Figure 1.
Histological scores of intestinal injury.
An animal model with intestinal injury similar to NEC was performed as described in “Materials and Methods”. Intestine from healthy dam-fed controls and experimental pups treated with vehicle or different CM were H&E stained and intestinal injury was evaluated on a “0–4” scale by a pathologist blinded to experiments. The healthy dam-fed pups show no intestinal injury and the three combined CM significantly decreased severity of the intestinal injury. “*” and “**” depict p<0.05 and 0.001, respectively, by one-way ANOVA with Bonferroni correction.
Table 1.
Combined conditioned media of probiotics lowered intestinal injury incidence.
Figure 2.
CM from L. plantarum, L. acidophilus, and B. infantis suppressed NF-κB signaling and preserved IκBα expression.
Intestines from healthy dam-fed controls and experimental intestinal injury pups treated with vehicle or different CM were IHC stained for the active/phosphorylated form of NF-κB p65 (p-NF-κB) and IκBα or lysed for immunoblotting for IκBα. A. Representative IHC staining (at 400×) shows that CM decreased IEC NF-κB activity/nuclear p-NF-κB (arrows) in experimental pups with or without intestinal injury (n = 3 per group). B. NF-κB activity from A. was quantified by percentage of nuclear p-NF-κB-positive IEC in three fields of each intestinal section (n = 3). C. Representative IHC staining shows La/Bi CM and all CM protection of IEC IκBα expression in experimental pups with or without intestinal injury, respectively, (n = 3 per group). D. IHC intensity of IκBα from C. was scored on a 0–4 scale to relatively quantify levels of IκBα. E. Immunoblotting shows CM protection of IEC IκBα expression in experimental pups. Intestines of vehicle- and CM-treated experimental pups with disease and healthy dam-fed pups were lysed and subjected to IκBα immunoblotting. IκBα protein levels were quantified by densitometry using ImageJ software and normalized to GAPDH and presented in F. Results are presented as mean of percentages, scores, or relative density ± SE. “*” depicts p<0.05 compared to the vehicle group by one-way ANOVA with Bonferroni correction.
Figure 3.
CM from L. plantarum (Lp CM) and L. acidophilus and B. infantis (La/Bi CM) inhibited proteasome activity.
CM inhibited proteasome activity in the experimental intestinal injury model. At the end of each experiment, liver was harvested, lysed, and proteasome was purified and subjected to a proteasome chymotrypsin-like activity assay as described in “Materials and Methods”. Five µM of the proteasome specific inhibitor epoxomicin (Epox) was added in separate wells as a control to confirm specificity of the assay. Proteasome chymotrypsin-like activity was determined by the rate of generation of fluorogenic product over time. Five µM of Epox was added in separate wells as a control to confirm specificity of the assay. Data (n = 4) are presented as mean ± SE. “*” and “**” depict p<0.05 and 0.001 compared to vehicle groups, respectively, by one-way ANOVA with Bonferroni correction.
Figure 4.
CM from L. plantarum, L. acidophilus, and B. infantis decreased the pro-inflammatory cytokine TNF-α in the intestine.
Intestine was collected from vehicle- or CM-treated experimental pups with intestinal injury and from healthy dam-fed controls and lysed in RIPA buffer. Fifty µgs of lysates were used to determine intestinal TNF-α levels using the xMAP technology. Data (n≥5) are presented as mean ± SE. “*” depicts p<0.05 compared to the vehicle group by one-way ANOVA with Bonferroni correction.
Figure 5.
CM from L. plantarum, L. acidophilus, and B. infantis protected intestinal barrier integrity.
A. Combined La/Bi and La/Bi/Lp CM protected intestinal barrier function. Experimental pups were fed with formula plus vehicle or CM from La, Bi, and Lp and stressed to induce intestinal injury. In vivo barrier function was determined in all surviving experimental as well as healthy dam-fed pups at the end of the animal experiment on day 5. Data from several animal experiments (n≥14) are presented as mean ± SE. Higher levels of the tracer FITC-dextran in blood indicate poorer intestinal barrier function. B. An image of sutured intestinal explant loop for ex vivo loop/barrier function assay. The loop was filled with trypan blue for demonstration. C. Ileum exhibited poor barrier function compared to jejunum/duodenum in three-day old experimental pups. Explants of ileum and jejunum/duodenum were harvested from euthanized rat pups, flushed, filled with dialyzed 10 kDa FITC-dextran (1 mg/mL in PBS), and ligated to form closed loops. Explanted loops were placed in PBS and fluorescence in PBS was measured at 30, 60, and 90 minutes. Barrier integrity of paired ileum and jejunum/duodenum from experimental pups (n = 5) at all time points (n = 15) was measured and compared to ileum. D. Combined La/Bi and La/Bi/Lp CM protected the localization and expression of the TJ protein ZO-1 in the intestine in experimental intestinal injury model. Intestines from animals used in A. were H&E and IF stained for ZO-1 and nuclei and then subjected to confocal microscopy. Representative immunofluorescence staining shows preserved expression and localization of ZO-1 at the TJs of healthy, La/Bi and Lp/La/Bi CM-treated experimental intestines but not in the vehicle- or Lp CM-treated ones. H&E staining demonstrates intestinal injury score “0” (disease free) for all sections. E. Combined La/Bi and La/Bi/Lp CM protected intestinal ZO-1 mRNA expression in experimental pups. Total RNA was isolated from intestine and reversed transcribed to obtain cDNA, which was then subjected to TaqMan real-time PCR to determine expression of ZO-1 and GAPDH mRNA in triplicate (n = 3). Normalized ZO-1 mRNA levels are presented in fold to dam-fed controls. F. CM from La, Bi, or La did not protect expression of the TJ protein occludin in the intestine in experimental intestinal injury. Intestines used as in D. were IF stained for occludin and nuclei and then subjected to confocal microscopy. Representative immunofluorescence staining shows increased occludin expression in the intestinal injury model compared to the healthy dam-fed control and all CM failed to normalize occludin levels to that in healthy controls. G. CM did not normalize occludin expression in experimental pups to that of healthy controls. Intestines from vehicle- and CM-treated experimental pups surviving to the end of the experiments (disease free) and healthy dam-fed pups were lysed and subjected to immunoblotting for occludin and GAPDH. Protein levels were quantified by densitometry using ImageJ software and GAPDH normalized occludin levels are presented as mean ± SE in H. “*” and “**” depict p<0.05 and 0.001, respectively, by t-test in C and compared to the vehicle group by one-way ANOVA with Bonferroni correction in A, E, and H.