Animals and experimental design

SCI and all experimental procedures were conducted in accordance with protocols approved by the National Institutes of Health (US) guidelines and the Guide for the Care and Use of Animals. The Animal Bioethics and Welfare Committee of the Universidad Anáhuac México approved all animal procedures (Registry number: 201702-CI) in accordance with the National Institutes of Health Guide for the care and use of laboratory animals, and the Mexican Official Norm on Principles of Laboratory Animal Care (NOM 062-ZOO-1999).

Twenty female Sprague-Dawley rats, aged eight to eleven weeks old, weighing between 200 g and 300 g, were housed into five groups of four rats each (n = 4/group) under a 12-hour light/dark cycle, in thermally controlled rooms. They were fed standard pellets and given water ad libitum. To minimize stress, all animals were handled at least once daily for 30 days prior to surgery, and environmental enrichment was provided. Sterile bedding and filtered water were replaced daily. Females were selected to minimize variability in body weight, stress-related autonomic responses, and post-injury urinary complications, which are more frequent in males and could confound gastrointestinal outcomes and fecal sampling. This approach also ensured consistency with prior SCI studies from our group that employed the same sex and strain, allowing for valid comparisons across experiments. The symbiont was administered daily by orogastric tube for four consecutive weeks (weeks 4–8 post-injury). To minimize potential stress or motility alterations due to the orogastric procedure, animals were gently handled and acclimated for three days before administration, and no signs of distress or reduced food intake were observed.

The rats were allocated into five groups using the STATS TM v.2.0 software: 1) Sham (n = 4), 2) severe injury at T5 level (thoracic five severe, T5S) (n = 4), 3) moderate injury at T5 level (thoracic five moderate, T5M) (n = 4), 4) severe injury at T9 level (thoracic nine severe, T9S) (n = 4), and 5) moderate injury at T9 level (thoracic nine moderate, T9M) (n = 4). The five groups were studied over a period of eight weeks. Four weeks post-injury (wpi), all groups received a symbiotic consisting of E. faecium and inulin. Stool samples were collected at 4 wpi (prior to symbiotic administration) and at 8 wpi (after four weeks of symbiotic treatment). Gut microbiota composition and butyrate concentrations were analyzed from fecal samples at both time points. Motor recovery was assessed weekly in all groups until the end of the experiment (8 wpi). The sample size calculation for studies evaluating the taxonomic composition of the gut microbiota was performed using the formula proposed by Scheff et al. [25], 2003 which estimates power based on pairwise distance matrix comparisons between groups. This calculation indicated that four rats per group would provide 90% power to detect significant differences in microbial composition between pre- and post-treatment conditions, taking into account both injury level and severity.

The study followed a pre- and post-test longitudinal design in which each rat served as its own control. Fecal samples collected at 4 wpi (pre-treatment) and 8 wpi (post-treatment) were compared within each group to assess the effect of the symbiotic. This approach minimized inter-animal variability and reduced total animal use in accordance with ethical principles.

Traumatic SCI

Rats were weighed and anesthetized with ketamine (90 mg/kg) and xylazine (10 mg/kg) administered intraperitoneally. After fifteen minutes, a laminectomy was performed at the thoracic vertebra to expose the dura mater. Stabilization clamps were used to access the T5 or T9 segment of the spinal cord. Injuries were inflicted using the Infinite Horizons SCI device 5.0, programmed for moderate injury at 200 kilovolts and severe at 300 kilovolts.

Post-surgery, all rats received 10 mg/kg of paracetamol orally twice a day and 5 mg/kg of ketoprofen, once a day via subcutaneous injection for one week, along with fluid replenishment with saline solution. Manual emptying of the bladder and rectum was performed three times daily until sphincter control was recovered. Sham-operated animals underwent all surgical procedures to expose the spinal cord but were not injured and served as controls. Animals were carefully monitored for signs of infection, dehydration, or self-mutilation, with appropriate veterinary care provided as needed. All experiments were designed and reported in accordance with established guidelines. At the end of the experiment (8 wpi), for euthanasia, the animals were first anesthetized intraperitoneally, with an injection of a ketamine (90 mg/kg) and xylazine (10 mg/kg) mixture. Once the animals no longer exhibited reflexes, an intraperitoneal injection of sodium pentobarbital (200 mg/kg, lethal dose) was administered.

Symbiotic administration

Treatment began four wpi, rats were weighed every third day, and a homogeneous mixture of E. faecium (0.8 g equivalent to 1.6 × 10¹⁰ CFU/kg body weight) plus agave inulin (0.86 g/kg body weight) was administered via an orogastric tube, diluted in drinking water until the end of the experiment (8 wpi). The symbiotic consisted of E. faecium [1 × 10⁷ CFU; Alimentos Esenciales para la Humanidad (AEH) S.A. de C.V., Ciudad de México, México] and agave inulin (8 g; AEH S.A. de C.V., Ciudad de México, México).

Characterization of the gut microbiota

Fecal samples were collected individually from each rat at 4 and 8 wpi using sterile forceps in a laminar flow hood. Each sample was stored at –80°C until processing. For microbial characterization, genomic DNA was first extracted individually from 0.25 g of feces using the QIAamp PowerFecal DNA Kit (QIAGEN, 11993). To obtain representative community profiles for each experimental group and time point, equal amounts of purified DNA from each individual within the same group were pooled prior to 16S ribosomal RNA (16S rRNA) gene amplification and sequencing on the Illumina MiSeq platform. This approach was chosen to ensure sufficient DNA yield and sequencing depth while reducing technical variability between samples. Consequently, microbial composition data reflect group-averaged taxonomic profiles, whereas metabolic (butyrate) and behavioral analyses were performed on individual samples. For sequencing, equal volumes of DNA extracted from each animal of the same experimental group and time point were pooled to obtain one composite sample per group. The pooled DNA was subsequently divided into four technical aliquots, each processed independently through the 16S V3–V4 library preparation (Nextera XT kit, Illumina) and sequenced on the MiSeq platform.

Stool samples were collected at 4 wpi (before symbiotic administration) and at 8 wpi (after four weeks of symbiotic administration) through manual extraction in a disinfected laminar flow hood. After handwashing and donning gloves, gentle pressure was applied to the rectal ampulla, and samples were collected with sterile Adson tissue forceps, previously sterilized with 12% Krit germicidal solution. Samples were stored in 2 mL centrifuge tubes at –80°C for further processing.

To isolate and purify bacterial DNA, 0.25 g of feces was processed using the QIAamp PowerFecal DNA Kit (QIAGEN Model: 11993), following the manufacturer’s protocol. Sample quality was assessed using the Thermo Scientific NanoDrop One Microvolume UV-Vis Spectrophotometer, measuring at a wavelength range 260–320 nm to obtain quantity (ng/mL) and assess purity using the ratios 260/280 nm and 260/320 nm (values above 1.8 were included). Electrophoresis (1% agarose gel) was performed to evaluate DNA integrity.

Sequencing

Sequencing was performed on MiSeq following the workflow for preparing 16S rRNA Gene Amplicons for the Illumina MiSeq System (Part # 15044223 Rev. B): amplicon primers, library preparation, and sequencing on MiSeq.

For the 16S library preparation, modifications were made. The conditions for the first stage PCR were 98°C for 30 seconds, followed by 5 cycles of 98°C for 20 seconds, 55°C for 30 seconds, and 72°C for 30 seconds, after which a final elongation step at 72°C for 5 minutes, and then held at 4°C. The second stage PCR involved 98°C for 30 seconds, followed by 8 cycles of 98°C for 20 seconds, 55°C for 30 seconds, and 72°C for 30 seconds, concluding with a final elongation step at 72°C for 5 minutes, and then held at 4°C.

Metagenomic analysis

Once genetic material was obtained, PCR was performed, and sequences were purified. Operational Taxonomic Units (OTUs) were identified. For classification analysis, we used the 16S Metagenomics software, available on the BaseSpace platform (Illumina, Inc.), utilizing a modified version of the Greengenes taxonomic database. The following metrics were reported for each sample: 1) alpha diversity: using the Shannon index and the number of identified species (R), and 2) beta diversity.

Chimeric sequences were discarded, and all non-chimeric input sequences were required to match at least one OTU with over 97% identity. Dissimilarity scores were calculated using the Bray-Curtis distance metric, considering relative abundance at the genus level for each rat and lesion group. Hierarchical clustering and principal coordinate analysis (PCoA) were employed to graphically summarize inter-sample relationships, with a heat map representing the relative abundance of the eight principal genera before and after symbiotic administration.

Butyrate determination

Fecal samples were collected by groups 4 wpi and 8 wpi. Prior to collection, sterile bedding was changed, and samples were harvested after an 8-hour waiting period using sterile tissue forceps. Pooled samples were stored in 50 mL centrifuge tubes at –80°C until processing.

Stool was lyophilized at –65°C under 1.33 × 10^–6 torrs. One gram was weighed into a 50 mL centrifuge tube and 5 mL of distilled, deionized water was added. The mixture was homogenized for 10 minutes at maximum speed in a vortex mixer. Subsequently, the pH was evaluated with a pH meter, and it was acidified with 5 M HCl to the range of 2 to 3. The sample was centrifuged at 4,000 rpm for 25 minutes at 16°C, and 2 mL of chloroform was added. After a second centrifugation at 4,000 rpm for 2 minutes at 16°C, the aqueous phase was transferred to sterile 50 mL centrifuge tubes.

Butyric acid was determined by gas chromatography (GC) with a Shimadzu GC2010 (Japan), equipped with a silica capillary column for free fatty acids (model DB-1701, 30 m length, 0.25 mm internal diameter and 0.25 µm thickness, Agilent, USA). GC conditions included helium as carrier gas with a flow of 4 mL/min. The oven temperature was set to 250°C, with the column maintained at 95°C. The flow rates for hydrogen and air were set at 40 and 400 mL/min, respectively. The flame ionization detector (FID) was operated at 300°C. The retention time for butyric acid was between 3.5 and 3.6 minutes, and the injected sample volume was 2 µL, with butyric acid concentration at 0.1% w/v and the molecular weight of 88.11 g/mol. Each analysis lasted 20 minutes, with butyrate levels determined using Sigma-Aldrich standard 107-92-6. For each pooled sample, butyrate concentration was reported in µg butyric acid/2 µL. All the pools were processed in triplicate.

Clinical assessment

To evaluate motor recovery, animals were examined weekly by blinded raters using the Basso, Beattie, and Bresnahan (BBB) locomotor rating scale [25, 26] from the time of surgery until the end of the study.

Statistical analysis

Statistical analyses were conducted using GraphPad Prism v9.0 and PAST v.4.11. Normality was evaluated using the Shapiro-Wilk test. Since most data (e.g., BBB scores, microbial diversity indices) did not follow a normal distribution, non-parametric analyses were performed. For alpha-diversity indices [Richness (R) and Shannon; Figure 1], comparisons between groups were analyzed by the Kruskal-Wallis test, and pre- vs. post-treatment comparisons within each group by the Friedman test. Beta-diversity analysis (Figure 2) was performed using Bray-Curtis dissimilarity matrices and analyzed with PERMANOVA (Adonis) to estimate the explained variation (EV, EV%) between groups. Butyrate concentrations were compared between groups using the Kruskal-Wallis test, and pre- vs. post-treatment differences were evaluated with the Wilcoxon signed-rank test. Locomotor performance (BBB scores) was reported as median ± interquartile range (IQR). Longitudinal analysis of BBB recovery was performed using the Friedman test, and between-group comparisons at final time points by the Kruskal-Wallis test. All analyses were two-tailed, and p < 0.05 was considered statistically significant.

Display full size

Figure 1. Comparison of means of Richness (R) and Shannon index before (4 wpi) and after (8 wpi) symbiotic administration. Box plots showing R (A) and Shannon index (B) across experimental groups (T5M, T5S, T9M, T9S, Sham) at 4 wpi (before symbiotic administration) and 8 wpi (after symbiotic administration). wpi: weeks post-injury; T5M: thoracic five moderate; T5S: thoracic five severe; T9M: thoracic nine moderate; T9S: thoracic nine severe.

Display full size

Figure 2. Relative abundance analysis of the gut microbiome (Beta diversity) before and after symbiotic administration. Agglomerative hierarchical clustering highlights the OTUs with the most significant differences in abundance between groups. Stacked bar plots showing the relative abundance of bacterial taxa across experimental groups (T5M, T5S, T9M, T9S, Sham) at 4 wpi (A—before symbiotic administration) and 8 wpi (B—after symbiotic administration). Each bar represents a technical replicate derived from pooled DNA samples of all animals within each group and time point. Colors indicate different bacterial taxa, and bar length corresponds to relative abundance. Following symbiotic supplementation, microbiome composition showed a trend toward homogenization across groups, with increased representation of beneficial genera such as Bifidobacterium and Lactobacillus. Enterobacteriaceae is presented as a family-level taxon (not genus-level) and is labeled accordingly in the figure. OTUs: Operational Taxonomic Units; wpi: weeks post-injury; T5M: thoracic five moderate; T5S: thoracic five severe; T9M: thoracic nine moderate; T9S: thoracic nine severe.

Characterization of the gut microbiota

A total of 40 samples were sequenced. The characterization of the microbiota was performed on individual samples from each lesion group, both before and after symbiotic administration (4 wpi and 8 wpi).

Alpha diversity

In Figure 1, we report the number of identified species (R) and the Shannon index. Notably, after the administration of symbiotics, the variability of (R) decreased at 8 wpi across all experimental groups, except the T9M group, where it only decreased in some cases. In contrast, the Shannon index showed greater dispersion following administration; however, no statistical differences were observed between the measurements taken before (4 wpi) and after (8 wpi) the administration of symbiotic.

Quantitatively, comparisons between 4 wpi (before symbiotic administration) and 8 wpi (after symbiotic administration) showed that the R index decreased by approximately 10–15% in T5S and T9S groups, remained stable in T5M, and increased by about 8% in T9M. Similarly, the Shannon index showed a slight increase in the T9M group (0.21 ± 0.04 units), suggesting a mild improvement in microbial evenness following symbiotic supplementation.

Thus, alpha-diversity analysis confirmed similar overall R between groups, with only mild trends toward increased evenness in the moderate injury models.

Although these results were not statistically significant, moderate injuries (particularly T9M) tended to show slightly higher R and Shannon values following symbiotic supplementation, suggesting a mild increase in evenness and species distribution compared with severe injuries.

Beta diversity

Relative abundance was calculated, and the most abundant genera per sample are presented below in Figure 2. The most prevalent genera were Bifidobacterium, Lactobacillus, Clostridium, and Alkaliphilus. After treatment (8 wpi), the microbiome composition across groups became more homogeneous. The predominant genera—Bifidobacterium, Lactobacillus, Clostridium, Alkaliphilus, Allobaculum, and Sarcina—together represented approximately 90% of the total abundance.

Quantitative analysis of microbial composition revealed time-dependent changes following symbiotic administration. In the T9M group, Clostridium decreased from 18.6% at 4 wpi to 7.0% at 8 wpi, while Bifidobacterium increased from 32.1% to 34.9%. In parallel, Lactobacillus increased from 17.1% to 32.8%, indicating a shift toward beneficial taxa after treatment.

In the T5M group, Bifidobacterium increased from 30.8% at 4 wpi to 37.1% at 8 wpi, whereas Lactobacillus decreased from 28.7% to 18.5%. Changes in Clostridium were modest, increasing from 9.3% to 12.1%. At the family level, Enterobacteriaceae showed a reduction after symbiotic administration across groups; this taxon is represented accordingly in Figure 2 (see legend).

In severe models (T5S and T9S), compositional shifts were less pronounced, indicating a severity-dependent microbial response. All comparisons correspond to changes between before (4 wpi) and after (8 wpi) symbiotic administration, consistent with the graphical representation.

The Shannon diversity index mirrored these findings, indicating greater evenness in T9M and T5M after symbiotic supplementation.

Before symbiotic administration at 4 wpi, the most abundant genera across groups were Bifidobacterium, Lactobacillus, Clostridium, and Alkaliphilus. After symbiotic administration at 8 wpi, all the groups exhibited a similar microbiome composition where Bifidobacterium, Lactobacillus, Clostridium, Alkaliphilus, Allobaculum, Turicibacter, Natronincola, Sarcina, Akkermansia, Johnsonella, and Blautia together constituted 89.96% of the relative abundance. The eight principal genera of each group were represented in a heat map, as seen in Figure 3.

Display full size

Figure 3. Analysis of species abundance clustering in the heat map. The longitudinal axis shows the group information, while the transverse axis represents the species annotation information. The clustering tree on the left of the figure displays species clustering, and the tree above illustrates the clustering among sample groups. The values in the heat map represent the z-scores obtained after standardization of the relative abundance of species in each row, reflecting gut microbial distribution at the genus level for each experimental group at 4 wpi and 8 wpi. The blue color indicates a higher percentage of relative abundance. wpi: weeks post-injury.

The main genera before and after symbiotic administration were grouped by category, as represented in Table 1, arranged from highest to lowest abundance.

To determine significant dissimilarities in microbiome composition between the experimental groups, PERMANOVA analysis was performed to assess the effects of injury intensity and level. First, we evaluated dissimilarities between the groups before symbiotic administration, which resulted in an explained variation (EV) of 23.5% with a significant difference (p = 0.004). After symbiotic administration, the dissimilarities between the groups showed an EV of 12.7% with a significant difference (p < 0.05), suggesting that symbiotic administration promotes homogenization of microbiota genera irrespective of injury level or severity.

Next, we compared each group to itself before and after symbiotic administration using PERMANOVA, yielding the following results: T5S with an EV of 25.6% (p = 0.2); T5M with an EV of 12.3% (p = 0.439); T9S with an EV of 41.4% (p = 0.058); T9M with an EV of 83.3% (p = 0.035); and Sham with an EV of 45.4% (p = 0.1). The principal observation of EV was noted in the T9 level and Sham. See Table 2.

Butyrate determination

Before symbiotic administration (4 wpi), butyrate levels varied across experimental groups and did not consistently decrease in all SCI conditions compared with the Sham group. While some injury groups showed lower concentrations, others (e.g., T5S) exhibited comparable or even higher baseline levels. In contrast, animals with moderate T9 injury showed relatively higher baseline concentrations overall (Figure 4).

Display full size

Figure 4. Butyric acid concentration before and after symbiotic administration. Bar graph showing butyric acid levels in pooled fecal samples from each experimental group at 4 wpi (before symbiotic administration) and 8 wpi (after symbiotic administration). Values are expressed as mean ± SD of technical replicates. Statistical significance is indicated as follows: ^* p < 0.05, ^** p < 0.01. Comparisons correspond to within-group differences between 4 wpi and 8 wpi. wpi: weeks post-injury; SD: standard deviation.

After four weeks of symbiotic supplementation (8 wpi), butyrate concentrations showed a heterogeneous response across experimental groups. While some groups (e.g., T9S) exhibited an increase, others, such as T9M, showed a decrease. Changes in T5 groups were modest, whereas the Sham group remained relatively stable across both time points. A significant increase in butyrate concentration was observed in the T9S group (p < 0.05), whereas other groups showed no consistent significant changes.

Overall, these results suggest that symbiotic supplementation may enhance butyrate production, particularly in animals with moderate T9 injury.

Clinical assessment: BBB

To verify the consistency of the contusion model, we evaluated motor recovery in all groups following injury. A mixed-effects model (REML) with Tukey’s multiple comparisons test was used to assess motor recovery over time, comparing the period before (4 wpi) and after (8 wpi) symbiotic administration.

During the first four weeks, we observed the expected spontaneous recovery of locomotor function across all groups. However, following treatment, this improvement continued to progress, despite the lesion having reached its chronic phase.

The recovery was statistically significant in the T9M (p < 0.0001), T9S (p = 0.0166), and T5M (p = 0.0013) groups (Figure 5).

Display full size

Figure 5. Motor recovery throughout the experiment. Statistical analysis was evaluated with REML with Tukey’s multiple comparisons to evaluate all animals per group over time. Values are expressed as mean ± SD. Significant improvements were observed during the symbiotic administration period, particularly in T9 groups. Asterisks indicate statistically significant differences as follows: ^* p < 0.05, ^** p < 0.01, ^**** p < 0.0001. Specifically, ^* indicates within-group differences during the early phase of treatment (around week 4–5), ^** indicates significant differences between groups at week 8 (notably between T5S and treated groups), and ^**** indicates highly significant improvements across time points during the symbiotic administration period (4–8 wpi), particularly in T9M and T9S groups. wpi: weeks post-injury; REML: mixed-effects model; SD: standard deviation; T5S: thoracic five severe; T9M: thoracic nine moderate; T9S: thoracic nine severe.

Overall, the symbiotic formulation with E. faecium and agave inulin promoted compositional and metabolic restoration of the gut microbiota primarily in moderate injuries. However, since no untreated SCI controls were maintained at 8 weeks, the observed effects should be interpreted cautiously as potential improvements rather than definitive treatment effects.

Conclusion

The present study demonstrates that symbiotic administration with E. faecium and agave inulin significantly influenced gut microbiota composition and metabolic function following SCI. Treatment led to the recovery of beneficial bacterial genera such as Bifidobacterium, Clostridium, and Lactobacillus. Locomotor recovery followed the expected pattern based on lesion level and injury severity, with better outcomes observed in moderate and lower thoracic injuries. Separately, the microbial and metabolic changes observed after symbiotic administration suggest a potential modulatory effect on the microbiota-gut-spinal cord axis. However, the observed changes in butyrate concentration were inconsistent with the expected relationship between butyrate levels and injury severity.

Although limited by the small sample size and pooled sequencing design, the consistent post-treatment trends indicate that symbiotic administration exerted a restorative and modulatory effect on intestinal microbial ecology and SCFA metabolism. These findings underscore the therapeutic potential of symbiotic supplementation as a complementary strategy to improve gastrointestinal and neurological function after SCI. Future studies with larger cohorts and individual-level analyses will help confirm these effects and explore their underlying mechanisms.