Probiotic microorganisms have been mainly studied in the context of bacterial infections of the gastrointestinal tract which is the natural target tissue of probiotics. However, there are a few reports which indicate that upon oral intake, probiotics can also affect infections of the respiratory tract 
. There are also reports in the literature where probiotics induce antiviral activity in vitro
and are even applied as a medical treatment against persistent virus infections in humans and animals.
We chose the zoonotic swine influenza viruses as a novel study object to test for the antiviral potential of the probiotic E. faecium and to elucidate its mechanisms of action and we present the results from in vitro experiments using porcine H1N1- and H3N2-influenza virus in MDBK- and 3D4/21 cells, respectively. Our results demonstrate that the probiotic E. faecium effectively protects host cells from swine influenza virus infection and are in support of the above author's hypothesis, that probiotics are not only useful to inhibit enteric viruses, but may also have potential for the control of respiratory viruses.
Two different SwIV strains were chosen which are currently circulating in the pig population, H1N1 (A/Swine/Greven/IDT2889/2004) and H3N2 (A/Swine/Bondelum/IDT5959/2007). As an established model for the present in vitro
study an epithelial- (MDBK-cells) and a porcine alveolar macrophage cell line (3D4/21-cells) were utilized. It can be argued that the concentration of the probiotic utilized may not reflect the situation in the target tissue in vivo
. However, the concentration chosen for treatment of the cell cultures (106
CFU/ml) reflects the same concentration which was determined in the gut of piglets fed E. faecium
as a supplement during previous feeding trials in our research consortium 
. In order to find out during which period of the SwIV replication cycle the probiotic has the most stringent effect, E. faecium
was added for a brief period of time (60 or 90 min) to the host cells either before, during or after virus infection (). The results indicate that the simultaneous addition of virus and E. faecium
to the host cell monolayer was the most effective timing for the inhibition of virus multiplication. As shown in and , this experimental setup (termed “competition”) resulted in a 4 log-unit reduction of virus titer and in a concomitant rescue of cell viability. Since a 1 h exposure of the monolayers to E. faecium
before SwIV-infection and a 1 h treatment after completion of virus infection also led to a 2–3 log-unit loss of virus titer, the probiotic must alter host cell factors which apparently results in an inhibition of influenza virus multiplication. Obvious candidates for such factors are mediators of cellular defence processes.
Experimental design of dose response study of probiotic effect on SwIV.
The expression of NO and its subsequent increased activity has previously been reported to play a role in the host response to multiple viral families, and in various host species 
. In addition to its antiviral properties, NO has been described to modulate intestinal barrier function, gut motility, iron transport, and has been implicated in numerous infections and non-infectious diseases of the intestine 
. We found that E. faecium
increased the expression of NO in both 3D4/21 and MDBK cells (). All the samples collected after treatment with E. faecium
showed significantly increased NO-values when compared to the non-treated counterparts on 3D4/21 cell line. This is consistent with the hypothesis that high NO levels are associated with decreased virus production.
In addition to stimulating NO-release, probiotics could also affect the expression of cytokines and other immune mediators relevant for the innate immune response to viral infections. We therefore determined the expression of selected mediators in SwIV-infected host cells (3D4/21-cells only). As seen from , E. faecium
promoted an increased expression of IFN-α. Since the difference between the values of non-treated and E. faecium
treated cells were found to be non-significant, IFN-α can be ruled out as the main immunoregulatory cytokine that could lead to E. faecium
induced inhibition of SwIV-infection. Another cytokine stimulated by the probiotic treatment was IL-10, which is a typical Th2 cytokine that is initially repressed in virus infected cells but then expressed at higher levels later in infection to control the initial inflammatory response to infection. Interestingly, this cytokine is clearly enhanced in the macrophage cell line upon E. faecium
treatment and thus could support cellular control of SwIV infection. Two pro-inflammatory cytokines were found to be clearly reduced in the SwIV-infected 3D4/21-cells treated with the probiotic, IL-6 and TNF-α. Secretion of IL-6 by macrophages is known to play an indirect immunoregulatory role in the immune response to viral infection 
, and TNF-α acts as an inflammatory cytokine by triggering a cascade of cytokine production 
Since both IL-6 and TNF-α are downregulated by E. faecium
in SwIV-infected 3D4/21 cells, the reduced inflammatory response caused by some cytokines at the cellular level may contribute to the antiviral effect of the probiotic. Toll-like receptor 3 (TLR-3) was the first identified antiviral TLR to have a central role in the host response to viruses 
. Our experimental data show that the treatment of SwIV-infected 3D4/21-cells with E. faecium
led to an decreased expression of TLR-3 at 2 h and 6 h post infections compare to virus alone which suggest that the probiotic induced modulation of this receptor may have a role in the antiviral function of E. faecium
. Since E. faecium
acts most inhibitory when it is added together with the virus particles, we assessed whether SwIV might be physically trapped or inactivated by the probiotic bacteria in a mixed incubation as detailed as “preincubation assay” in the experimental design shown in (lower panel). The results summarized in show that a substantial portion of the input virus particles are indeed trapped by the bacteria since virus infectivity is lost from the supernatants and viral genome equivalents are detected in the bacterial sediments after low speed centrifugation of the incubation mixture (data not shown). Thus under such experimental conditions two antiviral functions of the probiotic may operate synergistically and add up to produce a more severe inhibition of SwIV.
The results presented altogether show that the probiotic E. faecium quite effectively inhibits the multiplication of swine influenza viruses in relevant cell culture systems. The antiviral mechanism of this probiotic is probably manifold since it was found to act on both the virus particles and the host cells. However, at least a few inhibitory parameters could be identified: E. faecium bacteria are able to adsorb SwIV-particles and to alert the cells by mediating a rapid antiviral response through modulating the expression of defence relevant mediators. Amongst these IL-6, TNF-α, IL-10, IFN-α and TLR-3 were identified as entities modulated by the probiotic treatment. It is realized that E. faecium can induce much more complex reactions in a treated tissue since only a few mediators could be assessed in this study. One common denominator of probiotic action could be NO which is a mediator affected by many cellular signaling cascades. In line with publications for other virus-host cell systems, our results point to a central role of NO which is stimulated upon the treatment with the probiotic and which may mount an improved cellular defence response against SwIV-infection in tissues which were stimulated with a probiotic.
Based on the in vitro data shown here for a porcine influenza virus, we hypothesize that the use of E. faecium as a probiotic feed (or food) additive has the potential of reducing influenza virus infections in mammalian tissues. SwIV challenge experiments with piglets which are fed E. faecium as a supplement are presently in progress in order to test this hypothesis.