Lentilactobacillus kefiri SGL 13 (DSM 27331) was obtained from Sintal Dietetics Srl collection. The strain was grown in MRS broth (BD Difco, Franklin Lakes, NJ, USA) in aerobic condition and stored as stock cultures at −80°C in MRS broth supplemented with 15% (v/v) glycerol. In order to prepare L. kefiri SGL 13 biomass, lactobacilli were grown in MRS medium at 37°C and pH 5.5, in a 2 L fermentor (Omnitec srl, Milano, Italy) with an agitation speed of 200 rpm. Cells were harvested by centrifugation (17000 × g, 15 min at 4°C) and then lyophilized in a freeze-drier (Telstar Liobeta 3PS, Spain). After lyophilization, the number of viable cells/g was determined by plate count method.

Paniculin 13™ was composed of lyophilized L. kefiri SGL 13 (DSM 27331), dry extract of A. paniculata (50% andrographolide content) and inulin. Each daily dose per mouse consisted of SGL 13 (5 × 10⁸ CFU/kg bw) according to the formula for dose translation (30), A. paniculata (130 mg/kg bw) and inulin (1.25 mg/kg bw). Immediately before starting oral gavage procedure, the powder of Paniculin 13™ was suspended in a suitable volume of sterile phosphate buffered saline (PBS) (dose for each mouse = 0.2 ml/day). To assess the stability of Paniculin 13™, determination of the number of viable cells/vial of SGL 13, was carried out in triplicate by plate count method. The viability was determined on days 7, 14, 28, and 38; that is respectively after 0, 6, 20, and 30 days of Paniculin 13™ oral administration. The tolerance of SGL 13 to gastrointestinal conditions and its adhesion ability to epithelial cells HT-29, which are prerequisites for colonization and functional activity, were determined as described in a previous study (27).

Cells were harvested by centrifugation (17000 × g, 15 min at 4°C) and then lyophilized in a freeze-drier (Telstar Liobeta 3PS, Spain). After lyophilization, the number of viable cells/g was determined by plate count method.

Animal experimental procedures were approved by Italian Ministry of Health (approval no. 229/2018-PR) and were performed in accordance with national regulation for care and use of laboratory animals and following internal procedures of Experimental Zooprophylactic Institute of Abruzzo and Molise (IZSAM) in order to minimize animal suffering and distress.

Forty 7 weeks-old C57BL/6J (WT) mice, all male, were obtained from Charles Rivers Laboratories (Wilmington, MA, USA). The number of animals per group was statistically determined and approved by the Italian Ministry of Health. On the basis of well documented safety profile of both A. paniculata (31–33) and L. kefiri (28, 34, 35) and to adhere to the 3Rs rules (Replacement, Reduction and Refinement), we reduced the number of animals required to assess the safety of Paniculin 13™ (G1 and G2). In order to demonstrate the efficacy of Paniculin 13™ in reducing the severity of IBD, we used a larger sample size for experimental groups treated with DSS (G3 and G4), improving the statistical power. Moreover, to maintain the initial assumptions about the minimal differences that the test would be able to highlight as significant, we included extra mice (one for each experimental group) to prevent unexpected deaths. Therefore, animals were randomized into 4 experimental groups: control group called G1 (n = 6 + 1), Paniculin 13™ group called G2 (n = 6 + 1), DSS group called G3 (n = 12 + 1) and DSS + Paniculin 13™ group called G4 (n = 12 + 1).

Mice were housed in collective enriched cages in a controlled environment (12:12 h light/dark cycle, 22 ± 2°C room temperature, 55 ± 5% humidity) and were allowed to acclimate for 7 days. Mice were transferred to individual cages only before fecal samples collection. The whole experimental procedure lasted 38 days. All animals were fed the same diet (2016 Teklad global diet, Envigo) with access to food and water ad libitum.

G1 and G3 received the vehicle PBS for 31 consecutive days (days 8–38), G2 and G4 received Paniculin 13™ for 31 consecutive days (days 8–38). Paniculin 13™ and PBS were administered by daily oral gavage (0.2 ml/day).

In addition, G3 and G4 received 1.5% (w/v) DSS (DSS for colitis, TdB Consultancy, Sweden, MW 40,000) in drinking water for 9 consecutive days (15–23) to induce colitis, followed by 15 days of washout (days 24–38). Fresh DSS solution was prepared every day and administered to mice. The average amount of DSS solution taken was recorded daily.

Tolerability and safety of Paniculin 13™ were assessed. The wellbeing and state of health of mice were observed during the entire experimental period by the animal technologists and care staff, using a clinical score sheet system. As shown in Table 1, a specific pain scale was used in order to minimize animal pain and distress. Different parameters such as general appearance, food/water intake, behavior, nest building, were assessed as indicators of health and welfare in mice (36–39). Body weight was measured weekly for mice of G1 and G2 throughout the whole experimental period and daily during the DSS treatment period for mice of G3 and G4. Moreover, during DSS administration clinical symptoms of colitis were monitored daily such as weight loss, diarrhea and presence of blood in the stool (40, 41).

At the end of the study mice were euthanized by isoflurane inhalation followed by cervical dislocation, spleen and colon were collected and, respectively weighted and measured. Blood sampling (0.5–0.8 ml) were collected from each mouse. Blood glucose was measured by tail puncture using a glucometer (Accu-Chek^® Aviva, Roche Diagnostics, Germany) on days 23, 30, 38 for mice of G1 and G2 and on days 18, 23, 30, 38 for mice of G3 and G4. Total cholesterol (TC) triglycerides (TG) and high-density lipoprotein cholesterol (HDL-C) were measured at the end of the experiment (day 38) with enzymatic methods (42), using standard biochemical evaluations with Cobas c501 analyzer (Roche Diagnostic, Monza, Italy). The experimental design is depicted in Figure 1.

The DAI was calculated by the sum of scores assigned for each DAI category (body weight loss, stool consistency and bleeding) based on scoring system of De Fazio et al. (41, 43). All parameters were scored from day 15 to day 38.

Stool samples were collected by placing a single mouse in an empty cage without bedding material for 15 to 30 min and fecal pellet were checked for color, consistency and presence of blood. The criteria for DAI scoring are described in detail below.

Body weight loss: body weight loss 0%, 0 points; 5–10%, 1 points; 11–15%, 2 points; 16–20%, 3 points; >20%, 4 points. Weight loss percentage was calculated, using the formula:

Stool consistency: formed, 0 points; mild-soft, 1 points; very soft, 2 points; watery stool, 3 points. Bleeding: normal color stool, 0 points; brown color, 1 points; reddish color, 2 points; bloody stool, 3 points. The DAI score ranged from 0 to 10 (total score).

Mice (n = 7 for experimental group G1 and G2, and n = 8 for G3 and G4) were sacrificed on day 38 and the colons were excised immediately, rinsed in sterile PBS, fixed in 10% neutral buffered formalin, trimmed, and routinely processed for histology. After processing, samples were embedded in paraffin and 4 μm thick sections were obtained and stained with Mayer’s hematoxylin and eosin G (41, 43). Histological slides were then examined with a light microscopy.

Colon specimens were collected after sacrifice from each mouse and total RNA was extracted according to spin column (Qiagen, Milan, Italy). Tissue samples were homogenized by using TissueLyser II (QIAGEN, Milan, Italy) in ice cold commercial lysis buffer (RNeasy Micro Kit, Qiagen) supplemented with RNase-free DNase I (RNase-Free DNase Set, Qiagen) (44, 45). Reverse transcription-polymerase first-strand cDNA synthesis was performed by using maxima 1strand cDNA synthesis kit (Thermo, Life Technologies, Waltham, MA, USA). qPCR was then performed by using the Kit Thermo SYBR Green/ROX qPCR Master Mix (Thermo, Life Technologies, Waltham, MA, USA) and specific primers for selected genes. The analyses were performed with the 9,600 Biorad Real-Time PCR Detection Systems (Bio-Rad, CA, United States). The developed primer sequences have been listed in Table 2. PCR cycling conditions were as follows: 94°C for 5 min, 40 cycles at 94°C for 15 s, and 60°C for 30 s. Data were expressed as Ct values and used for the relative quantification of targets with the 2^–ΔΔCt calculation.

Fresh stool samples were collected on days 14, 30 and 38 for G1 and G2 and on days 14, 18, 23, 30 and 38 for G3 and G4. Samples, were immediately stored at −80°C until microbiota analyses. Total bacterial DNA was extracted from 100 mg of pooled fecal material from each group, using QIAamp PowerFecal DNA Kit (QIAGEN, Manchester, UK) with modified protocol (46). Amplification of V3-V4 region of 16S rDNA was performed by PCR in 25 μL of final volume mix containing 12.5 ng of microbial DNA and 200 nmol/L of S-D-Bact-0341-b-S-17/S-D-Bact-0785-a-A-21 primers (47), carrying Illumina overhang adapter sequences. After two amplification steps and their respective purification steps as previously reported (42) amplicons were used to build libraries which were purified and successively pooled at equimolar concentrations (4 nM), denatured, and diluted to 5 p.m. Samples were sequenced on Illumina MiSeq platform using a 2 × 300 bp paired end protocol with 20% of PhiX as run control, according to the manufacturer’s instructions (Illumina, San Diego, CA, USA). Paired-end reads, obtained by sequencing, were analyzed using the 16S Metagenomics GAIA 2.0 tool (Sequentia Biotech, Barcelona, Spain, 2017; Benchmark of Gaia 2.0), which performs the quality control of the reads/pairs (i.e., trimming, clipping and adapter removal) through FastQC and BBDuk. The reads/pairs are mapped with BWA-MEM against the custom databases (based on NCBI).

A non-parametric Mann–Whitney test was applied to verify any differences between 2 independent samples. Comparisons between k independent samples were made using the non-parametric Kruskal–Wallis test. If the Kruskal–Wallis test was significant, a post hoc analysis was performed to compare all possible pairs. Specifically, post hoc comparisons were carried out with the Dunn test (using the Bonferroni correction).

Prior to randomization, 3 mice were euthanized due to serious injuries. Thus, sample size of each experimental group was as follows: G1 (n = 7), G2 (n = 7), G3 (n = 13), G4 (n = 13).

Throughout the experimental period, neither death nor adverse effects were observed in any of the animals treated with Paniculin 13™ (G2). Paniculin 13™ had no effect on blood glucose at the considered time points (data not shown). Regarding serum lipid profile (Figure 2), no significant differences in HDL-C and TC were found among groups. On the contrary, mice treated with only DSS (G3) and those treated with DSS + Paniculin 13™ (G4), showed higher TG values, compared to mice assigned to G1 and G2, but only the differences between G4 and G1 and G4 and G2 were statistically significant (p < 0.0083). To support the effectiveness of the association between SGL 13 and A. paniculata (Paniculin 13™), the viability and stability of SGL 13 in the presence of herbal extract was assessed in vitro by plate count. No significant differences in cell concentration were found among the considered time points, namely, day 1, day 14, day 28, and day 38 (10.07 ± 0.08; 10.12 ± 0.03; 10.10 ± 0.02; 10.04 ± 0.04). The results were expressed in log10/vial.

Mice assigned to DSS group (G3) began to lose weight on day 17, only after 2 days of DSS, with a maximum mean weight loss of 24.35% on day 25. On the contrary, in mice assigned to DSS + Paniculin 13™ group (G4), a first reduction in body weight was observed on day 21, after 6 days of DSS, with a peak of 22.45% on day 25. Mice of G4 showed less body weight loss, compared to those from G3, with a maximum weight loss difference on day 32. Mice of G3 continued to lose weight until day 37 and began to gain weight on day 38. On the other hand, mice of G4 showed a progressive increase in body weight from day 35 to the end of the experimental period, with a maximum mean weight gain of 5.45% on day 38. Overall, all the differences in weight changes between G3 and G4, were statistically significant (p < 0.05), throughout the whole experimental period (Figure 3A).

The weekly mean body weight of G1 and G2 was similar during the whole experimental period: G1 day 0 (21.87 ± 1.40), G1 day 15 (23.35 ± 1.07), G1 day 38 (25.33 ± 1.54) and G2 day 0 (20.93 ± 1.28), G2 day 15 (22.76 ± 2.32), G2 day 38 (25.89 ± 2.09).

Furthermore, all mice from G3 and G4 showed changes in stool consistency, 7 days after the beginning of 1.5% DSS assumption (day 22). Stool consistency scores were higher in mice of G3, compared those of G4, these last exhibiting only soft or very soft stools. A normal stool consistency was reached starting from day 30 for both G3 and G4. The presence of blood in feces was observed after 8 days of DSS (day 23) and was noticed until day 28 for both mice assigned to G3 and G4. Mice of G3 reached a maximum score of 3 for bleeding (bloody stools), while mice of G4 showed only reddish stools (score of 2). Overall, the most severe clinical signs were observed between day 23 and 26 for all mice, reaching a peak on day 25, with an average DAI score of 9 for mice of G3 and of 7.75 for mice of G4. DAI values were significantly higher in mice of G3 compared to corresponding values of G4 (p < 0.05), throughout the entire DSS treatment period (day 38). DAI score results are all displayed in Figure 3B.

Mice of G3 started to show mild signs of pain and distress (slight piloerection, reluctance to move, reduction in nest building activity) on day 19, only after 4 days of DSS. On the contrary, mice of G4 showed similar signs after 9 days of DSS. Symptoms of colitis became severe at the end of DSS treatment, peaking on day 25, when the maximum weight loss was observed. Consequently, 3 mice of both G3 and G4 died on day 25 and 2 mice died on day 28 in G3 and on day 31 in G4. The mortality rate was 41.7%. Nevertheless, mice that survived DSS treatment in G4, showed in general a more rapid improving of health conditions than those of G3. Overall, Paniculin 13™ seems to delay signs and symptoms of colitis, e.g., abdominal pain, stool consistency, presence of blood in stools, and weight loss.

A non-parametric Mann–Whitney test was applied to verify any differences between 2 independent samples. The macroscopic changes associated with DSS-induced colitis classically include shortening of the colon (48), and splenomegaly (49). As colitis progresses and the epithelium begins to erode, the colon thins and becomes shorter and immune cells infiltrate the lamina propria (40). Changes in the inflammatory milieu driven by infiltration of immune cells results in an increased spleen weight. We found no significant differences in the mean colon length (cm) among control group G1, Paniculin 13™ group G2 and Paniculin 13™ + DSS group G4 (7.23 ± 0.46, 7.30 ± 0.50, 7.37 ± 0.58). On the contrary, the mean colon length of G3 was shorter (6.36 ± 0.63) than that observed in the other groups, even if this difference was not statistically significant. The mean spleen weight (mg) of mice in G1 and G2 was not significantly different, namely, 60.88 ± 9.65 and 76.83 ± 24.88, respectively. In comparison to G1 and G2, the mean spleen weight of DSS group was higher than that of DSS + Paniculin 13™ group (G3 = 125.01 ± 48.91; G4 = 100.51 ± 31.06), but only the differences between G3 and G1 were significant (p < 0.0083).

Histopathological examination was carried out on small intestine and colon. Lymphoid aggregates in lamina propria were found in all mice as a part of gut associated lymphoid tissue (GALT). Lymphoid aggregates were classified as small when composed of less than 100 cells and large when composed of 100 or more cells.

Histopathological changes in the colon after DSS treatment are shown in Figure 4. Mice of control group (G1) and Paniculin 13™ group (G2) revealed a fully intact colonic epithelium with no inflammatory infiltration. The tissue damage induced by DSS, tended to be limited to the terminal colon. Mice of DSS group (G3) showed moderate cellular infiltration in the colonic mucosa (granulocytes, lymphocytes, macrophages) in association to flattening of the intestinal epithelium. On the contrary, mice of Paniculin 13™ + DSS group (G4) showed reduced post ulcerative areas with a lower inflammatory infiltration and a flattened intestinal epithelium. In both G3 and G4 signs of gut epithelial regeneration were seen at day 38, after 14 days of washout. Overall, supplementation with Paniculin 13™ seems to ameliorate colon tissue injury induced by DSS.

Exposure to DSS significantly raised the mRNA levels of MPO (Figure 5A) and TNFα (Figure 5B) in mice treated only with DSS (G3), an effect which was counteracted in G4 receiving DSS + Paniculin 13™, even if a significant difference between G3 and G4 was not reached. In both cases, only the differences between G3 and G2 were found statistically significant (p < 0.0083). A similar profile of expression was found in the case of iNOS (Figure 5C), where G3 showed significantly higher iNOS mRNA levels (p < 0.0083) compared to both G2 and G4.

Contrary to previous findings (41, 50), administration of DSS failed to induce COX-2 over-expression, in comparison to untreated mice (Figure 5D). Moreover, a decreased expression of this inflammatory mediator was observed in mice of G2 treated only with Paniculin 13™, namely, significant lower mRNA levels (p < 0.0083) compared to G1 (control).

Firmicutes and Bacteroidetes were the most abundant phyla in all samples. DSS induced an increment in inflammatory Proteobacteria which was found 1 week after DSS washout. This effect was mitigated by Paniculin 13™ supplementation (day 30, G3 = 3.09% and G4 = 0.68%) (Figure 6A). Going down the taxonomic scale, the families (Figure 6B) of Enterobacteriaceae and Halomonadaceae, both belonging to Proteobacteria, were found enriched in DSS-treated mice, but were reduced by Paniculin 13™ supplementation. In details, Enterobacteriaceae increased upon DSS treatment in G3 (day 23 = 0.03328%, day 30 = 1.04478% and day 38 = 0.13112%) while remaining less abundant in mice of G4 receiving DSS + Paniculin 13™ (day 23 = 0%, day 30 = 0.02845% and day 38 = 0.00698%).

As for Halomonadaceae, they were measured during DSS treatment in G3 (day 18 = 0.00812% and day 23 = 0.00792%) and in lower amounts in G4 (day 18 = 0.00216% and day 23 = 0.00341%), but after DSS washout, they only remained measurable in non-supplemented mice (day 30, G3 = 0.00434% and G4 = 0%; day 38, G3 = 0.00334% and G4 = 0%). In addition, Bacteroidaceae, previously found increased in mice treated with DSS (51), showed higher levels in G3 (day 18 = 1.03% and day 23 = 1.38%) as compared to G4 (day 18 = 0.77% and day 23 = 0.66%). Moreover, the Carnobacteriaceae family, which is expanded in IBD patients (52, 53), were detected in G3 during and after DSS exposure, but disappeared in G4 from day 23 to day 30. At the genus level (Figure 6C), a progressive increase in the LPS-producing Bacteroides was observed in DSS-treated mice (G3 day 18, 1.00% and G3 day 23 1.34%) in agreement with previous reports (54). This increase, however, was mitigated by Paniculin 13™ supplementation in G4 (day 18 = 0.73% and day 23 = 0.66%). Among other interesting changes, Acidaminobacter, which is increased in colorectal cancer patients (55), was detected in all time-point of G3 but was completely absent in Paniculin 13™ -supplemented G4. Further Allobaculum, a genus enhancing intestinal barrier by producing short chain fatty acids (56) was measured in G1 and G2, and it remained undetectable in G3 during DSS treatment and after washout, while being recovered in G4.

Finally, at the species level (Figure 6D), some pro-inflammatory and colitis-associated bacteria were only measured in G3 but not in G4 during and after DSS treatment, such as Escherichia coli (day 30, G3 = 0.18483% and G4 = 0%; day 38 G3 = 0.11359% and G4 = 0%), and Enterococcus faecalis (day 23, G3 = 0.01981% and G4 = 0%; day 30, G3 = 0.02083% and G4 = 0%). On the other hand, some beneficial species, producing short chain fatty acids, resulted enriched in the same comparisons, including Ruminococcus bromii which was absent in G3 but was measured in G4 during DSS treatment (day 18 and day 23) and Eubacterium xylanophilum which was depleted in both groups during DSS treatment but was restored only in DSS + Paniculin 13™ G4 during washout (day 30 and day 38).