|Home | About | Journals | Submit | Contact Us | Français|
Solely in Europoe, Salmonella Typhimurium causes more than 100,000 infections per year. Improved detection of livestock colonised with S. Typhimurium is necessary to prevent foodborne diseases. Currently, commercially available ELISA assays are based on a mixture of O-antigens (LPS) or total cell lysate of Salmonella and are hampered by cross-reaction. The identification of novel immunogenic proteins would be useful to develop ELISA based diagnostic assays with a higher specificity.
A phage display library of the entire Salmonella Typhimurium genome was constructed and 47 immunogenic oligopeptides were identified using a pool of convalescent sera from pigs infected with Salmonella Typhimurium. The corresponding complete genes of seven of the identified oligopeptids were cloned. Five of them were produced in E. coli. The immunogenic character of these antigens was validated with sera from pigs infeced with S. Tyhimurium and control sera from non-infected animals. Finally, human antibody fragments (scFv) against these five antigens were selected using antibody phage display and characterised.
In this work, we identified novel immunogenic proteins of Salmonella Typhimurium and generated antibody fragments against these antigens completely based on phage display. Five immunogenic proteins were validated using a panel of positive and negative sera for prospective applications in diagnostics of Salmonela Typhimurium.
Salmonella spec. is a genus of the Enterobacteriaceae. Two species are in the genus Salmonella: S. bongeri and S. enterica. Salmonella enterica is classified in serogroups and serovars on the basis of their O- and H-antigens (somatic and flagellar antigens) [2,3]. So far, 2800 Salmonella enterica gene families and more than 2500 serovars are known. More than 1500 serovars belong to the subspecies Salmonella enterica subspecies enterica. These pathogens cause foodborne gastrointestinal infections, usually through raw poultry and pork, but it can also be found in non-alcoholic beer or seafood. The subspecies enterica is the cause of 99% of human Salmonella infections. The prevailent serovars are Typhimurium and Enteritidis [4-7]. The most reported phage types for Salmonella Typhimurium are DT193, U302 and DT104. Infections with the latter two phage types increased in 2009 . Human infections with phage type DT104 are particularly critical, because these strains are resistant to most of the commonly used antibiotics . In Europe, Salmonella caused more than 130,000 reported infections in 2008 and 108,614 cases in 2009. In the US more than a million cases are estimated to occur [5,8].
Improved detection o f livestock colonised with S. Typhimurium would be very helpful to prevent foodborne diseases. In particular, infections in swine are difficult to diagnose, because the animals develop either no or only slight symptoms . Only through continuous monitoring of the herds infections of humans can be prevented. Established methods for S. Typhimurium diagnostics are classically time-consuming, using microbiological cultures on different liquid and solid media [10,11], specific fluorescence labeled DNA probes , PCR  or recently, a quantum dot-based bead assay . Currently, high throughput diagnostic of S. Typhimurium is performed by indirect ELISA [9,15,16]. The commercially available ELISA kits e. g. SALMOTYPE®- or Enterisol®-ELISA use a mixture of O-antigens of Salmonella enterica subspecies entirca serovars. They are based on the system established by Nielson et al. . Because of this mixture, cross-reactions occur with other bacteria . In addition, the sensitivity varies between the different ELISA assays . For a sensitive and specific ELISA, immunogenic and species specific proteins are required . The improvement of detection methods, as well as the development of new vaccines would be facilitated by the identification, characterisation and validation of previously unknown immunogenic proteins.
The most common method for the identification of immunogenic proteins is 2D-PAGE of cultured bacterial pathogens and immunoblot using sera from infected patients or animals followed by mass spectrometry or microsequencing [19-24]. However, this method is limited. Differentially expressed proteins, e.g. dependent on pathogen-host interaction, can not be detected. Furthermore, weakly expressed antigens may also not be identified. In these cases, antigen phage display may circumvent these limitations. Our approach for the identification of immunogenic proteins is phage display. Phage display technology was invented by George P. Smith . This methods can be used both for the selection of antibodies [26-29] and for the identification of immunogenic proteins from genomic or cDNA libraries [30-34]. Here, the cloning of randomly fragmented genomic DNA or cDNA into phage display vectors should allow, in theory, the display of all polypeptides encoded by the genome of the donor or all polypeptides encoded by the transcriptome of the donor, respectively [35,36]. In this study, we combined the identification of immunogenic proteins by M13 phage display using genomic libraries from S. Typhimurium with the selection of open reading frames without any subcloning steps (Figure1 left part) in order to improve the library quality [37,38].
Afterwards, the genes corresponding to the identified immunogenic oligopeptides were cloned and produced in E. coli (Figure1 middle part). Using our phage display based pipeline for the generation of human antibodies , we were able to generate human, recombinant antibodies against these antigens (Figure1 right part).
Sonication of Salmonella DNA did not lead to clonable DNA fragments, whereas the sonication of E. coli DNA as a control could be cloned without problems (data not shown). Therefore, genomic DNA was digested with a mixture of the 4 base pair cutters DpnI and AluI and the 6 base pair cutter AfeI to construct the Salmonella Typhimurium genome library (Figure2A). The digested DNA was cloned into pHORF3  resulting in a library with 1.6x106 independent clones. The insert rate and size was analysed by colony PCR (Figure2B), which indicated that more than 90% of the clones contain an insert. The shortest inserts had a size of about 40bp while the longest inserts had a lenght of 2500–3000bp with an average length of about 500–800bp. For the selection of open reading frames, the library was packaged with Hyperphage [37,40,41]. The packaged library was also analysed by colony PCR (Figure2C) resulting in shorter DNA fragments compared to the initial library.
Pooled convalescent serum from pigs infected with Salmonella Typhimurium was used for the panning procedure. In total, two panning rounds were performed and 184 oligopeptide phage clones were analysed by ELISA for binding to serum IgGs. 58 oligopeptide phage clones with different gene fragments were found to bind to the porcine convalescent serum IgGs but not to the control serum IgGs. For 19 gene fragments, no homologous NCBI database hit was found. 31 of the identified proteins showed a higher similarity to other Salmonellaentericia serovars (Table1). One of the selected gene fragments showed the best NCBI database hit to Salmonella Typhimurium, but the gene fragment was not in frame (Table2). Seven oligopeptide phage clones with a good signal to noise ratio (three- to four-fold over background) in the ELISA and the highest homology to genes from Salmonella enterica spp. enterica serovar Typhimurium were selected for further characterisation (Table3). Binding of the oligopeptide phage to convalescent serum IgGs, including the non-Salmonella Typhimurium hits and the out of frame Salmonella Typhimurium hit, was verified in an additional ELISA with piglet serum as negative control (data not shown).
The correspoding complete protein coding sequence of the identifed gene fragments was amplified from genomic DNA (Figure3A) and cloned into pET21A+. When the ORF contained a leader peptide, it was replaced by the pelB leader peptide by cloning into pET21A+pelB. It was not possible to amplify the putative carbohydrate kinase encoding gene from the genomic DNA. The antigens were produced in 500mL scale shake flasks, purified by IMAC and verified by SDS-PAGE (Figure3B). Five of six antigens cloned into pET21A+were produced and purified successfully. It was not possible to produce the hypothetical protein STM14.
The five full size antigens were tested with sera from ten Salmonella Typhimurium positive classified pigs, using the commercial SALMOTYPE Pig Screen ELISA, and sera from four negative classified pigs. In addition, they were tested with the pooled immune positive sera used for the selection procedure and with a piglet serum as negative control (Figure4).
9 out of 10 positive sera bound better to the antigen putative dihydroxyacid dehydratase compared to the negative sera. This means, that 9 of the 10 positive sera had a higher ELISA O.D. value compared to the negative serum with the highest ELISA O.D. value.
7 out of 10 positive sera bound better to the antigen putative electron transfer protein alpha compared to the negative sera. The pooled immune sera (positive control) contained either more or better binders against the antigens than any of the individual positive classified samples.
7 out of 10 positive sera bound better to the antigen 2,4-dieonyl-CoA-reductase compared to the negative sera. Here, the piglet serum revealed a high antigen binding capacity compared to the mixture of positive sera.
8 out of 10 as positive classifed sera bound better to the antigen phage tail-like protein compared to the negative classified sera.
8 out of 10 as positive classified sera bound better to the antigen putative dimethyl sulphoxide reductase compared to the negative classified sera.
Not all individual positive sera bound significantly better than all four individual negative sera. However, in general the positive sera showed better binding to all identified immunogenic proteins compared to the negative sera.
Antibody fragments against all five antigens were selected using the human naive antibody gene library HAL7/8 . Monoclonal binders were identified by antigen ELISA using soluble scFv fragments (data not shown). These binders were sequenced to identify unique binders and analysed using VBASE2 (http://www.vbase2.org) . Human antibodies were successfully generated against all five antigens.
The best binders were recloned into the pOPE101-XP vector , produced in 1.6L scale in the LEX system and IMAC purified (Table4). The yields were between 0.5mg/L and 12mg/L. Afterwards, the purified scFv were analysed by titration ELISA (Figure5). The EC50 values of the scFv (monovalent) are given in Table4. For four scFv it was not possible to determine the EC50, since the maximal binding was not reached in the titrations ELISA.
Binding to linear epitopes was analysed by SDS-PAGE of the antigens, followed by a Western Blot and an immunostain using the purified scFv. All binders to putative dihydroxyacid dehydratase, phage tail-like protein and putative dimethyl sulphoxide reductase bound linear epitopes. Three of the four binders to putative electron transfer protein alpha and the binder against 2,4-dienoyl-CoA-reductase did not bind in the immunoblot (Table4).
Antibody phage display for generation of recombinant antibody fragments [39,44-47] and the identification of immunogenic proteins by phage display [30-32,38,48,49] are established methods. But in this work, for the first time a complete phage display based pipeline from antigen identification to the generation of the corresponding antibody fragments was shown. Oligopeptide phage display technology can expand the identification of immunogenic proteins compared to 2D-PAGE followed by mass spectrometry or microsequencing [32,38,48,50]. The identification of immunogenic proteins via oligopeptide phage display is independent of the natural expression rate of the immunogenic protein, which also allows the identification of low abundant proteins or proteins only produced in host-pathogen interactions. A disadvantage could be that only oligopeptides can be selected which can be secreted by the SEC pathway . Interestingly, when using sonicated S. Typhimurium DNA, the transformation rates were in the range of 102 - 104 clones per transformation. This is very low compared to the transformation rates of 105 for sonicated genomic DNA of Mycoplasma hyopneumoniae or 106 clones for E. coli. Hence, the sonication method appeared to be not applicable for some bacteria species or strains when constructing genomic libraries.
In this work, 58 different oligopeptides were bound by convalescent serum from pigs infected with Salmonella Typhimurium. Interestingly, many of the encoding gene fragments were not in frame with gIII and therefore, in theory, should not result in the production of functional oligopeptide-pIII fusion proteins. However, similar observations, that gene fragments encoding oligo- or polypeptides frequently contain frameshifts, have been described previously for selections by phage display [37,51]. For +1 frameshifts it is reported that oligo- or polypeptides are still displayed on phage particles with the same amino acid sequence as the corresponding constructs without a frameshift. One suggested explanation of this effect was the occurrence of RNA secondary structures. A second explanation could be the selection pressure against oligo- or polypeptides which are toxic for E. coli and thus may lead to a negative selection against these potential toxic proteins .
The most frequently identified oligopeptides did not show the best match with the Salmonella serovar Typhimurium (NCBI taxonomy IDs: 99287, 588858 and 568708), but instead with other Salmonella serovars. These antigens with a higher homology to other Salmonella serovars, could be interesting for further analyses. However in this work, we focused on the seven antigens with the highest homology to Salmonella Tyhphimurium. In contrast to former selections of immunogenic proteins using the pHORF system, where both new and known immunogenic proteins were selected [32,38], these seven antigens have not been described as immunogenic before. So far, five immunogenic proteins of S. Typhimurium were found using 2D-PAGE . Putative dihydroxyacid dehydratase, putative dimethyl sulphoxide reductase and hypothetical protein STM14 of Salmonella Typhimurium have not been described as immunogenic before. The putative electron transfer protein alpha  is located on a pathogenicity island . To date, 2,4-dienoyl-CoA-reductase of S. Typhimurium has not been identified as immunogenic, but interestingly, humans exhibiting anti-mitochondrial autoantibodies (AMA-positive), have also antibodies against the human 2,4-dienoyl-CoA-reductase . For the phage tail-like protein, a bactericidal activity is described for some bacteria, e.g. Pseudomonas. Immunogenic proteins from S. Typhimurium, which are used for diagnostics, are only rarely described in the literature. Described are OmpD  and a preparation of flagelates  for ELISA diagnostics. For nanobead based assays polyclonal antibodies against Salmonella were used whereupon the detailed antigens are unknown . The V genes of the selected scFv against the five immunogenic proteins are mainly derived from the HV families 1 and 3 and from the LV families 1 and 3. Member of these gene families are preferentially selected from naive scFv libraries [60,61]. Only scFv with a lambda VL but no kappa VL were selected. Interestingly, also one VL domain only binder was selected. This is an artefact from library cloning since the insert rate of HAL7 is not 100% . Functional VL domain dAbs have been described before .
The gold standard for diagnostics of Salmonella infections is microbiological culture . Currently, for high throughput detection of S. Typhimurium, ELISA is the ideal method [9,15,16,58]. The commercially available ELISA kits use a mixture of O-antigens (LPS) or total cell lysate of Salmonella enterica subspecies entirca serovars. This mixture of antigens causes, cross-reactions with other bacteria [9,18]. A comparison of four different ELISA detection systems showed “both sample matrices, blood sera and meat juice, are suitable for antibody detection. However, the test sensitivity mainly depends on the respective cut-off used for the specific test” and “our findings indicate that the currently used LPS-ELISA systems have diagnostic uncertainties…” . The use of one or a defined mixture of the selected immunogenic proteins and the corresponding antibody fragments will be useful to establish an ELISA based diagnostic kit with a higher specificity compared to the commercially available diagnostic kits.
Salmonella Typhimurium was cultivated in 2xTY medium  overnight at 34°C and 250rpm. For isolation of genomic DNA, 6x 3ml of the culture were used. The isolation was performed with the Quiaamp DNA Mini Kit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). After purification, the DNA was digested for 35min using three different blunt end-cutting restriction endonucleases (AluI, AfeI, DpnI) (NEB, Frankfurt, Germany). DNA fragments with a size up to 1200bp, were used for cloning into the PmeI-restricted vector pHORF3 . The ligated plasmids were transformed into E. coli Top10 F´ (Invitrogen, Karlsruhe, Germany) by electroporation.
The enrichment of ORFs in the S. Typhimurium genomic library requires the display of the corresponding polypeptides on phage particles for the panning. Therefore, the library was packaged using Hyperphage [40,41] as described previously [37,38].
E. coli clones bearing pHORF3 were analysed by colony PCR using the primers MHLacZPro_f (5' GGCTCGTATGTTGTGTGG 3'), MHgIII_r (5' GGAAAGACGACAAAACTTTAG 3'), and the following protocol: 94°C 1min, 56°C 0.5min, 72°C 1.5min, 25 cycles. The DNA was separated by 1% Agarose gel electrophoresis.
The panning was performed by following the protocol described before  with modifications. Six wells of a MaxiSorp® 96-well microtitre plate (MTP; Nunc, Wiesbaden, Germany) were coated with 150 μL 5μg/mL goat anti-swine IgG in PBS  overnight. The wells were washed with phosphate buffered saline (PBS) supplemented with 0.1% Tween20 (PBST) (Roth, Karlsruhe, Germany). Afterwards, they were blocked with PBST supplemented with 2% (w/v) skim milk powder (2% MPBST) for 1h. In parallel, several wells of a MaxiSorp® plate were coated with 150 μL 1×1011cfu/mL Hyperphage in PBS overnight and blocked with 2% MPBST for 1h. All washing steps were performed three times using PBST buffer and an enzyme-linked immunosorbent assay (ELISA) washer (Tecan Columbus, Crailsheim, Germany). A swine serum mixture (obtained from pigs after infection with S. Typhimurium and field sera) was diluted 1:10 in 2% MPBST and pre-incubated on MaxiSorp® MTP wells coated with Hyperphage for 1h at RT, to remove serum IgG binding to the helperphage. After pre-incubation, the swine serum was incubated in goat anti-swine IgG-coated MTP wells for 2h. After washing, 5×1010cfu polypeptide phage particles of the Hyperphage-packaged Salmonella Typhimurium genomic library were incubated on the captured swine IgGs for 2h. For the following panning rounds, 100 μL of amplified phage of the previous panning round were used. The non-binding polypeptide phage particles were removed by ten stringent washing steps. In the second and third panning round, the number of washing steps was increased to 20 and 30, respectively. Elution of bound phage particles was performed using 200 μL of 10μg/ml trypsin (10μg/mL trypsin in PBS) for 30min at 37°C. Ten microlitres of the eluted phage solution were used for titration. Twenty millilitres of the E. coli TOP10 F´ cells were grown to an OD600 of 0.4 - 0.5 which were then infected with the remaining 190 μL of the eluted phage solution and incubated for 30min at 37°C. Afterwards, the cells were harvested by centrifugation for 10min at 3.220×g. The bacterial pellet was resuspended in 250 μL 2×TY medium (1.6% [w/v] tryptone, 1% [w/v] yeast, 0.5%[w/v] NaCl) containing 100mM glucose and 100μg/mL ampicillin (2×TY-GA), plated onto 15cm 2×TY-GA agar plates and incubated overnight at 37°C. Grown colonies were harvested in 5mL 2×TY-GA medium using a Drigalsky spatula. Fifty millilitres of 2×TY-GA medium were inoculated with 200 μL bacteria culture and grown to an OD600 of 0.4-0.5 at 37°C and 250rpm in a shaking incubator. Five millilitres of bacterial culture corresponding to about ~2.5×109 cells were infected with 5×1010cfu Hyperphage, incubated at 37°C for 30min without shaking and another 30min with shaking at 250rpm. The infected cells were harvested by centrifugation for 10min at 3.220xg. The pellet was resuspended in 30mL 2xTY medium containing 100μg/mL ampicillin and 50μg/mL kanamycin (2×TY-AK). Phage particles were produced at 30°C and 250rpm overnight. On the following day, the supernatant containing phage particles was collected.
For phage production, polypropylene 96-well U bottom plates (Greiner bio-one, Frickenhausen, Germany) containing 175 μL 2xTY-GA per well were inoculated with single E. coli colonies from the phage titration plates of the third panning round and incubated at 37°C with constant shaking at 850rpm (thermo shaker PST60-HL4, lab4you, Berlin, Germany) overnight. A new plate with 165 μL 2xTY-GA per well was inoculated with 10 μL of the overnight cultures and incubated at 37°C and 850rpm for 2h. Afterwards, the bacteria were infected with 5x109 cfu Hyperphage/well and incubated at 37°C without shaking for 30min, followed by shaking at 850rpm for 30min. The MTP plate was centrifuged at 3,220xg for 10min and the supernatants were discarded. Afterwards, the bacterial pellets were resuspended in 175 μL/well 2xTY containing 100mg/mL ampicillin and 30μg/mL kanamycin (2xTY-AK) and incubated at 30°C at 850rpm overnight for phage production. The bacteria were pelleted again and the supernatants were transferred to a new plate. The phage were precipitated with 1/5 volume 20% PEG/2.5M NaCl solution at 4°C for 1h and centrifuged at 3,220xg for 1h. The phage pellet was dissolved in 150 μL PBS and residual bacteria were removed by another centrifugation at 3,220xg for 5min. The phage containing supernatants were stored at 4°C or directly used for ELISA.
For phage ELISA, the produced polypeptide phage particles were captured. Here, 100 μL of 250ng/mL mouse anti-M13 (B62-FE2, Progen, Freiburg, Germany) in PBS were coated at 4°C overnight. After coating the wells were blocked with 2% MPBST. Between each incubation step the wells were washed three-times with PBST using an ELISA washer. 150 μL of the monoclonal phage production were incubated for 2h. The pig convalescent serum was diluted 1:200 in 2% MPBST supplemented with 1/10 volume E. coli cell lysate and 1x1010 cfu Hyperphage/mL, added to the captured phage particles and incubated for 2h. The bound pig IgGs were detected with goat anti-swine IgG conjugated with horseradish peroxidase (HRP) (1:1000) for 1.5h and visualised with TMB (3,3´,5,5´-tetramethylbenzidine) substrate. The staining reaction was stopped by adding 100 μL 1N sulphuric acid. The absorbances at 450nm and scattered light at 620nm were measured and the 620nm value was subtracted using a SUNRISE microtiter plate reader (Tecan, Crailsheim, Germany).
The corresponding proteins of the identified immunogenic polypeptides were amplified by PCR using the genomic DNA and the following oligonucleotide primers: TM208-forward (5´ AAGGAGATATACATATGAGCCAAAAATGTCAACATGCT 3´), TM208-reverse (5´ CTCGAGTGCGGCCGCTTTAAGCCAGGCTCCGGCCATTAA 3`) for putative dihydroxyaciddehydratase, TM209-forward (5´ AAGGAGATATACATATGGCTTCTTTAGTTATTGCTGAACAT 3´), TM209-reverse (5´ CTCGAGTGCGGCCGCTAATTTATCGATAAGTTCAGGTAC 3´) for putative electron transfer protein alpha, TM210-forward (5´ AAGGAGATATACATATGAGCTACCCGTCGCTGTTCGCCCCG 3´), TM210-reverse (5´ CTCGAGTGCGGCCGCAATCTCCAGTGCCAGTCGGGTGCC 3´) for 2,4-dieonyl-CoAreductase, TM211-forward (5´ AAGGAGATATACATATGAATAGTCTGTTGCCGCCGGGTTCG3´), TM211-reverse (5´ CTCGAGTGCGGCCGCTGGATTCACTCTCATTGTGTCAAT 3´) for phage tail-like protein, TM212-forward (5´ CAGCCGGCCATGGCTTCAATGAATAAAGCAGTCAGTAGTGAG 3´), TM212-reverse (5´ CTCGAGTGCGGCCGCACGTGCCGGGCGGTATTCGCGCCA 3´) for putative carbohydratekinase, TM213-forward (5´ CAGCCGGCCATGGCTGCGGTGCAGCAGGCTATGCGCAACGAA 3´), TM213-reverse (5´ TGCGGCCGCAAGCTTGATTTTTTCGATTTCCACCAGATTTGT 3´) for putative dimethyl sulphoxide reductase chain A1, and TM214-forward (5´ AAGGAGATATACATATGATGTCAGCATGTTTTTTTGGCCGA 3´), TM214-reverse (5´ CTCGAGTGCGGCCGCATCAATTATTTTGGTGAGTGTTTG 3´) for hypothetical protein STM14. The PCR was performed using the following protocol: 98°C 30sec, 98°C 10sec, 60°C 20sec, 72°C 120sec, 24 cycles. The DNA was separated by electrophoresis in a 1% agarose gel. The bands of interest were cut out from the gel, the DNA was eluted and used for cloning into pET21A+or pET21A+pelB. After ligation the plasmids were transformed into E. coli BLR-DE3. Positive clones were identified by using colony PCR using the oligonucleotide primers MHpET21_f1 (5´ GAGCGGATAACAATTCCCC 3´) and MHpET21_r1 (5´ GCAGCCAACTCAGCTTCC 3´).
Five hundred mL 2xTY-GA medium were inoculated with 5mL overnight culture and cultivated to an O.D.600 of 1.0 at 37°C and 250rpm. The expression was induced with 1mM IPTG (final concentration) overnight. Cells were harvested by centrifugation at 7.500 xg for 15min. Lysis was performed with 1mg/mL lysozyme and 5μg/mL DNAseI in 15mL His-tag binding buffer pH7.4 (20mM Na2HPO4, 0.5M NaCl, 10mM Imidazol). For Isolation of inclusion bodies 8M Urea was added. The purification was performed under denaturing conditions with FastFlow Sepharose (GE Healthcare) loaded with nickel. The Sepharose was washed with 10mM, 30mM and 60mM imidazole (20mM Na2HPO4, 0.5M NaCl, 10, 30 or 60mM Imidazol). For elution, 5mL 100mM EDTA in PBS supplemented with 8M urea were used.
Antigens were analysed by 12% SDS-PAGE using a Protean II Minigel system (BioRad Inc, München, Germany) according to . Protein gels were stained with coomassie blue.
One μg of antigen was coated to 96 well microtitre plates (MaxiSorp, Nunc) in 50mM NaHCO3 pH 9.6 overnight at 4°C. After coating, the wells were washed three times with PBST and blocked with 2% MPBST for 1.5h at RT, followed by three washing steps with PBST. For serum ELISA, sera were diluted 1:200 in 100 μL 2%MPBST and incubated in the antigen coated plates for 1.5h at RT, followed by three PBST washing cycles. Bound pig IgGs were detected with goat anti-swine IgG HRP conjugate (1:10,000) (Dianova, Hamburg, Germany). The visualisation was performed with TMB (3,3´,5,5´-tetramethylbenzidine) as a substrate and the staining reaction was stopped by adding 100μl 1N sulphuric acid. Absorbance at 450nm was measured by using a SUNRISE™ microtitre plate reader (Tecan, Crailsheim, Germany).
The selection of recombinant antibodies was performed according to  with modificiations. In short, pannings were performed in 96 well microtitre plates (MaxiSorp, Nunc, Wiesbaden, Germany). One μg of antigen was coated in PBS pH 7.4  overnight at 4°C. The antigen-coated wells and wells for the preincubation of the library were blocked with 2% MPBST. In each case 2.5x1011 phage particles of the human naive antibody gene libraries HAL7 and HAL8  were diluted in PBST with 1% skim milk and 1% bovine serum albumin (BSA) and preincubated for 1h. The supernatant, containing the depleted library, was incubated in the antigen-coated wells at RT for 2h followed by 10 washing steps with PBST. Afterwards, bound scFv phage particles were eluted with 200 μL trypsin solution (10μg/mL trypsin in PBS) at 37°C for 30min. The supernatant containing eluted scFv phage particles was transferred into a new tube. Ten μL of eluted scFv phage were used for titration as described before . Twenty mL E. coli XL1-Blue MRF' (Agilent, Böblingen, Germany) culture in the logarithmic growth phase (O.D.600=0.4 - 0.5) were infected with the remaining scFv-phage at 37°C for 30min without shaking. The infected cells were harvested by centrifugation for 10min at 3220xg and the pellet was resuspended in 250 μL 2xTY medium  supplemented with 100mM glucose and 100μg/mL ampicillin (2xTY-GA), plated on a 15cm 2xTY agar plate supplemented with 100mM glucose and 100μg/mL ampicillin and incubated overnight at 37°C. Grown colonies were harvested with 5mL 2xTY-GA. Thirty mL of 2xTY-GA were inoculated with 100 μL of the harvested colony suspension and grown to an O.D.600 of 0.4 to 0.5 at 37°C and 250rpm. Five mL bacteria suspension (~2,5x109 bacteria) were infected with 5x1010 helperphage M13K07 (Stratagene), incubated at 37°C for 30min without shaking, followed by 30min at 250rpm. Infected cells were harvested by centrifugation for 10min at 3220 xg and the pellet was resuspended in 30mL 2xTY supplemented with 100μg/mL ampicillin and 50μg/mL kanamycin (2xTY-AK). Antibody phage were produced at 30°C and 250rpm for 16h. Cells were harvested by centrifugation for 10min at 3220xg. The supernatant containing the antibody phage (~1x1012 cfu/mL) were directly used for the next panning round or stored at 4°C for a few days.
The identification of monoclonal binders was performed as described before . In brief, 96-well MTPs with polypropylene (PP) wells (U96 PP 0.5mL, Greiner, Frickenhausen, Germany) containing 150 μL phosphate buffered 2xTY-GA  (2xTY-GA supplemented with 10% (v/v) potassium phosphate buffer (0.17M KH2PO4, 0.72M K2HPO4)) were inoculated with colonies from the titration plate of the third panning round. MTPs were incubated overnight at 37°C at 1000rpm in a MTP shaker (Thermoshaker PST-60HL-4, Lab4You, Berlin, Germany). A volume of 180 μL phosphate-buffered 2xTY-GA in PP-MTP well was inoculated with 10 μL of the overnight culture and grown at 37°C and 800rpm for 2h. Bacteria were harvested by centrifugation for 10min at 3220xg and 180 μL supernatant were removed. The pellets were resuspended in 180 μL buffered 2xTY supplemented with 100 μg/mL ampicillin, 100mM sucrose an 50 μM isopropyl-beta D thiogalacto pyranoside (IPTG) and incubated at 30°C and 800rpm overnight. Bacteria were pelleted by centrifugation for 10min at 3,220 xg and 4°C. The scFv-containing supernatant was transferred to a new PP-MTP and stored at 4°C before analysis.
Antigen coating was performed as described above (Enzyme linked immunosorbent assay (ELISA) for verification of immunogenic proteins). For identification of binders, supernatants containing monoclonal scFv were incubated in the antigen coated plates for 1.5h at RT followed by three PBST washing cycles. Bound scFv were detected using murine mAb 9E10 which recognises the C-terminal c-myc tag and a goat anti-mouse serum conjugated with horseradish peroxidase (HRP) (Sigma; 1:10,000).
The detection was performed as described above.
The large-scale expression system (LEX) (Harbinger Biotech, Toronto, Canada) was used for production of scFv. E. coli (XL1-Blue-MRF') was cultivated in 2L glass bottles up to a cultivation volume of 1.5L. To obtain sufficient oxygenation and mixing of the culture, the bottles were connected to an air manifold, which allows a general air flow rate of 4–6L/min. A thermostat-controlled water bath was used for regulating the temperature of the cultivation. 50mlTB supplemented with 100μg/mL ampicillin were inoculated with a glycerol stock of each scFv clone and the culture was grown over night at 37°C. Glass bottles with 1.5 L TB supplemented with 100μg/mL ampicillin and 500μL antifoam 204 (Sigma, München, Germany) were inoculated with the overnight culture. The O.D.600 was adjusted to 0.1 and incubated at 37°C until an O.D.600 of 1.5 to 5 was reached. The temperature of the water bath was then reduced to 25°C. After 1h, scFv expression was induced by addition of 50μM IPTG. The cultivation was continued for 3h resulting in a final O.D.600 of 2 to 7 depending on the antibody clone. E. coli cells were harvested by centrifugation at 4,400xg (Sorvall Zentrifuge RC6 Plus, Rotor F9S-4x1000Y) for 10min at 4°C. The pellet was resuspended in 60mL ice-cold PE-buffer pH 8 (20% (w/v) sucrose, 50mM Tris, 1mM EDTA) and was incubated on ice for 20min while shaking. Afterwards the sample was centrifuged at 20,000xg and 4°C for 30min (Sorvall Zentrifuge RC6 Plus, Rotor F12-6x500y). The supernatant (periplasmatic preparation) was filled into a fresh glass bottle and kept on ice. The pellet was re-suspended in 60mL ice-cold OS-buffer (5mM MgSO4 in dH2O) and was incubated on ice while shaking. After 20min the preparation was centrifuged at 20,000xg and 4°C for 30min (Sorvall Zentrifuge RC6 Plus, Rotor F12-6x500y). The supernatant (osmotic shock fraction) was combined with the periplasmatic preparation and was used for protein purification.
Antibody fragments were purified by affinity chromatography using IMAC. Chromatography using Profinia (BioRad) and 1mL FF-crude column (GE Healthcare, München, Germany) was performed according to the manufacturer's instruction. The protein solution was adjusted to 10mM imidazol containing buffer (20mM Na2HPO4, 500mM NaCl, 10mM imidazol) for loading. The column was washed one time with 10mM imidazol buffer (20mM Na2HPO4, 500mM NaCl, 10mM imidazol). Five hundred mM imidazol was used for elution, followed by desalting and storage in PBS.
For the scFv titration ELISA the antigen was coated as described above (Enzyme linked immunosorbent assay (ELISA) for verification of immunogenic proteins). The ELISA was performed as described above (Identification of monoclonal scFv using ELISA) with one modification: a dilution series of IMAC purified scFv was used instead of the scFv supernatant. The EC50 values (antibody concentration at the half maximal binding) are deduces from this titration ELISA.
Purified immunogenic proteins were separated by 12% SDS-PAGE. Western Blotting on PVDF (Polyvinylidenfluorid) membranes of gels was performed using the Mini Trans-Blot® system (BioRad). The membrane was blocked with 2% (w/v) skimmed milk powder in PBST over night.
The antigens were detected with 20μg/mL scFv for 1h at RT. The scFv myc-tag was detected with mouse anti myc-tag (9E10, Sigma, Taufkirchen, Germany) for 1h, followed by goat anti-mouse (Fc specific) (Sigma) conjugated with alkaline phosphatase (1:20,000) for 1h. The visualisation was performed by addition of BCIP (5-bromo-4-chloro-3-indolyl phosphate) and NBT (nitroblue tetrazolium).
A “pipeline” from antigen identification to the generation of recombinant antibodies using phage diplay was shown. Here, novel immunogenic proteins of Salmonella Typhimurium were identified using phage display and validated using a panel of positive and negative sera. Afterwards, recombinant human antibody fragments were generated against these marker proteins.
The authors declare that they have no competing interests.
TM and SM performed most of the experiments and helped to draft the manuscript. TS and AF performed some of the experiments and participated in the design and coordination of the study. JSS, SD, GFG and KSM particpated in the design and coordination of the study and helped to write the paper. MH and JSS wrote the grant application. MH particpated in the design and coordination of the study wrote the publication. All authors read and approved the final manuscript.
This project was supported by the BMBF (BioRegioN) and EFRE. Our special thanks go to Ronny Fischer from our EC office who helped with bureaucracy of EFRE/NBANK. We also thank to David Havlik, Jonas Zantow, Alex Pytka and David Becker for careful corrections on the manuscript.