Female BALB/c mice aged 8 weeks were purchased from Charles River. Influenza virus strain PR8 (A/Puerto Rico8/34, H1N1) was sourced from Virpur (San Diego). Viral infections were performed by intranasal administration of 100 PFU of virus in 50 μl of PBS. S. pneumoniae strain D39 glycerol stocks were inoculated into 5 ml of BHI Broth and incubated overnight at 37°C and 5% CO₂, without shaking. After 24 h, 50 ml of BHI broth was inoculated with 500 μl of overnight stationary culture in a 250 ml flask, and incubated at 37°C, 5% CO₂, without shaking. The culture was sampled at several time points (5, 6, 7, and 8 h), O.D. 620 nm was determined and CFU/ml controlled by plating bacteria on Mueller Hinton + 5% sheep blood agar plates. Mice were infected intranasally with 50 μl of the bacterial solution containing 1 × 10⁵ CFU. For preparation of K. pneumoniae, 5 ml of Nutrient Broth (NB) was inoculated with glycerol stock in a 14 ml round-bottomed tube and incubated at 37°C with shaking. After 24 h, 50 ml of NB was inoculated with 500 μl of the overnight stationary culture in a 250 ml flask followed by incubation at 37°C with shaking. The culture was sampled at several time points to measure O.D. at 620 nm and CFU/ml controlled by plating bacteria on NB plates. Mice were infected with 50 μl of the bacterial solution (endotoxin and DNA free content) containing 10¹ CFU of bacteria. For in vivo experiments, daily gavage of mice with 7.2 mg of OM-85-active principle (corresponding to 320 μL concentrate) or its equivalent without bacterial extract (referred to as “control solution”) was performed during 10 days. Animal experiments were performed in accordance with the Institutional Guidelines and Swiss Federal and Cantonal Laws on Animal Protection.

Different combinations of the following antibodies were used for flow cytometry. For analysis of dendritic cell (DC) subsets in the lung, a combination of CD11b PerCP-Cy5.5, CD11c APC-Cy7, B220 FITC, MHC II Alexa Fluor 700, CD40 PE, ICOS-L PE, CD80 PE, or CD86 PE was used. Analysis of B-cells subsets was performed with CD19 PE, CD23 PE-Cy7, CD21 Pacific Blue, IgD FITC, and IgM APC. Influenza-specific CD8+ T-cells were stained with NP-tetramer 366–374. All antibodies were purchased from Biolegend (San Diego) unless indicated. Stained cells were acquired on a BD FACS CANTO or LSRII and analyzed with FlowJo software (Tree Star, Inc.).

Total RNA was purified from cells obtained from whole lung and trachea of mice using Tri reagent (Sigma Aldrich). Real-time PCR was carried out according to manufacturers’ instructions using the quantifast SYBR green RT-PCR kit (Qiagen). The following primers were used: beta-Actin forward 5′-CCCTGAAGTACCCCATTGAAC-3′ and reverse 5′-CTTTTCACGGTTGGCCTTAG-3′; influenza matrix protein forward 5′-GGA CTG CAG CGT AGA CGC TT-3′, reverse 5′-CAT CCT GTT GTA TAT GAG GCC CAT-3′. Data are represented as the ratio of the Cq values from the influenza matrix protein gene to the house-keeping gene, beta-Actin.

Student’s t test (unpaired, two-tailed) or a one-way ANOVA was used to calculate significance levels between treatment groups, as indicated. Graph generation and statistical analysis were performed using Prism version 5 software (GraphPad, La Jolla, CA, USA). Standard deviation (SD) was used.

In order to gain insight into whether and how treatment of mice with OM-85 could enhance antiviral immunity, mice were treated orally with OM-85 for 10 days followed by infection with a sublethal dose of influenza A virus (Figure 1A). In this well-defined infection model, the peak of viral load is on day 5 post-infection and the virus is cleared by day 10 post-infection (24). Mice treated with OM-85 had a lower viral load in the lung tissue on day 5 post-infection as compared to mice treated with the control solution, which had undergone the same manufacturing process as OM-85, but without the bacterial extracts (Figure 1B). On day 10 post-infection, both groups had cleared the virus, as expected, showing that OM-85 treatment did not prolong the infection (Figure 1B). Analysis of the cellular infiltrates into the airways following infection showed that the OM-85 treated mice had a reduced proportion of neutrophils in the airways on day 5 post-infection (Figure 1C). These data corroborated the reduced viral load and indicated that the protective mechanism associated with OM-85 treated mice involved very early control of the infection, and as a consequence of this early control, the neutrophilic response resolved faster. Although the OM-85 mediated protective effect in this case was linked with time points associated with the innate antiviral response, an enhanced adaptive CD8+ T-cell response was noted on day 10 post-infection (Figure 1D). These data suggest that the actions of OM-85 were multifactorial, enhancing both innate and adaptive immune pathways, but the early innate response was likely to be responsible for the efficacy against influenza virus.

In both human beings and mice, a consequence of respiratory viral infections, particularly influenza, is an enhanced susceptibility to secondary bacterial infections (22, 23, 25). To place the OM-85-mediated enhanced antiviral immunity into this highly clinically relevant setting, we treated mice with OM-85 and on day 7 following an influenza infection mice were exposed to a sublethal dose of either K. pneumoniae or S. pneumoniae (Figure 2A). In the absence of any treatment, the infectious doses of the bacteria that were utilized resulted in mild infections with no weight loss, temperature loss, or signs of morbidity in the bacterial infection only groups, while the dose of influenza resulted in a transient weight loss (data not shown). However, when influenza infected animals were subsequently given the low-dose bacterial infection, bacteremia was present in control treated mice, while pre-treatment with OM-85 protected mice from the bacteremia (Figure 2B). Moreover, OM-85 treatment protected mice from the sequelae resulting from the secondary bacterial infections including weight/temperature loss, and morbidity scores (Figures 2C,D).

Given the protective effect of OM-85 against viral infections and consequently secondary bacterial infections, we next investigated how OM-85 shaped DC activation states, considering that DCs are one of the most potent cell types that influence immunity and have recently been shown to be activated by OM-85 (26). Accordingly, splenocytes were isolated from naïve mice and stimulated with different doses of OM-85 in vitro. OM-85 treatment led to a dose-dependent increase in the surface expression of MHC II (Figure 3A), CD40 (Figure 3B), and CD86 (Figure 3C) on both CD11b+ and CD11b− DC populations indicative of an increased activation state. Interestingly, not all markers were upregulated as a clear dose-dependent reduction in costimulatory ligand; ICOSL was found in these same cell populations (Figure 3D). We next investigated whether these markers were regulated in vivo and found similar trends in surface marker expression on lung DCs (Figures 3E,F), although no changes were evident on splenic DCs (data not shown), suggesting OM-85 elicited a mucosal tissue-specific effect. Given MHC II and the costimulatory molecules, CD40 and CD86 were increased by OM-85, it is likely that this effect was linked with the observed increase in influenza-specific CD8+ T-cell responses (Figure 1D), although it does not necessarily explain the rapid protection against the virus, which was already apparent in the first days post-infection prior to when the CD8+ T-cell response had developed.

To further dissect the possible mechanism through which OM-85 protects mice against influenza infection and subsequent secondary bacterial infection, we next characterized its effect upon B-cells. In an approach similar to the DC analysis, we stimulated splenocytes with different concentrations of OM-85 and assessed their activation state. Similar to the DC response, CD40 (Figure 4A) and CD86 (Figure 4B) were upregulated in a dose-dependent manner on B1 B-cells, Follicular B (FB) cells and marginal zone (MZ) B-cells. However, contrary to the DCs, these B-cell populations exhibited no significant changes in their expression of ICOSL except, to a minor extent, in MZ B-cells (Figure 4C). In order to determine whether these OM-85 driven changes in B-cell maturation had consequences in vivo, we administered OM-85 for 10 days by gavage as performed previously (Figure 1A), and then assessed IgG and IgA in the serum and airways of mice. OM-85 treatment alone led to a statistically significant increase in the levels of IgG detected in the serum of mice and trends toward increased IgA and IgG in the airways (Figure 5A). Of note, however, there was a statistically significant increase in IgA isolated from the serum or airways that bound to influenza antigens (Figure 5B). This was particularly intriguing because the OM-85 treated mice had not been infected with the virus in this setting. Moreover, antibodies that bound respiratory syncytial virus (RSV) were also detectible in the OM-85 treated groups (Figure 5C). These data indicate that OM-85 treatment had led to a polyclonal B-cell activation that resulted in release of antibodies against multiple antigens. Although we could detect comparably higher levels of these antibodies in OM-85 as compared to control treated mice, their relevance was unclear. Thus, to assess whether these polyclonal antibodies had functional implications, we tested their ability to limit influenza virus infection. Indeed, in an in vitro virus neutralization assay, heat-inactivated serum or BAL fluid from OM-85 treated mice was effective at limiting influenza infection (Figure 5D). Taken together, these data indicate that OM-85 can act to protect mice against influenza infection by enhancing B-cell activation and release of broadly protective antibodies that help to protect the host against infection.