Bacteriocin-transport-inspired oral peptide-probiotic delivery ameliorates IBD complications via autophagy and gut homeostasis

Abstract

Intestinal fibrosis (IF), a severe complication of inflammatory bowel disease (IBD), remains a critical unmet clinical need. Although the LL37 peptide and probiotics demonstrate therapeutic potential against IF, their clinical translation is hampered by enzymatic hydrolysis and rapid clearance. Here, inspired by the strategy of bacteriocin transport by bacteria (BTB), we developed an orally administered biotherapeutic platform [BTB-alginate (Alg)] featuring an "all-in-one" architecture that enables spatiotemporal coordination of LL37 and probiotics. The BTB-Alg effectively restored intestinal homeostasis through inflammation resolution, immune modulation, and gut microbiota reconstitution. Notably, integrated multiomics analysis and molecular dynamics simulations revealed that LL37 exerts antifibrotic effects by inducing adenosine 5'-monophosphate-activated protein kinase/mammalian target of rapamycin-mediated autophagy, a mechanism validated in clinical specimens. BTB-Alg exhibited potent therapeutic efficacy in three murine models of acute colitis, IBD-associated IF, and Clostridioides difficile-complicated colitis, highlighting its potential as an IBD treatment paradigm. This study offers a clinically translatable strategy for broad gastrointestinal applications.

Fig. 1.. Bioinspired construction of BTB-Alg for IBD therapy: Inflammation, fibrostenosis, and C. difficile infection.

(A) Bioinspired design of BTB biocomposites. (B) Protective mechanisms of orally delivered probiotics and LL37 in the GI tract via microbial mimicry. (C) Following oral administration, BTB-Alg undergoes sequential intestinal adhesion, payload release, tissue penetration, and lesion-specific targeting. (D) BTB-Alg alleviates AC, resolves fibrosis, and suppresses C. difficile infection (CDI). (E) Mechanism of BTB-Alg–mediated intestinal niche reprogramming. ROS, reactive oxygen species; MDS, molecular dynamics simulation; SCFA, short-chain fatty acid; BP, biological process; CC, cellular component; MF, molecular function.

Fig. 2.. In vitro characterization and systematic optimization of BTB-Alg.

(A to D) Antifibrosis efficacy of LL37 by quantitative polymerase chain reaction (qPCR) analysis of fibrotic markers (α-SMA, Col1α1, and FN) in TGF-β–stimulated CCD-18Co cells treated with LL37-1 (1.11 μM) or LL37-2 (11.1 μM). (E) Western blot assays of Col1α1 and FN protein expression. AU, arbitrary units; NC, negative control. (F to H) Cytotoxicity assessment of LL37 in (F) HT29 cells, (G) unstimulated CCD-18Co cells, and (H) TGF-β–pretreated (24 hours) CCD-18Co cells. (I and J) Cellular uptake of Cy5.5-labeled LL37 (20 μg ml⁻¹) in unstimulated versus TGF-β–stimulated intestinal fibroblasts (I), with fluorescence intensity quantification (J). Scale bars, 100 μm. (K) Chromatogram profiles of LL37 detected by high-performance liquid chromatography (HPLC) following different treatments. (L) Fabrication schematic of BTB-Alg. (M) Zeta potential of Lac, LL37, and Alg. (N) Growth kinetics of Lac with LL37 coculture. (O) Fluorescence imaging of LL37 loading optimization [37.5, 75, 150, 300, 600, 1000, 1500, and 2500 μg of LL37 per 1 × 10⁹ colony-forming units (CFU) of Lac]. (P) Fluorescence intensity of different BTB models. (Q) Zeta potential stabilization at saturation (Lac/LL₆ versus Lac/LL₇). n.s., not significant. (R) HPLC quantification of unbound LL37 in supernatants. (S) Alg coating optimization using Cy5.5-labeled Alg (0.032, 0.063, 0.125, 0.25, 0.5, 1, or 2 mg per 1 × 10⁹ CFU of Lac). (T) Fluorescence intensity of Alg deposition. (U and V) Protective efficacy of Alg against pepsin degradation. HPLC chromatograms (U) and residual ratio of LL37 quantification (V). In (B) to (D), (F) to (H), (J), (M), (N), (P), (Q), (T), and (V), data are shown as mean ± SD (n = 3 biological replicates). Statistical analysis was performed via a one-way analysis of variance (ANOVA) or an unpaired Student’s two-sided t test. Experiments in (E), (I), (K), (O), (R), (S), and (U) repeated three times with consistent results.

Fig. 3.. Structural characterization, GI stability, and targeted release of BTB-Alg.

(A) Surface charge evolution during the decoration. (B and C) Particle size (B) and transmission electron microscopy (TEM) images (C) showing structural transitions. Scale bars, 2 and 1 μm. (D) Triple-channel confocal imaging confirming core-shell architecture: FITC-labeled Lac [green; excitation/emission (Ex/Em): 490/520 nm], Cy3-labeled LL37 (red; Ex/Em: 550/570 nm), and Cy5-labeled Alg (purple; Ex/Em: 646/664 nm). Scale bars, 100 and 50 μm. (E) XRD patterns demonstrating LL37 crystalline structure attenuation after Alg encapsulation. (F) Growth curves of Lac and Lac/LL@Alg at 37°C recorded at 2-hour interval for 24 hours. (G) Schematic illustration of the physiological responsiveness behavior of Lac/LL@Alg. (H) Fluorescence images of Lac/LL@Alg after incubation with SGF supplemented with pepsin for 2 hours and SIF supplemented with pancreatin for 2, 4, 6, and 8 hours. FITC-labeled Lac (green), Cy3-labeled LL37 (red), and Cy5-labeled Alg (purple), respectively. Scale bar, 50 μm. (I and J) Survival of Lac and Lac/LL@Alg measured on the basis of plate counts after incubation with SGF supplemented with pepsin and SIF supplemented with pancreatin at 0 point (I) and indicated times (J). (K) Cumulative release profiles of LL37 from Lac/LL or Lac/LL@Alg after incubation with SGF and SIF. In (A), (B), (F), and (I) to (K), data are shown as mean ± SD (n = 3 biological replicates). Experiments in (C) to (E) and (H) repeated three times with consistent results.

Fig. 4.. In vivo GI retention and mucoadhesion profiling of BTB-Alg.

(A) The workflow of in vivo tissue distribution experiments. p.o., per os. (B to D) IVIS imaging of the GI tract (top) and major organs (bottom) after oral administration of (B) FITC-labeled native Lac, (C) Cy5.5-labeled free LL37, and (D) triple-labeled Lac/LL@Alg at specified time points (n = 3). (E) The quantification of total colonic fluorescence intensity for Lac and LL37 over time. (F) Time-dependent fluorescence changes of Lac, LL37, and Alg components in Lac/LL@Alg at 2, 4, 6, and 8 hours postadministration. (G and H) Colonic mucosal retention analysis of (G) representative images and (H) quantification of total intensity of different groups after 8 hours postadministration (n = 3). (I to K) Representative images (I), penetration ratios (J), and a schematic diagram (K) demonstrate the effective diffusivity of Lac, LL37, and Lac/LL@Alg in mucus. (L) Effective penetration of Lac/LL@Alg in intestinal tissues derived from patients with CD. Data in (E), (F), (H), and (J) presented as mean ± SD (n = 3 biological replicates). Significance determined by unpaired two-tailed Student’s t test. Experiments in (B) to (D), (G), (I), and (L) repeated three times with consistent results.

Fig. 5.. Oral administration of BTB-Alg alleviates DSS-induced AC.

(A) Schematic representation of DSS-induced AC in mice and therapeutic design. (B) Daily body weight changes throughout the study. (C) DAI changes during the whole study. (D to F) Colon length (D), weight length index (E), and morphology (F) across different groups. (G) Hematoxylin and eosin (H&E) and Alcian blue-periodic acid schiff (AB-PAS) staining of intestines from each group. Scale bars, 500 μm. (H) H&E-based histopathological scoring of colonic tissues posttreatment. (I) Spleen weight index in each group. (J to M) Measurement of TNF-α (J), IL-10 (K), MPO (L), and MDA (M) in different groups. prot, protein. (N) Analysis of Sobs index at the genus level. (O) Principal components analysis elucidated the similarity of gut microbial communities at the genus level. PC1, principal component 1. (P and Q) The composition of microbial communities was characterized at both phylum (P) and genus (Q) levels from the control, AC model, and Lac/LL@Alg. Bars represent relative abundance. (R) Relative abundance of Lac in feces from AC model and Lac/LL@Alg groups. (S) Linear discriminant analysis Effect Size (LEfSe) was implemented to identify differentially enriched taxa in gut microbiota among control, AC model, and Lac/LL@Alg-treated groups. Data in (B) to (M) are derived from n = 6 biological independent samples. In (B) to (E) and (H) to (M), data are shown as mean ± SD, with statistical analysis performed via one-way ANOVA. P values denote the statistical significance between AC model and LL37, Lac, and Lac/LL@Alg. In (N) to (S), data are shown as mean ± SD (n = 4 biological independent samples), with statistical analysis analyzed via one-way ANOVA or a Student’s two-sided t test.

Fig. 6.. Oral BTB-Alg alleviates DSS and C. difficile–induced colitis in mice.

(A) Schematic representation of DSS and C. difficile–induced colitis in mice and therapeutic design. (B) TcdB expression in feces from control, CD, and saline (DSS + CD)–treated groups on day 8 (n = 3). (C) Survival curves of mice with different treatments. (D) Representative colon morphology images from different groups (n = 5). (E to I) Body weight changes (E), colon lengths (F), and proinflammatory markers [IL-6 (G), IL-17 (H), and IL-1β (I) mRNA levels] between control and CD-treated groups. (J to N) Body weight changes (J), colon lengths (K), and proinflammatory markers [IL-6 (L), IL-17 (M), and IL-1β (N) mRNA levels] between CD and saline-treated groups. (O to S) Body weight changes (O), colon lengths (P), and proinflammatory markers [IL-6 (Q), IL-17 (R), and IL-1β (S) mRNA levels] from different treatment groups. (T) Western blot analysis of ZO-1 expression in different treatment groups (n = 3). (U) H&E staining of intestinal tissues from each group (n = 3). Scale bars, 500 μm. (V) Western blot analysis of NF-κB pathway in different treatment groups (n = 3). p–NF-κB, phosphorylated NF-κB. (W) Immunofluorescence staining of macrophages with distinct phenotypes (n = 3). CD206 staining (green): M2 macrophage; CD86 staining (red): M1 macrophage. Scale bars, 200 μm. In (G) to (I) and (L) to (N), data are shown as mean ± SD (n = 4 biological replicates), with statistical analysis conducted via an unpaired Student’s two-sided t test. In (Q) to (S), data are shown as mean ± SD (n = 4 biological replicates), with statistical analysis conducted via one-way ANOVA. P values denote the statistical significance among the saline, VAN, LL37, Lac, and Lac/LL@Alg groups.

Fig. 7.. Oral BTB-Alg alleviates DSS-induced IF.

(A) Experimental design of DSS-induced IF model and the timeline of therapeutic intervention. (B) Body weight changes during whole treatment. (C and D) Colon morphology (C) and length quantification (D) in different groups. (E) Histopathological analysis via H&E, Masson’s trichrome, and Van Gieson (VG) staining. Scale bars, 500 μm. (F and G) Quantitative histological scoring of inflammation (F) and fibrosis severity (G). (H and I) Immunofluorescence imaging of Col I (H) and α-SMA (I). Scale bars, 200 μm. (J to L) qPCR analysis of fibrotic markers α-SMA (J), Col1α1 (K), and FN (L) in CCD-18Co cells (C_LL37 = 1.11 μM). (M) EMT modulation with vimentin (green) and E-cadherin (red) costaining. Scale bars, 100 μm. In (B), (D), (F), and (G), data are shown as mean ± SD (n = 6 biological replicates). In (J) to (L), data are shown as mean ± SD (n = 3 biological replicates). Statistical analysis was studied via one-way ANOVA, with P values denoting the statistical significance among the IF model, LL37, Lac, and Lac/LL@Alg groups. The images in (E), (H), (I), and (M) are representative of five biological replicates.

Fig. 8.. Gut microbiome reconfiguration by BTB-Alg treatment.

(A) The workflow for metagenomic analysis. (B and C) Evaluation of species-level diversity using Venn diagrams (B) and the Sobs index (C). (D) Principal coordinate analysis of microbial community structures. (E and F) Abundance profiles of microbial communities at the phylum (E) and genus (F) levels following different treatments. (G) LEfSe analysis identifying differentially abundant taxa between the IF model and Lac/LL@Alg groups [linear discriminant analysis (LDA) score > 2.5]. (H) KEGG pathway enrichment analysis for IF model versus Lac/LL@Alg. CoA, coenzyme A. KO, KEGG Orthology. (I) KEGG functional contribution degree in control, IF model, and Lac/LL@Alg groups at the genus level (top 10). ABC, adenosine 5′-triphosphate–binding cassette. (J) Correlation coefficients between differentially enriched microbes and IF-related parameters. (K) Proposed mechanism illustrating the microbiome-mediated therapeutic effects of BTB-Alg treatment. Data are shown as mean ± SD (n = 5 biological replicates). Statistical comparisons between control, IF model, and Lac/LL@Alg groups were studied via one-way ANOVA.

Fig. 9.. Protein-level reconfiguration by BTB-Alg treatment.

(A) The workflow for proteomics analysis. (B) The DEPs evaluated using a Venn diagram. (C) Principal components analysis of proteomic profiles. (D) Volcano plots displaying DEPs in the IF group against the control or Lac/LL@Alg group. The red color indicated that the proteins were up-regulated; the blue color indicated that the proteins were down-regulated. FC, fold change. (E and F) GO pathway enrichment analysis of DEPs between the IF model and control (E) or Lac/LL@Alg (F) groups. (G) KEGG pathway enrichment analysis of DEPs between the IF model and control groups. JAK-STAT, Janus kinase–signal transducer and activator of transcription. (H) GSEA of the IBD pathway in colon tissues from control and IF model groups. (I) Immunofluorescence staining of macrophages from healthy versus fibrotic intestinal tissues (n = 3). CD86 staining (red): M1 macrophage; CD206 staining (green): M2 macrophage. Scale bars, 200 μm. (J) KEGG pathway enrichment analysis of DEPs between IF model and Lac/LL@Alg-treated group. (K and L) GSEA of IBD pathway (K) and ECM-receptor interaction (L) in colon tissues after Lac/LL@Alg treatment. (M) Macrophage polarization changes after different treatments (n = 3). (N) KEGG pathway enrichment analysis of DEPs elucidating therapeutic mechanisms of Lac/LL@Alg in IF. TRP, transient receptor potential. (O) GSEA demonstrating AMPK pathway activation by Lac/LL@Alg treatment in colonic tissues. In (B) to (H), (J) to (L), and (N) and (O), data are from n = 5 biological replicates. NES, normalized enrichment score.

Fig. 10.. Molecular investigation of antifibrotic mechanism.

(A) Workflow for antifibrotic mechanism investigation. (B) AMPK activation levels in intestinal stricture tissues versus normal tissues from patients with CD. (C) Western blot analysis of AMPK activation and fibrotic marker gene expression after treatment with LL37 or Lac/LL@Alg (C_LL37 = 20 μg ml⁻¹). (D and E) Grayscale analysis of Western blot bands for phospho-AMPK (pAMPK)/AMPK (D) and FN (E). (F to H) qPCR level of α-SMA (F), Col1α1 (G), and FN (H) in CCD-18Co cells after different treatments. (I) Expression levels of autophagy-related genes between patients with active-stage CD and healthy controls in the GSE75214 dataset. (J) Dysregulation of autophagy-related genes in structured intestinal tissues compared to nonstructured tissues from patients with CD (n = 9). (K and L) Autophagic flux dynamics were quantified through LC3B-II/I ratio modulation, p62 accumulation (K), and phospho-mTOR signaling activity (L). (M) GSEA of G protein signaling pathways and apelin signaling pathway activation in colon tissues after Lac/LL@Alg treatment. (N) Dynamic snapshot of the interaction between LL37 and apelin receptor (APJ). (O) Root mean square fluctuation (RMSF) analysis of the LL37-APJ and apelin-36–APJ complexes. (P) Average binding free energies of the simulated LL37-APJ or apelin-APJ complexes. In (B), (C), (K), (L), and (M) to (P), experiments were repeated three times with consistent results. In (D) to (H), data are shown as mean ± SD (n = 3 biological replicates). Statistical analysis was studied via one-way ANOVA comparing phosphate-buffered saline (PBS), Met, LL37, LL37 + CC, Lac/LL@Alg, and Lac/LL@Alg + CC.