γδ T cells are the majority peripheral blood T cells in young cattle. The role of γδ T cells in innate responses against infection with foot-and-mouth disease virus was analyzed on consecutive 5 d following infection. Before infection, bovine WC1+ γδ T cells expressed a nonactivated phenotype relative to CD62L, CD45RO, and CD25 expression and did not produce IFN-γ ex vivo. Additionally, CD335 expression was lacking and no spontaneous target cell lysis could be detected in vitro, although perforin was detectable at a very low level. MHC class II and CD13 expression were also lacking. Following infection with foot-and-mouth disease virus, expression of CD62L and CD45RO was greatly reduced on WC1+ γδ T cells, and unexpectedly, CD45RO expression did not recover. A transient increase in expression of CD25 correlated with production of IFN-γ. Expression of CD335 and production of perforin were detected on a subset of γδ T cells, and this correlated with an increased spontaneous killing of xenogeneic target cells. Furthermore, increased MHC class II expression was detected on WC1+ γδ T cells, and these cells processed protein Ags. These activities are rapidly induced, within 3 d, and wane by 5 d following infection. All of these functions, NK-like killing, Ag processing, and IFN-γ production, have been demonstrated for these cells in various species. However, these results are unique in that all these functions are detected in the same samples of WC1+ γδ T cells, suggesting a pivotal role of these cells in controlling virus infection.
There are two major types of T lymphocytes in vertebrates, those that express the αβ TCR and those that express the γδ TCR on their cell surface. In young ruminants, the percentage of cells expressing the γδ TCR is considerably higher (50–60% of PBLs) compared with other mammalian species. What human neonates lack in number of circulating γδ T cells compared with calves may be made up by their activation state (1). However, these cells reduce in number as the animals age and may ultimately account for 5–25% of circulating T lymphocytes (2), more similar to humans (1–20%) (3). The role of these cells in the response to the acute infection caused by foot-and-mouth disease virus (FMDV) is poorly understood.
γδ T cells in the bovine can be divided into two major subsets by expression of the WC1 molecule, a member of the scavenger receptor cystein-rich family of proteins. These subsets express either WC1+/CD3+/CD5+/CD2−/CD6−/CD8− cell surface proteins or WC1−/CD3+/CD5+/CD2+/CD6+/CD8+ (4, 5). Within the WC1+ population of γδ T cells, there are two subpopulations expressing different isoforms of the WC1 molecule, as follows: WC1.1 and WC1.2. The WC1.1+ γδ T cells appeared to be the main producers of IFN-γ, whereas the WC1.2+ γδ T cells have been reported to proliferate in response to stimulation with mitogens (6, 7). Although the two populations can be distinguished, there is little information to conclude that they differ functionally. However, several reports have pointed out differences in tissue distribution among γδ T cells of the skin and the intestinal mucosa (8, 9). γδ T cells express the CD3 molecule, but stimulation with anti-CD3 Ab showed only a small population that responded by proliferation, further suggesting a difference in subpopulations (10).
Although human and mouse γδ T cells do not share phenotypic properties with bovine γδ T cells, such as expression of WC1 molecule, studies of human γδ T cells show those cells expressing Vγ2 were capable of lysing HSV type 1 (HSV-1) or vaccinia virus–infected Daudi cells (11). Interestingly, blocking CD3 or the γδ TCR with specific Abs led to abrogation of cytotoxicity, although their cytotoxicity was reported to be independent of MHC I or II restriction, indicating this subset of γδ T cells may function like NK cells. In cattle, studies have shown a dramatic increase in γδ T cells in the peripheral blood of cattle within the first days of infection with bovine herpesvirus type I. Furthermore, cattle vaccinated against FMDV were reported to have increased numbers of γδ T cells that adopted a NK-like phenotype and, like human cells, were capable of lysing virus-infected target cells (12). Similarly, in bovine viral diarrhea virus infection, γδ T cells increased 100-fold (13).
A protective role of γδ T cells was demonstrated by Sciammas et al. (14) in a mouse model using HSV-1 infection as γδ-deficient mice succumbed to infection. An interesting observation in these studies was demonstration of Ag-specific reactivity of γδ T cells, indicating that the responses observed were a result of direct recognition of unprocessed HSV glycoprotein. Recently, following extensive description of CD4+/CD25high/Foxp3+ regulatory T cells in mice and humans, a similar function has now been ascribed to subpopulations of bovine WC1+ γδ T cells as opposed to bovine CD4+/CD25high/Foxp3+ cells (15). Taken together, γδ T cells have functions that are well described in other T cell subsets. However, γδ T cells are largely ignored in considerations of vaccine design. Understanding how these functions are executed in γδ T cells may indicate immunological potential of these cells in vaccination against various pathogens.
Foot-and-mouth disease, caused by FMDV, is a devastating disease of domesticated ruminants with large economic consequence. Although vaccination is available, there are drawbacks, including the lack of tests to distinguish between vaccinated and infected animals and short duration of immunity. Different vaccination strategies could be developed, but this requires a thorough understanding of the immune response against FMDV in cattle. In our studies, we have initially aimed at understanding the response of bovine WC1+ γδ T cells during the early phase of infection with FMDV. In a series of experiments, we now show that the γδ T cell subpopulation WC1+ is activated shortly postinfection with FMDV, strain O1 Manisa. Interestingly, these cells show an increased production of IFN-γ and were capable of lysing a tumor cell line in spontaneous cytotoxicity assays. Furthermore, a small subpopulation of these cells activates MHC class II proteins and is capable of processing protein Ags. All of these activated phenotypes are detectable only during acute infection, days 2–4 following inoculation of virus. Therefore, there may be multiple roles for γδ T cells in the acute-phase infection with FMDV, including induction of both innate and adaptive immune responses. Whether such activation status translates into a protective function of γδ T cells during infection with FMDV is now of great interest.
Materials and Methods
Holstein cattle aged 3–5 mo were purchased from Thomas Morris (Springfield, PA) and acclimated for at least 7 d before use in the experiments. Animals were tested in groups of three and were age, weight, and gender matched. Before infection, baseline analysis was performed. All animal protocols were reviewed and procedures were approved by the Institutional Animal Care and Use Committee.
The O1 Manisa strain of FMDV was used in all studies described. The O1 Manisa was initially harvested from vesicles of infected cattle, titrated, and kept at −70°C until use. Animals were first sedated, and 1 × 104 50% tissue culture-infective dose was used to infect each animal. The dose was divided into four parts and inoculated in four spots intradermal lingually, at the base of the tongue. Animals were examined daily for development of lesions.
Peripheral blood was drawn into vacutainer tubes from the jugular vein. Blood samples for cell separation were drawn into tubes containing sodium heparin. For cell separation, blood was diluted 1:1 with PBS in 50-ml conical tubes, underlayed with Lymphoprep (Axis-Shield, Oslo, Norway), and centrifuged at 1000 × g for 20 min. PBMCs were collected and washed at least three times in PBS to remove all platelets and finally resuspended in RPMI 1640 supplemented with 10% FBS and used immediately or were resuspended in sorting buffer (0.5% BSA, 2 mM EDTA in PBS [pH 7.2]) for cell separation.
In assays requiring isolated populations of WC1+ cells, PBMCs were incubated with anti-WC1 mAb (IL-A29; VMRD, Pullman, WA) for 20 min on ice, followed by a single wash in sorting buffer, and later incubated with anti-mouse IgG1 microbeads (Miltenyi Biotec, Gladbach, Germany) for another 20 min on ice. Finally, the cells were washed three times in sorting buffer and sorted through MACS columns. Cells were washed once in PBS and resuspended in RPMI 1640. In some assays, positively sorted cells were rested for 18 h before use.
Cell surface marker staining
To enumerate the γδ TCR+ cells in total PBMC, cells were labeled with anti-TCR δ-chain Ab (TcR1-N24, GB21A; VMRD), followed by anti-mouse IgG2b (BD Biosciences, San Jose, CA). To assess the expression of CD13, CD25, CD45RO, CD62L, CD335, and MHC II, cells were first labeled with primary Abs, anti-CD13, anti-CD62L (IgG1; AbD Serotec, Oxford, U.K.), anti-CD25, anti-CD45RO (IgG1; VMRD), or anti-CD335–Alexa Fluor 488 (AbD Serotec) for 20 min on ice, and then counterstained with fluorochrome-conjugated anti-Ig. Cells were finally fixed in 2% paraformaldehyde for 20 min on ice, washed, and resuspended in FACS buffer (0.3% BSA, 0.9% sodium azide, PBS). One hundred thousand events were acquired using the FACSCalibur and analyzed in CellQuest (BD Biosciences, San Jose, CA).
Intracellular cytokine measurement
Sorted WC1+ cells were incubated with 3 μg/ml brefeldin A for 5–6 h in RPMI 1640. Later, the cells were stained for IFN-γ or perforin using a protocol described previously (16), with the exception that anti-bovine IFN-γ (MCA1783PE; AbD Serotec) was used. Anti-human perforin Ab was used to detect bovine perforin (clone δG9; BD Biosciences). One hundred thousand events were acquired using the FACSCalibur and analyzed in CellQuest (BD Biosciences).
Spontaneous killing assay
To learn whether WC1+ cells were capable of spontaneously lysing a tumor cell target, we plated sorted WC1+ cells with K562-GFP cells (NK cell-sensitive tumor cell line) in 96-well plates together with 7-amino actinomycin D (live/dead discriminating dye; BD Biosciences) at E:T ratio of 50:1. PBMC isolated from the same animals and stimulated in vitro with human IL-2 and IL-12, a method previously reported to activate NK cells (17), were included as positive controls for K562-GFP killing. Cells were incubated for 4 h at 37°C. Later, 100,000 events were acquired using a FACSCalibur and corrected, as described earlier (16).
WC1+ cells were tested for their capability to respond to a mitogen during infection with FMDV. Sorted WC1+ cells were labeled with 5 μM CFSE for 15 min at room temperature, followed by addition of an equal volume of FBS to stop the labeling, and washed twice with PBS. A total of 1 × 104 cells was incubated with 5 μg/ml Con A for 72 h. The degree of proliferation was measured by flow cytometry (FACSCalibur; BD Biosciences) and analyzed in WinMDI version 2.9 (Joseph Trotter, http://facs/scripps.edu/software.html). One hundred thousand events were acquired and analyzed.
The capacity of WC1+ cells to ingest and process Ag was tested on sorted cells. DQ-OVA (Invitrogen, Carlsbad, CA) was added at 2 μg/ml to 106 cells and incubated at 37°C for 2 h. Control cells from both infected and noninfected animals were incubated on ice for 2 h. Cells were washed with PBS and resuspended in FACS buffer, and data were acquired in FACSCalibur (BD Biosciences) and analyzed in WinMDI version 2.9 (Joseph Trotter, http://facs/scripps.edu/software.html). One hundred thousand events were acquired and analyzed.
Where appropriate, the data were evaluated statistically with Student t test, and p values ≤0.05 were regarded as significant.
Percentage of γδ TCR-expressing cells in peripheral blood of cattle following infection with FMDV O1 Manisa
To examine the effect of virus infection on bovine γδ T cells, we initially looked at the number of γδ TCR-expressing cells as a percentage of total PBMCs. To this effect, cells were labeled with γδ TCR and analyzed by flow cytometry. Generally, there were only slight differences in the percentage of γδ TCR+ cells within the first 2 d following infection (Fig. 1). The only significant increase in γδ TCR+ cells was observed on day 3, and was statistically different (p = 0.016) from values generated on day 0 (before infection). This increase in cell number partially reflects the increase in activation of bovine γδ TCR+ cells in the early phase of infection, which is shown in the following analyses. No peripheral blood lymphopenia is observed during FMDV infection in cattle such that this is not due to a change in total lymphocyte counts. Contrarily, there is a marked lymphopenia in swine during the acute stage of infection, as we have previously reported (18).
Analysis of the proliferative response of WC1+ cells was also assessed. WC1+ cells isolated from the peripheral blood of cattle infected with FMDV were still responsive to the mitogen, Con A, showing normal in vitro proliferation equivalent to blood samples taken before infection (Supplemental Fig. 1).
Expression of selected cell surface markers following infection with 01 Manisa FMDV
We sought to determine any changes that FMDV may induce regarding expression of surface proteins associated with activation of WC1+ γδ T cells. Cells were enriched by positive selection for WC1 expression, and subsequent populations analyzed expressed both γδ TcR and WC1 on 96–99% of the cells (Supplemental Fig. 2). The cells were labeled with Abs against CD25, CD62, and CD45RO. As opposed to other T cells, there was a constitutive background staining of CD25 on WC1+ cells. Expression of CD25 (Fig. 2) on day 0 was at least 15–25%, but an increase was observed on day 1–3 postinfection. The increase, particularly for day 2, was statistically significant compared with cells on day 0 (p = 0.0106 in the first experiment; p = 0.02 in second experiment [day 3]; and p = 0.012 in the third). Increased expression of CD25 usually indicates activation of T cells.
CD62L is highly expressed on freshly isolated WC1.1+ γδ T cells (Fig. 3) (19). Following infection with FMDV, expression of CD62L was reduced on days 1–3 (Fig. 3), reaching the lowest expression level on day 2 in all three experiments that were done. This change in expression of CD62L, particularly on day 2, was statistically significant in comparison with the cell surface expression of CD62L on WC1+ cells on day 0 (p = 0.04, experiment 1; p = 0.32, experiment 2; and p = 0.02, experiment 3). By day 4, the expression levels had reached those on day 0 before infection with O1 Manisa FMDV. A similar downregulation pattern of CD62L expression has been reported in other species following virus infection.
CD45RO had a rather surprising expression pattern. WC1+ from noninfected animals expressed well over 80% of CD45RO. WC1+ cells from FMDV-infected animals drastically reduced expression of CD45RO (Fig. 3). Moreover, on days 2–5, virtually no cells expressed CD45RO. Recovery of CD45RO resumed by day 4 in experiments 2 and 3, but did not reach the levels observed before infection in the first trial. These data agree with previous reports analyzing bovine γδ T cells in viral infection, although we only followed expression for 1 wk, whereas these authors analyzed the development of a γδ T cell memory phenotype over many weeks (13). In contrast, in vitro results reported by Bembridge et al. (20) show a reduction of CD45RO on activated T cells in animals, but expression on WC1+ was intact in these secondary in vitro activation assays.
Activation to IFN-γ production
The percentage of WC1+ cells staining intracellularly for IFN-γ increased during acute infection with FMDV, peaking 2–3 d postinfection (Fig. 4). There were large variations in individual animal responses, and statistical significance could be demonstrated only in experiment 1 (Fig. 4, p = 0.031, day 2). Remarkably, WC1+ cells expressing the antiviral cytokine are not detected 5 d postinfection, indicating these cells go through a very rapid, but short-lived activation. Alternatively, activated cells may be migrating out of the circulation by 4–5 d postinfection.
NK cell-like cytolytic activity of bovine γδ T cells
NKp46 (CD335) is an activating receptor expressed on NK cells. These innate responding cells kill virus-infected and neoplastic cells in an Ag-independent manner via these receptor molecules. We tested for expression of NKp46 by WC1+ cells during the course of acute infection with FMDV. In the 5 d analyzed, a small population of these γδ T cells expressed the CD335 receptor, between 4 and 6% of WC1+ cells. Detection of CD335-expressing cells was very transient, on days 2 and 3, returning to preinfection levels by day 5 postinfection (Fig. 5). Production of the killer cell protein, perforin was also detected in a similar sized population concurrent with the detection of CD335 expression (Fig. 5). Gating on the WC1+ cells and analyzing CD335 and perforin expression shows that the cells expressing CD335 and perforin are the same population when we compare preinfection to 2 d following infection (Fig. 6). This is the peak of expression for both proteins, as shown in Fig. 5. This is a strong indication that these cells are capable of NK-like killing.
Given these results, we tested for killing activity against a tumor cell line, K562. In this study, we show sorted WC1+ cells’ capacity for killing K562, stably transfected with GFP (K562-GFP). In the two experiments, γδ T cells from all animals developed the capacity to kill target cells. The killing capability was highest on the second day following infection, and the difference in killing activity on day 2 differed significantly from activity observed before infection (p = 0.019, p = 0.013, and p = 0.041 for experiments 1, 2, and 3, respectively; Fig. 7). Attempts to develop a killing assay for foot-and-mouth disease virus-infected cells have to date been unsuccessful as even attenuated isolates of the virus were shown to be cytopathic in any bovine cell line we have analyzed.
Expression of cell surface proteins associated with Ag presentation
We analyzed the expression of MHC class II and show a baseline staining of 1–1.5% evident on day 0 that increased appreciably between days 1 and 3 postinfection with O1 Manisa FMDV (Fig. 8). This was significantly higher compared with day 0 (p = 0.04 in experiments 1 and 2). The expression gradually reached the preinfection levels by day 5. This increase in MHC class II expression indicated an activated status of WC1+ cells during infection with FMDV. Another unusual observation was the appearance of CD13 on a small population of WC1+ γδ T cells postinfection with O1 Manisa FMDV (Fig. 8). A considerably high level of expression was observed on consecutive 3 d beginning from day 1, although the expression level was not uniform in all animals, particularly in experiment 1. The expression reduced to background levels or was completely absent beginning from day 4. We do not know the function of CD13 on bovine WC1+ γδ T cells, but this molecule is expressed on splenic dendritic cells in cattle (21).
Ag processing by bovine γδ T cells during infection
In other species, γδ T cells have been reported to have functions of Ag processing and presentation (22–24). Data shown in Fig. 8 led us to examine this function by adding self-quenched DQ-OVA to bovine WC1+ cells and measuring the fluorescence resulting from cleavage of DQ-OVA by WC1+ cells. The DQ-OVA was successfully cleaved, resulting in fluorescence of the BODIPY FL dye. We observed a slight increase in fluorescence intensity of cells from noninfected animals on day 0 (Fig. 9), but the highest increase in fluorescence intensity was demonstrated by WC1+ cells originating from animals sampled 3 d following infection with FMDV (Fig. 9).
The relative abundance of γδ T cells in young ruminants has been well documented, although the role of these cells in viral infections remains poorly defined. In this study, we show that these cells expand and are activated during infection of Holstein calves with FMDV, strain O1 Manisa. In previous studies of pigs (18, 25), infection with FMDV caused immunosuppression such that lymphocytes isolated from such animals failed to respond to mitogens. We tested this aspect in bovine WC1+ cells and did not observe any profound changes in the proliferative capacity of γδ T cells from infected animals, besides a slightly increased reactivity to Con A on the second or third day of infection. This may suggest activation of γδ T cells following infection with O1 Manisa FMDV. A clearly activated status of bovine γδ T cells was shown by Amadori et al. (12) in their experiments with bovine herpes virus, in which a sharp increase of γδ T cells was detected on the first day postinfection. We show increase in the relative percentage of γδ TCR-expressing cells only on day 3 postinfection with FMDV relative to preinfection.
Analysis of cell surface molecule expression on these cells revealed low-level constitutive expression of CD25 on nonstimulated cells expressing WC1 isolated from PBMC of naive animals, as was reported in studies by Rogers et al. (7). These investigators showed an increased expression of CD25 following stimulation. In our analysis of animals infected with FMDV, we show a similar finding, WC1+ cells increased CD25 expression, indicating activation of these cells in vivo following viral infection.
In this study, we also report a detected loss of CD45RO on γδ T cells beginning from the second day following infection with FMDV. These results are similar to those reported for bovine viral diarrhea virus, although with different kinetics (13). Because currently we do not know whether the γδ TCR recognizes any foot-and-mouth disease viral Ags, we cannot speculate on the Ag-specific stimulation of γδ T cells in this study. However, antigenic stimulation of αβ T cells has been reported to lead to downregulation of CD45RO (20).
CD62L is a homing receptor that directs T cell trafficking into lymph nodes. In the current study, we show a reduction in CD62L expression that is most pronounced 2 d following infection, followed by a rapid recovery of CD62L expression, similar to what has been reported during bovine leukemia virus (26) or Mycobacterium bovis (27) infection. In contrast to our finding in this viral infection, Walcheck and Jutila (28) found a much slower downregulation of L-selectin following short-term activation of γδ T cells in vitro with PMA. In that model, they demonstrated that bovine γδ T cells do not appreciably extravagate from blood into parenchyma of lymph nodes. Our results indicate a transient activation of γδ T cells during acute infection with FMDV, and are remarkable not only for the rapid return to CD62L high expression, but that a majority of the circulating WC1+, γδ T cells show this phenotype. Likewise, the rapid increase in CD25 expression involves most WC1+ cells and all correlating with the peak of viremia. However, as these surface molecules are predicted to reflect activation state of the cells, it is interesting that only a small percentage (4–8%) of these γδ T cells are shown to express intracellular IFN-γ, albeit the kinetics of expression correlate to CD25 expression and inversely to CD62L.
CD2 is not normally found on WC1+ cells; however, Daubenberger et al. (29) reported that CD2 could be detected when WC1+ cells were cultured in the presence of Theileria parva. This was concordant with WC1+ cells acquiring a lytic phenotype and capacity to destroy T. parva– infected target cells. The role of CD2 in the target cell recognition was not defined. CD335 is a typical marker expressed on NK cells, including bovine NK cells, where it functions as a natural cytotoxicity-triggering molecule (30, 31). Therefore, expression of certain markers may be a result of the stimulus present in the environment that probably contributes to the activation status of cells involved (32). We did not detect CD335 on WC1+ cells from healthy animals. Following infection with O1 Manisa FMDV, a transient expression of CD335 was induced on WC1+ γδ T cells. In addition, we detected expression of intracellular perforin in these same cells with nearly identical kinetics and further show expression of these proteins is by the same subset of WC1+ cells. Increased expression of CD335 may have contributed to acquisition of the cytolytic function of WC1+ γδ T cells following infection with FMDV in our studies, and the capacity to kill tumor target cells peaked 2 d postinfection and was lost by day 5 after inoculation, identical kinetics to CD335/perforin expression.
The NK-like cytotoxic phenotype of γδ T cells in bovine was earlier reported by Brown et al. (33) in studies of infection with Babesia bovis. They showed that spontaneous killing was exhibited when parasite-stimulated WC1+ cells were presented with a xenogeneic NK-sensitive target cell line. More relevant to the current study, Amadori et al. (12) showed activation of bovine γδ T lymphocytes following vaccination of calves against FMDV. In those studies, γδ T cells were capable of lysing target cells in a manner similar to NK cells, indicating that either a spontaneous killing capability develops in this population of T cells or their reactivity could be driven by specific viral Ag recognition. The increase in killing capability observed in WC1+ cells in our study coincided with the presence of intracellular perforin, which indicated that the killing mechanism most likely involved perforin, as previously shown by Alvarez et al. (34) in SCID-bo mice. Perforin-bearing γδ T cells have also been reported in American bison with experimental sheep-associated malignant catarrhal fever (35).
Other viruses have been reported to use CD13 as a cell entry receptor (36, 37). Although CD13 is a receptor for porcine endemic diarrhea virus, permissiveness of porcine cell lines expressing CD13 is low (38). In the present studies, CD13 is not expressed on WC1+ cells, but is activated during acute infection with FMDV. CD13 has not been reported as a receptor for FMDV, and it is unlikely that induction of CD13 leads to γδ T cell infection, as only 2–4% of WC1+ cells express CD13. We did not test the WC1+ cells for the presence of newly synthesized viral proteins, but previous studies have shown FMDV does not infect lymphocytes in vivo. Conversely, CD13 is expressed on splenic dendritic cells in cattle, although no role in dendritic cell function such as Ag presentation was investigated (21).
Our observation that 5% of WC1+ cells also activated expression of class II MHC led to the hypothesis that these cells may be capable of Ag processing. Although all WC1+ cells from infected and noninfected animals ingested DQ-OVA in an Ag-processing assay, the fluorescence intensity was significantly more in cells originating from infected animals on day 3. This may indicate that this function does not require prior activation of γδ T cells and that virus infection does not adversely affect this function of γδ T cells, but rather enhances the cleavage of DQ-OVA, predicting enhanced Ag-processing capacity. The property of ingesting particulate matter was demonstrated for dermal γδ T cells in cattle (39). However, those γδ T cells could not be induced to produce IFN-γ, nor did they express MHC II or CD25. The γδ T cells may also be involved in Ag presentation independently from dendritic cells. In work performed by Howard et al. (40), WC1+ cells could not be stimulated to proliferate by dendritic cells, implying the lack of ligands for γδ T cells on allogeneic dendritic cells or a lack of costimulatory molecules for γδ T cells.
Juleff et al. (41) recently reported infection of two cattle with FMDV following depletion of WC1+ cells in vivo with Ab administered i.v. The authors showed lack of WC1+ cells in the peripheral blood did not have a measurable effect on the course of infection, concluding that WC1+ do not have a major role in the control of FMDV. However, in our studies, we performed functional analysis of WC1+ cells and found evidence of activation of this population of γδ T cells following infection with FMDV. These results are most likely not contradictory, as in gene knockout studies, if the WC1+ cells were eliminated, the breadth of immune responses to virus often compensates for lacking a particular function. The Juleff study did not include multiple Ab treatments to address what happens when multiple responding cell populations are blocked/eliminated in the same animal, most likely due to the very high cost of such a study. We performed no pretreatment of the young, healthy steers used in the current study, so this is a direct assessment of how FMDV infection alters the function of WC1+, γδ T cells.
Others have reported that it has been difficult to generate Ag-specific γδ T cells following vaccination (42). Reports of γδ T cell memory induction in the bovine (13, 43) may indicate that, in this species, specific Ag-dependent stimulation of γδ T cells may exist. It cannot be ruled out that γδ T cell responses at such an early stage as reported in this study result from nonspecific stimulation by endogenously derived ligands. Furthermore, data presented in this work raise the possibility that the γδ T cells can function as APCs. If confirmed, these cells could provide very early and efficient activation of CD4 T cell responses, initiating the classic adaptive immune response. Further studies are needed for clarification of whether such a response is linked to generation of adaptive and later memory immune responses against FMDV.
This rapid response to viral infection suggests that γδ T cells have a role in the innate immune response to viral infections and should be investigated further to understand the basis of this stimulatory effect. In FMDV infection in particular, viral clearance is most likely mediated by many aspects of innate and adaptive immune responses. Minimally, there is a significant type I IFN response mediated by various dendritic cell populations (44, 45), an early Ab response (41, 46, 47), and the γδ T cell response reported in this study. We suggest that all of these responses can work concurrently, consistent with the observation that the high viral load is cleared so early and effectively. Manipulation of these lymphocyte populations may be a way to develop biotherapeutics for the protection of animals against FMDV infection in advance of the development of adaptive immunity following vaccination.
Finally, previous studies in various species have shown the ability of γδ T cells to exert all of the functions we have characterized in this work, including expression of an activated cell phenotype, production of IFN-γ, expression of NK cell phenotype, production of perforin, Ag-nonspecific target cell killing, expression of MHC class II proteins, and Ag processing. This is the first demonstration of all of those functions ex vivo from single individuals and consistent over three trials and nine cattle. All of these activities of peripheral blood γδ T cells from FMDV-infected cattle occurred rapidly following infection, within 3 d, and were all no longer detected by 5 d postinfection in the circulation. This significant finding indicates a prominent role for activated γδ T cells among the various innate and adaptive immune responses that rapidly control this acute viral infection of livestock.
The authors have no financial conflicts of interest.
We thank the animal care staff at Plum Island Animal Disease Center for professional assistance with the cattle.
The online version of this article contains supplemental material.
Abbreviation used in this article:
- foot-and-mouth disease virus.
- Received November 1, 2010.
- Accepted February 10, 2011.