A total of 24 rabbits (meat lines M91 and P91) were weaned on day 35, including both sexes (equal male-to-female ratio per treatment), were divided into the experimental group (EG) and the control group (CG), 12 animals in each. The average body weight of rabbits at the beginning of the experiment was 1472 g. The experiment was conducted alongside colleagues in Nitra-Lužianky from the National Agricultural and Food Centre (NAFC). The guidelines in the Guide for the Care and Use of Laboratory Animals approved by the Slovak Veterinary and Food Administration and Ethical Commissions of both institutions (permission code: SK CH 17016 and SK U 18016) were accepted for care and experimental animal procedures. The animals were fed a commercial diet for growing rabbits (total energy value 10.99 MJ/Kg) using the following values: Dry matter content 895.38 g/kg, crude fiber 152.43 g/kg, fat 35.77 g/kg, N-free extract 153.64 g/kg, organic matter 834.97 g/kg, ash 60.41 g/kg, starch 139.78 g/kg, calcium 11.74 g/kg, phosphorus 5.89 g/kg, magnesium 2.56 g/kg, sodium 1.80 g/kg, kalium 10.17 g/kg, iron 260.39 mg/kg, and zinc 122.90 mg/kg. Rabbits were maintained in standard cages (type D-KV-72; 0.61 m $\times$ 0.34 cm $\times$ 0.33 m; Kovobel company Domažlice, Czech Republic), with two animals per cage. A cycle of 16 h light and 8 h dark was applied throughout the experiment. The temperature and humidity in the menagerie were recorded continuously by a digital thermograph positioned at the same level as the cages. The heating and ventilation systems allowed the menagerie air temperature to be maintained throughout the experiment at 16 $\pm$ 4 °C, with the relative humidity at about 70 $\pm$ 5%.

The strain was applied to the drinking water of the experimental rabbits at a dose of 500 µL per day per rabbit for 30 days. Feces for bacterial background control were sampled on day 0/1. Blood was also sampled for the appropriate analyses. Rabbits (n = 4) were culled on day 30, and four rabbits from each group (one rabbit/one replicate) were selected based on daily weight measurements to ensure similar animal weights. Animals were sacrificed electro-stunning (50 Hz, 0.3 A/rabbit/4 s) in an experimental slaughterhouse by dissecting the carotid and jugular veins, as previously described by Pogány Simonová et al. [9]. The caecum and appendix were removed and treated for microbiota enumeration, as described in Section 2.2. Cecal samples were also taken to analyze hydrolytic activities. Musculus longissimus thoracis et lumborum (MLTL) was separated by removing the skin, fat, and connective tissue before being chilled and stored at 4 °C for 24 h until analysis.

As previously introduced [3, 7], the strain Lactiplantibacillus plantarum LP17L/1 was isolated from stored ewes cheese. LP17L/1 was marked by rifampicin to differentiate it from the total lactic acid bacteria count. To prepare its rifampicin variant for application, the same protocol was used as previously reported for the strain Enterococcus faecium EK13 by Strompfová et al. [16]. Briefly, LP17L/1 was grown overnight in MRS broth (De Man Rogosa–Sharpe broth, pH, 6.0 $\pm$ 0.2, Merck, Darmstadt, Germany) to reach A${}_{600}$ = 1.00. Then, it was centrifuged at 10,000 $\times$g for 30 minutes. The supernatant was removed, and the obtained cells were re-suspended in Ringer solution (pH 7.0, Merck) to the requested cell concentration of 10${}^9$ colony forming units (CFUs)/mL. The strain count was checked after cell dilutions in Ringer solution, and their spreading on MRS agar (Merck, Darmstadt, Germany) was enriched with rifampicin (100 µL/L). The dose of the strain was selected based on our previous experiences of beneficial strain application [10]. The concentration of the LP 17L/1 strain cells applied in EG was 10${}^9$ CFUs/g per rabbit, as formerly mentioned.

Fecal samples (1 g, n = 6) from EG and CG were diluted in Ringer solution (pH 7.0, Darmstadt, Merck, Germany) using the standard microbiological dilution method (1:9 ratio [17]). The appropriate dilutions (10${}^{-1}$–10${}^{-6}$) were spread on the selective media according to the International Organization for Standardization (ISO). The LP17L/1 strain colonies on MRS agar enriched with rifampicin (100 µg/L, De Man–Rogosa–Sharpe agar, Merck, Darmstadt, Germany) were counted. The total LAB was enumerated using MRS agar (SO 15, 214, Merck, Darmstadt, Germany). Enterococci on M-Enterococcus medium (Difco Laboratories, Detroit, MI, USA) (ISO NF-V04503) were counted. M17 agar (Difco, Detroit, MI, USA) was enriched with starch (30 g/per L) to enumerate amylolytic streptococci. Mannitol Salt agar (MSA, ISO 688, Difco, MI, USA) was applied to count coagulase-negative staphylococci (CoNS), and Baird-Parker agar supplemented with tellurite and yolk egg was used to enumerate coagulase-positive staphylococci (CoPS, Detroit, MI, USA). Coliforms were counted on MacConkey agar (Difco, Detroit, MI, USA). Cecal and appendix samples were treated as was formerly described for feces and plated using the same cultivation media. The bacterial counts were expressed in colony-forming units per gram (CFUs/g) $\pm$ SD (standard deviation, log10).

Blood (n = 8) from the marginal ear vein (vena auricularis) was sampled from rabbits into Eppendorf tubes with micro-spheric hydrophilic (MSH) particles and heparin to test phagocytic activity (PA), as previously described by Pogány Simonová et al. [9]. Sampling was performed on day 30 (end of the strain application). The 50 µL of MSH particle suspension (ARTIM, Prague, Czech Republic) was mixed with 100 µL of blood in Eppendorf tubes and incubated at 37 °C for 1 h. Blood smears were prepared and stained using May–Gruenwald and Giemsa–Romanowski. To validate PA, the direct microscopic counting procedure was performed to calculate the number of white cells containing at least three engulfed particles per 100 white cells (monocytes/granulocytes). PA was expressed as a percentage (%). Moreover, an index of phagocytic activity was included in the analysis (IFA).

Glutathione–peroxidase (GPx (n = 8)) was determined by the colorimetric method (Spectrophotometer UV-2550, Shimadzu, Tokyo, Japan) using the commercial kit Randox RS504 (Randox Laboratory, Hong Kong, China) after blood sampling with heparin.

Hydrolytic activities of amylolytic, cellulolytic, xylanolytic, pectinolytic, and inulolytic (expressed in µmol/g/DM/min, meaning in micromol per gram of dry matter per minute) were processed as previously described by Lauková et al. [18]. These enzymes were extracted using the procedure described by Huhtanen and Khalili [19] and measured according to the method by Miltko et al. [20].

Blood samples were transferred into Eppendorf tubes, centrifuged (3000 $\times$g, 30 minutes), and delivered to Slovak-Laboratory (SK-Lab) company Lučenec (Slovakia) for analysis using previously validated methods. To assess the nitrogen profile, the total protein (TP in g/L), albumin (g/L), and creatinine (µmol/L) levels were tested. Regarding the enzymatic–hepatic profile, the following parameters were tested: alanine aminotransferase (ALT in µkat/L), aspartate transferase (AST in µkat/L), and alkalic phosphatase (ALP in µkat/L). To test the energy profile, glucose (mmol/L), cholesterol (mmol/L), and triglycerides (mmol/L) were evaluated. The following minerals were involved in mineral profile testing: sodium (Na), kalium (K), chlorides (CL), calcium (Ca), phosphorus (P), and magnesium (Mg), measured in mmol/L. The reference values for the rabbits are summarized in Table 1.

Body weight (BW) was measured every week during the experiment.

Animals were sacrificed, as it is indicated in Section 2.1. Musculus longissimus thoracis et lumborum (MLTL) was removed from each rabbit and processed by removing the skin, fat, and connective tissue, chilled, and stored at 4 °C for 24 h until analysis. The FA composition in the MLTL samples was determined using the previously described method by Ouhayoun [21], using gas chromatography of fatty acid methyl ester (FAME) on GC 6890 N (Agilent technologies, Swiss, AG Basel). Results are expressed as a percentage of total FA content. Morphometry testing was performed as previously described by Žitňan et al. [22].

The treatment effect regarding the tested parameters was statistically analyzed using a one-way analysis of variance (ANOVA) with Tukey’s posthoc test (unpaired). Data are expressed as the mean and standard deviation SD of the mean. Different superscript letters indicate a significant difference (p 0.05). Statistical analyses were performed using GraphPad Prism version 6.0 (Inc., San Diego, CA, USA).

On day 30, the LP17L/1 strain showed low colonization in feces, reaching up to 1.0 CFU/g (log10) (Table 2). However, the total LAB count in the EG increased significantly (p 0.001) compared to the LAB count on day 0/1. The total LAB count was also significantly increased in the EG (p 0.001) compared to CG (Table 1). On day 30, an increase in LAB was also noted in the CG (p 0.001) compared to day 0/1 (Table 2). The enterococci counts increased significantly (p 0.001) in both groups compared to day 0/1. The CoNS counts were not influenced. Enterococci and CoNS were well-balanced in both groups on day 30. Similarly, the total counts of amylolytic streptococci were well-balanced in both groups, and high counts were detected in both groups (Table 1). On day 30, amylolytic streptococci were significantly increased in the feces (p 0.001) of both groups compared to day 0/1. Regarding the CoPS, a mathematical decrease (0.31 log cycle) was noted on day 30 between the CG and the EG. However, a significant increase in CoPS was noted (p 0.01) in the EG and the CG compared to day 0/1. Coliforms were not influenced/reduced.

The LP17L/1 strain count in the caecum was the same as in the feces (Table 3). Bacterial counts were mostly lower than in the feces. LAB, enterococci, and CoPS were well-balanced in both groups on day 30, and CoNS and CoPS counts were even similar to those in the feces. Amylolytic streptococci were decreased in the caecum of the EG (p 0.01) compared to the CG. The coliform count in the caecum was not influenced, similarly to that in the feces.

Similarly, as in the feces and caecum, the LP17L/1 strain was found in the appendix at counts of up to 1.0 CFU/g log10 (Table 4). Notably, amylolytic streptococci were found at almost the same level in both groups and the caecum (Table 4). In general, bacterial counts in the appendix were not influenced. The counts of CoNS and CoPS were similar to those in caecum and feces.

On day 30, the phagocytic activity (PA) values were nearly identical in both groups (CG: 70.00 $\pm$ 1.41%; EG: 70.67 $\pm$ 1.75%) (Table 5). This means that applying the LP17L/1 strain did not stimulate PA but also did not negatively influence it. The same situation was noted for IPA (Table 5).

The GPx value was significantly lower in the EG (231.0 $\pm$ 35.09 U/gHb) compared to the GPx value in the CG (298.0 $\pm$ 58.18 U/gHb, p 0.05) (Table 5, Fig. 1). This result confirms that applying the LP17L/1 strain over 30 days did not affect the health of the rabbits and did not promote oxidative stress.

Fig. 1.

Glutathione-peroxidase (GPx) activity, *p 0.05 (CG:EG).

Regarding the biochemistry of the blood serum samples (see Table 1), the total protein value was measured at the lower level of the reference range (RR) (55.16 $\pm$ 1.76 g/L). In the EG, a slightly higher total protein value was noted (56.25 $\pm$ 3.49 g/L). The albumin, glucose, AST, and ALP values were not influenced. Albumin values were nearer the upper limit of the reference range (Table 1). The glucose value was at the top of the reference range. AST values were in the lower limit of the reference value (RV), and ALP values were well-balanced in both groups but higher than the upper limit of the reference value. The ALT values were also balanced; slightly higher in the EG (2.65 $\pm$ 0.39 µkat/L) than in the CG (2.95 $\pm$ 0.74 µkat/L) and in the lower level of RR. Triglycerides were measured in the physiological range up to the reference limit, whereas a lower value was measured in the EG compared to the CG (Table 6). Creatinine values were nearer the lower limit of the reference range, although slightly higher in the EG than in the CG (p 0.05). Cholesterol values were in the physiological range, nearer the upper limit. However, a lower value was measured in the EG (1.12 $\pm$ 0.16 mmol/L) than in the CG (1.61 $\pm$ 0.36 mmol/L, p 0.001). Regarding the mineral profile, chlorides were within the physiological range with a slightly lower value in the EG (95.68 $\pm$ 3.24 mmol/L) compared to CG (98.81 $\pm$ 3.65 mmol/L). K, Ca, P, and Mg values were not influenced; they were balanced in both groups, with slightly higher values in the EG than in the CG. A similar result was found for Na; however, a marginally lower value was measured in the EG than in the CG (Table 1).

Regarding the hydrolytic activity in the caecum on day 30 (Table 6), the EG exhibited increased activity compared to the CG. Amylolytic activity reached the highest value (p 0.001) in the EG compared to the CG. Higher amylolytic activity values are associated with higher amylolytic streptococci values. The xylanolytic activity was also higher in the EG (p 0.001) compared to the CG. Likewise, pectinolytic activity was higher in the EG (p 0.001) than in the CG (Table 7). A higher cellulolytic activity was found in the EG (p 0.001) than in the CG. The inulolytic activity was not influenced (Table 6).

On day 30, the live body weight of rabbits reached an average of 2791.8 g $\pm$ 44.67 in the EG and an average of 2469.16 $\pm$ 49.69 g in the CG. Thus, a higher body weight was noted in the EG (322.6 g higher), meaning that the LP17L/1 strain probably stimulated increased taste and hunger in the EG rabbits.

The total saturated fatty acids (SFAs) value reached 34.890 $\pm$ 2.149% from the total FAs in the EG on day 30. In the CG, the SFA value was 33.580 $\pm$ 0.849%. In the total monounsaturated fatty acids (MUFAs), the value for the EG measured 49.448 $\pm$ 0.975%, while 47.443 $\pm$ 1.731% was calculated for the CG. The total essential fatty acids (EFAs) value in the EG reached 8.808 $\pm$ 1.264%, while 9.200 $\pm$ 0.946% was measured in the CG. For the total polyunsaturated fatty acids (PUFAs), 10.858 $\pm$ 0.747% was measured in the EG, while 12.368 $\pm$ 1.106% was found in the CG.

Thus, the strain LP17L/1 beneficially influenced all tested morphometry parameters. Only in crypt depth was a slight difference found in the CG. However, the final villus height to crypt depth ratio was slightly higher in the EG than in the CG (3.92 vs. 3.83, Table 7), although it was not significantly different.