Rhodococcus equi Secreted Antigens Are Immunogenic and Stimulate a Type 1 Recall Response in the Lungs of Horses Immune to R. equi Infection

Bacteria.The following bacterial strains were used in experiments described in this paper: R. equi ATCC 33701, a virulent, plasmid-bearing strain; plasmid-cured R. equi ATCC 33701 PC, a strain lacking the virulence plasmid; and a clinical isolate of Corynebacterium pseudotuberculosis provided by the Washington Animal Disease Diagnostic Laboratory. All bacterial strains were stored as stabilates and were kept at −86°C prior to use. Each bacterium was reconstituted, and a single colony was selected and grown in 10 ml of brain heart infusion (BHI) medium (Difco Laboratories, Detroit, Mich.) for 7 h at 37°C and pH 7.0 with shaking. The 10-ml culture was added to an additional 90 ml of BHI medium and grown at 37°C and pH 7.0 for 16 h with shaking. The BHI medium culture was then rediluted 1:10 in 90 ml of Dulbecco's modified Eagle medium (DMEM) (Gibco/Invitrogen, Grand Island, N.Y.) (pH 7.0) and grown at 37°C for 4, 8, 24, 48, 96, or 144 h with shaking for the growth curve experiments. In the case of time zero in the growth curve experiments, the bacterial cultures were harvested immediately after 1:10 redilution in DMEM. For all other experiments, the bacterial cultures were prepared similarly and grown for 48 h with shaking. C. pseudotuberculosis was grown like R. equi for use as a control antigen (see below). R. equi ATCC 33701 and ATCC 33701 PC were also grown in 100 ml of BHI medium for 24 h at 37°C with shaking for use as positive and negative controls, respectively, in immunoblotting.

Preparation of CFS of R. equi ATCC 33701, R. equi ATCC 33701 PC, and C. psuedotuberculosis.Cultures were centrifuged at 5,200 × g for 15 min at 4°C with a Beckman-Coulter Avanti J-25, and the supernatants were sterile filtered by using a 0.22-μm-pore-size membrane (Millipore, Bedford, Mass.). The supernatants were concentrated 100-fold by using a 10,000-molecular-weight concentrator (Centricon Plus-80; Amicon, Danvers, Mass.) according to the manufacturer's instructions and were dialyzed by extensive washing with sterilized, distilled water while they were in the ultrafiltration device. DMEM containing 10% BHI was sterile filtered and concentrated exactly like the bacterial CFS preparations for use as a medium-alone negative control in immunoblotting, lymphoproliferation, and real-time reverse transcription (RT)-PCR analyses. The protein contents of all preparations were determined by performing a Micro-BCA protein assay with an albumin standard according to the manufacturer's instructions (Pierce, Rockford, Ill.).

Preparation of soluble R. equi antigen. R. equi ATCC 33701 was grown in BHI medium for 72 h at 37°C and pH 7.0 with shaking, harvested by centrifugation at 5,000 × g for 15 min, and washed with sterile phosphate-buffered saline. Two milliliters of the bacterial pellet was resuspended in 10 ml of phosphate-buffered saline and frozen at −20°C for 2 h. The pellet was thawed, sonicated for 3 min, centrifuged at 12,000 × g for 15 min, and then disrupted by three freeze-thaw cycles at −80°C. After the third cycle, the bacterial homogenate was centrifuged again at 25,000 × g for 1 h to separate the soluble components from the intact bacteria and debris. The supernatant was collected to obtain the soluble R. equi antigen used in proliferation assays and real-time RT-PCR. The protein content was determined by a Micro-BCA protein assay.

SDS-PAGE and immunoblotting.DMEM-BHI, CFS, and bacterial pellets were boiled with a solution containing 350 mM Tris-Cl, 33% glycerol, 3.3% sodium dodecyl sulfate (SDS), 3.3% β-mercaptoethanol, and 0.004% bromphenol blue; separated under nonreducing conditions by SDS-polyacrylamide gel electrophoresis (PAGE) by using a 12% acrylamide separation gel; and transferred electrophoretically to a nitrocellulose membrane (Hybond; Amersham Pharmacia Biotech Inc., Piscataway, N.J.). For the experiment whose results are shown in Fig. 5B, a 12% 2D-Prep gel (Bio-Rad Laboratories, Hercules, Calif.) was used, and 1.74 mg of 48-h CFS protein was loaded in a lane covering the entire gel in order to accurately interpret protein band comigration. For all other immunoblots, CFS and/or pellet antigen was loaded into individual lanes, as indicated below. Nitrocellulose membranes were incubated with sera from foals infected with R. equi (verified by isolation from transtracheal wash fluid), sera from immune horses 2 weeks postchallenge with R. equi ATCC 33701 (horses were shown to be immune by effective clearance of R. equi infection) (26), an anti-VapA monoclonal antibody (monoclonal antibody 10G5, provided by Shinji Takai) (38), or a monospecific rabbit antiserum to detect the presence of VapC, VapD, or VapE. VapA is a surface protein, and VapC, VapD, and VapE have been characterized previously as secreted proteins encoded by the R. equi ATCC 33701 virulence plasmid (11). Horseradish peroxidase-conjugated goat anti-horse, goat anti-rabbit, or goat anti-mouse immunoglobulin G (Kirkegaard & Perry, Gaithersburg, Md.) was used as a secondary antibody at a 1:20,000 dilution. The primary antibody dilutions used are given in the figure legends, and sera were obtained from the following foals and horses: foal 07 (see Fig. 2 and 5B); foals 136, CA11, and 151 (see Fig. 3A); and horses N124, 20, and 02 (see Fig. 3B and 4). Bound antibodies were detected by using a chemiluminescent substrate kit (ECL; Amersham Pharmacia Biotech Inc.).

Preparation of R. equi for bronchial challenge.Frozen R. equi stabilates were reconstituted, and a single colony was selected and grown in BHI medium for 14 h at 37°C and pH 7.0 with shaking. The culture was centrifuged at 800 × g, and the bacterial pellet was washed three times with saline. Then the organism was resuspended in saline at a final concentration of 2 × 10⁷ cells/ml. To confirm the presence of the virulence plasmid in the R. equi strains used in this study, bacteria were periodically tested for VapA expression by using immunoblots (R. equi soluble antigen [SRA] or bacterial pellet) probed with monoclonal antibody 10G5.

BAL and pulmonary challenge.All animal experiments were conducted in compliance with relevant federal guidelines and the Animal Care and Use Program of Washington State University. Previous studies have shown that peripheral blood mononuclear cells are not a sensitive indicator of the local pulmonary responses involved in clearance of R. equi (18); therefore, T-lymphocytes were isolated from the lungs of horses. Three clinically normal thoroughbred horses (horses 44, 71, and 02, which were 5, 17, and 28 years old, respectively) were studied. Two weekly bronchoalveolar lavages (BALs) were performed on each horse as previously described (18). Briefly, horses were mildly sedated with butorphanol and xylazine. An endoscope was passed nasally and directed into the right and left cranial lobar bronchi. A total of 540 ml of 0.9% sodium chloride-0.06% sodium bicarbonate (pH 6.5) was instilled into the right and left lungs in three 180-ml doses. After each instillation of 180 ml of saline, the BALF was aspirated for analysis. During the first BAL procedure (day 0) after both lungs had been completely aspirated, each lung was inoculated with 2 × 10⁷R. equi cells suspended in 5 ml of saline, and the endoscope was flushed with 60 ml of air just prior to removal. The second BAL (day 7) was performed by using the same procedure, except that no R. equi was instilled. On the day of each BAL procedure, blood was obtained via jugular venipuncture and submitted to the Washington State University Clinical Pathology Laboratory for determination of fibrinogen levels and complete blood counts. Following each BAL, the horses were placed in a stall for 1 week and monitored for fluctuations in rectal temperature and pulse and respiration rates. Previous work in our lab has shown that when infected in one lung, adult horses effectively clear the infection without major clinical signs in association with local, antigen-specific recall responses. Here we chose to infect both lungs in order to recover the maximum number of cells associated with a local recall response. Each horse infected in this study developed a transient fever of 39.4 to 40°C; however, all three horses effectively cleared the infection and had fibrinogen levels within the normal range 1 week postinfection. One horse developed a mild upper respiratory infection, and thus a more in-depth assessment was conducted, which included auscultation of the lungs with a rebreathing bag. This horse effectively cleared the respiratory infection 1 week after challenge, as determined by a lack of abnormal lung sounds and a negative BALF culture; thus, no antibiotics were administered.

BALF cell preparation.The BALF was centrifuged at 400 × g for 8 min. The pelleted cells were resuspended and washed three times with RPMI (Gibco) containing 0.05 mg of gentamicin per ml, 25 mM HEPES, 3.5 × 10⁻⁶ M β-mercaptoethanol, 100 mM l-glutamine (RPMI), and 2.5 μg of Fungizone per ml. If red blood cells were grossly apparent, cell pellets were incubated with 0.87% ammonium chloride for 5 min after the first wash. Cells were resuspended in RPMI containing 10% horse serum at a concentration of 2 × 10⁶ cells/ml for a 1-h macrophage adherence analysis in 175-cm² tissue culture flasks (Becton Dickinson Labware, Franklin Lakes, N.Y.). Nonadherent cells were washed and resuspended at a concentration of 2 × 10⁶ cells/ml for proliferation assays and quantification of IFN-γ mRNA.

Proliferation assay.BALF cells were assayed in quadruplicate wells in round-bottom 96-well microplates (Sarstadt Inc., Newton, N.C.) as previously described (18). Briefly, 4 × 10⁵ cells were incubated with R. equi CFS, C. pseudotuberculosis CFS, or concentrated DMEM-BHI (all at four concentrations, 1, 2.5, 10, and 25 μg/ml) or with soluble R. equi antigen (2.5 and 10 μg/ml) for 96 h at 37°C in 5% CO₂. Optimal stimulation was achieved with 10 μg of protein/ml (data not shown); therefore, 10 μg of DMEM-BHI per ml and each antigen were used in proliferation assays and in cytokine analyses. As a positive control, cells were also stimulated with pokeweed mitogen at a concentration of 5 μg/ml for 36 h at 37°C in 5% CO₂. The cells were incubated for an additional 21 h after addition of 0.25 μCi of [³H]thymidine and were harvested with a Tomtec cell harvester. The uptake of [³H]thymidine was measured by liquid scintillation counting with a β-plate reader (Wallac, Gaithersburg, Md.).

Quantification of IFN-γ mRNA.For antigen-specific IFN-γ expression, cells were incubated as described above for the proliferation assay. After incubation for 24 h in 25- and 75-cm² tissue culture flasks (Corning Inc., Corning, N.Y.) before and after intrabronchial challenge, respectively, the cultures were harvested as previously described (26). A culture period of 24 h was chosen to measure IFN-γ mRNA expression based on previous time course studies (15, 26; David Horohov, personal communication). RNA was immediately extracted with an RNeasy kit (Qiagen Inc., Valencia, Calif.) used according to the manufacturer's specifications. To eliminate contaminating DNA, RNA was subjected to a DNase treatment for 30 min at 37°C. The concentrations of RNA in samples were estimated by determining the optical densities at 260 and 280 nm, and cDNA was generated by using 1 μg of RNA with murine leukemia virus reverse transcriptase and random hexamers in a reaction for 20 min at 42°C. To confirm that no genomic DNA was present, portions of the DNase-treated RNA samples were used in cDNA reaction mixtures without reverse transcriptase (no-reverse transcriptase controls). The cDNA was then used to detect expression of IFN-γ by real-time PCR. BAL cells from a horse that was not challenged with R. equi were stimulated with pokeweed mitogen for 24 h under similar conditions and used to generate a linear standard curve for IFN-γ transcription in real-time PCR.

Expression of equine IFN-γ and of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in BAL cells was determined by performing real-time PCR with previously described primers (26). Briefly, the fluorescent dye SYBR Green was used to measure amplified products in a thermocycler (iCycler; Bio-Rad Laboratories). Reactions (volume of reaction mixtures, 25 μl) were performed in 96-well plates containing 50 ng of cDNA, 1× SYBR Green buffer, 1.5 mM MgCl₂, each deoxynucleoside triphosphate at a concentration of 1 mM, 0.04 U of uracil N-glycosylase (Amperase), and 0.2 U of AmpliTaq Gold polymerase. For each 96-well plate, a standard curve for IFN-γ and GAPDH was generated by using a dilution series of cDNA from a horse that was not challenged with R. equi. The resulting curves were used as standards to control for variation between plates. All samples were analyzed in triplicate, and amplification was performed by using the following procedure: 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of 95°C for 15 s, 55°C for 30 s, and 72°C for 30 s. Transcript levels were determined for each sample by comparing the threshold cycle values of IFN-γ and GAPDH to the corresponding standard curves. The transcript levels of IFN-γ were normalized by using the transcript level value for GAPDH from the same sample (IFN-γ/GAPDH). In each real-time PCR, reaction mixtures containing no cDNA were included to control for extraneous DNA within the reagents.

Statistical analyses.Statistical analyses were performed by using the SAS statistical package, version 8 (SAS Institute Inc., Cary, N.C.). Significant differences in proliferative responses and IFN-γ expression on days 0 and 7 were determined by using a randomized complete block design, followed by the Bonferroni multiple-comparison test. Significant differences in proliferative responses and IFN-γ expression between samples obtained before and after challenge were detected by using a repeated-measure randomized complete block design, followed by the Bonferroni multiple-comparison test. For all statistical analyses, the horse-antigen interaction was used as the error term for horse (block effect) and antigen. The blocking effect was significant in the analyses; thus, the randomized complete block design was used in this study. Data that were not normally distributed were transformed by using the square root transformation; transformation of data did not change the significance in any analysis. P values less than 0.05 were considered significant.

Growth of R. equi for CFS preparation.The growth of R. equi and the accumulation of proteins in the CFS were determined over a 144-h period (Fig. 1). The number of bacteria steadily increased for 96 h, and from 96 to 144 h the growth decreased. The total CFS protein content was at a plateau from 4 to 24 h and then began to increase until 96 h of growth. Between 96 and 144 h there was a substantial increase in the CFS protein content. This marked increase coincided with a decline in bacterial growth, which was consistent with passive release of nonsecreted proteins from dead and dying bacteria into the CFS.

• Open in new tab
• Download powerpoint

FIG. 1.

Initial characterization of R. equi CFS. R. equi ATCC 33701 was grown in DMEM-BHI for 4 to 144 h. Bacterial growth was determined by measuring the absorbance at 600 nm with a spectrophotometer. The accumulation of protein in CFS was measured by a BCA protein assay. The background protein content at time zero was subtracted from the CFS protein content at each time point.

Identification of immunogenic proteins in R. equi CFS.To identify proteins recognized by antibody during R. equi infection and to characterize the appearance of the proteins in CFS over time, immunoblots were probed by using serum antibodies from a foal with active rhodococcal pneumonia. The results after 4, 48, and 144 h of culture are shown in Fig. 2. Ten to twelve immunogenic proteins were identified in R. equi CFS. Importantly, the pattern of proteins detected in CFS as defined by the location and intensity of the bands in each lane (Fig. 2, lanes 4, 6, and 8) was consistently different from the pattern obtained for lanes loaded with the bacterial pellet (Fig. 2, lanes 5, 7, 9, and 10). CFS antigen did not include the characteristic 15- to 17-kDa doublet that represents the VapA protein in typical R. equi antigen preparations (41). VapA is an immunodominant, temperature- and pH-regulated surface molecule that is encoded by a gene on the R. equi virulence plasmid (37). The VapA doublet was consistently present in lanes loaded with pellet antigen (Fig. 2, lane 10). No proteins were recognized in the CFS medium control (Fig. 2, lane 3), which consisted of DMEM with 10% BHI.

• Open in new tab
• Download powerpoint

FIG. 2.

Detection of immunogenic proteins in the CFS at three time points. Immunoblots of CFS and bacterial pellets from R. equi ATCC 33701 grown in DMEM-BHI for 4, 48, and 144 h were probed with serum from a foal with active rhodococcal pneumonia (1:800 dilution). Lanes were loaded with virulent R. equi ATCC 33701, unless indicated otherwise, as follows: lane 1, SRA (positive control); lane 2, SRA of plasmid-cured strain R. equi ATCC 33701 PC; lane 3, DMEM-BHI (medium control); lane 4, 4-h CFS; lane 5, 4-h pellet; lane 6, 48-h CFS; lane 7, 48-h pellet; lane 8, 144-h CFS; lane 9, 144-h pellet; lane 10, 144-h pellet diluted 1:10. All lanes containing CFS were loaded with 153 μg of protein. Lanes containing pellets were loaded by volume as follows: 15 μl for 4- and 48-h pellet antigens and 5 μl for 144-h pellet antigen (undiluted). The arrow on the right indicates the position of the characteristic doublet attributable to VapA. The numbers on the left indicate molecular masses (in kilodaltons).

The protein bands identified in CFS by this immunoblot analysis were divided into two subgroups based on relative migration in SDS-PAGE gels. A group of five distinct proteins with calculated molecular masses of 12 to 24 kDa were present in CFS after 4 h of culture and at all subsequent times. The locations of these bands appeared to be slightly different after 48 and 144 h of culture than after 4 h, and the bands were slightly more intense at 48 h and slightly less intense at 144 h. A second group of CFS proteins had calculated molecular masses between 56 and 99 kDa. The migration patterns of the proteins in this molecular mass range were different after 4 and 48 h of culture, and the intensities of the bands in this part of the blot were significantly diminished after 144 h of culture. An additional broad band at 50 kDa was detected in the 144-h CFS (Fig. 2, lane 8) and was also present in the 4- and 48-h pellets. In accordance with the observed increase in the growth of bacteria between 48 and 144 h (Fig. 1), the amount of antigen in the pellet increased during this time (Fig. 2, lanes 7, 9, and 10). Media from cultures in the late logarithmic growth phase (e.g., 144 h) contained increasing amounts of nonsecreted, cytoplasmic proteins due to passive release from dead and dying bacteria (2). Since our goal was to identify proteins that are actively secreted by live bacteria, 48-h CFS was used for further characterization.

In order to determine whether the proteins identified in CFS were uniquely recognized by antibodies from the one foal or were more broadly recognized by serum antibodies from all infected horses, CFS and pellet antigen were probed by using sera from 12 additional foals with active rhodococcal pneumonia and by using sera from five adult horses. The results for three representative foal sera and three representative horse sera are shown in Fig. 3A and B, respectively. The adult horses had been challenged with virulent R. equi 2 weeks prior to serum collection and were shown to be immune (see Materials and Methods) (26). Similar to the initial serum tested, sera from all six animals reacted with five distinct protein bands in the12- to 24-kDa range of the 48-h CFS blot (Fig. 3, lanes CFS). When reacted with the 48-h pellet antigen, all sera recognized the characteristic 15- to 17-kDa doublet attributed to VapA (Fig. 3, lanes P). Again, the migration pattern of proteins identified in CFS was different from the pattern of proteins identified in the bacterial pellet. Sera from the foals, but not sera from the immune adults, also recognized a small number of CFS proteins that migrated at higher molecular masses, between 55 and 99 kDa (Fig. 3A, lanes CFS). However, the number and migration pattern of these higher-molecular-mass proteins were not consistent for all foals.

• Open in new tab
• Download powerpoint

FIG. 3.

Identification of proteins in 48-h CFS by using sera from R. equi-infected foals and immune adult horses. Immunoblots of CFS and bacterial pellets of R. equi ATCC 33701 grown in DMEM-BHI for 48 h were probed with sera from three foals infected with R. equi (A) or sera from three adult horses collected 2 weeks after intrabronchial challenge with R. equi ATCC 33701 (B). Lanes were loaded with 195 μg of a 48-h CFS or 3 μl of a 48-h pellet (P), as indicated. The antibody dilutions were as follows: foal 136 (#1), 1:800; foal CA11 (#2), 1:500; foal 151 (#3), 1:800; horse N124 (#4), 1:1,000; horse 20 (#5), 1:800; and horse 02 (#6), 1:800. The numbers on the left indicate molecular masses (in kilodaltons).

Presence of plasmid-associated proteins in R. equi CFS.The R. equi virulence plasmid is required for the bacterium to replicate and survive within the phagosome and to cause clinical disease in foals (14). Therefore, the 48-h CFS from both the plasmid-bearing strain R. equi ATCC 33701 and the plasmid-cured strain R. equi ATCC 33701 PC were probed with sera from immune adult horses to determine whether the low-molecular-mass immunogenic protein bands were associated with the virulence plasmid. The distinct 12- to 24-kDa protein bands that were present in CFS from the plasmid-bearing strain were not detected in the CFS from the plasmid-cured strain (Fig. 4). These results support the hypothesis that the 12- to 24-kDa immunogenic proteins in CFS are encoded by genes on the virulence plasmid or the hypothesis that the expression of these proteins is regulated by genes on the virulence plasmid. The CFS from both the plasmid-cured strain and the virulent strain produced two to four protein bands at calculated molecular masses between 45 and 98 kDa that reacted with the sera from adult horses, although the reactivity was inconsistent (Fig. 4).

• Open in new tab
• Download powerpoint

FIG. 4.

Immunogenic CFS proteins in R. equi lacking the virulence plasmid: immunoblot of R. equi ATCC 33701 CFS (+) and plasmid-cured R. equi ATCC 33701 PC CFS (PC) grown in DMEM-BHI for 48 h and probed with sera from three immune adult horses (as described in the legend to Fig. 3). Each lane was loaded with 195 μg of protein. The numbers on the left indicate molecular masses (in kilodaltons).

Identification of secreted proteins VapC, VapD, and VapE in R. equi CFS.To demonstrate that the VapA surface protein was not present in the 48-h CFS and to confirm that the 15- to 17-kDa protein in the 48-h pellet antigen recognized by the equine sera was in fact VapA, an immunoblot was probed with a monoclonal antibody to VapA (Fig. 5A). The characteristic 15- to 17-kDa VapA doublet was present in the lane containing the bacterial pellet (Fig. 5A, lane 5) but did not appear in the lane containing the CFS preparation (Fig. 5A, lane 4). Interestingly, however, the anti-VapA monoclonal antibody also reacted with a group of protein bands in the low-molecular-mass region of the CFS. The 48-h CFS contained six protein bands at 11, 12, 13, 17, 21, and 24 kDa, which are the predicted molecular masses of molecules that were reactive with the monoclonal antibody (Fig. 5A, lane 4). The monoclonal antibody did not react with any protein migrating in the higher-molecular-mass region of the immunoblot. As expected, no protein bands were identified when the anti-VapA antibody was used to probe CFS or pellet antigen from a plasmid-cured derivative of R. equi virulent strain ATCC 33701 (data not shown).

• Open in new tab
• Download powerpoint

FIG. 5.

Identification of plasmid-encoded Vap proteins in R. equi CFS: immunoblots of R. equi CFS grown in DMEM-BHI for 48 h. (A) The anti-VapA monoclonal antibody reacted with VapA in SRA and the pellet antigen and six protein bands in the CFS antigen. The membrane was probed with 10G5, a monoclonal antibody against VapA. Lanes were loaded as follows: lane 1, SRA from strain ATCC 33701 (positive control); lane 2, SRA of plasmid-cured R. equi ATCC 33701 PC; lane 3, DMEM-BHI (medium control); lane 4, 48-h ATCC 33701 CFS; lane 5, 48-h ATCC 33701 bacterial pellet. The arrow indicates the position of the characteristic VapA doublet in the pellet. (B) Antibodies in sera from an infected foal recognize proteins that comigrate with VapC, VapE (arrows), and possibly VapD. CFS antigen was probed with monospecific antisera against VapC (1:100 dilution), VapD (1:250 dilution), and VapE (1:800 dilution) or with serum from an R. equi-infected foal (1:800 dilution). The nitrocellulose membrane was cut into five sections; four sections are shown, and the fifth section was probed with normal rabbit serum and did not react with any bands (data not shown). The numbers on the left indicate molecular masses (in kilodaltons).

A previous study showed that VapA is not excreted into the medium and is therefore present only in bacterial pellet preparations (11). However, one explanation for the presence in R. equi CFS of proteins that are reactive with the anti-VapA antibody is shedding of VapA from the bacterial surface, possibly due to cleavage. An alternative explanation is active secretion of additional virulence-associated proteins related to VapA. The Vap proteins are encoded by a family of genes clustered in the pathogenicity island of the R. equi virulence plasmid. Recent work has shown that the 10G5 monoclonal antibody binds an epitope within a conserved region shared by multiple Vap proteins (42).

To further characterize R. equi CFS for the presence of secreted Vap proteins, the 48-h CFS was probed with monospecific rabbit antisera to VapC, VapD, and VapE (11) (Fig. 5B). These previously identified proteins are coordinately regulated with VapA (11). All three monospecific antisera recognized protein bands present in the CFS (Fig. 5B) and not in the pellet (data not shown). The anti-VapC and anti-VapD antisera reacted with bands at 12 and 15 kDa, respectively. Additionally, the anti-VapE antiserum identified a triplet, the dominant band of which migrated at 17 kDa. The protein bands identified by anti-VapC, anti-VapE, and possibly anti-VapD comigrated with protein bands recognized by equine serum (Fig. 5B).

R. equi CFS-specific lymphoproliferative responses.Immune adult horses effectively clear a challenge dose of virulent R. equi in association with a pulmonary recall response (18). The R. equi CFS specificity of lymphocytes recovered from BALF of three immune adult horses was evaluated in proliferation assays prior to challenge (day 0) and 1 week after challenge (day 7). Cells were stimulated with pokeweed mitogen, R. equi CFS, and (as a positive control) SRA, which contains VapA. The negative controls included DMEM-BHI (growth medium alone) and C. pseudotuberculosis CFS. Proliferation was evaluated by measuring incorporation of [³H]thymidine and was expressed in counts per minute. Prior to challenge, there was no significant difference in proliferation between antigen-stimulated BALF cells and nonstimulated BALF cells (Fig. 6A). This result also indicates that none of the antigens used had a mitogenic effect. In contrast, BALF cells obtained at day 7 postchallenge specifically proliferated in response to R. equi CFS and SRA, and the responses were significantly greater than the R. equi-specific responses before challenge (P < 0.01) (Fig. 6B). Additionally, the R. equi CFS-specific and SRA-specific responses after challenge were significantly greater (P < 0.05 and P < 0.01, respectively) than the proliferative responses of BALF cells stimulated with DMEM-BHI and C. pseudotuberculosis CFS, as well as the proliferative responses of nonstimulated BALF cells at day 7 (Fig. 6B). BALF cells from all the horses proliferated when they were stimulated with pokeweed mitogen (data not shown).

• Open in new tab
• Download powerpoint

FIG. 6.

Lymphoproliferation of BALF cells in response to R. equi CFS 1 week after challenge with R. equi ATCC 33701. BALF cells were cultured for 5 days before and after challenge and either not stimulated (n/s) or stimulated with 10 μg of DMEM-BHI per ml, C. pseudotuberculosis CFS (Cpstb CFS), R. equi CFS (Requi CFS), or SRA (Requi SRA). The data are the means for BALF cell cultures from all three horses. The error bars indicate the standard errors of the means. (A) Day 0; (B) day 7 after challenge. An asterisk indicates a significant difference (P < 0.05) compared to the values for the controls (not stimulated, DMEM-BHI, and C. pseudotuberculosis CFS).

R. equi CFS-specific IFN-γ expression by BALF cells.Prior to challenge and on day 7 after challenge with R. equi, BALF cells were analyzed by real-time RT-PCR to detect expression of IFN-γ transcripts. Cells were stimulated for 24 h with R. equi CFS, C. pseudotuberculosis CFS, R. equi SRA, and DMEM-BHI as the growth medium control. The efficiency of the real-time PCR was >0.90 (data not shown). The cytokine transcript levels were normalized by determining the level of GAPDH transcripts for each sample. The data were expressed as mean normalized levels of IFN-γ expression for each of three horses to more clearly demonstrate individual variation. On day 0, the mean normalized IFN-γ transcript levels for all three horses were low for all antigen-stimulated cells (Fig. 7). However, in one horse the IFN-γ mRNA levels for R. equi SRA-stimulated cells were significantly greater than the levels for cells stimulated with DMEM-BHI and C. pseudotuberculosis CFS, as well as the levels for nonstimulated cells (horse 44; P < 0.01) (Fig. 7B). Importantly, on day 7 postchallenge the mean normalized IFN-γ transcript levels for cells stimulated with R. equi CFS and R. equi SRA for all three horses were significantly greater than the levels on day 0 (P < 0.001). Furthermore, after challenge the R. equi CFS- and SRA-specific IFN-γ transcript levels were significantly greater than the levels for cells stimulated with DMEM-BHI, C. pseudotuberculosis CFS, and medium alone (nonstimulated) for each horse (P < 0.01) (Fig. 7).

• Open in new tab
• Download powerpoint

FIG. 7.

R. equi CFS-specific IFN-γ expression by BALF cells 1 week after challenge with R. equi ATCC 33701. The data are mean normalized levels of IFN-γ expression from each of three horses: horse 71 (A), horse 44 (B), and horse 02 (C). BALF cells were not stimulated (n/s) or were stimulated for 24 h with 10 μg of DMEM-BHI per ml, C. pseudotuberculosis CFS (Cpstb CFS), R. equi CFS (Requi CFS), or SRA (Requi SRA) before (day 0) and after (day 7) intrabronchial challenge. An asterisk indicates a significant difference (P < 0.05) compared to the values for the controls (not stimulated, DMEM-BHI, and C. pseudotuberculosis CFS) at the same time.