GRAM-POSITIVE BACTERIA OF THE SPECIES LACTOCOCCUS LACTIS OR STREPTOCOCCUS THERMOPHILUS HAVING A VERY LOW SURFACE PROTEOLYSIS, PROCESSES FOR OBTAINING THEM AND USES THEREOF

Abstract

The present invention relates to a bacterium, in which the expression and/or the activity of surface proteases is/are inhibited, to its preparation process and to the uses of this bacterium.

Claims

1. A gram-positive bacterium of the species Streptococcus thermophilus or Lactococcus lactis such that the endogenous surface protease comprising an amino acid motif and having at least 80% identity with the sequence SEQ ID No. 1, has a decreased or abolished expression and/or activity by mutagenesis or by the use of specific inhibitors of serine proteases or by the use of a strain such that the gene encoding said protease is absent or is present in a truncated form, wherein SEQ ID NO: 1 is defined as follows: TABLE-US-00009 I-A-G-T-G-T-I-E-X.sub.1-D-G-X.sub.2-X.sub.3-G-X.sub.4-I-G-G-X.sub.5-X.sub.6-X.sub.7-K with X.sub.1 is histidine (H) or lysine (K); X.sub.2 is serine (S), alanine (A) or threonine (T); X.sub.3 is isoleucine (I), leucine (L) or valine (V); X.sub.4 is aspartic acid (D) or glutamine (Q); X.sub.5 is alanine (A) or valine (V); X.sub.6 is aspartic acid (D) or tyrosine (Y); and X.sub.7 is lysine (K) or leucine (L).

2. The bacterium according to claim 1, wherein the Gram-positive bacterium is a Streptococcus thermophilus and the endogenous surface protease has at least 70% identity with the sequence of SEQ ID NO: 2 of Ster 1612 when the sequences are aligned along their entire length.

3. The bacterium of claim 2, wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 8 of HtrA or wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 9 of PrtS and has a decreased or abolished expression and/or activity by mutagenesis or by the use of specific inhibitors of serine proteases or by the use of a strain such that the gene encoding said protease is absent or present in a truncated form.

4. The bacterium of claim 2, wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 8 of HtrA and wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 9 of PrtS and has a decreased or abolished expression and/or activity by mutagenesis or by the use of specific inhibitors of serine proteases or by the use of a strain such that the gene encoding said protease is absent or present in a truncated form.

5. The bacterium according to claim 1, wherein the Gram-positive bacterium is a Lactococcus lactis and the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 3 of Ywdf or with the sequence SEQ ID NO: 4 of llmq 2442 when the sequences are aligned over their entire length.

6. The bacterium of claim 5, wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 10 of HtrA or wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID No. 11 of PrtP and has a decreased or abolished expression and/or activity by mutagenesis or by the use of specific inhibitors of serine proteases or by the use of a strain such that the gene encoding said protease is absent or present in a truncated form.

7. The bacterium of claim 5, wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 10 of HtrA and wherein the endogenous surface protease has at least 70% identity with the sequence SEQ ID NO: 11 of PrtP and has an expression and/or activity decreased or abolished by mutagenesis or by the use of specific inhibitors of serine proteases or by the use of a strain such that the gene encoding said protease is absent or present in a truncated form.

8. The bacterium according to claim 1, modified to express a heterologous protein of interest, said bacterium containing an expression vector containing a DNA fragment encoding the heterologous protein of interest or by a DNA fragment encoding the heterologous protein of interest inserted in its chromosome.

9. A method of preparing a low proteolytic bacterium, comprising mutagenizing or using specific inhibitors of serine proteases or using a strain such that the gene encoding said protease is absent or present in a truncated form for abolishing or decreasing in said bacterium the expression and/or activity of the endogenous surface protease of said bacterium comprising an amino acid motif having at least 80% identity with the sequence SEQ ID NO: 1; wherein SEQ ID NO: 1 is defined as follows: TABLE-US-00010 I-A-G-T-G-T-I-E-X.sub.1-D-G-X.sub.2-X.sub.3-G-X.sub.4-I-G-G-X.sub.5-X.sub.6-X.sub.7-K with X.sub.1 is histidine (H) or lysine (K); X.sub.2 is serine (S), alanine (A) or threonine (T); X.sub.3 is isoleucine (I), leucine (L) or valine (V); X.sub.4 is aspartic acid (D) or glutamine (Q); X.sub.5 is alanine (A) or valine (V); X.sub.6 is aspartic acid (D) or tyrosine (Y); and X.sub.7 is lysine (K) or leucine (L).

10. A method for producing a heterologous protein of interest, the method comprising culturing a bacterium according to claim 8, said bacterium being transformed by an expression vector containing a DNA fragment encoding the heterologous protein of interest or by a DNA fragment encoding the heterologous protein of interest inserted in its chromosome.

11. Use of a bacterium of claim 1, as a pre-maturation ferment for milk.

Description

[0105] The present invention will be better understood with the aid of the following description, which refers to the non-limiting examples illustrating the very low surface proteolytic activity of a bacterium according to the invention in which the 2 (strain CNRZ1066ΔhtrAΔster_1612) or 3 (strain LMD9ΔhtrAΔprtSΔster_1612) proteases as defined in the invention are inhibited compared to the parent bacterial strain from which it is derived (strain CNRZ1066 or strain LMD9), as well as the appended figures:

[0106] FIG. 1: Effect of inactivation of the two surface proteases PrtS and HtrA on the presence of peptides from the degradation of S. thermophilus LMD9 surface proteins in the culture medium. Mean of 3 experiments (black), with standard deviation of the mean (gray).

[0107] FIG. 2: Sequence alignment of the three proteins STER_1612, YwdF and llmg_2442. The multiple alignment was performed with MUSCLE (v3.8) available on the EMBL-EBI server lhttp://www.ebi.ac.uk/Tools/msa/muscle/). White characters on black background indicate amino acids located inside the cell, italic characters correspond to the transmembrane fragment and black ones are located outside the cell (HMMTOP predictions). Amino acids potentially constituting the catalytic diad are highlighted in gray.

[0108] FIG. 3: Growth of S. thermophilus LMD9 wild type strain and mutated strain for the three surface proteases PrtS, HtrA and STER_1612 (triple mutant) in chemically defined medium (CDM) containing only free amino acids as a source of amino nitrogen.

[0109] FIG. 4: Effect of inactivation of the 3 surface proteases PrtS, HtrA, and STER_1612 on the number of peptides from S. thermophilus LMD9 surface protein degradation. Average of 3 experiments (black), with standard deviation of the mean (gray).

[0110] FIG. 5: Coverage of IL-10 and elafin proteins by the degradation fragments (peptides) identified in the corresponding culture media of S. thermophilus LMD9 (wild-type strain producing IL-10 or elafin, respectively). The shaded background shows the regions of IL-10 and elafin proteins corresponding to the peptides identified in the S. thermophilus LMD9 culture media. These regions group all the identified peptides. Cysteines (C) are indicated in bold.

[0111] FIG. 6: Immunodetection of elafin in culture supernatant of S. thermophilus CNRZ1066 (wild type and ΔhtrAΔSTER 1612 double mutant strain).

[0112] FIG. 7: Effect of inactivation of surface proteolysis on elafin activity produced by S. thermophilus CNRZ1066 (wild-type strain in black and ΔhtrAΔSTER 1612 double mutant in gray).

[0113] FIG. 8: Immunodetection of elafin produced by S. thermophilus CNRZ1066 (wild type and ΔhtrAΔSTER 1612 double mutant strain) and L. lactis IL1403 (wild type and ΔhtrA mutant strain).

[0114] FIG. 9: Comparison of elafin activities produced by L. lactis IL1403ΔhtrA (denoted IL1403D) and S. thermophilus CNRZ1066ΔhtrAΔSTER_1612 (denoted CNRZ1066DD).

[0115] FIG. 10: Immunodetection of elafin produced by L. lactis IL1403 (wild-type and ΔhtrAΔywdF double mutant strain).

[0116] FIG. 11: Immunodetection of elafin produced by L. lactis MG1363 (wild-type and ΔhtrAΔllmg-2442 double mutant strain).

EXAMPLE 1: CONSTRUCTIONS OF MUTANTS OF THE STREPTOCOCCUS THERMOPHILUS LMD9 BACTERIUM STRAIN AND THE STREPTOCOCCUS THERMOPHILUS CNRZ1066 BACTERIUM STRAIN

[0117] 1) Construction of ΔhtrA, ΔprtS and ASTER 1612 Single Mutants of S. thermophilus Strain LMD9

[0118] General Principle

[0119] The single mutants ΔhtrA, ΔprtS and ΔSTER_1612 are constructed from the wild type S. thermophilus LMD9 strain by gene replacement. The technique used consists in producing a PCR fragment combining three amplicons: [0120] an upstream amplicon corresponding to a region located at the beginning of the gene to be replaced, [0121] an antibiotic resistance cassette and [0122] a downstream amplicon corresponding to a region located at the end of the gene to be replaced.

[0123] The three amplicons are produced separately and then combined via a final additional PCR. This association is made possible by adding extensions to the oligonucleotides used to amplify the upstream and downstream regions that allow binding to the antibiotic resistance cassette (PCR overlap). This final PCR fragment is then introduced into the strain by natural competence (WO2010/125091). The homology between the upstream and downstream fragments of the gene and the PCR fragment allows recombination at the site of the gene to be replaced. Colonies with the targeted gene replaced by the resistance gene can be easily isolated on a medium containing the corresponding antibiotic.

[0124] Mutant htrA

[0125] A kanamycin resistance cassette is used. The upstream and downstream regions of the htrA gene are amplified separately by PCR from S. thermophilus LMD9 chromosomal DNA; the kanamycin resistance cassette is amplified from the plasmid pKa (Trieu-Cuot et al., 1983) using the oligonucleotide pairs htrA-upstream-F/htrA-upstream-R, htrA-downstream-F/htrA-downstream-R, and aphA3F/aphA3-R-, the nucleotide sequences of which are shown in Table 1.

TABLE-US-00004 TABLE 1 SEQ ID Oligonucleotides Sequence No. htrA- GTA ATC ACG GTC ACC AAC C 16 upstream-F htrA- GAC ATC TAA TCT TTT CTG 17 upstream-R AAG TAC ATC CGC AAC AGT AAA CCA CCT AGT AAG CC htrA- ATA ATC TTA CCT ATC ACC 18 downstream-F TCA AAT GGT TCG CTG GGT AGT GTT CAG AAA GGT ATG CC htrA- GGA TTG AGA TTT GAT CGT TG 19 downstream-R aphA3-R GTT GCG GAT GTA CTT CAG 20 aphA3-F CCA GCG AAC CAT TTG AG 21

[0126] The three fragments are amplified, purified on column with the Clean-Up PCR kit (Promega) and the expected sizes (480 bp, 580 bp and 1350 bp) are verified by agarose gel electrophoresis. An additional PCR combining the three fragments is performed using the three fragments obtained and the htrA-upstream-F and htrA-downstream-R oligonucleotides to obtain a 2410 bp fragment which is purified on column and used to transform the LMD-9 strain according to the following natural transformation protocol A first pre-culture is performed during the day in M17 medium with 1% lactose (M17lac) (Terzagui et al., 1975) from a frozen stock of the strain; from which a second pre-culture in chemically defined medium (CDM) (Letort et al., 2001) is performed overnight. This second pre-culture is used to seed a culture in MCD at an optical density at 600 nm (OD6.sub.00) of 0.05. After 1 h incubation at 42° C., the PCR product combining the three fragments is added to an aliquot of the culture. After incubation at 42° C., the aliquot of culture is spread on M17lac agar plates containing 1 mg/ml kanamycin. The plates are incubated 48h at 42° C. in anaerobic jar. Several resistant clones are verified by PCR on colony and one clone is then verified by sequencing.

[0127] Mutant ΔprtS

[0128] An erythromycin resistance cassette is used. The upstream and downstream regions of the ptrS gene were separately amplified by PCR from S. thermophilus LMD9 chromosomal DNA; the erythromycin resistance cassette is amplified from plasmid pG+host9 (Biswas et al., 1993) using the oligonucleotide pairs prtS-upstream-F/prtS-upstream-R, prtS-downstream-F/prtS-downstream-R, and erm-F/erm-R, the nucleotide sequences of which are shown in Table 2.

TABLE-US-00005 TABLE 2 SEQ Oligo- ID nucleotides Sequence No. prtS- TGG TAA GCA CGT AGA CC 22 upstream-F prtS- CTA CTG ACA GCT TCC AAG 23 upstream-R GAG CTA AAG AGG TCC CAG GCT TGT CAA TTC ATC TG prtS- GCA AGT CAG CAC GAA CAC 24 downstream-F GAA CCG TCT TAT CTC CGA AAG CCA ACT TAG ATG G prtS- CGT ATG CTT ACC AAC AGA G 25 downstream-R erm-F GGG ACC TCT TTA GCT CCT TGG 26 erm-R GGA GAT AAG ACG GTT CGT GTT 27 CG

[0129] The three fragments were amplified, purified on column and the expected sizes (660 bp, 650 bp and 1060 bp) were verified. An additional PCR combining the 3 fragments is performed using the three fragments obtained and the oligonucleotides prtS-upstream-F and prtS-downstream-R to obtain a 2370 bp fragment which is purified on column and used to transform the LMD-9 strain according to the protocol described above. This culture is then plated on M171ac agar plates containing erythromycin (5 μg/ml) which are incubated 48h at 42° C. in anaerobic jar. Several resistant clones are verified by colony PCR and one clone is then verified by sequencing.

[0130] Mutant ΔSTER_1612.

[0131] A spectinomycin resistance cassette is used. The upstream and downstream regions of the STER 1612 gene are amplified separately by PCR from S. thermophilus LMD9 chromosomal DNA; the spectinomycin resistance cassette is amplified from plasmid pAT28 (Trieu-Cuot et al., 1990) using the oligonucleotide pairs STER_1612-upstream-F/STER_1612-upstream-R, STER_1612-downstream-F/STER_1612-downstream-R, and spec-F/spec-R, whose nucleotide sequences are shown in Table 3.

TABLE-US-00006 TABLE 3 SEQ Oligo- ID nucleotides Sequence No. STER_1612- CCC AAC AAC ACC AGG CTC ATT 28 upstream-F STER_1612- GAA AAA TTC TAT AGA AAC TTC 29 upstream-R TCT CAA TTA GGC TAA GGC TGA TCC GGA TGC CAA STER_1612- TAC AGA TTA ATA ATT ATT CTT 30 downstream-E TAT TAT ACA GAT CCA GAG TAA TTT CCA GTT GCC STER_1612- TTC GAG GCC TAC GCA ATG CG 31 downstream-R spec-F GAT CTG TAT AAT AAA GAA TA 32 spec-R AGC CTA ATT GAG AGA AGT TTC 33

[0132] The three fragments were amplified, purified on column and the expected sizes (596 bp, 957 bp and 566 bp) were verified. An additional PCR combining the 3 fragments is performed using the three fragments obtained and the oligonucleotides STER_1612-upstream-F and STER_1612-downstream-R to obtain a 2050 bp fragment which is purified on column and used to transform the LMD-9 strain according to the protocol described above. This culture is then plated on M17lac agar plates containing spectinomycin (150 μg/ml) which are incubated for 48h at 42° C. in anaerobic jar. Several resistant clones are verified by colony PCR and one clone is then verified by sequencing.

[0133] 2) Construction of the ΔhtrA ΔprtS Double Mutant of S. thermophilus Strain LMD9

[0134] The ΔhtrAΔprtS double mutant is constructed from the ΔprtS single mutant by natural transformation using chromosomal DNA from the ΔhtrA mutant. After breaking the cells with glass beads, the chromosomal DNA of the ΔhtrA mutant is extracted with phenol-chloroform and precipitated with ethanol. The ΔprtS mutant is transformed with purified chromosomal DNA from the ΔhtrA mutant, following the same natural transformation protocol as described above. Transformants are selected by plating on agar medium containing kanamycin (1 mg/ml). Some resistant clones are verified by colony PCR.

[0135] 3) Construction of the ΔhtrA ΔSTER_1612, ΔprtS ΔSTER_1612 Double Mutants and the ΔhtrAΔprtSΔSTER 1612 Triple Mutant of S. thermophilus Strain LMD9

[0136] The double mutants ΔhtrA ΔSTER_1612, ΔprtS ΔSTER_1612 and the triple mutant ΔhtrAΔprtSΔSTER 1612 are obtained by natural transformation of the strains LMD9ΔhtrA, LMD9ΔprtS and LMD9ΔhtrAΔprtS with the PCR fragment containing the upstream and downstream regions of the STER_1612 gene fused to the spectinomycin resistance cassette used to construct the single mutant ΔSTER_1612 according to the protocol described above. The mutant selection and control protocol is as described for the LMD9ΔSTER_1612 single mutant.

[0137] 4) Construction of Mutants of S. thermophilus Strain CNRZ1066

[0138] The general principle of the construction of the different mutants is identical to the one described above, as well as the protocol used. Only the differences are indicated below.

[0139] For single mutants, chromosomal DNA from strain CNRZ1066 is used as a template to amplify the upstream and downstream regions of the gene to be mutated. During natural transformation of strain CNRZ1066, the competence of the strain is stimulated by adding ComS competence peptide (LPYFAGCL) at a concentration of 1 μM for 10 minutes prior to addition of the PCR fragment to the culture.

[0140] The CNRZ ΔhtrAΔSTER_1612 double mutant is obtained by natural transformation of the CNRZ1066ΔSTER_1612 strain with chromosomal DNA from the CNRZ ΔhtrA strain. Some mutants were checked by colony PCR.

EXAMPLE 2: DELETION OF THE TWO PROTEASES PRTS AND HTRA DOES NOT ABOLISH SURFACE PROTEOLYSIS IN THE STREPTOCOCCUS THERMOPHILUS LMD9 STRAIN OF BACTERIA

[0141] Strain LMD9 naturally produces the two surface proteases PrtS and HtrA. To assess their respective roles in the formation of the exopeptidome of strain LMD9, two single ΔprtS and ΔhtrA mutants and one ΔhtrAΔprtS double-mutant were constructed by natural transformation (WO2010/125091), following the protocol described in Example 1.

[0142] 1) Material and Methods: Determination of the Exopeptidome of Strains

[0143] The exopeptidome of the resulting strains is determined under the same experimental conditions as that of the wild-type strain LMD9, as described below.

[0144] Culture of the Strain

[0145] 50 ml of MCD with the concentration of aromatic amino acids (Phenylalanine, Tryptophan and Tyrosine) reduced by a factor of 10 are seeded with 500 μl of standard MCD overnight preculture and incubated at 42° C. until OD.sub.600=1.0. The cells are then removed by centrifugation (5,000 rpm, 4° C.; 10 minutes), the supernatant is filtered on a 0.22 μm PVDF membrane.

[0146] Isolation and Concentration of Peptides.

[0147] The filtered culture supernatant is acidified with trifluoroacetic acid (TFA) to a final concentration of 0.1% (470 μl of a 10% TFA solution in 47 ml of culture supernatant), and stored overnight at 4° C.

[0148] Peptides present in the acidified supernatant are extracted by solid phase extraction (SPE) on a StrataX cartridge (Phenomenex) containing 200 mg of phase, at a flow rate of about 0.3 ml/min, according to the manufacturer's recommendations. The cartridge is first activated with 3 ml of methanol, then equilibrated with 6 ml of an aqueous solution containing 5% acetonitrile and 0.1% TFA. 1.75 ml of acetonitrile are added to 35 ml of acidified supernatant, so as to have a final concentration of 5% acetonitrile in the supernatant. 35 ml of the prepared supernatant is loaded onto the activated and equilibrated cartridge at a flow rate of 0.3 ml/min. The cartridge is then washed with 5 ml of the aqueous solution containing 5% acetonitrile and 0.1% TFA, and the peptides are eluted with 1.5 ml of an aqueous solution containing 50% acetonitrile and 0.1% TFA. The eluate is dried for 16 to 18 hours by vacuum evaporation (speed vac system) then stored at −20° C.

[0149] The dried eluate containing the peptides is taken up with 350 μl of an aqueous solution containing 0.1% TFA (final concentration), which corresponds to a concentration factor of 100. Peptide solubilization is obtained by vortexing and passing 5 minutes in an ultrasonic tank. The concentrated peptide solution is ultrafiltered through a 3 kDa membrane by centrifugation for 1 h at 13000 rpm.

[0150] The peptides are then separated by HPLC on a reverse phase column (Kinetex C18 column (Phenomenex), porosity 100 Å, particle size 2.6 μm, size 150×4.6 mm) with a linear gradient (slope 1.6%) of acetonitrile in ammonium formate (20 mM, pH 6.2) at a flow rate of 0.7 ml/min and a temperature of 40° C. The equivalent of 20 ml of culture (i.e. 200 μl of concentrated suspension) is injected on the column. The fractions eluted between 3.2% and 53.3% acetonitrile are collected and dried at speed vac.

[0151] Identification of Peptides

[0152] Peptide identification is done by mass spectrometry. The dried fractions are taken up in 30 μl of an aqueous solution containing 0.1% TFA and 2% acetonitrile, and a fraction of 4 μl is loaded on a Pepmap C18 column (150×0.075 mm, particle size 2 μm, porosity 100 Å). Peptides are eluted with a gradient of acetonitrile in formic acid (0.1%), and analyzed online by mass spectrometry (LTQ-Orbitrap Discovery, Thermo Fisher). Peptide ionization is done by electrospray (1.3 kV), and the parameters for analysis of the ionized peptides are as follows: measurement of mass/charge ratios (m/z) from 300 to 1600 with a resolution of 15000 on the Orbitrap mass analyzer, and fragmentation of the 6 most abundant parent ions on the LTQ linear trap. The doubly charged peptides are subjected to fragmentation, with a 40 second exclusion window and classical fragmentation parameters (collision energy: 35%).

[0153] The identification of peptides and the proteins from which they are derived is done with the X!Tandem search engine (version 2017.2.1.4) and the X!Tandem pipeline software suite (version 3.4.3, www.pappso.fr) using the protein sequence of the S. thermophilus LMD9 strain associated with a protein base of contaminants adapted to the analysis activity of the analysis platform (tryptic peptides of a eukaryotic protein sample containing in particular human keratins, bovine and murine proteins). The X!Tandem pipeline search parameters include the absence of tryptic cleavage for peptide identification, a minimum of one peptide per protein, identified with an E-value less than or equal to 0.01 and a mass tolerance of 10 ppm.

[0154] For each bacterial context (wild type, single and double mutants strain), all peptides identified as surface peptides (i.e. peptide resulting from the degradation of a surface protein) are counted. When a peptide is identified several times (identification redundancy), each identification is considered (counting of spectra). Thus a peptide identified once will have a count of 1, a peptide identified three times a count of three. The total number of surface peptides is totaled. As these peptides are derived from surface proteins, they result from the surface proteolytic activity of the strain. The number of peptides counted is therefore an indicator of the intensity of surface proteolysis: the higher the count, the greater the degradation activity, and therefore the surface proteolysis.

[0155] 2) Results

[0156] The results show that both HtrA and PrtS proteases are involved in the formation of the exopeptidome of S. thermophilus LMD9. In the double mutant, a reduction of more than half of the number of peptides from surface protein hydrolysis is observed (FIG. 1). A statistical test adapted to small sample sizes (non-parametric Kruskal-Wallis test) indicates that this decrease is statistically significant (P=0.049).

[0157] 3) Conclusion

[0158] The surface proteases HtrA and PrtS alone are not responsible for all the surface proteolytic activity of these strains.

EXAMPLE 3: IDENTIFICATION OF A NEW SURFACE PROTEASE IN STREPTOCOCCUS THERMOPHILUS LMD9

[0159] According to the MEROPS protease and peptidase specific database, the S. thermophilus LMD9 genome is reported to contain 45 proteolytic enzymes (http://merops.sanger.ac.ukkgi-bin/speccards?sp=sp000991; type=peptidase; strain=498). Genome annotation analysis of this strain (https://www.ncbi.nlm.nih.gov/genome/genomes/420?) suggests the presence of additional proteases, resulting in a total of 52 proteins that may contain a proteolytic domain.

[0160] Of these, 11 are predicted to be located on the cell surface based on the LocateP (http://www.cmbi.ru.nl/locatep-db/cgi-bin/locatepdb.py) or SecretomeP (http://www.cbs.dtu.dk/services/SecretomeP/) database. They include PrtS (STER_0846) (STER_RS04165) and HtrA (STER_2002) (STER_RS09790). The hypothesis is that at least one of the remaining 9 proteases is involved in surface proteolysis in S. thermophilus.

[0161] Of these nine proteases, six are present in L. lactis subsp. lactis IL1403 and L. lactis subsp. cremoris MG1363 (Table 4).

TABLE-US-00007 TABLE 4 LMD9 Size IL1403 MG1363 Protein (AAs) Protein Size % Identity Protein Size % Identity STER_0113 415 DacA 435 181/432 (42%) DacA 434 159/402 (39%) STER_0159 265 DacB 248 132/246 (54%) DacB 248 122/232 (52%) STER_0260 775 Pbp2A 743 359/628 (57%) Pbp2A 743 260/472 (55%) STER_1255 762 SrtA 287 134/255 (53%) SrtA 250 114/214 (53%) STER_1612 345 YwdF 342 169/338 (50%) Llmg_2442 343 147/318 (46%) STER_1741 207 SipL 208 109/206 (53%) SipL 208 99/206 (48%)

[0162] Peptides derived from the degradation of three of these 6 proteases (STER_0260, STER_1612 and STER_1741) were identified in the S. thermophilus LMD9 exopeptidome, indicating that these putative proteases were synthesized under these growth conditions. Of these three proteins, STER_0260 is annotated as a D-Ala-D-Ala carboxypeptidase and STER_1741 as a signal peptide peptidase.

[0163] The protease STER_1612 (STER_RS07910) (annotated YwdF in IL1403 and llmg_2442 in MG1363, respectively) thus appears to be the candidate protein to participate in surface proteolysis in L. lactis and S. thermophilus. The presence of a transmembrane fragment in the N-terminal region (www.enzim.hu/hmmtop/) is predicted for the protein in all 3 strains (FIG. 2).

[0164] The synthesis of these predictions leads to the hypothesis that the STER_1612 protein would be a cytoplasmic membrane-anchored protease whose active site would be oriented towards the outer side of the membrane.

EXAMPLE 4: DELETION OF THE THREE PROTEASES PRTS, HTRA AND STER_1612 ABOLISHES THE DEGRADATION OF ENDOGENOUS SURFACE PROTEINS

[0165] The Inventors hypothesized that the protease STER_1612 in the S. thermophilus strain LMD9 and its homologues in L. lactis strains, IL1403 and MG1363, were involved in the formation of the S. thermophilus and L. lactis exopeptidome.

[0166] 1) Materials and Methods

[0167] Validation of this hypothesis was undertaken in the LMD9 strain, by constructing the set of single, double and triple mutants by natural transformation, following the protocol described in Example 1.

[0168] In the previously constructed LMD9 wild type, LMD9ΔHtrA, LMD9ΔPrtS and LMD9ΔHtrAΔPrtS strains, the ster 1612 gene was replaced with a spectinomycin resistance cassette.

[0169] 2) Results

[0170] The growth of the strain mutated for the synthesis genes of the three surface proteases is not significantly affected in the culture medium used, a chemically defined medium containing only amino acids as a source of amino nitrogen (see FIG. 3), so that any differences observed between the wild type strain and the mutants cannot be attributed to a difference in growth between the strains.

[0171] Inactivation of only one of the three surface proteases reduces the number of peptides from surface protein proteolysis accumulated in the culture medium (exopeptidome surface peptides) by about a factor of 2. When all three proteases are inactivated, this reduction reaches a factor of 25, with only 24 peptides accumulated (average of three independent experiments; see FIG. 4). If we compare the number of peptides accumulated by the htrA_prtS double mutant and the prtS_htrA_ster_1216 triple mutant, the additional inactivation of the STER_1612 protease reduces it by a factor of 9.

[0172] 3) Conclusion

[0173] Surface proteolysis is nearly abolished in the S. thermophilus triple mutant, being only 5% of that of the wild type, based on the number of peptides present in the exopeptidome.

EXAMPLE 5: EXTENSION TO OTHER STRAINS OF STREPTOCOCCUS THERMOPHILUS NOT POSSESSING THE PRTS PROTEASE

[0174] 1) Materials and Methods

[0175] The role of the two surface proteases HtrA and STER_1612 on the surface proteolysis of S. thermophilus strains lacking the PrtS protease was evaluated using strain CNRZ1066 as a representative.

[0176] The CNRZ1066ΔhtrAΔprtS double mutant is constructed by natural transformation following the experimental protocol described in Example 1. The exopeptidome of the wild type CNRZ1066 and the resulting mutant are determined under the same experimental conditions as described for the LMD9 strain and its mutants (Example 2).

[0177] 2) Results

[0178] The exopeptidome of wild-type strain CNRZ1066 contains 240 spectra (peptide counts) from surface proteins. The double mutant exopeptidome contains only 15 spectra from surface proteins. Based on the number of spectra identified, the residual surface proteolysis of the CNRZ1066 double mutant can be estimated to be 6% of that of the wild-type strain, a reduction comparable to that obtained with the LMD9 strain.

[0179] 3) Conclusion

[0180] Even though the strain only possesses the two surface proteases HtrA and STER_1612, inactivation of the genes encoding these two proteases is sufficient to reduce the surface proteolysis of the bacteria by more than 90%.

EXAMPLE 6: OBTAINING STREPTOCOCCUS THERMOPHILUS STRAINS PRODUCING A HETEROLOGOUS PROTEIN

[0181] 1) Obtaining S. thermophilus Strains Producing IL-10 and Elafin

[0182] The L. lactis strains LL-pLB350 and LBH832 contain the plasmids pLB350 (Hossain et al., 2012) and pLB386, respectively, which carry the genes encoding IL-10 and elafin, respectively, placed under the control of a bile salt inducible promoter (pGroEL). Plasmids were extracted from these strains and purified using a commercial kit (Midikit, Quiagen).

[0183] The two plasmids pLB350 and pLB386 are then introduced into the wild-type S. thermophilus strain LMD9 and its triple mutant by natural competence, following the experimental protocol described in Example 1.

[0184] Plasmid pLB386 is introduced into wild-type strain CNRZ1066 and its surface protease mutant by natural competence.

[0185] Transformants are selected by plating on M17 agar medium containing 5 μg/ml chloramphenicol.

[0186] The presence of plasmid pLB350 was then verified by PCR on colonies using the two pairs of oligonucleotides pGroEL-F (ATAATGCCGACTGTACTTT of sequence SEQ ID No. 34)/IL-10-R (GGCCTTGTAGACACCTTGGTCTT of sequence SEQ ID No. 35) generating a band of 690 base pairs. The plasmid pLB386 was tested with the two pairs of oligonucleotides pGroEL-F and Elafin-R (TCACTGGGGAACGAAACAGGC of sequence SEQ ID No. 36) giving a band of 572 bp and the empty plasmid was tested using the two oligonucleotides Cm-F (GTTCAACAAACGAAAATTGG of sequence SEQ ID No. 37) and Cm-R (TTATAAAAGCCAGTCATTAG of sequence SEQ ID No. 38) giving a band of 807 bp.

EXAMPLE 7: A NON-PROTEOLYTIC BACTERIA STRAIN IMPROVES THE PRODUCTION EFFICIENCY OF HETEROLOGOUS PROTEINS

[0187] 1) Inactivation of Surface Proteases Reduces the Degradation of Heterologous Proteins

[0188] LMD9 Strain

[0189] Materials and Methods

[0190] Two heterologous protein models were chosen, interleukin 10 (IL-10) and elafin. Both proteins are candidate proteins in the treatment of chronic inflammatory bowel disease (Benbouziane et al., 2013 and Bermudez-Humaran et al., 2015). The plasmid carrying the gene encoding IL-10 was extracted from the lactococcus strain that contained it and introduced into the S. thermophilus strain LMD9 and its triple mutant lacking surface proteolytic activity by natural transformation, following the protocol described in Example 6. The same operation was performed for the plasmid encoding elafin. Four strains were obtained, two wild type strains producing IL-10 and elafin and two protease mutants producing the same two heterologous proteins.

[0191] To assess the stability of IL-10 and elafin, the two pairs of wild-type and mutant strains carrying plasmid pLB350 and plasmid pLB386, respectively, are grown for 4 h in MCD at 42° C. An equiperal mixture of cholic acid and deoxycholic acid is then added to the culture medium, at a final concentration of 150 μg/ml. The purpose of this addition is to induce the expression of genes controlled by the pGroESL promoter. After 15 minutes of induction at 42° C., the cells are removed by centrifugation and the peptides present in the culture supernatant are identified as shown in Example 2. The only difference is the modification of the interrogation library, to which the sequence of IL-10 or elafin was added, depending on the pair of strains considered.

[0192] Results

[0193] Several degradation fragments of the heterologous proteins are detected in the culture medium of the wild-type strains, covering in total more than 30% of their respective sequence (see Table 5 and FIG. 5). Under the same experimental conditions, no fragments are detected in the triple mutants. Thus, inactivation of the three surface proteases abolishes the degradation of the two heterologous proteins in S. thermophilus LMD9.

TABLE-US-00008 TABLE 5 Number of peptides from IL-10 or elafin degradation present in the culture medium of S. thermophilus LMD9 and its mutant of the three surface proteases PrtS, HtrA and STER_1612. Triple Strain Wild mutant protease Number of peptides from IL-10 9 0 Number of peptides from elafin 43 0

[0194] The results obtained with the two protein models being of the same nature, the rest of the work focused only on one model, elafin.

[0195] Strain CNRZ1066

[0196] The same experiments were conducted on strain CNRZ1066.

[0197] Materials and Methods

[0198] The plasmid encoding elafin is extracted from the lactococcal strain containing it, and introduced into the wild-type and mutant strains of CNRZ1066 by natural competence. To evaluate the stability of elafin, their degradation fragments are looked for in the culture medium after 4 h of growth, following the same approach developed for the LMD9 strain.

[0199] Results

[0200] Thirty-six spectra from elafin are identified in the supernatant of the wild-type strain. The region of elafin covered by the degradation products is the same as in the case of strain LMD9, and represents 40% of the total sequence (see FIG. 5). Under the same experimental conditions, no elafin degradation fragment is detected in the double mutant.

[0201] Conclusion

[0202] No heterologous protein degradation fragment is found in the supernatant of an S. thermophilus strain mutated for its surface proteases, regardless of whether this strain possesses the PrtS protease.

[0203] 2) Reduced Degradation of Heterologous Proteins Correlates with Increased Production of the Intact Protein

[0204] Since the results are comparable whether the S. thermophilus strain has two or three surface proteases, further work focuses on only one strain, CNRZ1066. Since no elafin degradation fragment is detected in the double mutant supernatant, it is necessary to ensure that the protein is produced in this strain, and at a higher level than in the wild type strain.

[0205] Materials and Methods

[0206] This was done by immunodetection of whole elafin in the supernatant (since the protein is secreted into the external environment by the streptococcal strain), using the following experimental protocol.

[0207] Culture and induction of elafin production by strains carrying plasmid pLB386 is performed as described in Example 6. After 15 min of induction, the cells are removed by centrifugation, and the supernatant containing elafin is filtered through a 0.22 μm pore size filter (low protein adsorption PVDF membrane). Ten ml of supernatant is concentrated by a factor of 20 by ultrafiltration on 3 kDa cut-off membrane (Amicon ultracell 3 k, MerckMillipore). Five μl of retentate are deposited on polyacrylamide gel (NuPAGE 4-12% Bis-Tris Gel, Invitrogen). After migration for 1 h at 110 mA and 200V, proteins are transferred onto a PVDF transfer membrane (Trans-Bot Turbo Mini transfer pack, Bio-Rad). Elafin, after being labeled with a mouse monoclonal anti-elafin antibody (SantaCruz Biotechnology), is detected by chemiluminescence (ECL Plus Western kit, Pierce).

[0208] Results

[0209] Several strains are used in these experiments, carried out several times and systematically giving the same results; illustrated for one of them in FIG. 6: [0210] The wild type and mutated S. thermophilus CNRZ1066 strains do not produce elafin (wells 3 and 4, labeled pls empty). No bands are observed, indicating that neither strain produces elafin (nor protein cross-reacting with elafin antibody), [0211] The wild type CNRZ166 strain producing elafin (well 5, noted WT pls elafin). Two bands are observed, the upper of which migrates at a very slightly smaller size than the mature form of commercial elafin. The second migrates at a significantly smaller size, and would be a truncated form of the degrading elafin, [0212] Strain CNRZ1066 mutated for its elafin-producing surface proteases (well 6, denoted elafin pls mutant). The lower band corresponding to the truncated form of elafin is no longer detected, only the band corresponding to mature elafin is revealed.

[0213] Conclusion

[0214] Inactivation of surface proteolysis abolishes elafin degradation in S. thermophilus.

[0215] 3) Inactivation of Surface Proteases Induces an Increase in the Enzymatic Activity Carried by Elafin

[0216] Materials and Methods

[0217] The activity of elafin accumulated in the supernatants of wild-type and mutant CNRZ1666 strains is assessed by determining the inhibitory potency of the activity of a control human protease (elastase), following the protocol described below.

[0218] S. thermophilus strains carrying the pLB386 plasmid are grown in MCD at 37° C. to an OD.sub.600 of 1.0.

[0219] L. lactis strains carrying the same plasmid are grown at 30° C. in lactococcal-specific MCD (Otto et al., 1983) to an OD.sub.600 of 1.0.

[0220] At this stage of growth, the cultures are centrifuged, and the cells resuspended in fresh MCD at an OD.sub.600 of 2.0.

[0221] Elafin production is obtained by induction with bile salts (15 ng/ml) and overnight incubation at 37° C. The cells are then centrifuged, and the elafin contained in the supernatant is concentrated by ultrafiltration by a factor of about 300.

[0222] Elafin is a protease inhibitor. It is therefore measured according to the following principle. The activity of a human protease (elastase) is measured by fluorescence using a labeled substrate (EnzCheck® Elastae assay kit, Molecular Probes), in the presence and absence of supernatant containing elafin. Inhibition intensity (reflecting elastase concentration) is measured by the difference in fluorescence between the measurement without and with elafin over time following the protocol delivered with the assay kit (Molecular Probes).

[0223] Results

[0224] Elafin activity produced by the strain lacking proteolytic activity increases approximately twofold compared to that produced by the wild type strain (FIG. 7).

[0225] Conclusion

[0226] Inactivation of surface proteolysis induces a doubling of the elafin activity produced by the strain.

EXAMPLE 8: PRODUCTION OF ELAFIN BY NON-PROTEOLYTIC BACTERIA STRAINS OF THE INVENTION

[0227] Elafin production is evaluated for L. lactis strain IL1403ΔHtrA and S. thermophilus strain CNRZ1066ΔHtrAΔster_1612.

[0228] Materials and Methods

[0229] Both strains are grown to an identical population level, and induction of elafin production is performed under the optimal conditions for each strain (see Example 7 point 3)).

[0230] Results

[0231] The results obtained by immunodetection of elafin produced by each of these strains show for strain CNRZ1066, the presence of a band corresponding to the intact protein detected in the protease mutant and the wild type strain, and a truncated form detected in the wild type strain only. For strain IL1403, truncated forms of elafin are still observed in the single ΔHtrA mutant (FIG. 8).

[0232] Thus, the lack of elafin degradation is only seen in S. thermophilus ΔHtrAΔSter_1612.

[0233] Furthermore, results on the elafin activity produced by each of these strains show that elafin activity is 5-10 times higher in S. thermophilus CNRZ1066ΔHtrAΔster_1612 than in L. lactis IL1403ΔHtrA (FIG. 9).

EXAMPLE 9: MUTANT CONSTRUCTIONS OF THE LACTOCOCCUS LACTIS BACTERIAL STRAIN

[0234] The L. lactis IL1403 and L. lactis MG1363 strains lack a plasmid and therefore do not produce the wall protease PrtP (whose gene is carried by a plasmid, Gasson, 1983). The other two surface proteases produced are therefore HtrA (Poquet et al., 2000) and YwdF in IL1403 (or its homolog llmg-2442 in MG1363). The IL1403ΔhtrAΔywdF double mutant, lacking all three surface protease activities PrtP, HtrA, and YwdF, was constructed from the IL1403ΔhtrA strain (Guillot et al., 2016) by homologous double recombination using the heat-sensitive plasmid pGhost9 following the established protocol (Biswas et al., 1993).

[0235] The L. lactis MG1363ΔhtrAΔllmg-2442 double mutant, lacking the three surface protease activities PrtP, HtrA and Ilgm-2442, was constructed in two steps from the wild type L. lactis MG1363 strain. The first step consisted in inactivating the htrA gene by double homologous recombination, the second in inactivating the Ilgm-2442 gene in the previously obtained single mutant MG1363ΔhtrA, following a strategy identical to that described above for L. lactis IL1403.

EXAMPLE 10: OBTAINING ELAFIN-PRODUCING STRAINS OF LACTOCOCCUS LACTIS

[0236] Plasmid pLB386 was purified from L. lactis strain LBH832 as described in Example 6. It was introduced into L. lactis strain IL1403, its L. lactis IL1403ΔhtrAΔywdf double mutant, and the L. lactis MG1363ΔhtrAΔllmg-2442 double mutant by electroporation. The L. lactis LBH832 strain is none other than the wild type L. lactis MG1363 strain carrying the plasmid pLB386.

[0237] Selection of transformants and the presence of plasmid pLB386 were performed as described in Example 6.

[0238] In parallel, control strains carrying a plasmid without elafin, called empty plasmid (pLB44) were constructed.

[0239] Plasmid pLB44 was purified from L. lactis strain LBH68 as described in Example 6. It was introduced into L. lactis strain IL1403, its double mutant L. lactis IL1403ΔhtrAΔwdf, and the double mutant L. lactis MG1363ΔhtrAΔllmg-2442 by electroporation. The L. lactis LBH68 strain is none other than the wild type L. lactis MG1363 strain carrying the pLB44 plasmid.

EXAMPLE 11: INACTIVATION OF SURFACE PROTEASES INCREASES THE AMOUNT OF HETEROLOGOUS PROTEIN PRODUCED BY LACTOCOCCUS LACTIS

[0240] Materials and Methods

[0241] Verification that inactivation of L. lactis surface proteases results in an increase in the amount of heterologous protein produced was done by immunodetection of whole elafin. The two pairs of wild type and mutant strains (i.e. not synthesizing PrtP, HtrA, or YwdF/llmg-2442) carrying the plasmid pLB386 or pLB44 were grown at 30° C. in chemically defined medium (Otto et al., 1983). Induction of elafin production and its immunodetection were performed as described in Example 7. It should be noted that the same amounts of supernatant are deposited for each strain, so differences in band intensity will reveal differences in protein concentration in the culture supernatant.

[0242] Results

[0243] Several strains are used in these experiments, performed twice and giving systematically the same results, illustrated respectively for L. lactis IL1403 and L. lactis MG1363 by FIGS. 10 and 11:

[0244] Wild-type and mutated L. lactis IL1403 and MG1363 strains not producing elafin (wells 4 and 6, labeled WT pls empty and Mutant pls empty, respectively). No bands are observed, indicating that neither strain produces elafin (or protein that cross-reacts with the elafin antibody),

[0245] The wild-type strain of L. lactis IL1403 producing elafin (FIG. 10, well 3, noted WT pls elafin). Two low intensity bands are observed, the upper of which migrates at a very slightly smaller size than the mature form of commercial elafin. The second migrates at a significantly smaller size (6 kDa), and is thought to be a truncated form of the degrading elafin,

[0246] The wild-type strain of L. lactis MG13633 producing elafin (FIG. 11, well 3, noted WT pls elafin). Two very low intensity bands can be seen, the upper of which migrates at a size very slightly smaller than that of the mature form of commercial elafin. The second migrates at a significantly smaller size (10 kDa), and would be a truncated form of elafin undergoing degradation. The low intensity of the bands reflects a high degradation activity of elafin,

[0247] The elafin-producing mutant strain of L. lactis IL1403 (FIG. 10, well 5, labelled Mutant pls elafin). The lower band corresponding to the truncated form of elafin is almost not detectable anymore, the band corresponding to mature elafin is revealed in a majority, with a much higher intensity than in the wild type strain, reflecting a very low proteolysis of elafin in the double mutant,

[0248] The elafin-producing L. lactis MG1363 mutant strain (FIG. 11, well 5, noted Mutant pls elafin). Three bands are observed, one of which is of very low intensity (6 kDa) corresponding to the truncated form of elafin also seen in the L. lactis mutant 1L1403. The upper band (in the form of a doublet) is the same as that found in the wild type strain, but with a much higher intensity, reflecting a much higher concentration of undegraded elafin in the supernatant of the mutated strain.

[0249] Conclusion

[0250] Inactivation of surface proteolysis greatly reduces the degradation of elafin, resulting in a higher concentration of the intact form of elafin in the supernatant of a L. lactis strain not producing surface protease than in that of a wild-type L. lactis strain.

REFERENCES

[0251] Benbouziane B. et al, J. Biotechnol, 2013, 168: 120-129 [0252] Bermudez-Humaran L. G. et al, Curr. Opin. Microbiol, 2013, 16:278-283 [0253] Bermudez-Humaran L. G. et al, Microb. Cell Fact. 2015, 14:26 [0254] Bermudez-Humaran L. G. and Langella P., Nat. Biotechnol, 2018, 36:816-818 [0255] Biswas I. et al, J. Bacteriol., 1993, 175:3628-3635 [0256] Botos I. et al, J. Biol. Chem., 2004, 279:8140-8148 [0257] Bron P. A. et al, Cur. Opin. Biotechnol, 2019, 56:61-68 [0258] Charrette M. F. et al, Proc. Natl. Acad. Sci. 1981; 87:4728-4732 [0259] Cossart P. and Joncquières R., Proc. Natl. Acad. Sci. USA, 2000, 97:5013-5015. [0260] Delorme C. et al, Appl. Environ. Microbiol, 2010, 76:451-460 [0261] Desvaux M. et al, Front. Microbiol, 2018, 9:100 [0262] Fernandez-Espla M. D. et al, Appl. Environ. Microbiol, 2000, 66:4772-4778 [0263] Gardan R. et al, J. Bacteriol, 2009, 191:4647-4655 [0264] Gasson M., J. Bacteriol, 1983, 154:1-9 [0265] Gottesman S. et al, J. Bacteriol, 1978, 133:844-851 [0266] Guillot A. et al, J. Proteome Res, 2016, 15:3214-3224 [0267] Hafeez et al, Appl. Microbiol. Biotechnol, 2013, 97, 9787-9799 [0268] Hafeez et al, J. Agri. Food Chem. 2015, 63, 7522-7531 [0269] Hossain M. S. et al, J. Bacteriol, 2012, 194:5886-5896 [0270] Letort C. et al, J. Appl. Microbiol, 2001, 91:1023-1029 [0271] Motta J. P. et al, Sci. Trans!. Med. 2012; 4:158ra144 [0272] Neirynck S. and Steidler L., Biotechnol. Genet. Eng. Rev, 2006, 22:253-266. [0273] Otto R. et al, FEMS Microbiol. Lett. 1983, 16: 69-74. [0274] Poquet I. et al, Mol. Microbiol, 2000, 35:1042-1051 [0275] Shamsi T. N et al, Int. J. Biol. Macromol, 2016, 91:1120-1133. [0276] Siezen R. et al, Ant. Leeuwenhoek, 1999, 76:139-155 [0277] Soualmia and El Amri, 2018. Expert Opin. Ther. Pat. 28:93-110 [0278] Terzagui B. E. et al, Appl. Microbiol, 1975, 29:807-813 [0279] Trieu-Cuot P. et al, Gene, 1983, 23:331-341 [0280] Trieu-Cuot P. et al, Nucleic Acids Res, 1990, 18: 4296 [0281] Wei C. et al, J. Genet. Genomics, 2013, 40:281-289