OUP user menu

Characterization of the cattle serum antibody responses against TEM β-lactamase and the nonimmunogenic Escherichia coli heat-stable enterotoxin (STal)

Astrid Zervosen, Claude Saegerman, Ingrid Antoniotti, Béatrice Robert, Nadia Ruth, Alfred Collard, Frédéric Schynts, Moreno Galleni, Paola Sandra Mercuri
DOI: http://dx.doi.org/10.1111/j.1574-695X.2008.00482.x 319-329 First published online: 1 December 2008

Abstract

In order to test the use of a subunit recombinant vaccine for its capacity to induce antibodies against the nonimmunogenic heat-stable enterotoxin STa from Escherichia coli and the TEM-1 β-lactamase, cattle were immunized with a hybrid protein created by insertion of the STa sequence at position 197 of the TEM-1 β-lactamase. Specific anti-STa IgG and IgG1 antibodies were detected at low levels, while no IgG2 antibodies were detected. In contrast, high levels of the different anti-TEM IgG subtypes were detected in cattle sera. In addition, β-lactamase activity was inhibited by the sera. The presence of antibodies against STa and TEM-1 β-lactamase was assessed in sera from 366 cattle taken from the field. No significant level of IgGs against the toxin or the TEM-1 was detected. A comparison of the antibody level between the immunized and the nonimmunized animals clearly demonstrated that STa was not able to induce a significant level of antibodies in the vaccinated animals. In contrast, a strong antibody response against TEM-1 β-lactamase was demonstrated.

Keywords
  • subunit recombinant vaccine
  • β-lactamase
  • thermostable enterotoxin

Introduction

Enterotoxin-producing Escherichia coli (ETEC) are the major cause of diarrheal disease in humans and in domestic animals (Nataro & Kaper, 1998). In newborn calves, an ETEC infection causes profuse fluid diarrhea, dehydration and even death (Holland, 1990). Hence, it is responsible for significant economic losses. The fluid secretions induced by ETEC strains have been shown to be mediated by two families of enterotoxin: the heat-labile enterotoxin (LT) and the heat-stable enterotoxin (ST) (Burgess et al., 1978; Greenberg & Guerrant, 1980). Two types of ST (STa and STb), which differ functionally and structurally, have been described (Kennedy et al., 1984; Lockwood & Robertson, 1984). DNA hybridization tests have indicated that STa can be divided into STaI and STaII. STaI and STaII have been purified and sequenced from human, bovine and porcine isolates (Chan & Giannella, 1981; Lallier et al., 1982; Dreyfus et al., 1983). The STaI and STaII sequences consist of 18 and 19 amino acids, respectively (Moseley et al., 1983). Twelve amino acids are preserved; six of these are cysteines, which form three disulfide bridges that are necessary for the toxicity of the peptides (Gariepy et al., 1978). Bovine ETEC produces the STaI enterotoxin (Blanco et al., 1988). STa is poorly antigenic. Several studies have been performed to develop its immunogenicity when coupled to an appropriate carrier such as bovine serum albumin (Lockwood & Robertson, 1984), B-subunits of cholera toxin (Sanchez et al., 1988) or to green fluorescent protein (GFP) (Wu & Chung, 2007). In the last case, Wu et al. created a fusion protein [GFP : STLT(B)] between the GFP, the thermolabile toxin (LT) and the thermostable toxin (ST). The GFP : STLT(B) product confirmed its ganglioside-binding ability. Oral inoculation of the Lactococcus reuteri-producing culture GFP : STLT(B) in mice elicited significant serum IgG and mucosal IgA antibodies against the STLT(B) antigen. These immunized mice were subsequently challenged with ETEC and showed full protection against the fluid influx response in the gut. In this work, we decided to insert the STaI nucleotide sequence into the amp gene encoding for the TEM-1 β-lactamase. In a previous study, we showed that the immunization of mice with plasmid DNA encoding hybrid TEM-STa protein can be used as a priming stimulus to provide T cell help in the antibody response against the nonimmunogenic STa peptide (Ruth et al., 2005). Based on these data, we decided to evaluate the possibility to use the hybrid protein to elicit the production of anti-STa and anti-β-lactamase antibodies in cattles. TEM-1 is a 263-amino acid protein that hydrolyzes the β-lactam ring (Jelsch et al., 1993; Frère, 1995). We chose TEM-1 as the carrier candidate because this enzyme is stable, highly soluble and can be produced in large quantities. In addition, TEM-1 is well characterized structurally and biochemically. Paus & Winter (2006) showed that the region 90–140, which corresponds to a flexible loop region of TEM β-lactamase, exhibited a high antigenicity. The TEM-1 permissive insertion sites had been characterized previously by Hallet (1997). In our study, the STaI was inserted into position 197 of the amino acid sequence of TEM-1. Large quantities of the hybrid protein TEM197STa were produced and were purified to homogeneity. The immune capacity of TEM197STa was tested in BALB/c mice and in cattle by vaccination of 20 cows in the presence of different adjuvants. Cattle antibody responses were measured using enzyme-linked immunosorbent assay (ELISA). Finally, the presence of TEM-1 and STa antibodies was investigated in sera collected from 366 cattle on different farms in Belgium.

Materials and methods

Bacterial strains, vectors and media

Luria–Bertani (LB) and Super-Broth (SB) media were prepared as described previously (Sambrook et al., 1989). Escherichia coli DH5α cells were used as the host for the construction of the recombinant proteins and E. coli JM109 cells for the production of proteins. Plasmid pFH197H, coding for TEM-1 with a unique KpnI insertion site, was kindly provided by Dr Hallet (Hallet et al., 1997). Plasmid pTAC11 was a gift from Dr Ptashne (Amann et al., 1983). Plasmid pTACTEM197H was generated by recombination between pTAC11 and pFH197H. The hybrid protein TEM197STa was produced from E. coli JM109 transformed with the pFHTEM197STa (Quinting & Ruth, pers. commun.).

Chromatography materials and antibiotics

CHT Ceramic Hydroxyapatite Type II was purchased from Bio-Rad Laboratories (B.V. KV Veendaal, NL). Q-Sepharose HP, a PD10 desalting column, Superdex G200 and Sephacryl-100 were from Amersham Bioscience (Little Chalfont, Buckinghamshire, UK). Tetracycline was supplied by Sigma (St. Louis, MO). Nitrocefin was purchased from Unipath Oxoid (Basingstoke, UK).

Production and purification of the TEM197H antigen

pTACTEM197H was introduced into E. coli JM109 competent cells. Colonies were selected from LB agar plates containing 12.5 µg mL−1 of tetracycline. Precultures were performed by inoculation of LB culture medium with colonies of a freshly transformed strain in the presence of 12.5 µg mL−1 tetracycline. They were grown for 16 h under agitation at 37 °C. The culture medium SB with 12.5 µg mL−1 tetracycline was inoculated with 5% preculture and was incubated for 24 h at 18 °C. Cells were harvested by centrifugation at 3500 g for 15 min. The protein was extracted from the cells by cold osmotic shock. Cells from culture medium (1 L) were suspended in 200 mL of 30 mM Tris-HCl, pH 8, 20% sucrose. After the addition of 400 µL of 0.5 M EDTA, the cell suspension was incubated for 10 min at room temperature under gentle agitation. After centrifugation at 16 000 g for 15 min at 4 °C, the sucrose solution was drained carefully and the cells were resuspended in 200 mL of cold 5 mM MgSO4. The cells remained on ice for 10 min and were centrifuged at 16 000 g for 15 min at 4 °C. The protein solution was dialyzed overnight against 10 mM 3-morpholino propane-1-sulfonic acid (MOPS) pH 6.5 (buffer A). The solution was loaded onto a Q-sepharose HP 26/10 column (volume: 60 mL) pre-equilibrated in buffer A. The protein was eluted by a linear salt gradient (0–0.5 M NaCl, 900 mL, flow: 6 mL min−1). The active fractions (6 mL) were loaded onto a Sephacryl 100 (volume: 100 mL) pre-equilibrated with phosphate-buffered saline (PBS) pH 7.2. The active fractions were concentrated to 1 mg mL−1 and the purity was verified using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE).

Production of the TEM197STa antigen

TEM197STa plasmid was introduced into E. coli JM109 competent cells. Colonies were isolated from LB agar plates containing 12.5 µg mL−1 of tetracycline. Precultures were performed by inoculation of LB culture medium with a colony of a freshly transformed strain in the presence of 12.5 µg mL−1 tetracycline. They were grown for 16 h under agitation at 37 °C. Three liters of SB medium containing 12.5 µg mL−1 of tetracycline was inoculated with 5% preculture and was incubated for 24 h at 18 °C. Cells were harvested after the centrifugation of the culture at 3500 g for 15 min. The protein was extracted from the cells by cold osmotic shock as described above.

Purification of the TEM197STa antigen

Periplasmic fractions were dialyzed overnight against 20 mM Tris-HCl, pH 7.5, and the solution was loaded onto a Q-sepharose HP XK50 (volume: 170 mL) pre-equilibrated with 20 mM Tris-HCl, pH 7.5. The protein was eluted by a linear salt gradient (0–0.3 M NaCl, 4 L, flow: 10 mL min−1). The active fractions were concentrated up to 30 mL and were dialyzed overnight against sodium phosphate 6 mM pH 7.2. The protein was loaded onto a hydroxyapatite column (volume: 100 mL) pre-equilibrated with the same buffer. The protein TEM197STa passed through the column but the contaminants were fixed to the matrix. The active fractions were concentrated to 1 mg mL−1 and the purity was verified using SDS-PAGE. The protein was stored at −20 °C.

Determination of the N-terminal sequence and the molecular mass of TEM197STa

The N-terminal sequence was determined with the help of a gas-phase sequencer (Prosite 492 protein sequencer; Applied Biosystem, Foster City). The Mr of the proteins was estimated using an electrospray mass spectrometer (VG Bio-Q) upgraded with a Platform source (Micromas, Altrincham, UK). The samples (100 pmol) were suspended in 0.05% formic acid–50% acetonitrile in water and were injected into the source of the mass spectrometer using a syringe pump (Harvard Instruments, South Natick, MA) at a flow rate of 6 µL min−1. The capillary was held at 2.7 K, and the cone voltage was set at 40 V. Fifteen scans covering 600–1500 amu were accumulated for 135 s and were processed using the masslynx software delivered with the instrument. Calibration was performed with horse heart myoglobin.

Determination of protein concentration

The protein concentration was determined as described by Bradford (1976). The protein purity was controlled using SDS-PAGE and was calculated with the program imagej1 31v (http://rsb.info.nih.gov/ij/Java1.3.1_03).

Enzymatic activity

The hydrolysis of the substrate nitrocefin (Δɛ482=+15 000 M−1 cm−1) was followed by monitoring the change in absorbance using an Uvikon 940 spectrophotometer equipped with thermostatically controlled cells at 482 nm. Initial rates were measured at 30 °C in 50 mM sodium phosphate buffer, pH 7.

Experimental design for cattle and mice

Twenty cows more than 4 months old were selected based on a lack of anti-TEM-1 and anti-STa antibodies in their serum (OD<0.2 with a dilution 1/50). They were immunized with the TEM197STa hybrid protein as follows: five groups of four animals were created (groups 1–5). Animals from groups 1 to 4 were injected by subcutaneous route of immunization with 1.8 mg (450 µg per injection) of hybrid protein TEM197STa in the presence of different adjuvants. In group 1, the adjuvant used was the mineral Montanide ISA 70 (Seppic, Paris, France), with an adjuvant/antigen ratio of 70/50. In group 2, the adjuvant used was Montanide ISA 206 (Seppic), with an adjuvant/antigen ratio of 50/50, and in group 3, the adjuvant used was Montanide IMS1313 (Seppic), with an adjuvant/antigen ratio of 50/50. In group 4, cows were injected with the same quantity of antigen in PBS. In the control group (group 5), each cow was injected with one adjuvant or PBS (Table 1). Antigens were injected four times in the neck on days 0, 28, 56 and 84. Blood samples were collected on days 0, 28, 56, 84, 112 and 140. The sera were collected following standard procedures and were stored at −20 °C until use. This experimental procedure was approved by the Ethical Committee of the CER Group of Marloie, Belgium. Three groups of four BALB/c mice (Charles River, France) were immunized with 50 µg per injection of TEM197STa hybrid protein in the presence of the different adjuvants ISA 70, ISA 206 and IMS 1313. The antigens were injected subcutaneously four times at 3-week interval. Sera were collected on days 0, 35, 56, 67 and 87.

View this table:
Table 1

In vivo experimental design used in this study

N1ImmunogenAdjuvant
Group 14TEM 197STaISA70 (70/50)
Group 24TEM 197STaISA206 (50/50)
Group 34TEM 197STaISA1313 (50/50)
Group 44TEM 197STaPBS
Group 5a1PBSISA70 (70/50)
Group 5b1PBSISA206 (50/50)
Group 5c1PBSISA1313 (50/50)
Group 5d1PBSPBS
  • N1 represents the number of cattle in the group.

Antibody measurement using ELISA

TEM-1- and STa-specific antibodies were measured using ELISA (Crowther, 2000). Ninety-six-well microtiter plates (Greiner Bio-One N.V./S.A., Wemmel, Belgium) were coated with 250 ng/50 µL of β-lactamase (TEM197H) per well for specific measurement of antibodies directed against carrier TEM-1 or with 400 ng/50 µL of synthetic STa (Eurogentec, Liège, Belgium) per well in 0.2 M carbonate buffer pH 9.6. The plates were incubated overnight at 4 °C and were washed 1 × with PBS. Subsequently, 100 µL of saturation buffer casein hydrolysate (CERgroupe, Marloie, Belgium) 36.5 g L−1 in PBS 0.05% Tween 20-pH 7.2–7.4 was added to each well and the plates were incubated at 37 °C for 1 h.

After washing with 1 × PBS buffer, 50 µL of the diluted sera (1 : 50 dilution) in saturation buffer were added per well and incubated for 1 h at 37 °C. After intensive washing (5 × with PBS 0.1% Tween 20/1 × PBS) the bound antibodies were detected by a 1-h incubation at 37 °C with polyclonal antibody sheep anti-bovine IgG (1 : 2000 dilution), IgG1 (1 : 10 000 dilution for TEM and 1 : 2000 for STa) and IgG2 (1 : 2000 dilution) affinity purified conjugated to horseradish peroxidase (Serotec Ltd, Oxford, UK). After incubation at 37 °C, the plates were washed 5 × with PBS 0.1% Tween/1 × PBS. The reaction was developed using tetramethylbenzidine (Sigma Chemical Co.) 4 mg mL−1 in 0.1% (v/v) HCl diluted, 1/20 in 50 mM citric acid, 0.05% H2O2, pH 4.25. The enzyme reaction was stopped by addition of 100 µL of 1 M H3PO4 and the absorbance was read at 450 nm.

Titers were determined as the highest serum dilution giving an absorbance higher than the mean absorbance value of preimmune sera (day 0) +3 SDs.

Sero-prevalence of anti-STa and TEM-1 antibodies in the Belgian field cattle population

Sera from a total of 366 cattle were collected randomly in different locations in Belgium. The investigation of the presence of IgG antibodies against STa and TEM-1 was carried out by ELISA as described above. The negative control group included sera (taken on experimental day 0) from cattle included in the in vivo experimental design described above and sera taken from cattle of control group 5 (which did not receive the hybrid protein). The number of sera constituting this group was 39. For a positive control group, sera from experimentally immunized animals were also taken (75 sera from cattle of groups 1, 2, 3 and 4 taken on D28, D56, D84, D112 and D140).

Statistical analysis

The mean of the highest titer serum dilutions for groups 1, 2 and 3 with unequal variance was compared with Welch's test (Dagnelie, 1998). The limit of statistical significance of the conducted tests was defined as P≤0.05.

Seroneutralization assays

The experiments were performed in PBS (50 mM sodium phosphate, 150 mM NaCl, pH 7.4).

Identical quantities (2 ng) of TEM-1 were preincubated or not with 500 µL of 10-fold dilutions of preimmune or 10–105-fold anti-TEM sera for 1 h at room temperature. Then, 10 µL of 5 mM nitrocefin was added. Its hydrolysis was recorded at 482 nm in order to determine the residual β-lactamase activity.

Results and discussion

Antigen production and purification

From 1 L culture media, 12.6 mg TEM197H were purified to a purity of 95% and a purification yield of 57%. The major loss of the enzyme occurred during the first step. We noted that a significant amount of the enzyme did not interact with the Q-sepharose. Thirty milligrams of TEM197STa were produced. After purification, 5.4 mg of TEM197STa with a purity of 90% were recovered. As already observed for TEM197H, the protein was lost during the first purification step.

This phenomenon could be linked to a high ionic strength of the protein solution. This can be avoided either by diluting the solution with a low ionic strength buffer or by increasing the number of occasions of dialysis.

Characterization of the proteins

The purified proteins TEM197H and TEM197STa were characterized using MS. For TEM197H, the measurement gave a molar mass of 29 372±0.8 g mol−1, a result that was in good agreement with the theoretical mass (29 372 g mol−1). In the case of TEM197STa, the measurement gave a molar mass of 31 582±1.2 g mol−1 (also in agreement with the theoretical mass of 31 580 g mol−1).

The N-terminal sequences correct for both proteins were NH2-HPETLV. No ragged N-terminus sequence was present. This result indicates that the presence of an additional domain — such as the STa sequence — yielded a hybrid protein, which was produced and processed as the parental TEM-1 enzyme. Finally, the integrity of the hybrid proteins was not affected by proteolysis. This observation indicates that the additional domain did not protrude strongly onto the β-lactamase scaffold and that it was not totally exposed to the solvent or to proteases either present in the bacteria or produced in the culture media. In addition, the presence of five amino acids of the STa sequence did not modify strongly the enzymatic activity of the β-lactamase toward nitrocefin, a good substrate of the native TEM-1. The steady-state kinetic parameters for TEM-1 (kcat=1410 s−1, Km=27 µM, kcat/Km=52 µM−1 s−1) and TEM197H (kcat=1450 s−1, Km=27 µM and kcat/Km=53 µM−1 s−1) were similar. The insertion of the STa sequence in position 197 of TEM-1 yielded an enzyme that possessed good but reduced enzymatic activity in comparison with TEM-1 (kcat=480 s−1, Km=47 µM and kcat/Km=10 µM−1 s−1). Although the Km value increased only twofold, kcat was mainly affected by the insertion of STa. The moderate modification of kcat and Km can be explained by the fact that the insertion site is located at 30 Å and on the opposite site of the active site.

Antibody responses and isotypes

ELISA was used to measure the levels of antibodies against the carrier protein TEM-1 and the toxin STa in the blood serum of each animal. One cow in group 4 died and was therefore excluded from the analysis. The preimmune sera (day 0) were used as a negative control. In each group, IgG, IgG1 and IgG2 levels were determined from serum diluted 50-fold. The immunization of cattle with TEM197STa induced a low level of the anti-STa IgG (Fig. 1a–e). Animals in group 1 and 2 were immunized with TEM197STa in the presence of adjuvants Montanide ISA70 and ISA206, respectively. They showed better responses than the cattle in group 4, who were immunized with the nonadjuvant protein. Nevertheless, in the first two groups of cattle, large discrepancies were observed in individual anti-STa response (especially on day 84 in group 1). We also observed from the cattle in group 3 that adjuvant IMS1313 did not induce an increase in IgG production. In summary, our results showed that TEM197STa is not the perfect antigen for inducing the production of IgG directed against the STa in different combinations of the antigens/adjuvants that we used in this study, even though some encouraging individual results were observed in some of the cattle in groups 1 and 2. In the case of the induction of anti-TEM IgG, immunization of the adjuvanted-hybrid protein (Fig. 1f–h) allowed the production of antibodies and this production was dependent on the adjuvant used (Fig. 1f–i). The comparison of the ELISAs also indicated that Montanide ISA70 was the best adjuvant for eliciting a strong IgG response. However, the presence of one of the three adjuvants used in this study was shown to be necessary, because the injection of TEM197STa alone did not allow the production of IgG (Fig. 1j) as observed for the negative controls (Fig. 1h).

Figure 1

Mean levels of IgG antibodies against STa and TEM-1 in blood sera of each group of cattle. (a) Group 1 sera, IgG against STa (n=4); (b) group 2 sera, IgG against STa (n=4); (c) group 3 sera, IgG against STa (n=4); (d) group 4 sera, IgG against STa (n=3); (e) group 5 sera, IgG against STa (n=4); (f) group 1 sera, IgG against TEM-1 (n=4); (g) group 2 sera, IgG against TEM-1 (n=4); (h) group 3 sera, IgG against TEM-1 (n=4); (i) group 4 sera, IgG against TEM-1 (n=3); (j) group 5 sera, IgG against TEM-1 (n=4); x-axis, days postinoculation; y-axis, the levels of increase indicates the difference between each level at a time postinoculation (respectively, D28, D56, D84, D112 and D140) and the initial level shown on day D0; Φ, mean with a 95% confidence interval (mean±1.96 SD).

Finally, we determined the anti-TEM-1 and anti-STa IgG1 antibody titers for the bovines in groups 1, 2 and 3 (Table 2). The data were compared with the immune response induced in mice using the same adjuvant. Interestingly, the adjuvants were not able to favor the production of antibodies directed against STa in the mice sera. We were not able to measure any relevant amount of IgG1 in the sera of mice immunized with the hybrid protein associated with Montanide ISA70 and Montanide IMS1313. In the case of Montanide ISA206, only one mouse developed the anti-STa antibodies (titer dilution 12 800). A previous study had indicated that the production of neutralizing antibodies against the enterotoxin could be generated when the hybrid protein was injected with the complete Freund adjuvant (N. Ruth et al., pers. commun.). The influence of the adjuvant was not detected for the anti-β-lactamase immunoglobulins. After immunization with Montanide adjuvant, the presence of anti-TEM IgG1 was detected at high level for all the tested sera. In cattle, our results indicated that the anti-IgG1 responses against the two antigens were observed. However, as already shown, the titers of anti-STa IgG1 were lower than those of the anti-TEM IgG1. Figure 2 shows the TEM-1 and STa antibody dilutions for groups 1, 2 and 3. The mean of the highest titer serum dilutions in IgG1 of TEM-1 is significantly higher than that of STa. For TEM-1 alone, the mean of the highest titer serum dilutions in IgG1 is significantly higher in group 1 in comparison with groups 2 and 3. In addition to the presence of IgG1, we also searched for IgG2 isotypes. Despite the poor response of the immune machinery against STa, the production of anti-STa IgG1 antibodies in groups 1, 2 and 3 was detectable after the second injection (Fig. 2a–e). For group 1, we measured the maximum titer of IgG1 anti-STa for two cattle after the third injection (day 84). For group 2, the highest response appeared after the last injection and was more homogenous than for the other groups. In group 3, the anti-STa response remained low.

View this table:
Table 2

Titers of the anti-TEM and anti-STa IgG1 present in the different bovine sera

Log2 (dilution)
Bovine sera titerMice sera titer
AdjuvantGroupTEM-1STaTEM-1STa
Montanide ISA 70115915ND
151017ND
16814ND
171014ND
Montanide ISA 2062171016ND
16915ND
179148
16812ND
Montanide IMS13133151013ND
15911ND
17915ND
15914ND
  • The highest titer serum dilutions giving an absorbance value of the mean values of three preimmune sera (day 0) +3 SDs. Measurements were made with sera collected at day 112. The starting sera for the titration (dilution=1) of the anti-TEM and anti-STa IgG1 were diluted 50-fold in saturating buffer.

Figure 2

Mean levels of IgG1 antibodies against STa and TEM-1 in blood sera of each group of cattle. (a) Group 1 sera, IgG1 against STa (n=4); (b) group 2 sera, IgG1 against STa (n=4); (c) group 3 sera, IgG1 against STa (n=4); (d) group 4 sera, IgG1 against STa (n=3); (e) group 5 sera, IgG1 against STa (n=4); (f) group 1 sera, IgG1 against TEM-1 (n=4); (g) group 2 sera, IgG1 against TEM-1 (n=4); (h) group 3 sera, IgG1 against TEM-1 (n=4); (i) group 4 sera, IgG1 against TEM-1 (n=3); (j) group 5 sera, IgG1 against TEM-1 (n=4); x-axis, days postinoculation; y-axis, the levels of increase indicate the difference between each level at a time postinoculation (respectively, D28, D56, D84, D112 and D140) and the initial level shown on day D0; Φ, mean with 95% confidence interval (mean±1.96 SD).

As already mentioned above, the IMS1313 adjuvant is less efficient than Montanide ISA206 and Montanide ISA70. In the case of TEM-1 β-lactamase, the anti-TEM-1 in groups 1, 2 and 3 were detected after the first immunization (Fig. 2f–j). For all immunized cattle, the levels of anti-TEM antibodies were high for the duration of the experiment. In group 2, we observed a decrease in antibody titer at the end of the experiment (day 140). The comparison of the impact of the adjuvant composition on the level of immune response indicates that Montanide ISA206 yielded a production of IgG1, which was very different between the four different groups of animals. With this adjuvant, a high variability was noted for all blood sera collected at days 28, 56, 84, 112 and 140. In order to follow the kinetic of the IgG1 response, we determined the IgG1 titers in all the sera collected from group 2. Figure 3 clearly indicates that, for the anti-TEM antibodies, the immune response was already present after the first injection of the chimeric β-lactamase. The two subsequent injections led to a small but significant increase in the IgG1 titers. Interestingly, we noted that the antibody level remained constant for the mice after the last injection. This behavior was not observed for the bovines, where the IgG1 level slowly decreased after the fourth injection. We also measured the kinetic of the appearance of the anti-STa IgG1 in the cattle sera. The highest titer was reached after the third injection and, as observed for the TEM-1, the titer decreased significantly after the last antigen injection. Our data confirmed again that the antibody level against STa remained low.

Figure 3

Specific anti-TEM-1 (○) and anti-STa (▵) IgG1 responses in cattle after immunization using Montanide ISA 206 as an adjuvant as a function of time. The specific anti-TEM IgG1 response in mice is also indicated (•). Data represent the log2 antibody titers, determined as the highest serum dilution giving an absorbance of the preimmune sera+3 SD. The arrows indicate the immunization schedule of the cows and mice.

Finally, we determined the level of the IgG2 isotype in the cattle sera. No detectable amount of this antibody isotype was found against STa (Fig. 4a–e). As already mentioned, the immune response against TEM was strong in the presence of the Montanides ISA70 (group 1) and 206 (group 2) (Fig. 4f–j). In the case of IMS1313 (group 3), the level of IgG2 was only detectable after the second injection (day 56). Furthermore, the IgG2 concentration decreased slightly after the last injection of TEM197STa.

Figure 4

Levels of increase in IgG2 antibodies against the toxin STa and the carrier protein TEM-1 in blood sera of each group of bovines. (a) Group 1 sera, IgG2 against STa (n=4); (b) group 2 sera, IgG2 against STa (n=4); (c) group 3 sera, IgG2 against STa (n=4); (d) group 4 sera, IgG2 against STa (n=3); (e) group 5 sera, IgG2 against STa (n=4); (f) group 1 sera, IgG2 against TEM-1 (n=4); (g) group 2 sera, IgG2 against TEM-1 (n=4); (h) group 3 sera, IgG2 against TEM-1 (n=4); (i) group 4 sera, IgG2 against TEM-1 (n=3); (j) group 5 sera, IgG2 against TEM-1 (n=4); x-axis, days postinoculation; y-axis, The levels of increase are the difference between each level at a time postinoculation (respectively, D28, D56, D84, D112 and D140) and the initial level shown at day D0; Φ, mean with a 95% confidence interval (mean±1.96 SD).

In the presence of the different sera diluted 10–1000-fold, we observed a 90% reduction in β-lactamase activity higher than or equal to 90% of the activity of TEM-1 incubated in buffer (Table 3). These preliminary results indicated the presence of neutralizing antibodies directed against TEM-1 β-lactamase.

View this table:
Table 3

Inhibition of the TEM-1 activity by cattle sera

Sera dilutionInitial rate (µM s−1)
Buffer0.3
Preimmune0.25
1/10<0.01
1/102<0.01
1/1030.04
1/1040.23
1/1050.20

Sero-prevalence of anti-STa and TEM-1 antibodies from cattle field sera

In order to estimate whether the hybrid protein TEM197STa could be used as a vaccine against one of the major class A β-lactamase and the enterotoxin STa produced by ETEC E. coli, we measured the sero-prevalence of IgG antibodies against STa and TEM in field sera from 366 cattle. The distribution of IgG antibodies against STa in cattle originating from the field, both negative and immunized groups, is presented in Fig. 5a. A weak but significant difference (Welch's test; P<0.01) was observed. This result indicates that the level of antibodies induced by injection of the hybrid protein was not very different from that of the STa antibody level observed in the other two groups of cattle. Therefore, the use of the TEM-1 β-lactamase as the carrier of the STa does not allow the production of a significant level of IgG antibodies directed against STa. Strikingly, the response of the bovine immune system is different from the one observed in mice. The injection of the subunit recombinant vaccine TEM197STa in mice has been shown to induce the production of neutralizing IgG1 antibodies against both the toxin and the β-lactamase (N. Ruth et al., unpublished data). In addition, the immunization of mice with plasmid DNA encoding a hybrid hapten-carrier protein can be used as a priming stimulus to provide T cell help in the antibody response against the nonimmunogenic STa peptide (Ruth et al., 2005).

Figure 5

Box plots of OD obtained by an ELISA test for IgG antibodies against STa (a) and TEM-1 (b) in each group of cattle. x-Axis: group of cattle, 1=field animals, 2=negative animals and 3=immunized animals (for details, see Materials and methods section); y-axis: OD obtained by the ELISA test; Box plot, horizontal line in box: median, upper hinge of the box: 75th%, lower hinge of the box: 25th%, upper adjacent value: largest value below the upper hinge +1.5 (difference of hinges) and lower adjacent value: smallest value above the lower hinge −1.5 (difference of hinges). The points represent the values outside the ranges.

However, for TEM-1, a significant and huge difference was observed between the vaccinated group and the others (Welch's test; P<0.003) (Fig. 5b). This finding suggests that the majority of the bovines in the field do not develop antibodies against the TEM-1 β-lactamase. We can conclude that our study clearly demonstrates that TEM-1 could be considered as a potential vaccine candidate. This hypothesis was already tested by Ciofu (2002). They showed that an antibody with β-lactamase inhibitory activity raised by immunization with β-lactamase can improve the outcome of treatment with ceftazidime of resistant Pseudomonas aeruginosa in a rat model of chronic lung infection. We are currently testing the possibility of restoring the susceptibility to classical β-lactam (ceftazidime and cefotaxime) of bacteria producing a TEM β-lactamase. The use of the class A β-lactamase as a carrier protein needs to be investigated using other adjuvants. We are also currently planning to perform a challenge experiment on vaccinated cattle. The basis for this new experiment is the fact that, in the present study, we were able detect the low presence of anti-STa antibodies in the vaccinated cattle sera.

Acknowledgements

We would like to thank Dr Hallet for the pFH197H plasmid and Dr Amman for the pTAC11plasmid. We also thank Prof. Devreese for the mass spectrometer measurements. Special thanks are due to N. Gerardin for the N-terminal sequences. This study was supported by a grant from the Ministry of the Walloon Region, Department of Technologies, Research and Energy (convention no. 114694), and the FRS-FNRS (FRFC grant 2.4561.07). P.S.M. is a postdoctoral fellow from the Walloon region.

Footnotes

  • Editor: Nicholas Carbonetti

References

View Abstract