Early studies by Cunliffe et al.  established the long duration of immunity following infection with FMDV in cattle. In contrast, current commercial vaccination protocols require regular re-inoculation to maintain immunity. Analysing the differences in the immune response in the natural host during these two processes may lead to the design of more effective vaccines.
By exposing six cattle to FMDV this study aimed to; investigate whether a generalised suppressive state is induced, compare early T cell responses to those of vaccinates, and examine the expression of key immune-modulatory cytokines.
After natural exposure to FMDV, viraemia is typically detectable within 2 to 3 days followed by clinical signs . Production of high titres of neutralising antibodies can be detected as early as six days following natural infection, by a process shown previously to be T cell independent in cattle . These previous observations were confirmed in all of the cattle used in this study. Compared to the rapid production of antibody, significant FMDV-specific T cell responses (assayed by proliferation of PBMC) were not observed following natural infection. These data are supported by other cattle studies, with no FMDV-specific T cell response detectable up to 32 days post exposure in some cases (M. Windsor unpublished data). We also investigated whether there was any loss of mitogen or specific antigen induced T cell proliferative responses during the course of acute infection. Abrogation of proliferation of PBMC to mitogen in pigs occurs between 2 and 8 days but most noticeably at day 5 after needle challenge . Following natural challenge, peak viraemia in cattle occurs at least two days later than needle challenge , which is consistent with the pattern of viraemia in our study. We did not observe any loss of mitogen or specific antigen induced T cell proliferative responses on day 6 post-infection. Even though a specific T cell response to FMDV could not be detected during the acute stages of infection, the T cell proliferative responses to mitogen and a third party antigen (BHV) were unaffected.
In a previous study using bovine dendritic cells (DC), Robinson et al.  showed that, in vitro, DC infection was enhanced by the formation of FMDV immune complexes. These data suggest that as FMDV rapidly loses its ability to infect susceptible cells due to increasing neutralising antibody titres, it gains the capacity to infect immune cells via CD32 (FcγR). These findings by Robinson et al. suggest that infection and killing of DC by immune-complexed FMDV may be responsible in part for the delayed FMDV-specific T cell proliferative response . Clearly, it is not the ability of the T cells to respond that is affected, as evidenced by the maintenance of established T cell responses to mitogen and third party antigen.
The FMDV-specific T cell proliferative response in the five vaccinates was variable. Where an effective response to antigen was seen, it was of a high magnitude and detectable from seven days post vaccination. It should be noted that although the initial response was low in some cattle, they all subsequently went on to develop a similar magnitude of T cell response following a booster vaccination, showing that there was no underlying impairment of the response in these animals.
Transient lymphopenia has been observed in swine following infection with FMDV which correlated with the virulence of the isolate used . In cattle, we were unable to detect any evidence of lymphopenia, with the numbers of leucocytes remaining constant and within physiological ranges during the period of viraemia and clinical FMD. Leukocyte subsets were also examined in whole lysed blood in two of the animals and no changes were observed, confirming our previous observations . This contrasts studies in pigs, where a significant loss of circulating CD4+ and CD8+ T cells was observed .
Studies in pigs have found high levels of type 1 IFN following experimental infection with several FMDV isolates [10, 36], which correlates with immune suppression. Due to the different methods of quantification used in these studies, it is problematic to compare the levels of IFN produced. In cattle we have measured the specific activity of type 1 IFN (in international units), whereas total circulating type 1 IFN protein was measured in pig serum. However, this IFN is not necessarily biologically active. Several sources, including manufacturers' data sheets, give type 1 IFN specific activity (cross species) at between 1 and 3 × 108 IU/mg of total IFN. Using this calculation we can speculatively compare type 1 IFN in the sera of pigs and cattle during FMDV infection, and postulate that pigs do indeed produce more IFN than the cattle in this study. The most compatible study in pigs was performed by Nfon et al., who used O1Campos challenge . At least 9 fold more type 1 IFN was produced in pigs compared to our results in cattle. In pigs, IFN clearly has an important role to play in the resolution of infection, as the administration of type 1 IFN in a viral vector protects against subsequent challenge with FMDV .
Our study implies that cattle produce very little type 1 IFN in comparison to pigs, which is in keeping with previous studies , yet cattle still resolve the disease effectively. Depletion of CD4+ cells in cattle during infection further reduces circulating type I IFN, but infection is still resolved , indicating that the presence of type 1 IFN in the circulation may not have any bearing on the resolution of disease. The IFN detected in the circulation during FMDV infection in cattle is thought to be produced by CD4+ PDCs interacting with immune-complexed virus . The PDCs found in cattle secondary lymphoid tissue are capable of producing large amounts of IFN in vitro . It is possible therefore, that IFN does play a local role at the sites of infection such as lesions, or in the lymph nodes, and that the IFN found in the circulation is derived from these high concentration sources. Summerfield et al. and Nfon et al. found large numbers of circulating PDCs in pigs [10, 40] in comparison to the low numbers detectable in cattle. Nfon also found that these PDCs were functionally impaired during the latter stages of infection with FMDV.
Low levels of IL-10, corresponding with the peak of clinical signs, were found in the serum of the cattle studied. Whilst IL-10 does appear to play a role in immune suppression in pigs during infection with FMDV , it is most commonly associated with the maintenance of chronic infections such as Hepatitis C in humans [41, 42] and Mycobacterium Bovis in cattle . Studies in mice have shown that FMDV infected DC can stimulate splenic CD9+ B cells to produce T-independent neutralising IgM antibodies, via an IL-10 dependent process . We propose that the absence of leucopenia and immunosuppression in cattle, during acute FMDV infection, is associated with the low levels of type 1 IFN and IL-10. These differences in cytokine profile between pigs and cattle may also explain why, in general, more severe clinical signs are seen in pigs infected with FMDV .