US 20050169937 A1
The invention relates to surface display of proteins on microorganisms via the targeting and anchoring of heterologous proteins to the outer surface of cells such as yeast, fungi, mammalian, plant cells, and bacteria. The invention provides a proteinaceous substance comprising a reactive group and at least one attaching peptide including a stretch of amino acids having a sequence corresponding to at least a part of the consensus amino acid sequence listed in FIG. 10 and further includes a method for attaching a proteinaceous substance to the cell wall of a microorganism comprising the use of the attaching peptide.
22. A proteinaceous substance comprising:
at least one stretch of amino acids derived from a first microorganism, wherein the proteinaceous substance is capable of attaching to a cell wall of a second microorganism, said stretch of amino acids comprising at least one AcmA repeat or a derivative thereof, and
a reactive group selected from the group consisting of antibiotics, hormones, aromatic substances, reporter molecules, antigenic determinants, enzymes, single-chain antibodies or fragments thereof, polyhistidyl tags, and fluorescing proteins.
23. The proteinaceous substance of
24. The proteinaceous substance of
25. The proteinaceous substance of
26. A pharmaceutical composition comprising the proteinaceous substance of
27. A vaccine comprising the proteinaceous substance of
28. A composition that attaches to a bacterial cell wall, said composition comprising:
an amino acid sequence comprising at least one AcmA repeat, and, associated therewith, a reactive group.
This application is a divisional of U.S. Ser. No. 09/554,354, filed May 12, 2000, which is a national entry of PCT International Patent Application No. PCT/NL98/00655, filed on Nov. 12, 1998, (corresponding to PCT International Publication No. WO 99/25836, the contents of which is incorporated by this reference as is the sequence listing in U.S. Ser. No. 09/554,354).
The invention relates to surface display of proteins on microorganisms via the targeting and anchoring of heterologous proteins to the outer surface of cells such as yeast, fungi, mammalian, plant cells, and bacteria.
Heterologous surface display of proteins (Stahl and Uhlen, TIETECH May 1997, 15, 185-192) on recombinant microorganisms via the targeting and anchoring of heterologous proteins to the outer surface of host-cells such as yeast, fungi, mammalian and plant cells, and bacteria has been possible for several years. Display of heterologous proteins at these cells' surfaces has taken many forms, varying from the expression of reactive groups such as antigenic determinants, heterologous enzymes, (single-chain) antibodies, polyhistidyl tags, peptides, and other compounds. Heterologous surface display has been applied as a tool for research in microbiology, molecular biology, vaccinology, and biotechnology, and several patent applications have been filed.
Another application of bacterial surface display has been developing live-bacterial-vaccine delivery systems. The cell-surface display of heterologous antigenic determinants has been advantageous for inducing antigen-specific immune responses when using live recombinant cells for immunization. Another application has been the use of bacterial surface display in generating whole cell bioadsorbents or biofilters for environmental purposes, microbiocatalysts, and diagnostic tools.
In general, one has used chimeric proteins consisting of an anchoring or targeting part specific and selective for the recombinant organism used and has combined this part with a part comprising a reactive group as described above. A well known anchoring part, for example, comprises the so-called LPXTG box that binds covalently to a Staphylococcus bacterial surface, i.e., in the form of a fully integrated membrane protein. In this way, chimeric proteins are composed of at least two (poly)peptides of different genetic origin joined by a normal peptide bond. For example, in PCT International Publication Number WO 94/18830 relating to the isolation of compounds from complex mixtures and the preparation of immobilized ligands (bioadsorbents), a method has been claimed for obtaining such a ligand which comprises anchoring a binding protein in or at the exterior of the cell wall of a recombinant cell. The binding protein is essentially a chimeric protein produced by the recombinant cell, and is composed of an N-terminal part, derived from, for example, an antibody, that is capable of binding to a specific compound joined with a C-terminal anchoring part, derived from an anchoring protein purposely selected for being functional in the specific cell chosen. In patent application WO 97/08553 a method has been claimed for the targeting of proteins selectively to the cell wall of Staphylococcus spp, using as anchoring proteins long stretches of at least 80-90 amino acid long amino acid cell wall-targeting signals derived from the lysostaphin gene or amidase gene of Staphylococcus which encode for proteins that selectively bind to Staphylococcus cell wall components.
Vaccine delivery or immunization via attenuated bacterial vector strains expressing distinct antigenic determinants against a wide variety of diseases is now commonly being developed. Recently, mucosal (for example nasal or oral) vaccination using such vectors has received a great deal of attention. For example, both systemic and mucosal antibody responses against an antigenic determinant of the hornet venom were detected in mice orally colonized with a genetically engineered human oral commensal Streptococcus gordonii expressing the antigenic determinant on its surface (Medaglini et al., PNAS 1995, 2; 6868-6872). Also, a protective immune response could be elicited by oral delivery of a recombinant bacterial vaccine wherein tetanus toxin fragment C was expressed constitutively in Lactococcus lactis (Robinson et al., Nature Biotechnology 1997, 15; 653-657). Especially mucosal immunization as a means of inducing IgG and secretory IgA antibodies directed against specific pathogens of mucosal surfaces is considered an effective route of vaccination. Immunogens expressed by bacterial vectors are presented in particulate form to the antigen-presenting cells (for example M-cells) of the immune system and should, therefore, be less likely to induce tolerance than soluble antigens. In addition, the existence of a common mucosal immune system permits immunization on one specific mucosal surface to induce secretion of antigen-specific IgA, and other specific immune responses at distant mucosal sites. A drawback to this approach is the potential of the bacterial strain to cause inflammation and disease in itself, potentially leading to fever and bacteraemia. An alternative approach avoids the use of attenuated bacterial strains that may become pathogenic themselves by choosing recombinant commensal bacteria as vaccine carriers, such as Streptococcus spp. and Lactococcus spp.
However, a potential problem with such recombinant organisms is that they may colonize the mucosal surfaces, thereby generating a long-term exposure to the target antigens expressed and released by these recombinant microorganisms. Such long term exposure can cause immune tolerance. In addition, the mere fact alone that such organisms are genetically modified and contain recombinant nucleic acid is meeting considerable opposition from the lay public as a whole, stemming from a low level of general acceptance for products containing recombinant DNA or RNA. Similar objections exist against the use of even-attenuated strains of a pathogenic nature or against proteins or parts of proteins derived from pathogenic strains. However, as explained above, present techniques of heterologous surface display of proteins in general entail the use of anchoring or targeting proteins that are specific and selective for a limited set of microorganisms which in general are of recombinant or pathogenic nature, thereby greatly restricting their potential applications.
The invention provides substances and methods to anchor or attach the substances to a cell wall or cell wall component of a wide range of microorganisms. A preferred embodiment of the invention provides substances and methods to attach the substances to non-recombinant microorganisms. The substances provided by the invention—are not limited to (chimeric) proteins alone, but can be fully or only partly of a peptide nature, whereby a peptide part is covalently joined to a non-peptide moiety. The invention provides a proteinaceous substance comprising at least one stretch of amino acids derived from a first microorganism which substance is capable of attaching to a cell wall of a second microorganism. The substance according to the invention is, for example, produced by a first microorganism (for example a microorganism from which the knowledge about the sequence of the stretch of amino acids originates, but another recombinant microorganism can produce the substance as well). After its production, the substance is harvested, optionally stored for future use, and then brought in contact with the second microorganism, where it attaches to its cell wall. Alternatively, the substance is produced synthetically by using established peptide synthesis technology. A preferred embodiment of the invention provides a substance wherein the second microorganism is a non-recombinant microorganism. With a substance provided by the invention, it is now possible to attach or anchor, for example, a heterologous or chimeric protein produced by a recombinant microorganism to an innocuous non-recombinant microorganism.
A preferred embodiment of the invention provides a proteinaceous substance wherein the stretch of amino acids has a sequence corresponding to a consensus sequence listed in
Yet another preferred embodiment of the invention provides a proteinaceous substance wherein the second microorganism is selected from any of the group of Gram-positive bacteria and Gram-negative bacteria. Examples are microorganisms, such as Bacillus subtilis (SEQ ID NO:75 through SEQ ID NO:78, SEQ ID NO:81 through SEQ ID NO:87, SEQ ID NO:104, SEQ ID NO:107, SEQ ID NO:109 and SEQ ID NO:110), Clostridium beijerinckii, Lactobacillus plantarum, Lb. buchneri, Listeria inocua, Streptococcus thermophilus, Enterococcus faecalis (SEQ ID NO:23 through SEQ ID NO:27), E. coli (SEQ ID NO:64, SEQ ID NO:66, SEQ ID NO:72, SEQ ID NO:73, SEQ ID NO:79, SEQ ID NO:106), and others.
The invention provides a proteinaceous substance which is additionally comprising a reactive group. For example, the invention provides a proteinaceous substance comprising a reactive group such as an antigenic determinant, heterologous enzyme, single-chain antibody or fragment thereof, polyhistidyl tag, fluorescing protein, luciferase, binding protein or peptide, or another substance such as an antibiotic, hormone, non-peptide antigenic determinant, carbohydrate, fatty acid, aromatic substance and reporter molecule, and an anchoring or targeting protein or part thereof (herein also called attaching peptide) useful in heterologous surface display which is both broadly reactive with cell wall components of a broad range of microorganisms.
For example, the invention provides a substance wherein the reactive group is a non-protein moiety, for example, is selected from the group of antibiotics, hormones, aromatic substances and reporter molecules. The substance is constructed by binding, for example, an antibiotic, such as penicillin or tetracycline, but various other antibiotics can be used, or a hormone, such as a steroid hormone, or any other compound to an attaching peptide provided by the invention. Such binding can be achieved by various techniques known in the art, and thereby can label or “flag” the attaching peptide. A preferred example is the binding of an attaching peptide to a reporter molecule such as FITC, or HRPO, whereby tools are generated that can be used in diagnostic assay whereby microorganisms having peptidoglycan are detected. Similarly, an attaching peptide with an antibiotic bound thereto can be used in vivo by, for example, parenteral administration into the bloodstream of humans or animals or in vitro to bind to such microorganisms having peptidoglycan, thereby increasing the concentration of antibiotic around the organism, which then gets killed by the antibiotic action.
The invention provides a substance wherein the reactive group is a protein moiety, for example, selected from the group of antigenic determinants, enzymes, single-chain antibodies or fragments thereof, polyhistidyl tags, fluorescing proteins, binding proteins or peptides. For example, the invention provides a protein which comprises as a reactive group a protein or polypeptide. Also, the invention provides a nucleic acid molecule encoding a protein provided by the invention. Such a nucleic acid molecule (being single- or double-stranded DNA, RNA or DNA/RNA) at least comprises nucleic acid sequences specifically encoding an attaching peptide as well as nucleic acid sequences specifically encoding the reactive group polypeptide, but can additionally also comprise other nucleic acid sequences, which, for example, encode a signal peptide, or comprise, for example, promoter and/or regulatory nucleic acid sequences. The invention also provides a vector comprising a nucleic acid molecule encoding a protein provided by the invention.
The invention provides a proteinaceous substance comprising a reactive group joined with or bound to at least one attaching peptide which comprises a stretch of amino acids corresponding to the consensus amino acid sequence listed in
“Corresponding to” is defined as having an amino acid sequence homologous to the consensus amino acid sequence listed in
Preferably, the attaching peptide is derived from any one of the proteins listed in
The invention provides a method for attaching a substance to the cell wall of a microorganism comprising the use of an attaching peptide which comprises a stretch of amino acids having a sequence corresponding to at least a part of the consensus amino acid sequence provided in
A preferred method according to the invention comprises the use of an attaching peptide which is derived from the major peptidoglycan hydrolase of Lactococcus lactis. Another method according to the invention is provided wherein the substance is a (poly)peptide or a protein, for example, being part of a protein provided by the invention. Since post-translational modifications occurring to such a polypeptide or protein are inherent to the host cell or expression system used, a post-translationally modified protein as provided by the invention is, therefore, also provided. However, yet another method according to the invention is provided wherein the compound is selected from the group composed of antibiotics, hormones, antigenic determinants, carbohydrate chains, fatty acids, aromatic compounds and reporter molecules. The substance is constructed by binding, for example, an antibiotic, such as penicillin or tetracycline (but various other antibiotics may be used), or a hormone, such as a steroid hormone, or any other compound to an attaching peptide provided by the invention. Such binding can be achieved by various techniques known in the art and thereby can label or “flag” the attaching peptide. A preferred example is the binding of an attaching peptide to a reporter molecule such as FITC or HRPO, whereby tools are generated that can be used in a diagnostic assay whereby microorganisms having peptidoglycan are detected. Similarly, an attaching peptide with an antibiotic bound thereto can be used in vivo by, for example, parenteral administration into the bloodstream of humans or animals or in vitro to bind to such microorganisms having peptidoglycan, thereby increasing the concentration of antibiotic around the organism, which than can get killed by the antibiotic action. The microorganism is preferably selected from any of the group of yeast, molds, gram-positive bacteria and gram-negative bacteria. For example, the experimental part of this description describes mixing of β-lactamase::cA fusion protein with lactococcal cells which resulted in binding to the cells whereas this was not the case when mature β-lactamase not joined with an attachment protein was added. Also, fusion of β-lactamase of E. coli and α-amylase of Bacillus licheniformis to the attaching peptide provided by the invention and subsequent production of these fusion proteins resulted in active, secreted proteins which were located (attached) in L. lactis cell walls. Binding of AcmA and the β-lactamase::cA fusion protein was also demonstrated to isolated lactococcal cell walls and SDS-washed cell walls (the major part of this fraction is peptidoglycan).
Anchoring of recombinant proteins to non-recombinant microorganisms such as lactococci or other, bacteria or fungi, is especially attractive if the use of recombinant bacteria is not desired, e.g., in food processes or as pharmaceuticals for medical use such as in vaccines or in antibacterial therapy. The invention provides, for example, vaccine delivery or immunization via microorganisms which are labeled with distinct antigenic determinants and which may be directed against a wide variety of diseases. A protective immune response can, for example, be elicited by oral delivery of a bacterial vaccine provided by the invention wherein tetanus toxin fragment C is attached via a protein provided by the invention to a non-recombinant Lactococcus lactis. Such immunogens expressed by microorganisms provided by the invention are presented in particulate form to the antigen-presenting cells (for example M-cells) of the immune system and are, therefore, less likely to induce tolerance than soluble antigens. In addition, the existence of a common mucosal immune system permits immunization on one specific mucosal surface to induce secretion of antigen-specific IgA and other specific immune responses at distant mucosal sites. The invention solves the drawback of earlier bacterial vaccines whereby the potential to flourish on mucosal surfaces of the attenuated or recombinant bacterial strain used can cause problems such as inflammation and disease in itself, potentially leading to fever and bacteraemia or to the induction of immune tolerance. Also, the invention avoids the potential risks that are involved when using recombinant DNA-containing bacterial vectors for vaccination. In yet another possible vaccine and vaccine use provided by the invention, certain (killed) micro-organisms with adjuvant properties (such as the mycobacteria used in BCG) are labeled or loaded with a protein or substance composed of an antigenic determinant and an attaching peptide. These microorganisms then function as adjuvant, thereby greatly enhancing the immune response directed against the specific antigenic determinant. Yet another use provided by the invention comprises anchoring proteins from the outside to a microorganism which provides a means to present proteins or peptides which normally cannot be overexpressed and/or secreted by the microorganism. For example, the subunit B of cholera toxin (CTB) can be overproduced in E. coli but expression in L. lactis has been unsuccessful until now. The adjuvant activity of CTB in experimental recombinant vaccines is well documented and the ability of CTB or part thereof to bind to GM1 ganglioside on eucaryotic cell surfaces is of interest with respect to the use of L. lactis or other gram-positives in vaccines which specifically require targeting to mucosal surfaces. Yet another medical use provided by the invention is the addition of purified antigen::cA fusion proteins in vivo by parenteral administration into the bloodstream of humans or animals to combat bacterial infections. In this case, the antigen::cA fusion protein is used as a “flag” for the immune system. The antigenic determinant of a protein provided by the invention being a subunit of a vaccine regularly used for the immunization of humans (preferably children) or animals, e.g. a subunit of the Rubella, Pertusis, Poliomyelitis, tetanus or measles vaccine. After delivery into the bloodstream, the “flag” will bind through the AcmA repeats to the pathogenic bacterium present in the blood. A “flag” protein provided by the invention will then activate a memory response, i.e., the response to the antigenic determinant present in the protein. The antibodies thus produced recognise the “flag”-labeled bacteria, which will then be neutralised by the immune system. In this way, the protein is used to stimulate a pre-existing memory immune response, non-related to the bacterial infection, to clear bacterial infections from the system. Yet another use (which alternatively may be considered medical use or food use) provided by the invention is the use wherein a protein provided by the invention has the ability to bind to cells, such as mucosal cells, e.g. of the gut. The reactive group of such a protein is in such a case, for example, partly or wholly derived from a fimbriae protein or another gut attachment protein as is present, for example, in various E. coli strains. Microorganisms to which such a protein is attached will specifically home or bind to certain areas of the gut, a property which, for example, is beneficial for certain bacterial strains (i.e., lactococcal or lactobacillar strains) used as a probiotic. In another food or use of food provided by the invention, the protein or substance is composed of a food additive (such as an enzyme or flavor compound) which affects quality, flavor, shelf-life, food value or texture, joined with an attaching peptide, and subsequently attached or anchored to a microorganism which is then mixed with the foodstuff. The anchoring of such proteins to a bacterial carrier offers the additional advantage that the additive can be targeted to a solid bacteria-containing matrix (e.g. curd) in a process for the preparation of food, e.g. cheese or tofu. Yet another use of a proteinaceous substance or microorganism provided by the invention is the use of bacterial surface display in generating whole-cell bioadsorbents or biofilters for environmental purposes, microbiocatalysts, and diagnostic tools.
The invention is further explained in the experimental part which cannot be seen as limiting the invention.
The major autolysin AcmA of Lactococcus lactis subsp. cremoris MG1363 is an N-acetylmuramidase which is required for cell separation and is responsible for cell lysis during the stationary phase (5, 6). The 40.3-kDa secreted mature protein produces a number of activity bands in a zymogram of the supernatant of a lactococcal culture. Bands as small as that corresponding to a protein of 29 kDa were detected. As no clearing bands are produced by an L. lactis acmA deletion mutant, all bands represent products of AcmA (6). From experimental data and homology studies, it was inferred that AcmA likely consists of three domains: a signal sequence followed by an active site domain and a C-terminal region containing three highly homologous repeats of approximately 45 amino acids which are involved in cell wall binding. As the smallest active protein is 29 kDa, it was suggested that the protein undergoes proteolytic breakdown in the C-terminal portion (5, 6).
Cell wall hydrolases of various bacteria and bacteriophages contain repeats similar to those present in AcmA (4, 9, 10, 17). Partially purified muramidase-2 of Enterococcus hirae, a protein similar to AcmA, containing six similar repeats, binds to peptidoglycan fragments of the strain (11). The p60 protein of Listeria monocytogenes contains two such repeats and was shown to be associated with the cell surface (24). However, which parts of these enzymes contained the binding capacity was not assessed in any of these studies.
Nearly all cell wall hydrolases examined so far seem to consist of a catalytic domain and usually, although not always, a domain containing a number of specific amino acid repeats. In several studies it has been shown that only a part of some of the cell wall hydrolases is required for enzymatic activity (13, 14, 17, 19, 22, 34). Rashid et al. reported the cloning of the gene encoding a 90-kDa glucosaminidase of Bacillus subtilis of which the C-terminus shows significant similarity with the glucosaminidase domain of the S. aureus autolysin (23). The protein contains two repeated sequences in its N-terminus and two different repeats in the middle domain. A deletion derivative lacking the C-terminal 187 amino acids remained tightly bound to the cell walls, but no catalytic activity was observed when expressed in B. subtilis. By making deletions from the N-terminus it was shown that nearly two-thirds of the protein could be removed without complete loss of cell wall-hydrolyzing activity in E. coli, although loss of more than one repeat drastically reduced lytic activity.
The N-terminal domain of the major autolysin LytA of Streptococcus pneumonia provides the N-acetylmuramyl-L-alanine amidase catalytic function, whereas the C-terminal domain, which contains six repeated sequences, determines the specificity of binding to the cell wall (for review: see reference 18). The protein lacks a signal sequence and requires choline-containing teichoic acids to fully degrade pneumococcal cell walls. Furthermore, it was shown that at least four of the six repeats were needed for efficient recognition of the choline residues of pneumococcal cell walls and the retention of appreciable hydrolytic activity (7).
LytA, pneumococcal phage lysins, as well as clostridial and lactococcal cell wall hydrolases, have been used for the construction of active proteins, such that the activity domain and cell wall recognition domains were exchanged. The N-terminal half of the lactococcal phage enzyme was fused to the C-terminal domain of LytA (28). The chimeric enzyme exhibited a glycosidase activity capable of hydrolyzing choline-containing cell walls of S. pneumonia. This result showed that the lactococcal phage lysin consisted of at least two domains with a glucosidase activity contained in its N-terminus and two repeats similar to those in AcmA in the C-terminus (6). A tripartite pneumococcal peptidoglycan hydrolase has been constructed by fusing the N-terminal catalytic domain of the phage CPL1 lysozyme to HBL3, a protein with an amidase activity and a choline-binding domain (27). The three domains acquired the proper conformation as the fusion protein behaved as an amidase, a lysozyme and as a choline-dependent enzyme.
Also from nature, an enzyme is known as having two separate functional activity domains: the autolysin gene from Staphylococcus aureus encodes a protein that contains an amidase and an endo-β-N-acetylglucosaminidase domain separated by three highly similar repeats (20). This protein is processed posttranslationally into the two constituting activity domains.
The aim of the present study was to investigate the modular structure of AcmA. This was done by consecutively deleting the C-terminal repeats and by fusing the repeats to heterologous proteins. On the basis of cell fractionation and binding studies involving whole cells, it is concluded that the C-terminal repeats in AcmA bind the autolytic enzyme to the cell wall of L. lactis.
Materials and Methods
Bacterial strains, plasmids, and growth conditions. The strains and plasmids used in this study are listed in Table 1. Lactococcus lactis was grown at 30° C. in two-fold diluted M17 broth (Difco Laboratories, Detroit, Mich.) containing 0.5% glucose and 0.95% β-glycerophosphate (Sigma Chemical Co., St. Louis, Mo.) as standing cultures (½M17). Agar plates of the same medium contained 1.5% agar. Five μg/ml of erythromycin (Boehringer GmbH, Mannheim, Germany) was added when needed. Escherichia coli was grown at 37° C. with vigorous agitation in TY medium (Difco), or on TY medium solidified with 1.5% agar. When required, the media contained 100 μg of ampicillin (Sigma), 100 μg erythromycin or 50 μg kanamycin (both from Boehringer) per ml. Isopropyl-β-D-thiogalactopyranoside (IPTG) and 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) (both from Sigma) were used at concentrations of 1 mM and 0.002%, respectively.
General DNA Techniques and Transformation.
Molecular cloning techniques were performed essentially as described by Sambrook et al. (25). Restriction enzymes, Klenow enzyme and T4 DNA ligase were obtained from Boehringer and were used according to the instructions of the supplier. Deoxynucleotides were obtained from Pharmacia (Pharmacia Biotech, Uppsala, Sweden). All chemicals used were of analytical grade and were from Merck (Darmstadt, Germany) or BDH (Poole, United Kingdom). Electrotransformation of E. coli and L. lactis was performed by using a gene pulser (Bio-Rad Laboratories, Richmond, Calif.), as described by Zabarovsky and Winberg (37) and Leenhouts and Venema (16), respectively. Plasmid DNA was isolated using the QIAGEN plasmid DNA isolation kit (QIAGEN GmbH, Hilden, Germany) or by CsCl-ethidiumbromide density gradient centrifugation and DNA fragments were isolated from agarose gels using the QIAGEN gel extraction kit and protocols from QIAGEN.
Primer Synthesis, PCR and DNA Sequencing.
Synthetic oligo deoxyribonucleotides were synthesized with an Applied Biosystems 392 DNA/RNA synthesizer (Applied Biosystems Inc., Foster City, Calif.). The sequences of the oligonucleotides used are listed in Table 2.
Polymerase chain reactions (PCR) were performed in a Bio-Med thermocycler 60 (Bio-Med GmbH, Theres, Germany) using super Taq DNA polymerase and the instructions of the manufacturer (HT Biotechnology Ltd., Cambridge, United Kingdom). PCR fragments were purified using the nucleotide removal kit and protocol of QIAGEN.
Nucleotide sequences of double-stranded plasmid templates were determined using the dideoxy chain termination method (26) with the T7 sequencing kit and protocol (Pharmacia) or the automated fluorescent DNA sequencer 725 of Vistra Systems (Amersham Life Science Inc., Buckinghamshire, United Kingdom).
Nucleotide and amino acid sequences were analyzed with the PC/GENE sequence analysis program (version 6.8, IntelliGenetics, Inc., Geneva, Switzerland). Protein homology searches in the SWISSPROT, PIR, and Genbank (release September 23, 1996) databases were carried out with the BLAST program (1).
Construction of AcmA Derivatives.
A stop codon and EcoRI restriction enzyme site were introduced in acmA at the end of nucleotide sequences encoding the repeats and at the end of the sequence specifying the active site domain by PCR using the primers REPDEL-1 (SEQ ID NO:1), REPDEL-2 (SEQ ID NO:2), and REPDEL-3 (SEQ ID NO:3) and plasmid pAL01 as a template. Primer ALA-4 (SEQ ID NO:4), annealing within the sequence encoding the signal peptide of AcmA, was used in all cases as the upstream primer. All three PCR products were digested with SacI and EcoRI and cloned into the corresponding sites of pBluescript SK+ leading to pDEL1, pDEL2, and pDEL3. Subsequently, the 1,1 87-bp PflmI-EcoRI fragment of pGKAL1 (5) was replaced by the 513, 282 and 76-bp PflmI-EcoRI fragments of the inserts of pDEL1, 2 and 3, respectively. The proper plasmids specifying proteins containing one, two or all three repeats (pGKAL5, 4, and 3, respectively) were obtained in L. lactis MG1363acmAΔ1. pGKAL1 was cut with SpeI. The sticky ends were flushed with Klenow enzyme and self-ligation introduced a UAG stop codon after the Ser 339 codon of acmA. The resulting plamid was named pGKAL6.
A DNA fragment encoding half of the first repeat until the SpeI site in the middle of the second repeat was synthesized by PCR using the primers REP-4 A (SEQ ID NO:5) and B (SEQ ID NO:6). The NheI and SpeI sites at the ends of the 250-bp PCR product were cut and the fragment was cloned into the unique SpeI site of pGKAL1 resulting in plasmid pGKAL7.
Overexpression and Isolation of the AcmA Active Site Domain.
A DNA fragment encoding the active site domain of AcmA was obtained using the primers ACMHIS (SEQ ID NO:7) and REPDEL-3 with plasmid pAL01 as a template. The 504-bp PCR fragment was digested with Bg/II and EcoRI and subcloned into the BamHI and EcoRI sites of pET32A (Novagen R&D Systems Europe Ltd, Abingdon, United Kingdom). The proper construct, pETAcmA, was obtained in E. coli BL21(DE3) (30). Expression of the thioredoxin/AcmA fusion protein was induced in this strain by adding IPTG (to 1 mM final concentration) at an OD600 of 0.7. Four hours after induction, the cells from 1 ml of culture were collected by centrifugation and the fusion protein was purified over a Talon™ metal affinity resin (Clontech Laboratories Inc., Palo Alto, Calif.) using 8 M ureum-elution buffer and the protocol of the supplier. The eluate (200 μl) was dialyzed against a solution containing 50 mM NaCl and 20 mM Tris (pH 7) after which CaCl2 was added to a final concentration of 2 mM. One unit of enterokinase (Novagen) was added and the mixture was incubated at room temperature for 20 hours. The protein mixture was dialyzed against several changes of demineralized water before SDS-PAGE analysis and cell binding studies.
Construction of β-lactamase and α-amylase Fusions to the AcmA Repeat Domain.
For the introduction of a unique NdeI site at the position of the stop codon of the E. coli TEM-β-lactamase, the oligonucleotides BETA-1 (SEQ ID NO:8) and BETA-2 (SEQ ID NO:9) were used in a PCR with plasmid pGBL1 (21) as a template. The 403-bp PCR fragment was cut with NdeI and PstI and cloned as a 311 -bp fragment into the same sites of pUK21. The resulting plasmid, pUKblac, was digested with NdeI, treated with Klenow enzyme and subsequently digested with XbaI. The β-lactamase encoding fragment was ligated to a 1,104-bp PvuII-XbaI DNA fragment from pAL01 containing the acmA part encoding the repeat region of AcmA. The resulting plasmid, pUKblacrep, was digested with PstI and DraI and the 1349-bp fragment was inserted into the PstI-SnaBI sites of pGBL1, leading to plasmid pGBLR. After digestion of pGAL9 (21) with ClaI and HindIII, the 1,049-bp fragment encompassing the 3′-end of the Bacillus licheniformis α-amylase gene was subcloned into corresponding sites of pUK21. According to the paper of Perez Martinez et al. (21), this fragment should be 1,402-bp, but after restriction enzyme analysis, it turned out to be approximately 350-bp smaller. The resulting plasmid was called pUKAL1. A unique EcoRV restriction enzyme site was introduced by PCR at the position of the stop codon of the B. licheniformis α-amylase gene using the oligonucleotides ALFA-A (SEQ ID NO:10) and ALFA-B (SEQ ID NO:11) with plasmid pGAL9 as a template. After restriction of the 514-bp PCR fragment with SalI and EcoRV, the 440-bp fragment was cloned into the same sites of pUKAL1 resulting in plasmid pUKAL2. The EcoRV and XbaI sites of this plasmid were used to clone the 1,104-bp PvuII-XbaI fragment of pAL01 encoding the repeats of AcmA. The 1,915-bp ClaI-HindIII fragment of the resulting plasmid pUKALR was used to replace the corresponding 1,049-bp fragment of pGAL9 (pGALR). All cloning steps described above were performed in E. coli NM522. The plasmids pGBL1, pGBLR, pGAL9 and pGALR were used to transform L. lactis MG1363 and MG1363acmAΔ1.
SDS-polyacrylamide Gel Electrophoresis (SDS-PAGE) and Detection of AcmA and α-amylase Activity.
Two ml of end exponential phase L. lactis cultures were subjected to centrifugation. 0.5 ml of the supernatant fractions were dialyzed against several changes of demineralized water, lyophilized, and dissolved in 0.25 ml of denaturation buffer (3). Cell pellets were washed with 2 ml of fresh ½M17 medium and resuspended in 1 ml of denaturation buffer. Cell extracts were prepared as described by van de Guchte et al. (32).
AcmA activity was detected by a zymogram staining technique using SDS-PAA (12.5% or 17.5%) gels containing 0.15% autoclaved, lyophilized Micrococcus lysodeikticus ATCC 4698 cells (Sigma) as described before (6). For the analysis of α-amylase activity, 1% starch was included into 12.5% PAA gels. After electrophoresis, proteins were renatured using the AcmA renaturation solution (3) and the gel was stained with an I2/KI solution (at final concentrations of 12 and 18 mM, respectively) (33).
SDS-PAGE was carried out according to Laemmli (15) with the Protean II Minigel System (Bio-Rad) and gels were stained with Coomassie brilliant blue (Bio-Rad). The standard low range and prestained low and high range SDS-PAGE molecular weight markers of Bio-Rad were used as references.
Fractionation of mid- and end-exponential phase cultures of L. lactis was performed according to the protocol of Baankreis (2).
Binding of AcmA and its Derivatives to Lactococcal Cells.
The cells of 2 ml of exponential phase cultures of MG1363acmAΔ1 were gently resuspended in an equal volume of supernatant of similarly grown MG1363acmAΔ1 carrying either plasmid pGK13, pGKAL1, -3, -4, -5, -6 or -7 and incubated at 30° C. for 20 m Subsequently, the mixtures were centrifugated. The cell pellets were washed with 2 ml of ½M17 and cell extracts were prepared in 1 ml of denaturation buffer as described above, while 0.4 ml of the supernatants were dialyzed against demineralized water, lyophilized and dissolved in 0.2 ml of denaturation buffer.
To analyze competitive binding between AcmA derivatives containing 1 or 2 repeats, equal volumes of the supernatants of MG1363acmAΔ1 containing pGKAL3 or pGKAL4 were mixed prior to incubation with the MG1363acmAΔ1 cells. The samples were treated for SDS-PAGE as described above.
Three 500 μl samples of a mid-exponential phase culture of MG1363acmAΔ1 were centrifugated. From one sample, 50 μl of the supernatant were replaced by 50 μl of a solution containing the AcmA active site domain (see above). 100 μl of the supernatant of sample two were replaced by 50 μl demineralized water and 50 μl of the supernatant of a mid-exponential phase culture of MG1363acmAΔ1 (pGKAL4). Of the third sample, 100 μl of the supernatant were replaced by 50 μl of the solution containing the AcmA active site domain and 50 μl of the supernatant of MG1363acmAΔ1 (pGKAL4). Subsequently, the three samples were vortexed to resuspend the cells and incubated for 15 minutes at 30° C. After centrifugation, cell and supernatant fractions were prepared in 500 μl of denaturation buffer for analysis of AcmA activity in SDS-(1 7.5%)PAGE as described above.
Binding of the β-lactamase/AcmA fusion protein was studied by growing MG1363acmAΔ1 containing pGK13, pGBL1 or pGBLR until mid-exponential phase. The cells of 1 ml of MG1363acmAΔ1 (pGK13) culture were resuspended in an equal volume of supernatant of either of the other two cultures. The mixtures were prepared in duplo and one series was incubated at 30° C. for five minutes while the other was kept at that temperature for 15 minutes. Then, cell and supernatant fractions were treated as described for the AcmA-binding studies, resuspended in denaturation buffer in half of the original volume, and subjected to SDS-(12.5%)PAGE followed by Western blot analysis.
Western Blotting and Immunodetection.
Proteins were transferred from SDS-PAA gels to BA85 nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany) as described before (31). α-amylase and β-lactamase antigen was detected with 2000-fold diluted rabbit polyclonal anti-ampicillinase antibodies (5 prime→3 prime, Inc., Boulder, Colo.), and alkaline phosphatase-conjugated goat anti-rabbit antibodies (Promega Corporation, Madison, Wis.) using the western-Light Chemiluminescent detection system and protocol (TROPIX Inc., Bedford, Mass.).
Enzyme Assays and Optical Density Measurements.
AcmA activity was visualized on ½M17 agar plates containing 0.2% autoclaved lyophilized M. lysodeikticus cells as halos around colonies after overnight growth at 30° C.
α-amylase activity was detected by spotting 10 μl of an overnight culture onto a ½M17 agar plate containing 1% of starch (Sigma). After 18 hours of incubation at 30° C., halos were visualized by staining with an iodine solution according to the protocol of Smith et al. (29). A similar method was used for the detection of β-lactamase activity (29).
X-prolyl dipeptidyl aminopeptidase (PepX) was measured using the chromogenic substrate Ala-Pro-p-nitroanilid (BACHEM Feinchemicalien AG, Bubendorf, Switzerland). After two minutes of centrifugation in an eppendorf microcentrifuge, 75 μl of a culture supernatant was added to 50 μl substrate (2 mM) and 75 μl Hepes buffer (pH 7). The mixture was pipetted into a microtiter plate well and color development was monitored in a THERMOmax microtiter plate reader (Molecular Devices Corporation, Menlo Oaks, Calif.) at 405 nm during 20 minutes at 37° C. Optical densities were measured in a Novaspec II spectrophotometer (Pharmacia) at 600 nm.
Two of the Three Repeats in AcmA are Sufficient for Autolysis and Cell Separation.
Several mutant AcmA derivatives were constructed to investigate the function of the three repeats in the C-terminus of AcmA. A stop codon was introduced behind the codon for Thr-287 (pGKAL4) or Ser-363 (pGKAL3) (see
Halo formation. On a ½M17 plate containing cell wall fragments of M. lysodeikticus, halos were absent when MG1363acmAΔ1 carried pGK13 or pGKAL5. All other strains produced a clear halo that differed in size. The halo size was clearly correlated with the number of full-length repeats present, although the addition of an extra repeat resulted in a reduced halo size (see Table 3). Apparently, for optimal cell wall lytic activity, a full complement of repeats is required.
Cell separation and sedimentation. The deletion of one and a half, two and all three repeats had a clear effect on the chain length and on sedimentation of the cells after overnight growth (see Table 3). Thus, efficient cell separation requires the presence of at least two repeats in AcmA.
Enzyme activity. Cells and supernatants of overnight cultures of all strains were analyzed for AcmA activity by SDS-PAGE. In the cell fractions, no activity was detected for A0, not even after one week of renaturation of the protein (Table 3). Of the other derivatives, two major activity bands were present in this fraction. In each case, their positions in the gel corresponded to proteins with the calculated molecular weights of the unprocessed and the processed form. (Table 3 and not shown.) As shown in
Autolysis. To analyze the effect of the repeats on autolysis during the stationery phase, overnight cultures of all strains were diluted an hundred-fold and incubated at 30° C. for six days. The decrease of optical density (OD600) was followed. All cultures exhibited similar growth rates, reached the same maximal optical densities and did not lyse during the exponential phase of growth. After approximately 60 hours of incubation, maximal reduction in OD600 was reached in all cases. The results are presented in Table 3 and show that the reduction in OD600 is correlated with the reduction of the number of AcmA repeats. To investigate whether the decrease in OD600 really reflected autolysis, the activity of the intracellular enzyme PepX was measured. After 60 hours of incubation, PepX activity in the culture medium was also maximal in all samples, decreasing in all cases upon further incubation. Hardly any PepX activity was detected in the supernatant of the acmAΔ1 mutant and in cultures producing A0, A1 or A.1.5. In contrast, a considerable quantity of PepX had released into the supernatant of cultures producing A2 and A3. Thus, two repeats in AcmA are sufficient for autolysis of L. lactis. A2 or A4 production led to reduced lysis of the producer cells. Taken together, these results indicate that the repeats in AcmA function in efficient autolysis and are required for cell separation.
The Active Site Domain of AcmA Resides in the N-terminal Part.
To examine whether the active site is located in the N-terminal domain of AcmA, a DNA fragment starting at codon 58 until codon 218 of acmA was synthesized by PCR and fused to the thioredoxin gene in plasmid pET32A. The fusion protein comprises 326 amino acids. A protein with the expected molecular mass (35 kDa) was isolated from a culture of E. coli BL21(DE3) (pETAcmA) (
Fusion of the Repeats of AcmA to α-amylase and β-lactamase Yields Active Enzymes.
The three C-terminal repeats of AcmA (cA) were fused C-terminally to B. licheniformis α-amylase and E. coli TEM β-lactamase as described in Material and Methods and shown in
The activities of α-amylase and the αcA fusion protein were also detected in a renaturing SDS-(12.5%)PAA gel containing 1% starch. The primary translation product of the α-amylase gene is a protein of 522 amino acid residues which contains a signal sequence of 37 amino acids (21). It is secreted as a 55-kDa protein. αcA consists of 741 amino acids and, if processed and secreted, would give rise to a 78-kDa protein. Cell and supernatant fractions of L. lactis MG1363 and MG1363acmAΔ1 carrying pGAL9 or pGALR were analyzed after overnight growth of the strains. The results are presented in
The β-lactamase Fusion Protein is Predominantly Present in the Cell Wall.
To examine whether the presence of the C-terminal domain of AcmA resulted in binding of βcA to the cell wall, mid-exponential phase cultures of L. lactis MG1363acmAΔ1 containing pGBL1, encoding β-lactamase or pGBLR specifying βcA were fractionated and subjected to Western blot analysis (
The C-Terminal Repeats in AcmA are Required for Cell Wall Binding.
Although the results presented in the previous section strongly suggest that the C-terminal repeats are required for the retention of protein in the cell wall, definite proof was obtained by mixing the supernatant fractions of end-exponential phase cultures containing AcmA, or one of its deletion derivatives (see
Binding of AcmA or βcA to Lactococcal Cells at Different pHs.
The supernatant fraction of a mid-exponential phase L. lactis MG1363acmAΔ1 culture was replaced by the supernatant of a mid-exponential phase L. lactis MG1363 culture. This mixture was incubated at 30° C. for five minutes. Thereafter the supernatant was removed by centrifugation and the cell pellet was washed with M17. The cell pellets were dissolved in M17 with pHs ranging from 2 to 10 and incubated at 30° C. for 30 minutes. The cell and supernatant fractions were separated and treated as described before and analyzed for the presence of AcmA activity. A similar experiment was executed with mid-exponential phase L. lactis MG1363acmAΔ1 cells with the supernatant of an L. lactis MG1363acmAΔ1 (pGBLR) culture. The presence of βcA was analyzed by western blotting and immunodetection as described.
At all different pHs, ACmA and βcA was found to be bound to the lactococcal cells. The binding of both AcmA and βcA was better at low pH as judged from the activity in a zymogram and the visual presence of the amount of βcA fusion protein in the cell extracts after immunodetection.
Proteolytic Breakdown of AcmA by Pronase and Trypsin.
The supernatant fraction of a mid-exponential phase MG1363acmAΔ1 culture was replaced by the supernatant of a mid-exponential phase MG1363 culture. This mixture was incubated at 30° C. for 15 minutes. Thereafter, the supernatant and the cell fractions were separated and the cell pellet was dissolved in an identical volume of M17. To both fractions, Pronase and Trypsin (1 mg/ml) dissolved in 10 mM NaPi buffer (pH=7) was added to an end concentration of (10 μg/ml) and the mixtures were incubated at 30° C. Samples were taken after 5 and 30 minutes and two hours of incubation. The cell and supernatant fractions of each sample were separated and prepared for zymographic analysis as described above.
A complete hydrolysis of AcmA by pronase was observed in the supernatant fraction after two hours of incubation while activity was still present in the cell extract at this time point. The hydrolysis of AcmA by trypsin was slower and activity was still present in the supernatant after two hours of incubation. In time, the portion of activity present in the cell extracts was always higher than that observed in the supernatant. These results indicate that the ACmA protein is protected when it is bound to the cell.
Binding of AcmA to Different Types of Bacterial Cells.
The strains Bacillus subtilis DB104, Lactobacillus plantarum 80, Streptococcus faecalis JH2-2, Streptococcus thermophilus ATCC 19258, Listeria P, Lactobacillus buchneeri L4, Clostridium beijerinckii CNRZ 530 and Escherichia coli NM522 were grown overnight in GM17. Two fractions of each overnight culture were centrifuged and the supernatants were replaced by the supernatant of an overnight-culture of L. lactis MG1363acmAΔ1 (pGKAL1) or MG1363acmAΔ1 (pGK13). The mixtures were incubated at 30° C. for 15 minutes. Subsequently, the cell and supernatant fractions were separated and the cells were washed once with M17 and were prepared for SDS-PAGE as described before and analyzed for AcmA activity.
In all cell extracts AcmA activity was present while such an activity was absent in extracts of cells which had been incubated with the supernatant of MG1363acmAΔ1 (pGK13) which lacks the presence of AcmA.
To investigate the effect of repeat numbers on binding, equal volumes of the supernatants of cultures of MG1363acmAΔ1 (pGKAL3, encoding A2) and MG1363acmAΔ1 (pGKAL4, specifying A1) were mixed. The undiluted and a ten-fold diluted mixture were incubated with the AcmA-free cells. Analysis of zymograms of serial dilutions showed that the two activities were equally distributed over the cell and supernatant fractions, indicating that both proteins bind equally well (results not shown).
To examine whether the C-terminal repeat sequences of AcmA had the capacity to bind a heterologous, extracellular enzyme to lactococcal cells, binding of βcA was assessed by incubation of AcmA-minus L. lactis cells with culture supernatants containing either secreted wild-type β-lactamase or βcA. As
The results presented in this work indicate that the mature form of the N-acetylmuramidase AcmA of L. lactis consists of two separate domains. The overproduced and purified N-terminus, from amino acid residue 58 to 218 in the pre-protein, is active on M. lysodeikticus cell walls and, thus, contains the active site of the enzyme. This is in agreement with the finding that the repeat-less AcmA mutant A0 can still hydrolyze M. lysodeikticus cell walls, albeit with severely reduced efficiency. Prolonged renaturation was needed to detect the activity of the enzyme in vitro while colonies producing the protein did not form a halo. Enzymes A1 and A2 had in vitro activities which were nearly the same as that of the wild-type protein, although in the plate assay A1 produced a smaller halo than A2 which, in turn, was smaller than the wild-type halo. A strain producing A1 grew in longer chains than cells expressing A2 and, in contrast to A2-producing cells, sedimented and did not autolyze. Taken together, these results indicate that, although the N-terminus of AcmA contains the active site, the presence of at least one complete repeat is needed for the enzyme to retain appreciable activity. Second, only cultures producing AcmAs containing two or more full-length repeats are subject to autolysis and produce wild-type chain lengths. It is tempting to speculate that this apparent increase in catalytic efficiency of AcmA is caused by the repeat domain by allowing the enzyme to bind to its substrate, the peptidoglycan of the cell wall. As was postulated by Knowles et al. (12) for the cellulase-binding domains in cellobiohydrolases, such binding would increase the local concentration of the enzyme. The repeats could be involved in binding alone or could be important for proper positioning of the catalytic domain towards its substrate. The increase in AcmA activity with an increasing number of repeats to up to 3 in the wild-type enzyme, suggests an evolutionary process of repeat amplification to reach an optimum for proper enzyme functioning. The binding of A1, A1.5 and A4 was comparable with that of wild-type AcmA, but these enzyme varieties caused only little or no autolysis. These observations seem to support the idea that three repeats are optimal for proper functioning of AcmA. The presence of five and six repeats in the very similar enzymes of E. faecalis and E. hirae, respectively, may reflect slight differences in cell wall structure and/or the catalytic domain, requiring the recruitment by these autolysins of extra repeats for optimal enzyme activity.
The hypothesis that the C-terminal domain of AcmA is involved in cell binding (6) was corroborated in this study. First of all, it was shown that AcmA is indeed capable of cell binding. AcmA and its derivatives A1, A1.5, A2 and A4 all bound to cells of L. lactis when added from the outside. To prove that it was the C-terminus of AcmA that facilitated binding and not some intrinsic cell wall-binding capacity of the N-terminal domain, the repeat domain was fused to two heterologous proteins which do not normally associate with the cell wall. The smaller halos produced by αcA and , βcA compared to the wild-type proteins and the presence of most of βcA in the cell wall fraction are indicative of cell binding of the fusion proteins via the AcmA-specific repeats.
The βcA-binding studies clearly show that it is the AcmA repeat domain that specifies cell wall-binding capacity: whereas wild-type β-lactamase (and, for that matter, repeat-less AcmA) did not bind to lactococcal cells, βcA did bind to these cells when added from the outside. The results obtained with A1 in the binding assay show that only one repeat is sufficient to allow efficient binding of AcmA. In a separate study (5), it was shown that AcmA can operate intercellularly: AcmA-free lactococcal cells can be lyzed when grown together with cells producing AcmA. Combining this observation with the results presented above, it was concluded that AcmA does not only bind when confronting a cell from the outside but, indeed, is capable of hydrolyzing the cell wall with concomitant lysis of the cell.
AcmA-like repeats were found to be present at different locations in more than 30 proteins after a comparison of the amino acid sequences of the repeats in AcmA with the protein sequences of the Genbank database (release 23). Not all of these proteins with repeats varying from one to six are cell wall hydrolases. Alignment of the amino acid sequences of all the repeats yielded a consensus sequence similar to that postulated by Birkeland and Hourdou et al. (4, 9). Interestingly, if a limited number of modifications are allowed in the consensus repeat, the repeat is also present 12 and 4 times, respectively, in two proteins of Caenorhabditis elegans, which both show homology with endochitinases (Gene accession numbers U64836 and U70858) (36). Possibly, these repeats anchor these enzymes to fungi ingested by this organism. The presence of similar repeats in proteins of different bacterial species strongly suggests that they recognize and bind to a general unit of the peptidoglycan. An interesting goal for the future will be to elucidate the unit to which they bind and the nature of the binding.
As has been reported earlier for intact AcmA (5) and as shown here for its C-terminal deletion derivatives, the enzyme is subject to proteolytic degradation. None of the degradation products were present in cell extracts of whole cells indicating that they are not formed inside the cell (data not shown). The degradation pattern of each AcmA derivative is specific and very reproducible. Based on the sizes of the degradation products, a number of the proteolytic cleavage sites probably residing in the intervening sequences. One such site (1 in
All degradation products of AcmA and those of the two fusion proteins are mainly present in the supernatant and, to some extent, in the cell wall fraction but not in the cells. As none of the L. lactis strains used produced the cell wall-anchored proteinase PrtP, this enzyme cannot be held responsible for the specific degradation of AcmA or the fusion proteins. Apparently, an extracellular proteinase exists in L. lactis that is capable of removing the repeats, which may represent a mechanism for the regulation of AcmA activity.