All animal experiments were conducted according to the European Guideline for animal welfare (2010/63/EU) with approval from the commission for animal experiments headed by the “Landesamt für Gesundheit und Soziales” (LaGeSo, Berlin, Germany, registration number G0184/12 and G0097/12).

Animals were bred and maintained in the facilities of the “Forschungseinrichtungen für Experimentelle Medizin” (FEM, Charité – Universitätsmedizin, Berlin, Germany) under specific pathogen-free (SPF) conditions. Female age matched C57BL/6j wildtype mice were used.

To eradicate the murine intestinal microbiota 8–10 week old mice were transferred to sterile cages and treated with a quintuple broad-spectrum antibiotic cocktail as previously described (Heimesaat et al., 2006). Three days prior recolonization experiments the antibiotic cocktail was withdrawn and replaced by autoclaved drinking water. For FMT, fresh murine fecal samples were collected from 10 individual female 3 months old naive mice (harboring a conventional SPF microbiota), pooled, dissolved in 10 ml sterile phosphate buffered saline (PBS; Gibco, life technologies, Paisley, UK), and bacterial loads were evaluated by cultural and molecular methods before peroral challenge of mice with 0.3 ml of the suspension by gavage. Another group of mice received an oral suspension of VSL#3 bacteria. VSL#3 is a commercially available probiotic mixture (Manufacturer: SIIT S.r.l, Trezzano sul Naviglio, Italy; distributed by Actial Farmaceutica, Funchal, Madeira, Portugal) consisting of the following eight bacterial species: Streptococcus thermophilus, Bifidobacterium breve, Bifidobacterium longum, Bifidobacterium infantis, Lactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus paracasei, and Lactobacillus delbrueckii subsp. Bulgaricus. A total of 4.5 × 10¹¹ probiotic bacteria were dissolved in 50 ml PBS. By gavaging 0.3 ml, each mouse received ~10⁹ viable probiotic bacteria as confirmed by cultural analyses of the suspensions. Mice were continuously kept in a sterile environment (autoclaved food and drinking water or sterile antibiotic cocktail) and were handled under strict aseptic conditions to prevent contaminations.

To survey clinical signs of inflammation, a standardized cumulative clinical score (maximum 12 points), addressing the occurrence of blood in feces (0: no blood; 2: microscopic detection of blood by the Guajac method using Haemoccult, Beckman Coulter/PCD, Krefeld, Germany; 4: overt blood visible), diarrhea (0: formed feces; 2: pasty feces; 4: liquid feces), and the clinical aspect (0: normal; 2: ruffled fur, less locomotion; 4: isolation, severely compromized locomotion, pre-final aspect) was applied daily as described earlier (Haag et al., 2012).

Mice were sacrificed by isofluran treatment (Abbott, Greifswald, Germany) at day (d) 28 post recolonization. Tissue samples from spleen, MLN, ileum and colon, and luminal samples from colon were removed under sterile conditions. Ileal and colonic ex vivo biopsies were collected from each mouse in parallel for immunological, microbiological, and immunohistochemical analysis. For immunohistochemical stainings, ileum and colon samples were immediately fixed in 5% formalin and embedded in paraffin, and sections (5 μm) were stained with the respective antibodies as described below.

Total intestinal loads of VSL#3 bacteria were quantified in serial dilutions of fecal and large intestinal luminal samples streaked onto Columbia-Agar supplemented with 5% sheep blood and Columbia-CNA Agar supplemented with colistin and nalidixic acid (both Oxoid) in parallel and incubated under aerobic (with 5% CO₂), microaerophilic (in jars using CampGen gas packs; Oxoid) and obligate anaerobic (in jars using Anaerogen gas packs; Oxoid) conditions for at least 2 days. Bacterial species were identified according to their typical morphological appearances. The total VSL#3 bacterial loads of intestinal samples were approximated as the sum of identified colony forming units (CFU) derived from the respective culture conditions. The detection limit of viable bacteria was ≈100 CFU per g.

The complex intestinal microbiota composition in conventionally colonized SPF mice and mice subjected to FMT was assessed by quantitative 16S rRNA based real time PCR as described previously (Heimesaat et al., 2010, 2014; Rausch et al., 2013; Thoene-Reineke et al., 2014).

In situ immunohistochemical analysis of ileal and colonic paraffin sections was performed as previously described (Heimesaat et al., 2007, 2010; Alutis et al., 2015). Primary antibodies against cleaved caspase-3 (Asp175, Cell Signaling, Beverly, MA, USA, 1:200), Ki67 (TEC3, Dako, Glostrup, Denmark, 1:100), CD3 (#N1580, Dako, 1:10), FOXP3 (FJK-16s, eBioscience, San Diego, CA, USA, 1:100), and B220 (eBioscience, 1:200) were used. For detection, the LSAB method was applied with FastRed (Dako) as chromogen. For each animal, the average number of positively stained cells within at least six high power fields (HPF, 400× magnification) was determined microscopically by a blinded investigator.

Single cell suspensions were generated from spleens and MLN, and erythrocytes were removed from splenic samples by 1.66% ammonium chloride. All samples were resuspended in defined volumes of PBS/0.5% BSA and subjected to further processing (Cording et al., 2013).

Segments of the murine gut were removed and freed from fat, connective tissue and PP, cut longitudinally and cleared from luminal content and mucus with ice-cold PBS. The isolation of LPL followed a standard protocol with minor modifications (Sheridan and Lefrancois, 2012). Briefly, the intestines were cut into 0.5 cm pieces and incubated twice with 25 ml Hanks balanced salt solution (HBSS; Gibco) containing 1 mM dithioerythritol (DTE; Carl Roth) for 20 min at 37°C and 220 rpm. Afterwards the intestines were introduced to HBSS containing 1.3 mM ethylenediaminetetraacetic acid (EDTA; Life Technologies, Eugene, Oregon, USA). Subsequently the cells were placed in digestion solution, containing 0.5 mg/ml collagenase A (Roche, Mannheim, Germany), 0.5 mg/ml DNAse I (Roche), 10% FCS, and 1 mM of each CaCl2 and MgCl₂ (both Carl Roth). Digestion was performed through incubation for 45 min at 37° and 220 rpm. After the incubation the digested tissues were washed with RPMI containing 5% FCS, resuspended in 5 ml 44% Percoll (GE Healthcare, Uppsala, Sweden), and overlaid on 5 ml 67% Percoll in a 15 ml Falcon tube. Percoll gradient separation was performed by centrifugation at 600 g for 20 min at room temperature. LPL were collected from the interphase, washed once and suspended in PBS/0.5% BSA.

Surface staining was performed using the following antibodies: FITC-anti-CD4 (Clone RM4-5; 1:200), PerCP-anti-CD8 (Clone 53-6.7; 1:100), PacBlue-anti-B220 (Clone RA3-6B2, 1:200), APC-Cy7-anti-CD25 (Clone PC61, 1:200), PE-anti-CD44 (Clone IM7, 1:200), APC-anti-CD86 (Clone B7-2, 1:200; all from BD Biosciences, San Jose, CA, USA).

For intracellular staining cells from spleen, MLN and intestinal LP were restimulated for 5 h with 10 ng/ml phorbol myristate acetate (PMA) and 1 μg/ml ionomycin, in a tissue culture incubator at 37°C (both Sigma-Aldrich). Ten micrograms per microliter of brefeldin A (Sigma-Aldrich) was added to the cell suspensions after 1 h of polyclonal restimulation. Then cells were treated with LIVE/DEAD Fixable Aqua Dead Cell Stain kit (life technologies) and hereafter fixed with 2% paraformaldehyde (PFA; Sigma-Aldrich) for 20 min at room temperature. Cells were stained in 0.5% saponin (Sigma-Aldrich) using the following antibodies: PacBlue-Anti-CD4 (Clone RM4-5; 1:400), PE-Cy7-anti-IFN-γ (Clone XMG 1.2; 1:400), (both from BD Biosciences) FITC-anti-IL17A (Clone TC11-18H10.1; 1:200, BioLegend, San Diego, CA, USA), PE-anti-IL10 (Clone JESS-16E3; 1:100), APC-anti-IL22 (Clone IL22JOP; 1:100) (both from eBioscience). All data were acquired on a MACSQuant analyzer (Miltenyi Biotec, Bergisch Gladbach, Germany) and analyzed with FlowJo Software v10.1 (Tree star, Ashland, OR, USA).

RNA was isolated from snap frozen ileal and colonic ex vivo biopsies, reverse transcribed, and analyzed for cytokine specific mRNA as described earlier (Munoz et al., 2009). Murine IL-22, IL-17A, IL-10, and IFN-γ mRNA expressions were analyzed using Light Cycler Data Analysis Software (Roche). Expression levels were calculated relative to the housekeeping gene for hypoxanthine-phosphoribosyltransferase expression and indicated as “Arbitrary Units” (fold expression).

Medians and significance levels using appropriate tests as indicated (Mann Whitney U-test and one-way ANOVA with Tukey's post-hoc test for multiple comparisons) were determined using GraphPad Prism Software v6 (La Jolla, CA, USA). Two-sided probability (p) values ≤ 0.05 were considered significant. All experiments were repeated at least twice.

Given a better acceptance by the patients and more practical mode of peroral application of a probiotic formulation as compared to FMT in both ambulant and hospital settings, we compared the immunopathological impact of the probiotic compound VSL#3 and complex microbiota in the with broad-spectrum antibiotics treated host. Therefore, we virtually depleted the intestinal microbiota of conventional mice by quintuple antibiotic treatment. These secondary abiotic (ABx) mice then received either 10⁹ viable VSL#3 bacteria via the oral route or were subjected to FMT by gavage, whereas naive, conventionally colonized and ABx mice served as positive and negative controls, respectively. From day 3 until day 28 following VSL#3 peroral challenge, mice could be stably colonized as indicated by 10⁹ CFU of VSL#3 bacteria per g feces (Figure 1A). In order to exclude that upon cessation of antibiotic treatment and peroral reconstitution with VSL#3 remnant commensal bacteria might grow back, we surveyed the intestinal microbiota composition applying highly sensitive and culture-independent 16S rRNA based molecular quantitative RT-PCR. In fact, only bifidobacteria and lactobacilli as main bacterial constituents of the applied probiotic compound increased in fecal samples until day 28 following probiotic challenge (p < 0.001; Figure 1B), whereas the other intestinal bacterial commensals remained virtually unchanged (n.s.; Figure 1B). Moreover, mice subjected to FMT showed at day 28 after recolonization a complex large intestinal microbiota composition that was comparable to the microbiota in conventional control animals (Figure 1C).

Given that mice were clinically/macroscopically uncompromised upon antibiotic treatment as well as following respective recolonization regimens as assessed by a clinical scoring system on a daily basis (not shown), we next assessed potential microscopic changes in the intestinal tract. To address this, we quantitatively surveyed apoptotic cell numbers in small and large intestinal epithelia applying in situ immunohistochemistry. In line with the uncompromised clinical conditions, apoptosis was neither induced by broad-spectrum antibiotic treatment nor by respective peroral reassociation (n.s.; Figure 2A; Figure S1). Numbers of epithelial cells positive for Ki67, a sensitive marker for cell proliferation and regeneration (Scholzen and Gerdes, 2000), however, were significantly reduced in both, ileum and colon following broad-spectrum antibiotic treatment. Notably, administration of either VSL#3 or complex SPF microbiota was sufficient to restore regenerative epithelial properties as indicated by higher small as well as large intestinal Ki67+ cell numbers as compared to ABx mice (p < 0.001), that were comparable to those observed in naive SPF mice (Figure 2B; Figure S1).

To further dissect the role of VSL#3 in modulating adaptive immune responses following microbial depletion, we quantified cell numbers of distinct immune cell populations in both small and large intestines of mice at day 28 post recolonization with either VSL#3 or FMT by in situ immunohistochemical staining of paraffin sections (Figure 3; Figure S1). Broad-spectrum antibiotic treatment was associated with reduced numbers of CD3+ T lymphocytes (p < 0.001; Figures 3A,D; Figure S1), B220+ B lymphocytes (p < 0.001; Figures 3B,E; Figure S1), and Foxp3+ Treg (p < 0.001; Figures 3C,F; Figure S1) in mucosa and lamina propria of both ileum and colon. Application of either VSL#3 or complex SPF microbiota, however, was sufficient to restore adaptive immune cell populations in the colon as indicated by T and B cell as well as Treg numbers that were comparable to naive mice at day 28 following respective recolonization (Figures 3D–F; Figure S1). Numbers of T and B lymphocyte numbers increased in small intestines of secondary abiotic mice following FMT, but not VSL#3 recolonization (p < 0.001; Figures 3A,B; Figure S1). Small intestinal Treg numbers were elevated upon either reassociation regimen and to the highest extent by FMT (Figure 3C; Figure S1).

To exclude that the observed effects were microbiota driven and not due to antibiotic withdrawal per se, we quantitatively assessed respective intestinal immune cell populations in secondary abiotic mice 4 weeks following cessation of broad-spectrum antibiotic treatment (Figure S2). However, withdrawal of antibiotic treatment (ABx%) did not restore any of the analyzed small or large intestinal immune cell populations as indicated by comparable cell numbers in ABx and ABx% mice (n.s.; Figure S2).

Hence, depending on the intestinal compartment, peroral application of VSL#3 or complex murine microbiota could sufficiently reverse antibiotics-induced decreases in intestinal immune cell populations with the most prominent effect in the colon.

We next elaborated the capacities of peroral VSL#3 application or FMT to induce, maintain and modulate distinct immune cell populations in mucosal, peripheral, and systemic immunological sites of mice that had been treated with broad-spectrum antibiotics. To address this, we performed flow-cytometric analyses of lymphocytes derived from the lamina propria of small and large intestine, MLN and spleen of mice at day 28 post-recolonization. The gating strategies are depicted in Figures S3A–F. We firstly focused on relative abundances and absolute numbers of the main lymphocytic groups, namely CD4+ (Figures 4A–H) and CD8+ (Figures 5A–H) T lymphocytes as well as B220+ B lymphocytes (Figures 6A–H). Antibiotic treatment resulted in decreased relative abundances and absolute numbers of CD4+ T helper cells in both the small and large intestines (p < 0.05–0.001; Figures 4A–D), whereas VSL#3 administration could sufficiently restore respective cell numbers at either mucosal site. Furthermore, abundances of CD4+ cells increased in MLN upon VSL#3 treatment (p < 0.05; Figure 4E). Following FMT, CD4+ cell frequencies were higher in small and large intestines as compared to ABx mice (p < 0.001; Figures 4A,C). Whereas frequencies of splenic CD4+ cells were rather unchanged upon antibiotic treatment and peroral reassociation (n.s.; Figure 4G), increased CD4+ cell numbers could be observed in the spleens of mice that had undergone antibiotic therapy, regardless whether subsequently recolonized or not (p < 0.05; Figure 4H).

We further analyzed CD8+ cytotoxic T cell responses in intestinal and systemic compartments upon antibiotic treatment and subsequent bacterial reassociation. Both VSL#3 treatment and FMT could sufficiently restore the antibiotics induced CD8+ cell frequency reduction in the small and large intestines as indicated by higher small and large intestinal CD8+ cell abundances as compared to ABx mice (p < 0.01; Figures 5A,C), which also held true for absolute CD8+ cell numbers upon VSL#3 treatment (p < 0.05–0.001; Figures 5B,D). Whereas CD8+ cells were rather unchanged in MLN of ABx mice with and without bacterial recolonization, both frequencies and absolute numbers of splenic CD8+ cells increased upon broad-spectrum antibiotic treatment (p < 0.001; Figures 5G,H). Interestingly, VSL#3, but not FMT could reverse this effect on CD8+ cell abundances (p < 0.001 vs. ABx mice; Figure 5G). Hence, again peroral VSL#3 application or FMT were able to reverse antibiotics induced decreases in T cells, depending on the respective immunological compartment.

We next expanded our comprehensive survey on lymphocyte populations during antibiotic treatment and bacterial recolonization to B220+ B cells. Whereas decreased B220+ cell counts were detected in small and large intestines as well as in MLN following broad-spectrum antibiotic treatment (p < 0.05–0.01; Figures 6B,D,F), VSL#3, but not FMT resulted in elevated small intestinal B lymphocytes (p < 0.001 vs. ABx; Figure 6B). In the splenic compartment, B220+ cell numbers increased following antibiotic treatment, but also upon additional VSL#3 challenge (p < 0.05 and p < 0.01, respectively; Figure 6H). In addition, B220+ cells were more abundant in the spleen of VSL#3 as compared to untreated ABx mice (p < 0.05; Figure 6G). Taken together, our data indicate that an intestinal VSL#3 microbiota is capable of inducing and modulating distinctive immune cell populations, thus antagonizing immunological consequences of antibiotic treatment not only at mucosal site, but, to some extent, also on a systemic level of the immune system.

In the following, we addressed whether recolonization with VSL#3 or FMT following broad-spectrum antibiotic treatment was associated with changes in defined T cell subsets and in the activation status of distinct cell populations. We therefore stained CD4+ cells with antibodies against CD25, a surface protein characteristic for Treg. Microbial depletion by antibiotic treatment led to reduced abundances of the CD4+CD25+ subpopulation in all intestinal and systemic immunological compartments under investigation (p < 0.001; Figure 7). Remarkably, VSL#3 administration as well as FMT enhanced Treg abundances and completely reversed the antibiotics induced effect (p < 0.01–0.001 vs. ABx; Figures 7A,C,E,G). VSL#3-induced splenic abundances of CD4+CD25+ cells were, however, less distinct than in naive controls (p < 0.001; Figure 7G). Broad-spectrum antibiotic treatment was further accompanied by a down-regulation of the surface marker CD86, a co-stimulatory molecule marking activated DC (Wallet et al., 2005), in small intestine, colon, MLN and spleen of ABx mice, whereas both VSL#3 treatment and FMT virtually reversed these immune-depressive effects (p < 0.001; Figures 7B,D,F,H). Hence, the activation status of distinct cells in intestinal as well as systemic compartments is down-regulated by broad-spectrum antibiotic treatment, but can be restored upon VSL#3 application or FMT.

We then investigated the impact of VSL#3 and FMT on the memory/effector CD4+ and CD8+ cells by evaluating high expression of CD44, a surface marker expressed (on both, CD4+ and CD8+ cells) upon previous antigen contact (Sprent and Surh, 2002). ABx mice exhibited a significant reduction in abundances of both, CD4+CD44+ and CD8+CD44+ cells in all examined intestinal and systemic lymphoid organs (p < 0.001; Figure 8). In the small and large intestines, VSL#3 recolonization and FMT resulted in a strong up-regulation of CD44 expression on CD4+ as well as CD8+ cells (p < 0.05–0.001; Figures 8A–D). The same held true for CD8+ cells in MLN and spleen (p < 0.001; Figures 8F,H), whereas FMT alone resulted in higher frequencies of memory CD4+ cells in MLN and spleen as compared to antibiotic treatment (p < 0.001; Figures 8E,G). Hence, microbial depletion resulted in reduced abundances of memory/effector T cells in intestinal and systemic lymphoid compartments, that could, however, at least in part be restored by VSL#3 treatment or FMT.

We further assessed the cytokine producing properties of CD4+ T lymphocytes following VSL#3 administration or FMT in ABx mice. Therefore, we determined the frequencies of IFN-γ and IL-10 (Figure 9) as well as of IL-17 and IL-22 (Figure 10) producing CD4+ cells in small and large intestines, MLN, and spleens. Gating strategies are depicted in Figures S3G–I and representative dot plots shown in Figure S4. Small intestinal IFN-γ producing CD4+ cells were depressed in ABx mice (p < 0.01–0.001; Figure 9A) and could not be fully recovered by either bacterial recolonization regimen. Notably, small intestinal CD4+IFN-γ+ cells were more frequently abundant following FMT as compared to ABx mice (p < 0.05; Figure 9A). Moreover, IFN-γ producing CD4+ cells were less abundant in large intestines of ABx and VSL#3 treated, but not with fecal microbiota transplanted mice (p < 0.05–0.001; Figure 9C), whereas CD4+IFN-γ+ cells did not differ between respective groups in MLN and spleen (n.s.; Figures 9E,G). A strong reduction of CD4+ lymphocytes producing the anti-inflammatory cytokine IL-10 could be determined in all immunological sites following antibiotic therapy (p < 0.001; Figures 9B,D,F,H). Notably, reassociation with either VSL#3 or complex murine microbiota could fully restore the frequencies of CD4+IL10+ cells in all compartments (p < 0.001; Figures 9B,D,F,H) and thus reestablish the pre-antibiotic (naive) status. Interestingly, intestinal as well as systemic CD4+ cells producing the pro-inflammatory cytokines IL-17 or IL-22 were less abundant upon antibiotic and also subsequent VSL#3 treatment (p < 0.05–0.001; Figures 10A–H), but not upon FMT, except for small intestinal CD4+IL17+ cells (p < 0.05; Figure 10A).

These findings were further underlined by results obtained from mRNA analysis of respective cytokines measured in ileal and colonic ex vivo biopsies (Figure 11). IL-10 as well as IL-17, IL-22, and IFN-γ mRNA were all down-regulated in both small and large intestines of ABx mice (p < 0.05–0.001; Figure 11). In the small intestine, FMT (p < 0.05), but not VSL#3 treatment could sufficiently up-regulate IL-10 expression back to naive levels, whereas the other way around, recolonization with VSL#3 (p < 0.001), but not with complex SPF microbiota reversed antibiotics-induced colonic IL-10 down-regulation (Figure 11A). Furthermore, IL-17 and IL-22 mRNA expression were down-regulated upon antibiotic treatment and also subsequent VSL#3 administration (p < 0.001; Figures 11B,C), whereas respective cytokine levels were comparable in mice subjected to FMT and naive controls, which also held true for ileal IFN-γ mRNA expression (Figure 11D). Notably, colonic IFN-γ mRNA levels were highest in mice following FMT (p < 0.05), but did not differ between secondary abiotic, VSL#3 treated and naive mice (Figure 11D).

Taken together, peroral probiotic VSL#3 treatment and FMT are both sufficient to induce regulatory, anti-inflammatory mechanisms of the peripheral and central immune system and to restore intestinal as well as systemic immunological collateral damages of broad-spectrum antibiotic treatment.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.