Two critical determinants of FMD pathogenesis are the ability of the virus infection to spread through the different tissues and, the clearance rate by the host immune response. This is of particular relevance regarding the possibility of FMDV persistence in pigs. In the present study we show evidences that, at least during 17 dpi, FMDV can be isolated from different lymphoid tissues, not only as viral RNA but also as infectious virus. FMDV was also detected in coronary band epithelium even at time points when viremia was already cleared. To the best of our knowledge, this is one of the first reports where infectious virus has been isolated from swine lymphoid tissues, at later times than 3-5 dpi. The results of these studies have important implications for understanding both the viral spread in pigs and the likelihood for the virus to persist in the animals.
The evolution of the epithelial lesions suggests that one of the first sites of viral replication is the stratum spinosum, spreading later on to all epithelial layers. The appearance of lesions is correlated with the presence of FMDV RNA in coronary band epithelium in which viral load was within the range of 108-1010 molecules/mg. However, by day 17 pi no FMDV RNA was detected, indicating that virus was cleared from epithelium at this time. Interestingly, the specific infectivity of FMDV RNA in this tissue remains high at 14 dpi (3.2 × 10-5 PFU/FMDV RNA mol as compared with the specific infectivity showed by the virus used for challenge, FMDV C-S8c1, 2.7 × 10-6 PFU/FMDV RNA ), suggesting that although the amount of virus detected is not remarkable, the infectivity of this virus is even higher than that of the parental virus. Alexandersen et al.  proposed that a cycle of FMDV replication in pigs is of 12-24 h duration, initially replicating in epithelial cells and resulting in a significant amplification of the virus, producing high viremia and the appearance of clinical disease. After several replication cycles, in which the virus infects other susceptible cells, the host immune response is likely to prevent the development of higher level of virus. The fact that viremia has been cleared but virus was still found in epithelium might indicate that the virus is sequestrated from the circulation trying to escape immune surveillance as has been previously suggested for cattle . We are currently investigating whether this might be the case in pigs. On the other hand, a number of experiments have been carried out in cattle, in which FMDV transmission from carriers to susceptible animals was studied [25, 26]. Although virus transmission from carriers to susceptible animals in experimental model has not been described it would also be interesting to rule out the possibility of transmission of disease from convalescent pigs to other natural susceptible animals.
FMDV replication in tonsils has been found until the last day tested in this work (17 dpi), being unexpectedly high at day 17. Zhang et al.  described the detection of FMDV RNA in pig tonsil at 28 dpi, although no infectious virus was isolated in this report. In addition, Carrillo et al.  described virus isolation from pig tonsil at day 26 post-contact. These results, together with our data, indicate that virus is likely to persist in pigs to some degree following infection, resembling what has been described for cattle [1, 28] but not yet described for pigs. Nevertheless, it would be interesting to further investigate the presence of infectious virus in tonsil at later time points as a mechanism of persistence in pigs.
The predilection for the palatine tonsil showed in our work correlates with high viral replication in RTF LN, in which infectious virus can be detected at 14 dpi, and viral RNA until 17 dpi. Given that the RTF LN constitutes the draining pathway of the palatine tonsil , the high amount of virus in both anatomical locations until late during the infection might indicate that virus is being shed from the tonsil to the RTF LN. Since all the oropharynx area has been related to persistence of FMDV in cattle , our data support the possibility of a similar persistent stage in pigs, although additional data regarding later times post-inoculation are needed to confirm this hypothesis.
Some reports  have pointed out the presence of FMDV RNA in blood for a period of at least 28 dpi. In our study, the spleen and thymus, two major lymph organs, showed high amounts of FMDV RNA at 17 dpi and replicating virus measured by infection in BHK-21 cells. It is unlikely that the virus is produced elsewhere and filtered in these organs since by this time point in our study the viremia is completely cleared. A similar pattern is found in lymph nodes, in which the amount of FMDV RNA is high by 17 dpi and infectious virus can be rescued in all lymphoid tissues. The fact that FMDV RNA was rescued in these tissues suggests that viral replication has taken place at some point during viral spreading. One possibility could be the sequestration of immune complexes of FMDV particles within lymphoid tissue that has been suggested as a possible source of infectious material detected in tissues samples [13, 30, 31]. Different cell types such as B cells, macrophages and dendritic cells (DC) are able to support virus replication cycle at some level in the presence of high amounts of neutralising antibodies [31–33], and these cell types may also act as a source for viral spreading to other sites. This mechanism could also account for the detection of viral RNA 28 dpi in blood. Indeed, these cell types might act as a reservoir of virus in a replicative or non-replicative stage that eventually gets activated giving as a consequence a new virus production and, possible, an outbreak. Recently, FMDV RNA has been detected in germinal centers (GC) of cattle lymphoid tissues, suggesting that GC might be a reservoir area for infectious virus in lymphoid tissue . More interestingly, DC network might trap virus and serve as a repository for maintenance of persistence. Nevertheless, to distinguish sequestration of non-replicating virus from replicating virus (that eventually might be a source of infectious virus), minus strand RNA have to be measured. We are currently carrying out experiments to determine the role of DC in this pseudopersistence stage described for pigs.
Evolutionary patterns of virus replication and distribution in lymphoid tissue during 17 days pi with FMDV in pigs has shown that replication, or at least FMDV RNA detection, remains elevated in these lymphoid tissues despite the fact that viremia is significantly low or cleared. However, identification of infected internal structure from each tissue and of infected cell types deserves further investigation. These findings have implications for a putative persistence in pigs although with clear differences to that observed in cattle. Indeed, a longer follow-up time of infectivity in pigs will be needed to determine a true carrier state as described in cattle. Nevertheless, based on the evidences presented here, the paradigm that FMDV does not persist in pigs should be reconsidered.