamase: purification, properties and folding kinetics. EMBO J 8:14691477 Li P, Beckwith J, Inouye H (1988) Alteration of the amino terminus of the mature sequence of a periplasmic protein can se-

verely affect protein export in Escherichia coli. Proc Natl Acad Sci USA 85:7685-7689 Liu G, Topping TB, Cover WH, Randall LL (1988) Retardation of filding as a possible means of suppression of a mutation in the leader sequence of an exported protein. J Biol Chem 263:14790-14 793 Liu G, Topping TB, Randall LL (1989) Physiological role during export for the retardation of folding by the leader peptide of maltose-binding protein. Proc Natl Acad Sci USA 86:92139217 Maclntyre S, Henning U (1990) The role of the mature part of secretory proteins in translocation across the plasma membrane and in regulation of their synthesis in Escherichia coli. Biochimie 72:157-167 O'Callaghan CH, Morris A, Kirby SM, Shingler AH (1972) Novel method for detection of fl-lactamase by using a chromogenic cephalosporin substrate. Antimicrob Agents Chemother 1:283-288 Ohmura K, Nakamura K, Yamazaki H, Shiroza T, Yamane K, Jigami Y, Tanaka H, Yoda K, Yamasaki M, Tamura G (1984) Length and structural effect of signal peptides derived from Bacillus subtilis a-amylase on secretion of Escherichia coli fllactamase in B. subtilis cells. Nucleic Acids Res 12:53075319 Oliver D (1985) Protein secretion in Escherichia coli. Annu Rev Microbiol 39: 615-648 Palva I (1989) Engineering for secretion of proteins by bacteria. In: Baumberg S, Hunter IS, Rhodes PM (eds) Society for General Microbiology Symposium, 44. Cambridge University Press, Cambridge, UK, pp 255-269 Palva I, Sarvas M, Lehtovaara P, Sibakov M, Kii~iriainen L (1982) Secretion of Escherichia coli fl-lactamase from Bacillus subtilis by the aid of a-amylase signal sequence. Proc Natl Acad Sci USA 79:5582-5586 Park S, Liu G, Topping TB, Cover WH, Randall LL (1988) Modulation of folding pathways of exported proteins by the leader sequence. Science 239:1033-1035 Pliickthun A, Knowles JR (1987) The consequences of stepwise deletions from the signal-processing site offl-lactamase. J Biol Chem 262:3951-3957 Randall LL, Hardy SJS (1989) Unity in function in the absence of consensus in sequence. Science 243:1156-1 | 59 Sibakov M (1986) High expression of Bacillus licheniformis a-amylase with a Bacillus secretion vector. Eur J Biochem 155:577581 Stephens MA, Ortlepp SA, Ollington JF, McConnell DJ (1984) Nucleotide sequence of the 5' region of the Bacillus licheniformis a-amylase gene: comparison with the B. amyloliquefaciens gene. J Bacteriol 158:369-372 Takkinen K, Pettersson RF, Kalkkinen N, Palva I. S6derlund H, Ka~iri~inen L (1983) Amino acid sequence of a-amylase from Bacillus amyloliquefaciens deduced from the nucleotide seq u e n c e of the cloned gene. J Biol Chem 258:1007-1013 Udaka S, Tsukagoshi N, Yamagata H (1989) Bacillus brevis, a host bacterium for efficient extracellular production of useful proteins. In: Russell GE (ed) Biotechnology and genetic engineering reviews, vol 7. Intercept, Andover, UK, pp 113-146