|After a stimulation period of 15 h at 37 C in 5% CO2, culture supernatants were collected and stored at-80 C until cytokine analysis. IL-10 were analyzed using commercially available ELISA kits (BD PharMingen) according to manufacturer's instructions.
As shown in Fig. 22 the ESP from L. plantarum 299v induces IL-10 production in dendritic cells in a concentration dependent matter. The lack of IL-10 induction from PBS alone (Fig. 22) or from the culture supernatants (data not shown) of L. plan- tarum 299v shows that a component in the ESP is responsible for the induction of IL-10. These results demonstrate that part of the immunomodulating component (s) of L. plantarum 299v is present at and non-covalently bound to the cell surface.
Example 14: High throughput screening of L. plantarum 299v mutant strains for low extracellular amounts of GAPDH A library of random mutants of Lactobacillus plantarum 299v was generated by a modified version of the method of Ibrahim and O'Sullivan, 2000. Strain 299v was
grown in MRS broth (Oxoid, Basingstoke, Hampshire, England) at 30 C for 24h. The optical density of the culture (at 600 nm) was 8.0. The culture was harvested (5000 RPM, 20 C, 10 min) and resuspended in 10 mL of 100 mM K2HPOIKH2PO4 buffer, pH 7.5. 100 p. L of the cell suspension was withdrawn, and the remaining cell sus- pension was mixed with 1,2 mL ethyl methanesulfonate (EMS) (Sigma Co. , St.
Louis, MD, USA). Samples of the cell suspension (1 mL) was withdrawn after 15 min, 30 min, 1h, 1h 30 min, 2h, 2h 30 min, 3h, 4h, 5h, 7h 30 min and 10h. Cells were harvested (10,000 g, 4 C, 30 sec) and washed twice in 1 mL 100 mM K2HP04/KH2PO4 buffer, resuspended in 40 mL of pre-warmed MRS and allowed to grow at 30 C for two hours. To determine kill rates, dilutions of cultures were plated on MRS agar (Oxoid) and grown for two days at 30 C and the number of colonies was counted. Cells withdrawn before addition of EMS were not allowed to grow before plating. Ten mL 75% (v/v) Glycerol was added to cultures, and 12 aliquots (1 mL) from each culture were frozen and stored at-80 C.
If it is assumed that EMS treated cells were dividing once before plating, kill rates are :
Oh 15 min EMS treatment : (490-14. 6/2)/490x100% = 98. 5%
Oh 30 min EMS treatment: (490-1, 28/2)/490x100% = 99, 87%
1 h EMS treatment: (490-0, 073/2)/490x100% = 99,993%.
Mutant libraries generated by 15 min EMS treatment and 30 min EMS treatment were used in the screening for strains with lower extracellular amounts of GAPDH.
Identification of isolates with lower amounts of extracellular GAPDH Dilutions of a thawed aliquot of 299v mutants generated by 15 min EMS treatment were plated on Genetix Qtrays (Genetix, New Milton, Hampshire, UK) containing 200 mL MRS agar (Oxoid) in order to obtain 2000-3000 colonies per tray. Clones were transferred to 150 tL of sl\ARS in Nunc 96 microwell plates (Nunc, Roskilde, Denmark) using a Qpix colony-picking robot (Genetix). Microwell plates were incu- bated 21-27h at 30 C in a gas mixture containing 10% H2, 10% CO2, and 80% N2 in a MK3 Anaerobic work station from DW scientific (Shipley, West Yorkshire, UK).
The assay for extracellular GAPDH activity was modified for microwell plates (ex- ample 9). Cultures were mixed using a multi channel pipette to resuspend precipi-
tated cells, and 5 IlL of the cultures were transferred to new microwell plates. The assay reactions were initiated by the addition of 150 pi-reaction mixture. Plates were incubated 45 min at room temperature and photographed on an UV transi- luminator to record fluorescence of NADH at 450 nm. During incubation, the optical densities at 595 nm (OD595) were determined using a microwell plate reader.
Photographs from plates were visually inspected and cultures resulting in lower fluorescence, indicating lower levels of extracellular GAPDH, were selected. For selected isolates, OD595 readings were examined to estimate, whether low flores- cence was a result of poor growth or low levels of extracellular GAPDH. If low fluo- rescence was estimated to be due to low levels of extracellular GAPDH, then iso- lates were selected for further analysis. An example, showing photographs and OD595 readings from two assay plates, is shown in Fig. 23. Based on photographs, the cultures in plate 53, well H12 and in plate 52, well D4 were selected. However, OD595 readings revealed that the low GAPDH activity in plate 52, well H12 was due to poor growth. The culture in plate 52, well D4 grew normally, and this clone was therefore selected for further analysis.
A total of 161 microwell plates containing more than 15, 000 mutants were screened.
All screened microwell plates were stored at-80 C after addition of 50 IlL 75% (v/v) glycerol (to each well) until the analysis of isolates were completed Initially, 47 clones were selected. For selected clones, new assays for extracellular/surface- located GAPDH were performed as described in example 9. Nine clones displayed lower GAPDH activity in these assays and were further characterised (example 15).
In conclusion example 14 demonstrates generation of random mutants by EMS mutagenesis of L. plantarum 299v. Furthermore, 15000 mutant strains could be investigated for the presence of extracellular/surface-located GAPDH by use of a high-through-put screening method. Of the 15000 screened clones the high-through- put screening produced nine final candidates with apparent low amounts of extra- cellular/surface-located GAPDH.
Example 15: GAPDH and LDH activities in culture supernatants and ESP- fraction of selected isolates
Overnight cultures of L. plantarum 299v, L. plantarum WCFS1, and nine selected mutants of L. plantarum 299v, were assayed for GAPDH and LDH activity in the ESP-fraction and in culture supernatants as described in example 9. The result is shown in the table below.
GAPDH and LDH activities (U/mL) in culture supernatants (SN) and ESP-fraction
Isolate 299v WCFS1 8-C8 14-A5 30-H9 75-H12 91-F11 92-A4 102-B3 108-F4 149-D7
LDH SN 0. 00 0. 00 0. 00 0.00 0.01 0.00 0.00 0. 00 0. 00 0. 22-0. 01
LDH ESP 0. 02 0. 00 0. 00 0. 02 0. 00 0. 00 0. 00 0. 01 0. 00 0. 14 0. 00
GAPDH SN 0. 00 0. 00 0. 00 0. 01 0. 00 0. 00 0. 00 0. 00 0. 00 0. 08 0. 00
GAPDH ESP 0. 18 0. 01 0. 00 0. 15 0. 02 0. 05 0. 04 0. 09 0. 07 0. 20 0. 04
Isolate 8-C8 was deselected. This mutant showed very slow growth and had no LDH activity in a cell lysate prepared as in example 9 (not shown). It was assumed to be an LDH mutant. Several of the other clones showed a high extracellular LDH activity indicating a high degree of lysis.
Isolate 149-D7 was selected for further work because this strain showed normal growth, and reproducible low GAPDH activity in culture supernatants and in ESPfractions. Lactobacillus plantarum strain 149-D7 was deposited at the DSMZ (Deut- sche Sammlung von Mikroorganismen und Zellkulturen GmbH), and has been registered under number DSM 16241.
New activity assays were made to confirm the low GAPDH activity in culture super- natants and ESP-fractions of strain 149-D7, and to study the activity levels for other strains.
Fig. 24 shows a comparison of GAPDH and LDH activities in the culture supernatants and ESP-fractions of L. plantarum strains 299v, WCFS1, and 149-D7. The GAPDH activity in the ESP fraction of 149-D7 is significantly lower than in equivalent fraction from the wild type 299v, indicating that the mutation in 149-D7 has affected genes involved in surface display of GAPDH.
Also shown in Fig. 24 is assay results from strains 149-D7/129 and UP102. Strain UP102 was isolated from a library of WCFS1 containing DNA fragments from L. plantarum 299v during screening for clones that displayed higher levels of extracel- lular/surface-associated GAPDH than the host strain WCFS1 (see example 17).
Lactobacillus plantarum strain UP-102 was deposited at the DSMZ (Deutsche
Sammiung von Mikroorganismen und Zelikulturen GmbH), and has been registered under number DSM 16240. The plasmid in strain UP102, pUP102 was transformed into strain 149-D7 to obtain the strain 149-D7/129.
The GAPDH activities of culture supernatants and ESP-fractions of strain UP102 were higher than the corresponding activities for the host strain, WCFS1, but lower than for 299v. In contrast to 299v, approximately half the extracellular GAPDH activity was found in the culture supernatant. This shows that although pUP102 was able to increase the levels of GAPDH outside cells of WCFS1, it did not give the host the same phenotype as 299v with respect to binding GAPDH to the cells.
The question was whether the pUP102 plasmid could also increase extracellu- lar/surface-located GAPDH in the 299v-mutant 149-D7. As seen in Fig. 24, strain 149-D7/129 had higher GAPDH activities than 149-D7 in both culture supernatants and ESP-fractions. Although the activities did not reach the levels of strain 299v, the presence of pUP102 partially complement the mutation in 149-D7 with respect to surface display of GAPDH. A high proportion of the extracellular activity was found in the ESP fraction, indicating that the complemented 149-D7 mutant has retained the 299v wild type phenotype of binding GAPDH to the surface.
LDH activity was high in the culture supernatant of UP102, but low in culture super- natants from other strains. This indicated a higher degree of lysis in this strain.
Strain 149-D7/129 also had higher LDH activity outside cells (compared to 149-D7), but in this case LDH adhered to cells. It is therefore possible, that some of the higher GAPDH activities seen outside cells of strain UP102 and strain 149-D7/129 were due to higher lysis. Similarly, the level of LDH outside strain 149-D7 cells was lower than the levels seen for strain 299v.
In conclusion, we have obtained a mutant of L. plantarum 299v that shows a significantly lower surface-located activity of GAPDH. Obviously, the mutation may also affect the surface-location of other proteins that are not synthesised with secretion signals, like ENO and PGK. This was investigated by immunoblotting (see example 16). The mutant will be useful in investigation of the role of GAPDH in the strains adhesive properties and its ability to colonise and stimulate the host immune sys-
tem. Furthermore, complementation studies with the mutant as described above can be used to isolate the genes involved in surface display of GAPDH.
Example 16: Immuno-detection of GAPDH, ENO and PGK from selected mutants of 299v.
In order to confirm the results obtained in activity assays, SDS-PAGE and immu- noblotting was also used to evaluate levels GAPDH in cell extracts, ESP (eluted surface proteins) and culture supernatants.
ESP and culture supernatants were prepared as described (Example 9). Cell extracts were prepared as follows : Equal amounts of cells, glass beads (from Sigma, 106 pm and finer) and PBS buffer were mixed. The cells were disrupted using a Fastprep FP120 from Qbiogene (Carlsbad, CA, USA) in three cycles at maximum speed for 25 seconds followed by cooling on ice. Western blot of cell extracts, ESPfractions, and culture supernatants from strains 299v, WCFS1, and the selected mutant, 149-D7 were prepared as described in example 10. The volume loaded on gels corresponds to 10 uL culture for the cell extracts, 1 RL culture for the ESPfraction and 15, uL culture for supernatants. The western blots are shown in Fig. 25.
The lower amounts of GAPDH in the ESP-fraction and culture supernatant of strain 149-D7, compared to strain 299v, are clearly seen. A Coomassie stained gel loaded with ESP-fractions did also demonstrate a reduced amount of GAPDH on the surface of strain 149-D7.
Proteins in culture supernatants and ESP-fractions (surface proteins) from the L. plantarum strains 299v, WCFS1,149-D7, 149-D7/129 and UP102 were separated by SDS-PAGE and blotted onto nitrocellulose membranes as described above.
Loaded sample volumes corresponding to 150 FL for culture supernatants and 10 I1L for ESP-fractions. Three blots were prepared, and as primary antibodies were used anti-GAPDH, anti-ENO, and anti-PGK, respectively. Alkaline phosphataseconjugated goat anti-rabbit antibodies from Dako Cytomation were used as secondary antibodies. Blots were developed using NBT/BCIP tablets from Roche Diagnostics. The results are shown in Fig. 26. The amounts of GAPDH, ENO and PGK were lower in culture supernatants and in ESP-fractions of strain 149-D7 compared to strain 299v. However, the levels of all three proteins were even lower for strain WCFS1. Strain UP102, described in example 18 had higher levels of all three proteins in both culture supernatants and at cell surfaces compared to the host strain
WCFS1. The levels of GAPDH, ENO and PGK were also higher for strain 149- D7/129 than for strain 149-D7. This confirms the results from activity assays for GAPDH in culture supernatants and in ESP-fractions from these strains (example 16). Estimated from immunoblots in Fig. 26, levels of ENO and PGK correlate to the levels of GAPDH in these strains.
To summarise the observations, activities of GAPDH, ENO and PGK in ESP frac- tions decreased in concert in the 149-D7 mutant. Likewise they increased simula- neously in the complemented strain 149-D7/102. This indicates that a common mechanism is responsible for the surface display of these glycolytic enzymes.
Example 17: Screening for genes involved in surface display of GAPDH L. plantarum 299v is able to display the normally intracellular located enzyme glyc- eraldehyde-3-phosphate dehydrogenase (GAPDH) on the cell surface. In contrast, GAPDH levels were close to or below the detection limit on the surface of L. plan- farum WCFS1. With the aim of identifying genes involved in GAPDH surface dis- play, a 299v genomic DNA library was screened in strain WCFS1.
Genomic DNA was isolated from L. plantarum 299v and partially digested with Sau3AI. The partially digested DNA fragments were separated on agarose gel and fragments with a minimum size of 5 kb were isolated. These fragments were ligated with the vector pTRKL2 (OSullivan and Kiaenhammer), which had been digested with BamHI. The ligation mixture was transformed into E. coli and transformants selected on LB agar plates containing erythromycin (200 lig/mL). Pools of transfor- mant colonies were washed off the selective agar plates and plasmid DNA was isolated from these transformant pools. The obtained plasmid DNA pools were used for transformation of L. plantarum WCFS1 and transformants were selected on MRS agar in Qtrays (Genetix) containing erythromycin (5 lig/mL). Individual transformants were picked from the Qtrays using a Qpix colony-picking robot and inoculated into the wells of microtiter plates containing 150 IlL MRS with erythromycin. After over- night incubation at 37 C, glycerol was added to a final concentration of 20%, and the cultures were stored at-80 C. Later, transformants were inoculated into new microtiter plates containing 150 IlL or 200 L sMRS with erythromycin. These mi- crotiter plates were incubated and assayed for GAPDH activity as described in example 9.
Three clones with extracellular/surface-associated GAPDH activity were identified.
Plasmid DNA was isolated from these clones and restriction enzyme analysis using EcoRl indicated that the three plasmids contained the same insert. One of the plas- mids, pUP0102, has been further characterised. The plasmid contains an insert of 6.2 kb (Fig. 27) and sequence analysis indicate the presence of the 3'end of a regulatory gene, the entire rpoB gene including the promoter region and the 5'end of the rpoC gene. Deletions in the rpoB sequence were made by digestion with Fspl (pUP0164), Nrul (pUP0165) or Bglll (pUP0163) (Fig. 27). The resulting plasmids were transformed into L. plantarum WCFS1 and the transformants analysed for surface-associated GAPDH activity. None of these transformants displayed extra- cellular/surface-associated GAPDH activity (Fig. 27).
In summary, these results indicate that the rpoB gene from L. plantarum 299v is able to mediate surface display of GAPDH when transformed into L. plantarum WCFS1. The rpoB gene encodes the subunit present in the core enzyme (a2pp') of the RNA polymerase complex. The 8 subunit is implicated in the binding of nu- cleotides needed for RNA polymerisation.
Example 18: Analysis of eluted surface proteins by 2D-PAGE In order to determine the identity of the most abundant proteins displayed on the cell surface of L. planzarum eluted surface proteins (ESP) were analysed by D PAGE.
ESP from L. plantarum 299v were prepared as described in example 9. The ESP were precipitated from the solution by adding 4 volumes of ice-cold acetone and incubation for 2 hours at-21 C. The resulting pellet was collected by centrifugation (15,000 x g, 10 min) at 4 C. The pellet containing the cell surface-associated pro- teins were resuspended in a solution containing 8 M urea, 2% Chaps, 0.002% bro- mophenol blue, 50 mM dithiothreitol (DTT), 0.2% w/v carrier ampholyte, pH 3-10 (Bio-Rad, Laboratories Ltd., Hemel Hempstead, Hertfordshire, UK) to give a final protein concentration of approximately 0.1 llg/lli and used to rehydrate 11-cm pH 4 to 7 linear immobilised pharmalyte gradient (IPG) strips. Strips were rehydrated over-night under passive conditions and focused for 30,000 Volt-hours according to manufactures instructions.
Prior to loading on the second dimension, focused IPG strips were equilibrated sequentially in a buffer (Tris-HCI buffer containing 6 M urea, 30% [vol/vol] glycerol, 2% sodium dodecyl sulphate [SDS] ) containing 2% DTT or 2.5% iodoacetamide for 15 min each and applied to 10% polyacrylamide Tris-HCI gels. SDS-PAGE was carried out with a Protean II cell (Bio-Rad), and proteins were resolved at a constant voltage of 200 V for 1 hour. Proteins were visualised by silver staining (Shevchenko et al. ; 1996). The 2-D PAGE of L. plantarum 299v analysis is shown in Fig. 28. The most abundant gel spots were excised from the gel,"in-gel"digested with trypsin, and analysed by nano-ESI-MS/MS as described in example 3. The box below summarises the sequences and protein identities obtained from the MS analysis of the tryptic digests of proteins isolated by 2D-PAGE.
In summary all the identified ESP are homologous to proteins normally considered as being located intracellular. Nine of the ten identified are enzymes of the metabolic pathways of the cell.
Peptide sequences and protein identities/assignments from outer cell surface associated proteins separated by 2D-PAGE and analysed by MS. Protein spots were in-gel digested with trypsin, and the resulting peptides were extracted and sequenced by nano ESI-MS/MS analyses. Proteins were identi- fied by searching the non-redundant Blast-protein-protein sequence database.
Spo ton Peptide t measured no.
1 40 (+2) DQLPLNQNW (SEQ ID : 36) Maltose phosphorylase
2 : 37) Pyruvate Kinase 3 41 (+2) DTSLDWGGMQFDR (SEQ ID : 38) GroEL chaperonin 3 60 (+2) GRVLEQSYGSPTLT (SEQ ID : 39) GroEL chaperonin 4 5 50 (+2) VNVNTENQVAFANATR : 40) Fructose bisphosphate aldolase 6 88 (+2) VMPFWEDFF (SEQ ID NO : 41) Phosphoglycerate mutase 7 44 (+2) LVESLTPDDVLPGMK (SEQ ID NO 42) 7 37 (+2) LTDYFEGWDPA (SEQ ID NO : 43) ss-Phosphoglucomutase 8 43 (+2) TATSDQAEE (I/L) : 44) Triosephosphate isomerase 9 38 (+2) NVGVDN (I/L) DVPTVK (SEQ ID D-Lactate dehydrogenase 10 657. 87 (+2) (I/L) NDG (I/L) AE (I/L) (SEQ ID : 46) L-Lactate dehydrogenase 10 589. 80 (+2) (I/L) GSGTS (I/L) DSSR (SEQ ID NO : 47) L-Lactate dehydrogenase sequence Identification: For the ions measured and peptide sequences please see Example 3.
Example 19: Identification of different isoforms of GAPDH The 2-D PAGE analysis revealed three distinct protein spots all identified as GAPDH (Fig. 28). This indicated the presence of different isoforms of GADPH. A theoretical tryptic digest of the polypeptide sequence of GAPDH showed two putative peptides (ALGLVIPELNGK (SEQ ID NO : 48); MW 1223.74 Da) corresponding to residue 221- 232, and (YDSTHGTLNADVSATDDSIWNGK (SEQ ID NO : 49); MW 2478.16 Da), corresponding to residue 51-74 that could not be assigned by ESI-MS analysis of a tryptic digest of GAPDH isolated by gel electrophoresis.
Two peptides, a doubly charged peptide at m/z 612.87 and a triply charged peptide at m/z 827.38 (Fig. 29) with MWs of one dalton higher than the corresponding pep- tides from the theoretical digest (MW 1223.74 and 2479.14 Da for the analysed peptides versus 1222.73 and 2478.16 Da obtained from the theoretical digest) were identified in the GAPDH tryptic digest. Nano-ESI MS/MS analysis of the ions at m/z 612.87 (Fig. 30) and m/z 827. 38 (Fig. 31) revealed sequences identical with the sequences obtained from the theoretical digest of GAPDH except from the aspara- gine residues N72 and N that were deamidated to aspartate residues, D and D230. These results account for the three isoforms of GAPDH corresponding to non- modified GAPDH, a single deamidated form of either residues of N72 or N230 and the third isoform where both residues N72 and N230 are deamidated. This post- translational modification of GAPDH may be unique to the extracellular GAPDH and may be essential for the non-glycolytic function e. g. adhesion or signaling of GAPDH on the cell surface.
Example 20. Gene inactivation in L. plantarum 299v.
Construction of integration plasmid and transformation into L. plantarum.
Plasmid pTN1 was recently developed and successfully used for gene inactivation in L. gasseri (Nue and Henrich, 2003). The pTN1 vector replicates at 35 C whereas replication is efficiently shut down at 42 C, allowing the use of the vector for single copy integrations in L. gasseri. The present example describes the use of pTN1 for construction of a threonine auxotroph mutant in L. plantarum 299v. Inactivation of genes needed for threonine biosynthesis serves as an example of inactivation of specific genes. A similar approach can be used for inactivation of genes that are expected to be essential for probiotic activity.
The complete genome sequence of L. plantarum WCFS1 revealed the presence of a threonine biosynthetic pathway. Based on the threonine biosynthetic genes identified in L. plantarum strain WCFS1 we assumed that the same genes are present in strain 299v. Furthermore we assumed that the gene sequences between the two subspecies are almost identical allowing primer design that is based on the published WCFS1 genome sequence (Kleerebezem et al. ; 2003). We attempted to construct a threonine auxotrophic strain by deletion of the C-terminus of the gene encoding Hom2 (homoserine dehydrogenase) and the N-terminus of the gene encoding ThrB (homoserine kinase). A 500 bp PCR fragment covering an internal region of the hom2 gene was obtained using the primers hom2-thrB-1 (5'GAGGATATTGCGGAAGCTC 3' (SEQ ID NO : 50) ) and hom2-thrB-2 (5'GCGCCGGTCAAT- CATTCATGGCATGGGTAATG 3' (SEQ ID NO : 51) ) and genomic L. plantarum 299v DNA as template. Similarly, a 500 bp PCR fragment covering an internal region of the thrB gene was obtained using the primers hom2-thrB-3 (5'CATGAATGATT- GACCGGCGCAACG CGCTCTTC 3' (SEQ ID NO : 52)) and hom2-thrB-4 (5' CTTGGCTCAATTGTGC CTGC 3' (SEQ ID NO : 53)) and genomic L. plantarum 299v DNA as template.
The following PCR profile was used to amplify both 500 bp fragments:
94 C 5 min 94 52 C 30 sec 25 72 72 C 2 min The two primers hom2-thrB 2 and hom2-thrB 3 contain 5'ends that are complementary to each other. The two synthesised PCR fragments containing overlapping regions were allowed to anneal to each other before extension and amplification using the outer primers hom2-thrB 1 and hom2-thrB 4.
The following extension profile was used: 94 C 5 min
60 C 30 sec 72 C 10 min The following PCR profile was used to amplify the extended 1000 bp fragment:
94 C 5 min 94 52 C sec 25 72 72 C 2 min The extended PCR product was purified using the GFXTM PCR DNA and gel band purification kit (Amersham Biosciences) and inserted into the pCR2. 1#-TOPO vector (Invitrogen) resulting in plasmid pPSM1081. The polylinker region of pCR2. 1;- TOPO contains two EcoRl restriction sites that flank the hom2-thrB insert in pPSM1081. Plasmid pPSM1081 was digested with EcoRl, the 1000 bp fragment was purified and inserted into plasmid pTN1 (integration vector), which was pre- digested with EcoRI and treated with bacterial alkaline phosphatase.