Pigs are preferential targets of many respiratory diseases caused by bacteria such as Pasteurella multocida . Different swine respiratory viruses such as African swine fever virus (ASFV)  or porcine reproductive and respiratory syndrome virus (PRRSV)  also target alveolar macrophages for their replication. It is well established that when incorporated in feed or given in daily doses, T-2 toxin decreases resistance against viral or bacterial infections [12, 15]. Moreover, Ziprin et al.  showed that the T-2 toxin effect on the course of in vivo bacterial infection depends on the nature of the infectious agent. Based on these different data, the present study is of further interest to address whether T-2 toxin can similarly reduce immune resistance depending on the nature and composition of microbial pathogens, using TLR4 and TLR6-agonists, structurally conserved molecules derived from these pathogens.
The current study indicates that the immune response implemented by alveolar macrophages is impaired by T-2 toxin. T-2 toxin is known to be cytotoxic and to induce apoptosis on immune cells like human monocytes, human dendritic cells and human and rat macrophages [10, 28, 38, 39]. We observed that after 16 h exposure to T-2 toxin, the cell viability of PAM is affected by 10 nM of T-2 toxin and the IC50 value was evaluated near 20 nM. This concentration was in accordance with other studies on immune cells [10, 40]. Cell death may be due to necrosis or apoptosis. The T-2 toxin is known to induce the apoptosis pathway at varying concentrations depending on the targeted cells [38, 39, 41, 42]. For example, in primary culture cells of hematopoietic progenitors, T-2 toxin induces apoptosis after 3 h of incubation with 10 nM of the toxin and a maximum reached at 12 h  while in macrophage cell lines (RAW 267.7), T-2 toxin induces less than 5% of apoptotic cells after 6 h of incubation with 5 nM of T-2 toxin . The latter results are comparable to our data. A significant induction of apoptosis was shown with the exposure of 30 nM of T-2 toxin on PAM after 16 h of treatment. In order to avoid possible adverse effects due to cytotoxic and apoptotic process, a non cytotoxic and non apoptotic concentration of 3 nM of T-2 toxin was used for PAM activation studies. Moreover, the dose of 3 nM of T-2 toxin was relevant as a low dose and close to the provisional tolerable daily intake (TDI) of 100 ng / Kg bw / d .
Activated macrophages are the primary source of pro-inflammatory cytokines, including IL-1β and TNF-α [27, 45]. Production of these two cytokines, in response to TLR-agonists and notably LPS, could be observed in the present study. Interestingly, after pre-treatment with low doses of T-2 toxin, a reduction of pro-inflammatory cytokine concentrations appeared dependent on the nature of the TLR-agonists. The IL-1β production was reduced after activation by the TLR2-agonist and significantly decreased after TLR4- and TLR6-agonist activation. By contrast, we did not observe a reduction of IL-1β production after activation of Pathogen Associated Molecular Patterns (PAM) by the TLR7-agonist. Similar results were observed regarding TNF-α response.
TLR expressed on antigen presenting cells (APC) such as macrophages, are critical for recognition of the PAMP on microorganisms and their abilities to induce appropriate immune responses. Extensive research on TLR has been performed in humans and rodents while only a few studies on porcine TLR have been completed. However, most of the studies on swine TLR have clearly demonstrated a correlation between the expression levels of TLR2 and TLR9 expressed in several tissues (lung, heart, spleen, small intestine, mesenteric lymph nodes…) at the first stages of life and the capacity for newborn piglets to have an early efficient immune function [46, 47]. Moreover, similar distribution of TLR on human and swine cells, unlike rodents, has been previously described . Human alveolar macrophages have low TLR2 expression leading to a low production of TNF-α and IL-6 cytokines . Similarly, the TLR2 expression level measured in human bronchoalveolar cells by Juarez et al.  could explain the low TNF-α protein concentration obtained in response to the TLR2-agonist in this study.
It is known that activation of macrophages by agonists, such as LPS, triggers the enhancement of expression of iNOS leading to NO production . In the present investigation, a T-2 toxin pre-treatment induced a significant decrease of NO production in PAM activated by TLR4 and TLR2/6 agonists. However, this toxin did not affect NO production after activation of macrophages by either IFN-γ/LPS or TLR2 and −7 agonists. DON and nivalenol also affect the immune response by inhibiting NO production in mice macrophage cell lines in vitro (RAW264) . In addition, this study demonstrated that there is a parallel between the decreasing of NO production and the decreasing of iNOS mRNA gene expression which could partially explain the effect of T-2 toxin on NO production.
Taken together, these results demonstrate the complexity of mycotoxin action depending on the cell activation status and mycotoxin doses. The TLR agonist-sensitized macrophage responds to subsequent toxin challenge with a robust and potentiated cytokine response. By contrast, our study showed that T-2 toxin pre-exposure inhibited naive macrophage to initiate an inflammatory response to TLR-ligands. This inhibition could be attributed to a defect of pathogen pattern recognition due to a T-2 toxin effect on TLR2, -4 and −2/6 expression. Due to a lack of commercially available antibodies for pigs, we could not verify the cell surface expression of these TLR. However, the present study showed that the importance of this inhibition seems to depend on the nature of the TLR. Indeed, the expression of TLR7 mRNA seemed less affected by the toxin. These results were consistent with the less inhibitory effect of T-2 toxin on the inflammatory response when PAM were activated by TLR7-agonist. This one is specifically located on the endosomal membrane and could explain that a T-2 toxin pre-exposure did not appear to disrupt the pro-inflammatory response.
The alteration of TLR expression could possibly influence the functions and efficacy of alveolar macrophages in terms of antigen processing and release of cytokines. It is tempting to speculate that the sensitized macrophage phenotype depends on both appropriate surface TLR expressions enabling the recognition of pathogen motifs and the appropriate induction of the cytokine signaling pathway. However, the effects of T-2 toxin seem to be TLR-receptor specific since neither co-receptor CD14 nor mannose receptor (CD206) were affected by the toxin. Our group has already shown in a previous study that DON can also inhibit cell surface expression of activation markers of human macrophages such as CD14, CD54, CD119 and HLA-DR .
Trichothecenes have the ability to modulate immune function by disrupting intracellular signaling pathways within leukocytes responsible for driving both altered immune-regulatory gene expression as well as apoptosis . A review by Pestka et al.  demonstrated the preferential targeting of DON on the activation of mitogen-activated protein kinases (MAPK), thereby leading to the modulation of immune responses and apoptosis . In addition to specific intracellular localization of TLR7, the absence of T-2 toxin action on immune responses after TLR7 agonists activation can be explained by focusing on the specificity of TLR7-signaling pathways. In contrast to other TLR which involved both MAPK and canonical IKK complex for the transcription of inflammatory cytokine genes, TLR7 agonist activates only the IKKα signaling pathway to induce inflammatory cytokines .
Based on our finding that the T-2 toxin down-regulated the activation of TLR, it could be hypothesized that exposure to low concentrations of T-2 toxin may increase the susceptibility of humans and animals to opportunistic infections. The present study sheds new light on the potential of mycotoxins to interfere with the immune system by decreasing the expression pattern recognition of pathogens, and thus the initiation of immune responses against bacterial and viral infections. These results may help to explain the immunosuppressive effect of T-2 toxin observed in vivo during viral respiratory infection such as reovirus .
Given that food and feed are sometimes contaminated by T-2 and other trichothecenes, serious questions remain regarding risks from chronic ingestion of these toxins. Understanding the molecular mode of action of a mycotoxin can assist in predicting potential adverse human and animal health effects.