|Home | About | Journals | Submit | Contact Us | Français|
Susceptibility of foals to Rhodococcus equi pneumonia is exclusive to the first few months of life. The objective of this study was to investigate the immediate immunologic response of foal and adult horse antigen-presenting cells (APCs) upon infection with R. equi. We measured the activation of the antigen-presenting major histocompatibility complex (MHC) class II molecule, costimulatory molecules CD40 and CD86, the cytokine interleukin-12 (IL-12), and the transcriptional factor interferon regulatory factor 1 (IRF-1) in monocyte-derived macrophages (mMOs) and dendritic cells (mDCs) of adult horses and foals of different ages (from birth to 3 months of age) infected with virulent R. equi or its avirulent, plasmid-cured derivative. Infection with virulent or avirulent R. equi induced (P ≤ 0.01) the expression of IL-12p35 and IL-12p40 mRNAs in foal mMOs and mDCs at different ages. This response was likely mediated by the higher (P = 0.008) expression of IRF-1 in foal mDCs at birth than in adult horse mDCs. R. equi infection promoted comparable expression of costimulatory molecules CD86 and CD40 in foal and adult horse cells. The cytokine and costimulatory response by foal mDCs was not accompanied by robust MHC class II molecule expression. These data suggest that foal APCs detect the presence of R. equi and respond with the expression of the Th1-inducing cytokine IL-12. Nevertheless, there seems to be a limitation to MHC class II molecule expression which we hypothesize may compromise the efficient priming of naïve effector cells in early life.
Infectious respiratory diseases constitute one of the major causes of death and economic losses in the horse industry (59). Prevention and treatment of Rhodococcus equi pneumonia in young foals constitute a significant clinical challenge (3, 6). Susceptibility of foals to R. equi pneumonia is exclusive to the first few months of life (4, 16). R. equi causes severe pyogranulomatous pneumonia, enteritis, and occasionally bacteremia and joint infection, involving expensive treatments and concerns about the health and well-being of the foal. The bacterium is prevalent in the horse environment (e.g., feces, soil), and foals are exposed to it immediately after birth. Whereas young foals are susceptible to R. equi infection, pneumonia caused by this pathogen has also been described in immunocompromised adult horses and human patients (39). These observations strongly suggest that limitations in the immune responses of foals in early life affect the clearance of infection (41).
R. equi is a gram-positive, facultatively intracellular organism capable of multiplying in macrophages (21). Of the 40 genera in the actinomycete group, the genus Rhodococcus is placed among the members of the mycolata taxon, along with Mycobacterium, Nocardia, Gordonia, Tsukamurella, and Corynebacterium. In vitro, R. equi binding to macrophages can be mediated by complement receptor CR3 (CD11b/CD18 or MAC-1), mannose receptor (which binds lipoarabinomannan), and potentially Toll-like receptor 2 (9, 14, 25). Mycolic acids are long fatty acids found in the lipid-rich cell wall envelope of these bacteria, and they form a protective barrier that increases resistance to chemical damage, dehydration, oxidative stress, and low pH (47). Once inside alveolar macrophages, the bacterium is found within membrane-enclosed vacuoles that do not fuse with lysosomes, which allows intracellular survival (10, 58, 63). Therefore, R. equi has evolved a mechanism to escape bactericidal activity in macrophages, which, paradoxically, are important cells of the immune system that perform surveillance, removal, and killing of microorganisms. Virulent strains of R. equi contain a large plasmid that encodes a family of eight virulence-associated proteins (VapA and VapC to VapI) (46, 52, 55). This plasmid is critical for intracellular replication within macrophages and for the development of disease in foals (18, 26, 30).
Phagocytes participate in the early innate immune response by removing and killing pathogens. Nitric oxide and superoxide, in combination as peroxynitrite but not as individual compounds, mediate intracellular killing of R. equi, which is dependent on the activation of macrophages by immune mediators, importantly, gamma interferon (IFN-γ) (8). In both mice and adult horses, the activation of helper and cytotoxic T lymphocytes is also an essential mechanism by which this intracellular pathogen infection is controlled (22, 23, 34). Presumably, the source of IFN-γ for macrophage activation would come from activated lymphocytes, and the latter would respond to antigen-presenting cell (APC) signaling through interleukin-12 (IL-12) (15). Although macrophages have been a major focus of studies of R. equi pathogenesis, dendritic cells (DCs) may be involved in the early events of infection in order to signal lymphocytes for the effective removal of infected cells (via cytotoxic T cells) and activation of macrophages for intracellular bacterial killing (via IFN-γ production). This seems to be the case in infection with Mycobacterium tuberculosis, which also infects and replicates inside human macrophages and requires the priming of lymphocytes for protection (57).
In this context, we questioned if the activation of DCs in young foals would be comparable to that of adult horse cells upon R. equi infection. Notably, DCs, rather than macrophages, are necessary for priming of naïve T cells in a first encounter with the pathogen (1). Our hypothesis was that DCs of adult horses, but not of young foals, become activated upon R. equi infection to prime the effector cells of the acquired immune system. Susceptibility of young foals to R. equi may be associated with maturation of DCs in priming lymphocytes; therefore, immune competence of these cells may develop in the first few months of life. Our main objective was to measure the expression of molecules in APCs that are essential for antigen presentation and costimulation of lymphocytes and their priming into a Th1 type of response upon R. equi infection. These factors were measured in foals from birth to 3 months of age and in adult horses for comparison. In addition, we measured distinct effects of infection with virulent versus avirulent R. equi on cell activation.
These experiments were approved by the Cornell University Center for Animal Resources and Education and Institutional Animal Care and Use Committee for the use of live vertebrates in research. Forty-milliliter peripheral blood samples were collected from healthy adult horses (n = 8) and foals (n = 8) belonging to the Equine Park, Cornell University. Foal blood samples were collected in Vacutainer heparinized tubes within the first 5 days of life and monthly up to 3 months of age via jugular venipuncture (27, 40). Adult horse blood samples were collected once for comparison. The foals and their dams had access to pasture and stalls. The Equine Park does not have a history of R. equi disease in foals. All parturitions were observed, and the absorption of colostral immunoglobulins was assessed by using the SNAPTest (Idexx, Westbrook, ME) within 14 h of birth. Daily physical examination in the first month of life and complete blood cell counts were performed monthly to evaluate potential inflammatory or infectious conditions.
Peripheral blood mononuclear cells were isolated with single Ficoll-Paque Plus 1077 (GE Healthcare Bio-Sciences AB, Piscataway, NJ) density centrifugation at 700 × g for 15 min, washed with phosphate buffer solution (PBS; 300 × g for 10 min), resuspended in Dulbecco modified Eagle medium-F12 with 10% bovine growth serum (HyClone Laboratories, Inc., Logan, UT) plus antibiotics-antimycotics (Gibco-Invitrogen, Grand Island, NY), and incubated in tissue culture-treated plates (Becton Dickinson Labware, Franklin Lakes, NJ) for 2 h at 37°C in 5% CO2 (12, 48). The loosely adherent and nonadherent cells were removed by warm PBS washes. The number of cells left in the well was calculated by subtracting from the number of cells originally plated the number of nonadherent and loosely adherent cells removed by PBS washes. For the generation of DCs, monocytes were cultured in fresh medium with added recombinant equine IL-4 (10 ng/ml; generously provided by David Horohov, University of Kentucky) and recombinant human granulocyte-monocyte colony-stimulating factor (50 ng/ml; Genzyme, Boston, MA) for 4 days before infection. For the generation of macrophages, monocytes were cultured with no additives for 4 days before infection. Cells in tissue culture plates were incubated in fresh medium at 37°C in 5% CO2 for 4 days.
Virulent R. equi strain ATCC 33701, originally obtained from a pneumonic foal, or its avirulent, plasmid-cured derivative (kindly provided by Shinji Takai, Kitasato University, Japan) was transported in chocolate agar to our laboratory and subsequently cultured on blood agar at 30°C for 48 h. One colony was inoculated into 10 ml BBL Trypticase soy broth (Becton, Dickinson and Company, Sparks, MD) and incubated at 30°C. When the bacteria reached an optical density (OD) at 600 nm (Bio-Rad, Hercules, CA) between 0.4 and 0.5 (exponential growth phase), 100-μl aliquots were frozen in 10% glycerol (Amresco, Solon, OH) stocks at minus 80°C. For each of the experiments presented herein, a frozen bacterium aliquot was inoculated in 10 ml broth at 37°C and used when OD values reach between 0.4 and 0.5. The bacterial concentrations were estimated from OD values by using an equation previously determined from a polynomial curve fit between CFU and OD. We controlled for virulence plasmid retention with primers for the vapA gene and PCR (Cornell University Diagnostic Laboratory, Ithaca, NY).
The bacterial suspensions were added to mMO or mDC cultures at a multiplicity of infection of 5 bacteria per cell, and the cultures were incubated for 1 h at 37°C in 5% CO2. Noninfected cells were cultured under the same conditions for baseline. Extracellular bacteria were removed by three washes with warm PBS. The cell cultures were subsequently incubated with warm Dulbecco modified Eagle medium-F12 with 10% bovine growth serum and 20 μg/ml gentamicin (Invitrogen, Carlsbad, CA) to prevent extracellular growth of bacteria. Infected cells were harvested at 0, 8, and 24 h with the chelating agent EDTA (Gibco-Invitrogen). The time of harvesting was determined based on kinetic analysis (at 0, 8, 12, 16, and 24 h) for peak protein and gene expression in virulent R. equi-infected mMOs and mDCs and other publications (12, 17). Cells were pelleted by centrifugation at 500 × g and subsequently labeled for flow cytometric analysis or frozen at minus 80°C in Lysis Buffer RLT (Qiagen, Valencia, CA) for RNA expression experiments.
To confirm intracellular R. equi infection, virulent bacteria were labeled with carboxyfluorescein succinimidyl ester (CFSE). Briefly, pelleted bacteria were washed twice and resuspended in PBS (pH 7.8) with 3 μg/ml CFSE (Molecular Probes-Invitrogen, Eugene, OR). After a 30-min incubation in a rotator in the dark at 37°C, the bacteria were pelleted and washed twice with PBS (pH 7.2)-0.025% Tween 20 (Sigma-Aldrich, Saint Louis, MO). mMOs or mDCs were cultured with CFSE-labeled virulent R. equi bacteria (multiplicity of infection, 5 bacteria per cell) for 2 h to ensure that there was enough time for the bacteria to reach the intracellular space. For mMOs, adhering cells were washed twice with warm PBS to remove excess bacteria; the cells were subsequently removed from the plate with the chelating agent EDTA. For mDCs, only the cells in suspension were washed to remove excess bacteria by centrifugation at 300 × g. APCs were stained with Alexa 647-conjugated (Molecular Probes-Invitrogen) equine major histocompatibility complex (MHC) class II monoclonal antibody (Second Equine Leukocyte Antigen Workshop, WSII no. 43; clone cz11). After cell washes with PBS, CFSE-labeled R. equi and APCs were observed by confocal microscopy (Leica TCS SP5 confocal and multiphoton imaging with LAS AF software; Leica Microsystems, Bannockburn, IL) at appropriate wavelength settings.
Infected and noninfected mMOs or mDCs were tested for cell surface molecules at 24 h after infection with fluorophore-conjugated monoclonal antibodies for equine MHC class II (fluorescein isothiocyanate, clone cz11, as described above) and CD86 (phycoerythrin, clone 2331 monoclonal antibody; BD Biosciences, San Jose, CA) by two-color flow cytometry (12). After cell washes to remove excess reagents, leukocyte subpopulations were displayed in a dot plot and gated according to size based on forward light scatter and granularity based on 90° side light scatter with a FACScalibur (BD Biosciences). Cells compatible with APCs (based on size, granularity, and a CD172a positive marker) were gated for mean fluorescence expression analyses (12).
Isolation of total RNA from infected and noninfected mMOs and mDCs was performed with an RNeasy Mini Kit (Qiagen). The RNA product was treated with DNase (Invitrogen) to eliminate possible genomic DNA from the samples. Forward and reverse primers and fluorogenic probes (Applied Biosystems, Foster City, CA) were designed for the equine β-actin housekeeping gene, the cytokine IL-12 (IL-12p40 and IL-12p35), the costimulatory molecule CD40, and the transcriptional factor IRF-1 (interferon regulatory factor 1) (Table (Table1).1). The RNA expression of target genes was measured in stimulated and nonstimulated mMOs and mDCs with the ABI Prism 7500 Sequence Detection System (Applied Biosystems). For cytokine IL-12p35/40 and transcription factor IRF-1 expression, cells were tested at 0 and 8 h after infection. For costimulatory molecule expression, cells were tested at 0 and 24 h after infection. Reactions were performed in replicates in 96-well plates with the TaqMan One-Step RT-PCR Master Mix Reagents (Applied Biosystems) and standard thermal cycling parameters (12). The efficiencies of the target and reference gene amplifications were similar, with a linear curve slope of the log RNA amount versus delta cycle threshold (ΔCT; target gene CT minus housekeeping gene CT) of <0.1. Analyses of data were performed by normalizing the target cytokine amplification value (target gene CT) with the corresponding endogenous control (β-actin reference CT). The quantity of cytokine gene expression in each sample (n-fold difference) was calculated relative to the respective noninfected cells harvested at time zero.
The data for mMOs or mDCs from adult horses (n = 8) or foals at different ages (n = 8 at birth and 1 month of age; n = 6 at 2 months; n = 7 at 3 months) were summarized and are reported in dot plots with medians depicted by horizontal lines. Comparisons between APCs from adult horses and foals at independent time points (e.g., CD86 expression in virulent R. equi-infected mDCs from adult horses and foals at birth or CD86 expression on noninfected adult mMOs and mDCs) were analyzed by using nonparametric Wilcoxon rank sum tests because these data often had nonnormal distributions (49). The Kruskal-Wallis test was used for comparisons of three independent groups (e.g., CD86 expression in noninfected, virulent R. equi-infected, and avirulent R. equi-infected adult mMOs). Observations from adult horses and foals at different time points after different treatments were assumed to be independent because our hypothesis implied that foals were immunologically different at distinct ages. All descriptive and inferential statistical analyses were performed with commercially available software (KaleidaGraph version 4.01 [Synergy Software, Reading, PA] and SAS version 9.1 [SAS, Inc., Cary, NC]). The type I error rate was set at 1.3% after Bonferroni adjustment (P = 0.05/4) to account for the multiple comparisons.
Bacterial replication in broth demonstrated exponential growth between 8 and 20 h. The OD correlated well with the CFU values, with R = 0.98 measured by linear fit. The presence of the virulence plasmid in strain ATCC 33701 was confirmed by PCR amplification of the vapA gene after growth in broth at 37°C.
To confirm infection of mMOs and mDCs, virulent R. equi was labeled with CFSE and cultured cells were observed by confocal microscopy. After 2 h of culture and washes with PBS to remove extracellular bacteria, green fluorescent bacteria were seen inside cells (see Fig. S1 in the supplemental material). mMOs became more vacuolated and showed increased motility upon infection based on their elongated shape (in contrast to their rounded shape before infection) and filopodium formation. mDCs also became more vacuolated and detached from plate more easily (see Fig. S1 in the supplemental material).
MHC class II molecule expression on the surface of mMOs was comparable among noninfected and R. equi-infected mMOs from foals at all ages and those from adult horses, whether the cells were infected with virulent or avirulent bacteria (P ≥ 0.3) (Fig. (Fig.1,1, left). Although not statistically significantly different (P = 0.02 to 0.08), the median values of expression in adult horse mDCs were greater than those of expression in foals at all ages, independently of treatment (noninfected cells or virulent or avirulent R. equi-infected cells) (Fig. (Fig.1,1, right).
R. equi-infected mMOs and mDCs from foals of all ages had CD86 expression comparable to that in virulent (P ≥ 0.04) and avirulent (P ≥ 0.03) R. equi-infected cells from adult horses (Fig. (Fig.2,2, left). Virulent or avirulent R. equi-infected mDCs from foals of all ages had CD86 expression comparable to that in adult horse cells (P ≥ 0.2) (Fig. (Fig.2,2, right).
Virulent or avirulent R. equi slightly affected the expression of CD40 mRNA in mMOs relative to that in noninfected cells, and foal median values at different ages were comparable to adult horse median values (P ≥ 0.08) (Fig. (Fig.3,3, left). The expression of CD40 mRNA in virulent or avirulent R. equi-infected foal mDCs was comparable to that in adult horse mDCs (P = 0.04) but higher (P = 0.01) at 2 months of age when the cells were infected with virulent R. equi (Fig. (Fig.3,3, right).
Virulent or avirulent R. equi infection induced the expression of IL-12p40 mRNA in foal mMOs and mDCs at most ages (Fig. (Fig.4).4). Values in mMOs at birth and 3 months of age were greater than those in adult horse mMOs infected with virulent (P = 0.001 and 0.004, respectively) or avirulent (P = 0.003 and 0.0003, respectively) R. equi. Likewise, median values in mDCs at 2 and 3 months of age were greater than those in adult horse mDCs infected with virulent (P = 0.0001 and 0.0007, respectively) or avirulent (P = 0.0001 and 0.0007, respectively) R. equi. For mDCs, the comparison of values between adult horse cells and foals at birth and 1 month of age with the virulent R. equi infection resulted in P values of 0.06 and 0.03, respectively; they were 0.6 and 0.01, respectively, for infection with avirulent R. equi.
An effect of R. equi infection on IL-12p35 mRNA expression was more marked in foal mDCs at most ages than in mMOs (Fig. (Fig.5).5). The expression of this cytokine subunit in foal mMOs was comparable (P ≥ 0.03) to that in adult horse mMOs at birth and 3 months of age when cells were infected with virulent or avirulent R. equi. Foal mDCs infected with virulent or avirulent R. equi revealed greater expression of IL-12p35 mRNA at birth (P = 0.003 or 0.004, respectively) and 1 (P = 0.04 or 0.004, respectively), 2 (P = 0.001 or 0.004, respectively), and 3 (P = 0.008 or 0.01, respectively) months of age than did adult horse cells.
When comparing IRF-1 mRNA expression in adult horse and foal mMOs infected with virulent or avirulent R. equi, the P values were 0.1 and 0.2, respectively, at birth; 0.03 at 1 month of age; 0.01 and 0.008, respectively, at 2 months of age; and 0.03 at 3 months of age (Fig. (Fig.6,6, left). Foal mDCs infected with virulent or avirulent R. equi responded with greater IRF-1 mRNA expression at birth (P = 0.008) than did those of adult horses, and values were comparable at the other ages (P = 0.02 to 0.08) (Fig. (Fig.6,6, right).
As expected, and confirming the phenotypic differences between cells, noninfected and avirulent and virulent R. equi-infected adult horse mMOs had lower MHC class II expression than did mDCs (P = 0.007, 0.0003, and 0.0006, respectively). Importantly, this difference was not detected in foal cells. Although the difference was modest, the expression of IL-12p35 was greater in avirulent and virulent R. equi-infected adult horse mDCs than in mMOs (P = 0.003 and 0.0009, respectively) but not in foal cells. For IL-12p40 expression, avirulent and virulent R. equi-infected foal mMOs had greater expression than did mDCs at birth (P = 0.0006 and 0.0006, respectively) and 3 months of age (P = 0.0006 and 0.004, respectively).
In this study, we measured the direct effects of virulent R. equi or its avirulent, plasmid-cured derivative on mMOs or mDCs from foals at different ages (from birth to 3 months of age) and adult horses. Foal APCs at all ages responded to R. equi infection with comparable or higher expression of cytokines and costimulatory molecules that favor a Th1-type immune response. Nonetheless, only adult horse cells revealed a statistically significant difference in MHC class II expression between infected and noninfected mMOs and mDCs. This observation is important because the higher MHC class II molecule expression is a major factor that indicates mDC differentiation and maturation. Therefore, our hypothesis was partially rejected because even though foal APCs became activated upon infection with R. equi, further investigation is necessary to determine the function of MHC class II in this context.
Specific host factors that determine susceptibility to R. equi infection have not been fully identified in foals. Kanaly et al. (33, 34) have highlighted the importance of IFN-γ in the clearance of R. equi because mice treated with IFN-γ monoclonal antibodies failed to clear a pulmonary infection. One of the significant roles of IFN-γ is the enhancement of macrophage microbicidal activity via the secretion of both reactive oxygen and nitrogen intermediates into peroxynitrite (8, 37, 45, 50). Experimentally infected adult horses rapidly clear the organism without developing disease; the immune response includes an increase in both CD4+ and CD8+ T-lymphocyte counts in bronchoalveolar lavage fluid and a greater postinfection proliferative response to R. equi antigen or recombinant VapA in vitro (22). In addition, clearance of virulent R. equi from the lungs of adult horses is associated with IFN-γ production by CD4+ and CD8+ T lymphocytes, which suggests a recall immune response (23). Although R. equi replicates in macrophages, resistance to disease seems to be closely associated with T-cell function.
Primary activation of naïve lymphocytes, which would happen immediately after birth in foals exposed to R. equi, depends on antigen presentation of peptides and strong costimulation from professional APCs. Maturation of DCs is essential for the priming of T cells (32, 38, 43). Murine DCs become mature upon M. tuberculosis infection; i.e., they increase MHC class II expression for peptide presentation and upregulate the expression of costimulatory molecules and cytokines for T-cell activation (15, 31, 53). IL-12 produced by DCs upon encountering microbial products induces cell-mediated immunity with IFN-γ production by NK and T cells, promotes the differentiation of naïve CD4+ T cells into Th1, and increases the activity of cytotoxic T lymphocytes. Therefore, we investigated if mDC infection with R. equi would create the necessary signals to activate lymphocytes into a Th1 immune response. Our data indicated that foal APCs expressed high levels of the transcriptional factor IRF-1 and the cytokine IL-12p40/p35 upon R. equi infection. However, this activation response was not accompanied by robust expression of the APC MHC class II molecule, which calls into question the capacity of these cells to effectively signal to the acquired immune system in early life. This finding agrees with our previous study that described a physiologic increase with age in MHC class II molecule expression in foal APCs (12). Neither CpG-oligodeoxynucleotide nor R. equi affected the expression of this essential molecule, suggesting a greater threshold of maturation of APCs in the foal.
In order for DCs to initiate naïve T-cell responses against presented peptides (MHC-mediated signal 1), costimulation is required (signal 2). The interaction between the CD40 and CD86 molecules (expressed on APCs) and CD40L and CD28, respectively (expressed on T cells) stimulates kinase activation, gene expression, expression of surface molecules, and production of cytokines in both DCs and lymphocytes (51). R. equi infection slightly increased the expression of CD40 and CD86 in APCs. It is possible that R. equi infection alone, without the interaction with lymphocytes in vitro, was not sufficient to promote higher CD40 and CD86 expression in foal mDCs; in contrast, the infection of cells was sufficient to promote signaling for transcriptional factor IRF-1 and cytokine IL-12 expression (7). Partial activation of APCs may be achieved by contact with pathogens, yet appropriate expression of costimulatory molecules may require interaction with the effector cells. If this is true, this condition reveals the importance of cell-cell interaction and the activation threshold in early life.
IRFs may play an important role in DC differentiation and function during R. equi infection for T-lymphocyte priming and memory (13, 56). IRF-1 directly regulates the production of IL-12, promotes differentiation of lymphocytes into Th1 cells, and inhibits IL-4 transcription in lymphocytes (36, 62). In a positive feedback loop, IFN-γ that is produced by Th1 cells further increases the synthesis of IRF-1 in APCs, which extends the differentiation into Th1 cells. IRF-1 also augments the inducible nitric oxide synthase, which mediates upregulation of the IL-12 receptor in lymphocytes and the bactericidal activity in phagocytes (5, 54). Virulent or avirulent R. equi-infected foal mMOs and mDCs responded with high IRF-1 mRNA expression, which agrees with the observed high IL-12 expression. The IRF-1 response in foal APCs was most likely mediated by recognition of R. equi pathogen-associated molecular patterns via Toll-like receptors on the cell surface (2, 24, 35).
APC-derived IL-12 consists of two covalently linked subunits, p35 and p40, and belongs to a cytokine family that includes IL-23 and IL-27 (60). Synchronized expression of the p35 and p40 genes is crucial to form immunologically active IL-12p70 (42). Although several IRF-regulatory factors participate in the synthesis of IL-12p70 and IL-23, IRF-1 activation is required for IL-12p35 expression (20). Foal APCs responded with significant expression of IRF-1, IL-12p35, and IL-12p40, whereas adult horse cells revealed modest expression. In addition, the expression of IL-12p40 was greater in foal mMOs than in foal mDCs and further investigation is necessary to determine the mechanisms by which R. equi promotes contrasting cytokine responses in mMOs and mDCs and foal and adult horse APCs.
We did not identify a distinct difference in cell response between infections with virulent and avirulent R. equi, suggesting that plasmid-encoded products do not directly affect phagocytic cell expression of molecules essential for antigen presentation and costimulation of lymphocytes into a Th1 type of response. Our study agrees with a previous report that demonstrated no differences in cytokine production in murine macrophages infected with virulent or avirulent R. equi (17). Likewise, virulent R. equi infection still induced an immune response characterized by translocation of NF-κB into the nucleus and subsequent secretion of the cytokines tumor necrosis factor and IL-12, despite the fact that macrophages failed to impede intracellular bacterial replication (9, 61).
An age-dependent IFN-γ production measured in foal peripheral blood mononuclear cells has been suggested as a condition of susceptibility to diseases (2). In contrast, Hietala and Ardans (21) demonstrated that macrophages from R. equi-exposed foals had a 100% increase in their R. equi killing capacity when cocultured with autologous lymphocytes (activated during infection), whereas macrophages cocultured with lymphocytes from nonexposed foals showed only a 50% increase. Other studies support the latter observation, including the response of bronchial lymph node lymphocytes from foals with a significantly higher ratio of IFN-γ to IL-4 than adult horses when stimulated with recombinant Vap proteins (19, 28). Perhaps individual foals do have a delayed IFN-γ response that increases their susceptibility to disease in early life; however, once the infection is established, foals seem to respond with a robust Th1 immune response (11, 29). It is important to consider that a protective immune response is the result of combination and interaction among the many arms of the immune system (innate and acquired mechanisms). Indeed, foal mMOs revealed a measurable response to R. equi infection and other studies have shown their competence at interacting with cytotoxic lymphocytes (44, 64).
In summary, our study suggests that foal mDCs responded to R. equi infection with robust expression of both IL-12 subunits, likely coordinated by IRF-1 expression. Nevertheless, there seems to be a limitation to MHC class II molecule expression and perhaps additional activation stimuli are necessary to promote efficient maturation. If lymphocytes are responsible for the source of IFN-γ for the activation of macrophages and the maturation of DCs, the understanding of the mechanisms of activation of lymphocytes during R. equi infection in early life is critical. An individual delayed efficiency in lymphocyte priming would agree with the individual temporal risk of disease in early life.
We thank Carol Collyer and staff at the Cornell University Equine Park for facilitating the handling and care of foals. We are also grateful to David Horohov, Gluck Equine Research Center, University of Kentucky, and Shinji Takai, Kitasato University, Japan, for sharing their reagents.
This study was supported by the Harry M. Zweig Memorial Fund for Equine Research.
None of us have commercial affiliations or consultancies, stock or equity interests, or patent-licensing arrangements that could pose a conflict of interest regarding this report.
Published ahead of print on 24 December 2008.
†Supplemental material for this article may be found at http://cvi.asm.org/.