Identification of region-specific signaling modules governing Crohn’s pathogenesis via mass spectrometry-based integrative acetyl-proteomics

Abstract

BACKGROUND

Crohn’s disease (CD), a subtype of inflammatory bowel disease, is characterized by chronic inflammation affecting the entire gastrointestinal tract. Intestinal damage and subsequent disability caused by CD represent a growing global health concern, yet the molecular basis underlying its segmental heterogeneity remains poorly understood.

AIM

To construct a regional acetylome within different colonic segments and to determine the acetylation response to ileocolonic inflammation.

METHODS

Using colon tissue from distinct anatomical segments of CD rat models and normal controls, we performed quantitative acetyl-proteomics based on antibody enrichment coupled with mass spectrometry, as well as 16S sequencing of the gut microbiota. Targeted parallel reaction monitoring (PRM) was employed to validate the reliability of the discovery proteomics.

RESULTS

We constructed a spatial landscape of the colonic acetylome, identifying 4457 acetylation sites within 1476 acetylated proteins. Immune-regulatory signaling pathways were enriched, including Nucleotide-binding oligomerization domain-containing protein 2-receptor-interacting serine/threonine-protein kinase 2, mitogen-activated protein kinase, nuclear factor kappa-B and signal transducer and activator of transcription 3-Janus Kinase. Bioinformatic analyses integrating acetylome and microbiome data further revealed acetylation-associated alterations and their interactions with the gut microbiota, which may contribute to dysbiosis, providing new insights into CD’s segmental heterogeneity. PRM validation confirmed significant changes of selected acetylation sites including Ywhaz K3ac and SMRD2 K185ac.

CONCLUSION

Our findings suggest that multiple signaling cascades and acetylation events may contribute to disease progression, and these acetylation patterns may serve as a resource for future mechanistic and translational investigations.

Key Words: Crohn’s disease; Lysine acetylation; Proteomics; Histone modification; Regional heterogeneity; Gut microbiota; Nuclear factor kappa-B signaling; 2,4,6-trinitrobenzenesulfonic acid-induced colitis

Core Tip: This study presents a segment-resolved acetyl-proteomic landscape in a rat model of Crohn’s disease, revealing distinct acetylation signatures across the proximal, mid, and distal colon. Fifteen hub proteins, including receptor-interacting serine/threonine-protein kinase 2, signal transducer and activator of transcription 1, nuclear factor kappa B subunit 2, Ywhaz and SMRD2 were identified from the acetyl-protein interaction network. Site-specific validation by parallel reaction monitoring confirmed the reliability of the discovery proteomics. These region-specific acetylation patterns and hub proteins provide a foundation for biomarker discovery and targeted therapeutic strategies in Crohn’s disease.

INTRODUCTION

Inflammatory bowel disease (IBD), with over 6.8 million patients globally, has shown a continuous upward trend in incidence and prevalence and has emerged as a significant global health challenge in recent decades[1,2]. IBD primarily refers to ulcerative colitis (UC) and Crohn’s disease (CD)[3,4]. Specifically, CD is a chronic and relapsing-remitting inflammatory bowel condition characterized by transmural inflammation that can affect various segments of the entire gastrointestinal tract. Common initial manifestations include diarrhea, abdominal pain and systemic symptoms such as fatigue and weight loss with a peak incidence occurring between 20 years and 30 years of age[5,6]. Compared to UC, inflammatory lesions in CD exhibit a distribution pattern within the intestinal tract, whereby affected segments (inflammation/ulceration zones) alternate with normal bowel regions. This transmural nature of CD often leads to more complex complications, such as strictures, fistulas, and abscesses. This characteristic highlights the importance of considering regional variation across intestinal segments of colonic tissues when investigating CD. Chronic gastrointestinal inflammation in CD is associated with dysregulated expression of pro- and anti-inflammatory molecules and tissue destruction within the innate and adaptive immune systems[7,8]. The innate immune system serves as the first line of defense against invading pathogenic microorganisms, executing essential functions through pattern recognition receptors (PRRs). Nucleotide-binding oligomerization domain-containing protein 2 (NOD2), a well-characterized PRR within the NOD-like receptor family, represents the first identified genetic susceptibility locus for CD[9,10]. Receptor-interacting serine/threonine-protein kinase 2 (RIPK2/RIP2) functions as a signaling transducer and integrator for both innate and adaptive immunity, serving as an indispensable downstream signaling component and critical element of the NOD2 signaling complex[11,12]. Monomeric NOD2 undergoes oligomerization via caspase recruitment domain (CARD)-CARD domain interactions to recruit RIPK2, which subsequently engages with tumor necrosis factor receptor-associated factors to activate the inhibitor of nuclear factor kappa-B (NF-κB) kinase complex. This cascade initiates downstream pathways including NF-κB, mitogen-activated protein kinase (MAPK), and autophagy, thereby modulating inflammatory responses[13]. The NOD2-RIPK2 interaction promotes effective inflammatory signaling and plays a pivotal role in orchestrating immune homeostasis and fine-tuning inflammatory responses[14].

On one hand, the colon exhibits substantial regional heterogeneity along its proximal-to-distal axis. The ascending, transverse, and descending colon differ in their epithelial, stromal, and immune cell compositions, as well as in their spatial organization and cellular adjacency relationships[15]. During mucosal healing following injury, this molecular regionalization becomes even more pronounced: The distal colon activates transcriptional programs dominated by acute injury and stress responses, whereas the proximal colon remains relatively homeostatic. Different regions display markedly distinct pathway activities. For example, areas with reduced p53 activity spatially colocalize with proliferating crypts[16]. This inherent spatial heterogeneity of the colon is intimately linked to the enigmatic skip lesions of CD, which similarly manifest as regional differences within the colonic mucosa[17]. Therefore, deciphering the location-specific molecular drivers within the colon itself is critically important for revealing the basis of skip lesions and ultimately for understanding the pathogenesis of CD[15-17].

On the other hand, lysine acetylation is a key epigenetic regulatory process governing gene expression[18]. Investigations of horizontal epigenetic differences between UC and CD patients have, revealed significant elevations of specific histone modifications in natural killer cells of CD patients[19]. Inhibiting BRD4 binding to acetylated histones has been shown to suppress MYC oncogene expression and pro-inflammatory cytokine production[20]. Even more, evidence suggests that in CD, altered gut microbial protein lysine acetylation may regulate disease metabolism by impacting enzymes involved in short-chain fatty acid synthesis[21].

This study aims to investigate spatial heterogeneity in acetylome profiles across distinct intestinal segments and their differential signaling implications via high-resolution liquid chromatography-tandem mass spectrometry (LC-MS/MS) (Figure 1A). By characterizing region-specific acetylation patterns, we seek to delineate how spatially stratified protein acetylation modulates inflammatory factor responses and dysregulated signal transduction pathways in CD, including its regulation of microbiota interactions within the gut microenvironment. Our work further elucidates the regulatory role of post-translational modifications in CD pathogenesis, thereby identifying pathophysiological relevant therapeutic targets critical for guiding precision management strategies in CD patients.

Figure 1 Experimental workflow and characterization of the colonic acetylome. A: Schematic workflow of the study, including colonic tissue segmentation, acetyl-proteomics analysis, 16S rRNA sequencing, and bioinformatic integration; B: Transmission electron microscopy of colonic epithelia. Left: Control group; right: Crohn’s disease (CD) model group. Scale bars: 5.0 μm (overview), 1.0 μm (detail); C: Hematoxylin and eosin staining of colonic sections. Left: Control group; right: CD model group. Scale bar: 50 μm; D: Functional classification of all identified acetylated proteins across Kyoto Encyclopedia of Genes and Genomes pathways, biological process, cellular component, and molecular function. TNBS: 2,4,6-trinitrobenzenesulfonic acid; P-colon: Proximal-colon; M-colon: Mid-colon; D-colon: Distal-colon; WB: Western blot; MS: Mass spectrometry; NGS: Next-generation sequencing; CD: Crohn’s disease; KEGG: Kyoto Encyclopedia of Genes and Genomes; GO: Gene Ontology; MAPK: Mitogen-activated protein kinase; PI3K/Akt: Phosphatidylinositol 3-kinase/protein kinase B; HTLV: Human T-cell leukemia virus.

MATERIALS AND METHODS

CD rats model construction

24 male rats (2 months old, 280-350 g) were obtained from the Experimental Animal Center of Henan University of Chinese Medicine, randomly assigned to two experimental groups: Normal control group (control group, n = 12) and 2,4,6-trinitrobenzenesulfonic acid (TNBS)-induced model group (CD group, n = 12). All rats were housed under specific pathogen-free conditions with a 12-hour light/dark cycle, ambient temperature of 25 ± 2 °C, humidity of 60% ± 2%, and had unlimited access to food and water for 1 week prior to experimentation. Before model induction, the rats were fasted for 24 hours but had free access to water to ensure colonic emptying and prevent the enema solution from being expelled by fecal material, which would compromise induction efficacy. For the surgical procedure, animals were anesthetized with 5% isoflurane (in 1 L/minute oxygen) for induction and then maintained on 2.5% isoflurane delivered via a nose cone. Following the stabilization of anesthesia, the CD model was established by intrarectal administration of TNBS (Sigma-Aldrich, United States; CAS 2508-19-2) at a dose of 3 mL/kg once every 7 days for a total of four enemas. Under surgical-plane anesthesia (confirmed by the absence of pedal reflex), two randomly selected rats were euthanized by cervical dislocation in compliance with American Veterinary Medical Association Guidelines. Colonic tissues were collected and evaluated by macroscopic and microscopic pathological examination to confirm successful CD model establishment, which defined the study endpoint[22]. After successful modeling was verified, the remaining animals were euthanized using the same procedure, and colon tissue samples were collected for subsequent experiments.

Animal health and behavior were monitored twice daily (9: 00-10: 00 and 17:00-18:00). Humane endpoints were defined as: (1) Body weight loss > 20%; (2) Severe diarrhea or hematochezia for > 48 hours; (3) Rectal prolapse; (4) Lethargy, piloerection, reduced response to stimuli, or inability to eat or drink; and (5) Pronounced abdominal distension. If any criterion was met, the animal would be immediately euthanized. No animal died unexpectedly or was euthanized prematurely. Death was confirmed by cessation of breathing, loss of heartbeat, and absence of pupillary reflex; after cervical dislocation, absence of respiration, heartbeat, and corneal reflex was reconfirmed.

The sample size was determined based on the classic TNBS-induced CD rat model established by Morris et al[22], which has been widely validated for high reproducibility and stable success rates, and on a > 90% induction success rate observed in our preliminary experiments[23]. This sample size was considered sufficient to obtain statistically meaningful data while minimizing animal use. All animal procedures were conducted in accordance with protocol approved by Laboratory Animal Welfare and Ethics Committee of Henan University of Traditional Chinese Medicine (Approval No. IACUC-202310037).

Ultrastructural screening

Colonic specimens (1 mm³) were fixed in 2.5% (v/v) glutaraldehyde in 0.1 M phosphate buffer [potential of hydrogen (pH) = 7.4] for 4 hours at 4 °C, followed by three 10-minute washes in phosphate-buffered saline (PBS). Samples were post-fixed in 1% (w/v) osmium tetroxide for 2 hours. After three PBS washes (10 minute/wash), samples were dehydrated through a graded ethanol series and embedded in epoxy resin (Sigma-Aldrich). Polymerization was conducted at 60 °C for 48 hours. Ultrathin sections (70 nm) were prepared using a diamond knife (Leica UC7), stained with uranyl acetate and Reynolds’ lead citrate, and examined under a transmission electron microscope (Hitachi HT7800) at 80 kV accelerating voltage.

Histological analysis

Segments of approximately 1 cm were fixed in 10% neutral-buffered formalin for 24-48 hours, followed by paraffin embedding. Tissue sections (4-5 μm thick) were cut using a rotary microtome, deparaffinized in xylene, and rehydrated through graded ethanol. Then hematoxylin and eosin (HE) staining was performed with nuclei staining using Mayer’s hematoxylin for 5 minutes and cytoplasmic counterstaining with eosin Y for 1 minute[24].

Protein extraction

Segmental dissection of colonic tissues was performed aseptically along the proximal-distal axis, dividing specimens into three anatomically defined regions: (1) Proximal-colon (P-colon): Segment adjacent to the cecum; (2) Mid-colon (M-colon): Transverse segment between hepatic and splenic flexures; and (3) Distal-colon (D-colon): Terminal segment encompassing the sigmoid-rectal junction. Colonic segmental tissue from each individual animal was minced separately in liquid nitrogen and lysed in lysis buffer containing 100 mmol/L ammonium hydrogen carbonate (pH = 8), 6 M urea and 0.2% sodium dodecyl sulfate (SDS), followed by 5 minutes of ultrasonication on ice. The lysate was centrifuged at 12000 g for 15 minutes at 4 °C and the supernatant was transferred to a clean tube. Protein concentration was determined by Bradford protein assay. Extracts from each sample were reduced with 10 mmol/L dithiothreitol (DTT) for 1 hour at 56 °C, and subsequently alkylated with sufficient iodoacetamide for 1 hour at room temperature in the dark. Then 4 volumes of precooled acetone were mixed with samples by vortexing thoroughly and incubated at -20 °C for at least 2 hours. Colonic tissue was then centrifuged and the precipitate was collected. After washing twice with cold acetone, the pellet was dissolved in dissolution buffer containing 0.1 M triethylammonium bicarbonate (TEAB) (pH = 8.5) and 6 M urea. Protein concentration was re-determined by Bradford protein assay.

Acetylated peptide preparation

The supernatant containing precisely 0.12 mg of protein from each tube was digested with trypsin gold (Promega) at an enzyme-to-substrate ratio of 1:50. After 16 hours of digestion at 37 °C, all samples were desalted using a C18 cartridge to remove urea. The peptides were fractionated using an LC-15AC chromatograph with a high pH C18RP column (Bridge Peptide BEH C18, 130Å, 3.5 μm, 4.6 mm × 250 mm, Waters). Acetylated enrichment magnetic beads (Cell Signaling 13416, rabbit mAb) were added to the loading buffer and mixed thoroughly with the sample to be enriched, and incubated at room temperature for 20 minutes. After incubation, the supernatant was retained. The acetylated peptides were eluted with 5% ammonium hydroxide, and the eluate was acidified with formic acid, then lyophilized under vacuum for mass spectrometry analysis.

LC-MS/MS analysis

The purified peptides were re-dissolved and analyzed on an LC-MS/MS system consisting of an Ultimate 3000 RSLCnano system (Thermo Fisher Scientific) coupled online to a Q Exactive HF mass spectrometer (Thermo Fisher Scientific) equipped with a Nanospray Flex ion source (Thermo Fisher Scientific). The sample was loaded onto a 150 mm analytical column (in-house packed, 75 μm inner diameter, 3 μm C18 Reprosil-HD packing material). The peptides were eluted by a nonlinear solvent gradient, with the proportion of solvent B [0.1% formic acid aqueous solution/acetonitrile (2:8, v/v)] increasing from 2% to 56% within 70 minutes at a constant flow rate of 300 nL/minute, followed by a 20-minute flush with 99% solvent B and re-equilibration to solvent A [0.1% formic acid aqueous solution/acetonitrile (2:8, v/v)]. The mass spectrometer was operated in data-dependent acquisition mode, automatically switching between MS and MS/MS scans, and collecting 16 strongest ion peaks for each MS spectrum. The full-scan MS spectra (400-1800 m/z) were acquired in the Orbitrap analyzer with a resolution of 60000 and an accumulation target value of 3 × 10⁶. The MS/MS spectra (100-2000 m/z) were acquired in the Orbitrap analyzer with a resolution of 15000. Full width at half maximum was set to 400 m/z with an automatic gain control (AGC) target of 1e6 ions, and the first mass was fixed at 100 m/z.

Identification of Kac sites

Raw data analysis was completed using MaxQuant software (version 2.0.3.1) in conjunction with the Andromeda search engine, with the default search parameter settings, including a false positive rate of 5% at both the peptide and protein levels. The mass spectrometry data were searched against the rat (Rattus norvegicus) proteins in the UniProt/Swiss-Prot database. The mass tolerances for precursor ions and fragment ions in the main search stage were set to 6 ppm and 20 ppm, respectively. The enzyme cleavage specificity was set to cleavage at the C-terminus of arginine and lysine, allowing cleavage at proline bonds, with a maximum of 2 missed cleavages. The variable modifications were set to oxidation of methionine residues and acetylation (K), while carbamidomethylation of cysteine residues was set as a fixed modification. The false discovery rate (FDR) was set to 0.01 at both the protein and peptide levels. Peptides with a minimum length of seven amino acids were considered. The match between runs function was enabled, with a match time window of 1 minute and an alignment time window of 20 minutes.

Kac motif analysis

Amino acid sequence context surrounding identified Kac sites was analyzed and visualized using WebLogo 3.7 (https://weblogo.berkeley.edu/)[25]. Sequence windows spanning 11 residues (centered on Kac sites) were generated for all identified acetylated peptides. Statistically overrepresented Kac motifs were identified using pLogo 1.3 (https://plogo.uconn.edu/)[26]. Specifically, 15-mer sequences (Kac site ± 6 flanking residues) were extracted from the UniProtKB/Swiss-Prot database (release 2023_03) and submitted to the Motif-X algorithm (v1.2) for significant motif extraction with occurrence threshold P < 1 × 10^-6.

Taxonomic and functional analysis

Gene Ontology (GO) and InterPro (IPR) analyses were conducted using the interproscan-5 program against the non-redundant protein database (including Pfam, PRINTS, ProDom, SMART, ProSiteProfiles, PANTHER). The Clusters of Orthologous Groups and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases were used to analyze the protein families and pathways. The probable interacting partners were predicted using the STRING-db server (http://string.embl.de/) based on the related species. STRING is a database of both known and predicted protein-protein interactions. The enrichment Python program (version 2.7) and R package (version 3.3.2) were used to perform enrichment analysis and visualization of GO, IPR and KEGG, respectively.

16S sequencing

Following aseptic longitudinal dissection of excised colonic segments from each animal, luminal contents were collected individually by gentle scraping with sterile spatulas. Mucosal contamination was rigorously avoided during sampling. Genomic DNA was extracted from intestinal samples (100 ± 20 mg) using the MOBIO PowerSoil^® Pro Kit (QIAGEN, Germany). Mechanical lysis was performed using 0.1 mm zirconia beads (10 minutes vortexing at 3200 rpm), followed by chemical disruption in 1% SDS buffer (65 °C, 10 minutes). The V6 hypervariable region was amplified with primers 967F (5’-CNACGCGAAGAACCTTANC-3’) and 1046R (5’-CGACRRCCATGCANCACCT-3’)[27]. Polymerase chain reaction conditions: 95 °C/3 minutes; 30 cycles of 95 °C/30 seconds, 55 °C/45 seconds, 72 °C/60 seconds; Final extension 72 °C/5 minutes. Purified amplicons underwent paired-end sequencing (2 × 250 bp) (Illumina MiSeq). Processed sequences were classified using the RDP Classifier (v2.13, 80% confidence threshold) against SILVA 138 database, and operational taxonomic units (OTUs) were clustered at 97% identity using VSEARCH[28].

Western blot

Protein samples were boiled and denatured in 5 × SDS-polyacrylamide gel electrophoresis (PAGE) protein loading buffer (Beyotime) and then separated by SDS-PAGE. Subsequently, the proteins were transferred to a methanol-activated polyvinylidene difluoride (Beyotime) membrane by electro-transfer. The membrane was rinsed with tris-buffered saline with tween-20 buffer (Beyotime), blocked with 5% milk at room temperature for 30 minutes, and incubated with anti-RIPK2 (Proteintech, 15366-1-AP, 1:1500), anti-NOD2 (Proteintech, 30628-1-AP, 1:1500), anti-magnesium dependent phosphatase 1 (MDP1) (Proteintech, 13929-1-AP, 1:1500), anti-ATG16 L1 (Proteintech, 29445-1-AP, 1:1500) and anti-acetylated lysine (pan-acetyl) antibody (ImmuneChem, ICP0380, 1:1000). After overnight incubation with primary antibodies at 4 °C, the membrane was incubated with a horseradish peroxidase-labeled goat anti-rabbit immunoglobulin G secondary antibody (Proteintech) diluted 1:1500 at room temperature for 30 minutes. Finally, protein detection was performed using an enhanced chemiluminescence chemiluminescence kit (CWBio).

Parallel reaction monitoring verification

Colon tissues were ground in liquid nitrogen, lysed in buffer [50 mmol/L Tris-hydrogen chloride pH = 8.0, 1% Triton X-100, 0.5% SDS, 1 mmol/L ethylene diamine tetraacetic acid (EDTA), 1 mmol/L phenylmethylsulfonyl fluoride (PMSF), 20 mmol/L DTT] with ultrasonic crushing, centrifuged (12000 g, 4 °C, 15 minutes) and the supernatant was acetoneprecipitated three times before redissolution in 8 M urea, 2 mmol/L EDTA, 10 mmol/L DTT, 1 mmol/L PMSF, followed by another centrifugation. Protein concentration was measured by a modified Bradford assay. Then 100 μg protein per sample was reduced with 10 mmol/L DTT (37 °C, 30 minutes), alkylated with 25 mmol/L iodoacetamide (dark, room temperature, 20 minutes), diluted with 100 mmol/L TEAB, and digested with trypsin (protein: Trypsin = 50:1 overnight). After desalting and lyophilization, peptides were redissolved in 0.1% formic acid, and 2 μg per sample underwent LC-parallel reaction monitoring (PRM)/MS analysis on a Q-Exactive HF with a 78 minutes gradient (full scan: 60000 resolution, 350-2000 m/z, AGC 1e6, max fill 35 ms; scheduled scan: 15000 resolution, AGC 2e5, max fill 80 ms, 2 Da isolation window, normalized collision energy 27%). Raw data were searched with MaxQuant (Andromeda, trypsin/P, two missed cleavages, fixed carbamidomethylating, variable oxidation and acetylation, 20 ppm mass tolerance, 1% FDR). A spectral library was built in Skyline (cut-off = 0.90), and the six most intense product ions per peptide were selected as transitions; manual peak verification was performed. For normalization, each sample’s average base peak intensity (full scan) was extracted with RawMeat (V2.0), and a sample specific factor (relative to median intensity of all samples) was applied to the area under the curve (AUC) of each transition; relative acetylation site abundance was defined as the intensity of a given peptide at the peptide-level AUC.

Statistical analysis

Differentially acetylated sites were identified based on a fold change > 1.2 and an un-adjusted P value < 0.05 (Student’s t test, two-tailed). To control the FDR, the Benjamini-Hochberg procedure was applied for multiple testing correction, with an FDR < 0.01 considered statistically significant. This relatively permissive threshold was adopted to maximize the retention of proteins potentially associated with TNBS-induced colitis at the discovery stage, while maintaining acceptable control of false positives. For 16S rRNA sequencing data, alpha diversity indices (Chao1, ACE, Shannon, Simpson, and PD whole tree) were calculated, and intergroup differences were compared using the non-parametric Mann-Whitney U test, with P < 0.05 considered statistically significant. Principal coordinate analysis based on the Bray-Curtis distance matrix was performed to visualize sample relationships, and permutational multivariate analysis of variance (PERMANOVA, using the adonis function in the vegan package, 999 permutations) was used to test for differences in community structure between groups. The Wilcoxon rank-sum test combined with linear discriminant analysis (LDA) (threshold > 2.0) was used in linear discriminant analysis effect size (LEfSe) to identify differentially abundant microbial taxa. Krona charts were generated to display multi-level taxonomic composition. Western blotting and PRM data were analyzed using Student’s t-test (two-tailed), with a more stringent threshold of P < 0.01 applied to ensure the reliability of the confirmed proteins and acetylation sites during the targeted validation phase.

RESULTS

Acetylome landscape of colonic tissues in Crohn’s rat model

Electron microscopy observations of the intestines revealed that in the normal group, intestinal villi were neatly arranged with a plump rounded shape, free from cracks or defects, and cellular organelles remained intact. In contrast, the model group exhibited ruptured surfaces of intestinal villous cells, partial damage, disrupted intestinal cells, and vacuolization (Figure 1B). HE staining showed transmural inflammation with crypt architectural distortion, con-firming successful establishment of the colitis model (Figure 1C).

For acetyl-proteomic analysis, colonic tissues from four individual animals per group were processed and analyzed as biological replicates. The peptide length distribution reflected appropriate trypsin digestion efficiency, and the mass error distribution centered near zero indicated stable instrument performance, collectively confirming that the LC-MS/MS data acquisition met quality control standards (Supplementary Figure 1). Quantitative acetylproteomics identified a total of 4457 Lysine acetylation (Kac) sites on 1476 proteins across the P-, M-, and D-colon. The acetylated proteins were distributed across diverse functional categories, including chromatin organization, signal transduction, and metabolism (Figure 1D). Abundant histone acetylation sites were detected, with modification profiles mapped onto core histones H2A, H2B, H3, and H4 (Figure 2). High reproducibility across LC-MS/MS replicates further supported data reliability (Figure 2).

Figure 2 Segmental heterogeneity of the colonic acetylome. A: Principal component analysis of acetylated peptides from three colonic segments of Crohn’s disease and control groups; B: Spearman correlation heatmap of acetylated protein abundances across all samples. Color scale represents Spearman’s correlation coefficient ranging from 0 (blue) to 1 (red); C: Venn diagram showing overlap and uniqueness of differentially acetylated proteins identified in proximal-colon, mid-colon, and distal-colon tissues; D: Schematic illustration of the identified acetylation sites on histones, with the positions of all identified acetylated lysine residues mapped onto the corresponding histone proteins (H3, H4, H2A and H2B). Numbers indicate the count of acetylated proteins in each region; E: Venn diagram of acetylated proteins overlapping between rat model and human compendium of protein lysine modifications. Acetylome data were derived from n = 4 biological replicates per group. PCA: Principal component analysis; P-colon: Proximal-colon; M-colon: Mid-colon; D-colon: Distal-colon; CD: Crohn’s disease.

Motif analysis of the sequences flanking the identified Kac sites (± 10 residues) revealed segment-specific preferences in enriched motifs (Figure 3A). The P-colon was pre-dominantly characterized by the lysine-alanine-lysine motif, the M-colon by the methionine-lysine-lysine (MKK) motif, and the D-colon by glycine-lysine-alanine and MKK motifs. Despite these regional differences, the high-frequency amino acids surrounding the acetylated sites, namely lysine (K), methionine (M), glycine (G), and alanine (A), remained consistent across all three regions (Figure 3B). This con-served amino acid preference may reflect the activity of a common set of acetyltransferases operating throughout the colon.

Figure 3 Motif analysis and abundance patterns of acetylated sites. A: Sequence logos showing conserved amino acid motifs surrounding lysine acetylation (Kac) sites (± 10 residues) in proximal-colon, mid-colon, and distal-colon tissues; B: Enrichment scores of overrepresented motifs identified in each colonic segment; C: Heatmap showing normalized abundance of quantified acetylated sites across 24 tissue samples. Each row represents an acetylated protein, and each column represents an individual sample. Color scale indicates z-score-normalized abundance (red: High, blue: Low). P-colon: Proximal-colon; M-colon: Mid-colon; D-colon: Distal-colon; CD: Crohn’s disease; JAK: Janus Kinase; STAT: Signal transducer and activator of transcription; Y: Tyrosine; W: Tryptophan; V: Valine; T: Threonine; S: Serine; R: Arginine; Q: Glutamine; P: Proline; N: Asparagine; M: Methionine; L: Leucine; K: Lysine; I: Isoleucine; H: Histidine; G: Glycine; F: Phenylalanine; E: Glutamic acid; D: Aspartic acid; C: Cysteine; A: Alanine.

Region-specific acetylome alterations in colonic segments of the CD model

The three colonic segments formed relatively independent acetylome datasets, with up to 60% of the differentially acetylated sites in each segment being distinct from those in the other two segments. Nevertheless, the analysis revealed 115 quantitatively altered acetylation sites that were conserved across all three segments (Figure 2). All three regions exhibited substantial acetylome remodeling, with the primary differences in acetylation levels concentrated in the Janus Kinase (JAK)-signal transducer and activator of transcription (STAT), Ras, and insulin signaling pathways, as well as microRNAs in cancer and viral carcinogenesis (Figure 3C). Functional enrichment analysis further revealed that each region displayed a distinct molecular signature in response to TNBS-induced colitis: P-colon was characterized by chromatin organization and cell cycle regulation, M-colon by NF-κB signaling and ubiquitin/proteasome pathway activation, and D-colon by metabolic reprogramming and cell proliferation.

In P-colon, 1046 differentially acetylated sites in 453 proteins (258 up, 195 down) were identified. These proteins were predominantly enriched in chromatin structure and dynamics, and signal transduction. Biological process enrichment highlighted regulation of cell cycle G2/M phase transition and histone phosphorylation, while KEGG pathway analysis revealed enrichment in viral carcinogenesis, Wnt signaling, and hypoxia inducible factor-1 signaling (Supplementary Figure 2). In M-colon, 796 regulated acetylated sites were identified in 385 proteins (129 up, 252 down), with distinct enrichment in negative regulation of mitotic cell cycle and canonical Wnt signaling. Notably, ubiquitination and proteasomal catalysis pathways were significantly activated. Unsupervised clustering resolved four functional modules, with NF-κB showing the strongest upregulation (Supplementary Figure 3). The D-colon demonstrated the most pronounced acetylome alterations, with 764 differentially acetylated proteins (DAPs) (352 up, 412 down), enriched terms included positive regulation of cell proliferation and digestive system development, along with the Fanconi anemia pathway and cell adhesion molecules (Supplementary Figure 4). Key enriched terms in the three regions are presented in Table 1.

Table 1 Key enriched terms in the three colonic regions under 2,4,6-trinitrobenzenesulfonic acid-induced colitis.

Category	M-colon	P-colon	D-colon	Shared
Domain	IPR000219 DH domain	IPR023780 chromo domain; IPR000330 SNF2-related	IPR023780 chromo domain; IPR000330 SNF2-related	IPR009072 histone-fold
GO-CC	GO: 0005815 microtubule organizing center	GO: 0005667 transcription factor complex	GO: 0005815 microtubule organizing center	GO: 0005694 chromosome; GO: 0000786 nucleosome
GO-MF	GO: 0005089 Rho GEF activity	GO: 0003712 transcription cofactor activity; GO: 0019787 ubiquitin-like transferase	GO: 0008270 zinc ion binding	GO: 0003677 DNA binding
GO-BP	GO: 0060070 Wnt signaling; GO: 0032434 proteasomal regulation	GO: 0051568 histone H3-K4 methylation	GO: 0032496 response to lipopolysaccharide	GO: 0043933 macromolecular complex organization; GO: 0006325 chromatin organization; GO: 0051276 chromosome organization
KEGG	map05322 SLE; map05034 alcoholism	map05203 viral carcinogenesis; map04917 prolactin signaling	map05203 viral carcinogenesis

Enrichment analysis was performed using Fisher’s exact test. Terms with a Fisher’s exact test P < 0.05 were considered significantly enriched. The listed terms represent the top enriched entries for each category [domain, Gene Ontology (GO)-cellular component, GO-molecular function, GO-biological process, Kyoto Encyclopedia of Genes and Genomes] in the proximal, mid, and distal colon. P-colon: Proximal-colon; M-colon: Mid-colon; D-colon: Distal-colon; GO: Gene Ontology; CC: Cellular component; MF: Molecular function; BP: Biological process; KEGG: Kyoto Encyclopedia of Genes and Genomes; DH: Dbl homology; GEF: Guanine nucleotide exchange factor; SLE: Systemic lupus erythematosus; SNF2: Sucrose non-fermenting 2.

Dynamic alterations of upregulated acetylated proteins across intestinal pathological regions

M-colon exhibited significantly fewer upregulated acetylated proteins compared to others, with its upregulated protein count also markedly lower than its downregulated counterparts. To elucidate functional and characteristic differences in upregulated acetylation among the colonic segments, we subsequently performed cluster analysis on upregulated DAPs from model groups and control groups across all three regions.

While all three regions shared enrichment in histone-related domains (histone fold, H2A/H2B/H3) and nucleic acid binding, striking differences emerged (Figure 4A-E). Specifically, histone H2A-associated domains were prominently clustered in M-colon and P-colon, whereas D-colon was uniquely enriched in homeobox domains, phosphatidyl-inositol 3-/4-kinase catalytic/conserved site domains, and phosphatidylinositol kinase-related kinase domains (Figure 4D). M-colon group showed notable enrichment in NF-κB signaling pathway; P-colon group enriched in viral carcinogenesis, cell cycle, and basal transcription factors; D-colon group enriched in insulin signaling pathway, cell cycle, and Legionellosis pathways. Besides, the pathways enriched among the downregulated proteins also showed distinct regional differences (Figure 4F).

Figure 4 Segmentspecific enrichment clustering of differentially acetylated proteins. A-E: Heatmaps showing upregulated enrichment clustering for cellular component (A), molecular function (B), biological process (C), domain architecture (D), and Kyoto Encyclopedia of Genes and Genomes pathways (E) across proximal-colon, mid-colon, and distal-colon tissues (Crohn’s disease model vs corresponding control). Color intensity represents the significance level [-log10 (P value)] of enrichment, with red indicating higher significance; F: Schematic summary of significantly up- and down-regulated enriched pathways in the three colonic segments under 2,4,6-trinitrobenzenesulfonic acid-induced colitis. KEGG: Kyoto Encyclopedia of Genes and Genomes; GO: Gene Ontology; CC: Cellular component; MF: Molecular function; BP: Biological process; P-colon: Proximal-colon; M-colon: Mid-colon; D-colon: Distal-colon; PI3K/Akt: Phosphatidylinositol 3-kinase/protein kinase B; HTLV: Human T-cell leukemia virus.

Characteristics of hub acetylated proteins in the colon

Further clustering analysis based on protein-protein interaction networks of acetylated differential proteins in CD colon tissues revealed that, in addition to the MAPK, cell cycle, and oxytocin signaling pathways, key signaling transduction pathways including necroptosis, viral carcinogenesis, Salmonella infection, and cancer-related microRNAs also occupy central regulatory positions. This suggests a crosstalk between protein acetylation, microRNA-mediated regulation, and gut microbiota dysbiosis, which collectively regulate the onset and progression of CD (Figure 5A).

Figure 5 Protein interaction networks and validation of key acetylation events. A: Protein-protein interaction network of differentially acetylated proteins. Nodes represent acetylated proteins, and edges represent known or predicted interactions. Modules enriched in key pathways are highlighted by distinct colored regions; B: Western blot analysis of innate immune regulators receptor-interacting serine/threonine-protein kinase 2 (RIPK2), nucleotide-binding oligomerization domain-containing protein 2, magnesium dependent phosphatase 1, and ATG16 L1. Data are presented as mean ± SEM. P < 0.01 indicates significant upregulation in Crohn’s disease (CD) model group compared to control. n = 3 biological replicates per group; C: Immunofluorescence staining of RIPK2 (green) in colonic sections. Left: Control group; right: CD model group. Nuclei were counterstained with 4’,6-diamidino-2-phenylindole (blue). Scale bar: 50 μm. n = 2 biological replicates per group; D: Representative mass spectrometry (MS)/MS spectrum of RIPK2 depicting fragment ions containing lysine 508 (K508ac), with a characteristic mass shift of +42.01 Da indicative of acetylation, the Y-axis represents the relative abundance of fragment ions. ^aP < 0.01. CD: Crohn’s disease; MAPK: Mitogen-activated protein kinase; RIPK2: Receptor-interacting serine/threonine-protein kinase 2; NOD2: Nucleotide-binding oligomerization domain-containing protein 2; MDP1: Magnesium dependent phosphatase 1.

15 hub proteins occupied the most central network positions: General transcription factor TFIIB (Gtf2b), Stat1, non-specific serine/threonine protein kinase, nuclear factor kappa B subunit 2 (Nfkb2), heat shock protein 90 alpha family class B member 1), tyrosine-protein kinase (Abl1), M-phase inducer phosphatase (Cdc25b), receptor protein-tyrosine kinase (Erbb4), Ripk2, histone-lysine N-methyltransferase (SETDB2), glyceraldehyde-3-phosphate dehydrogenase, stathmin (Stmn1), inhibitor of NF-κB kinase subunit beta, receptor-interacting serine/threonine-protein kinase 3, and DNA (cyto-sine-5)-methyltransferase 1. Among these, Stat1, Nfkb2, and Ripk2 have been previously reported to be significantly dysregulated in CD or IBD patients[29-31]. The acetylation changes detected in the present study further suggest that alterations in the modification status of these proteins may influence intestinal inflammation. Moreover, our findings raise the possibility that acetylation changes in additional hub proteins may also participate in the pathogenesis of TNBS-induced colitis or CD.

To further elucidate the inflammation-mediating signaling pathways changes and potential immunological disparities in Crohn’s colitis, Western blotting and Immunofluorescence were applied to inflammatory-related proteins RIPK2, p65, MDP1, and ATG16 L1 (Supplementary Figure 5, Figure 5B and C), which were moderately elevated in the model group.

Alterations in the colonic microbiome

For 16S rRNA sequencing, luminal contents from three individual animals per group were analyzed. Alpha diversity analysis showed no statistically significant differences between the control and CD groups in Shannon index (3.94 ± 0.38 vs 3.64 ± 0.22, P = 0.27), Chao1 index (300.0 ± 0.0 vs 317.5 ± 48.8, P = 0.51), and phylogenetic diversity index (26.08 ± 2.99 vs 21.97 ± 3.42, P = 0.22). No intergroup differences were detected in Simpson and ACE indices either. OTU intersection analysis revealed 324 OTUs shared between the two groups, with 141 OTUs unique to the control group and 54 OTUs unique to the CD group (the number of unique OTUs in the model group was significantly lower than that in the control group). Core microbiota analysis showed that 110 OTUs were detected in all six samples, accounting for 21.2% of the total OTUs, and these were mainly assigned to Bacteroidetes and Firmicutes. Given the small sample size and the limited resolution of sequencing, separate analysis of trends at the phylum and genus levels is therefore warranted.

In Figure 6, at the phylum level, both groups were dominated by Bacteroidetes and Firmicutes, whose combined relative abundances accounted for > 95% of the total sequence counts in each group (Figure 6A). Krona plots (Figure 6D and E) further confirmed that the model group exhibited a significant increase in Bacteroidetes and a decrease in Firmicutes and Lachnospiraceae in the gut. At the genus level, Lactobacillus was the most abundant genus, with a slightly higher mean relative abundance in the CD group (12.9%) than in the control group (11.1%), mainly contributed by the increased abundance of OTU3 (Lactobacillus prophage Lj928) (Figure 6B). The LDA bar plot (Figure 6F) showed that at the genus level, Parabacteroides (LDA score = 2.4) and Lactobacillus (LDA score = 2.1) were relatively enriched in the CD group, whereas Alistipes (LDA score = 2.3) was relatively enriched in the control group. The LEfSe cladogram (Figure 6C) illustrated the phylogenetic relationships of the differential taxa across taxonomic hierarchies. At the species level, Parabacteroides distasonis was detected only in the CD group (relative abundance 2.6%); Lactobacillus prophage Lj928 was significantly enriched in the CD group (12.6% vs 3.1%); and Alistipes uncultured bacterium showed a higher relative abundance in the control group. In addition, microbes associated with colitis and colorectal cancer (e.g., Escherichia coli, Clostridium tertium) also exhibited significant changes.

Figure 6 Compositional analysis of gut microbiota in control and Crohn’s disease model groups. A and B: Bar charts showing the relative abundance of dominant bacterial phyla (A) and genera (B) in control and Crohn’s disease (CD) model groups; C: Linear discriminant analysis effect size cladogram displaying the taxonomic hierarchy of differentially abundant taxa between groups; D and E: Krona charts illustrating the taxonomic composition from phylum (D) to species (E) level in control and CD model groups, respectively; F: Linear discriminant analysis (LDA) bar plot showing differentially abundant genera between groups (|LDA| > 2). 16S rRNA sequencing data were derived from n = 3 biological replicates per group. CD: Crohn’s disease; LDA: Linear discriminant analysis.

Targeted validation of site-specific acetylation by PRM

Although the mass spectra showed the presence of modified sites, including RIPK2-K508, mitogen-activated protein kinase kinase kinase kinase (MAP4K)-K239, NF-κB-K353, Stat3-K239, and Jak3-K551 (Figure 5D and Supplementary Figure 6), to independently confirm the acetylation status of specific lysine residues identified in the discovery acetylome, we performed targeted PRM analysis on colonic tissues from the CD model and control groups with four biological repeats. Among the shared DAPs in the three tissue sites, Ywhaz (14-3-3 protein zeta/delta) is considered ubiquitously expressed and recognizes a conserved phosphoserine/threonine binding motif on partner proteins, regulating the function of hundreds of partner proteins, including cell cycle progression, cellular proliferation, signal transduction, and metabolite signaling. When 14-3-3 is dysregulated, these interactions can facilitate the development and progression of disease[32,33]. Switch/sucrose non-fermentable (SWI/SNF) is the first described adenosine triphosphate (ATP)-dependent chromatin-remodeling complex. In mouse embryonic stem cells, approximately 200 SWI/SNF-interacting proteins have been described, including histone- and DNA-modifying enzymes and regulators (e.g., HDAC1, JARID2, DNMT3b, and DNMT3 L). Mutations in several subunits of SWI/SNF have been linked to malignant transformation and progression across a broad range of tumor types. Epigenetic modifications also contribute to SWI/SNF and chromatin associations[34,35]. Thus, Ywhaz and the SWI/SNF subunit SMRD2 were chosen as representative proteins for PRM verification. Consistent with our LC-MS/MS discovery data, PRM quantification with four biological repeats confirmed a significant decrease in Ywhaz K3ac levels in the CD model group and a marked upregulation of SMRD2 K185ac. Representative extracted ion chromatograms and quantification of acetylation abundances are shown in Figure 7 and Supplementary Figure 7.

Figure 7 Targeted parallel reaction monitoring validation of Ywhaz K3ac acetylation in colonic tissues. A: Representative extracted ion chromatograms of fragment ions after collision-induced dissociation of the Ywhaz K3ac precursor peptide from control (n = 4) and 2,4,6-trinitrobenzenesulfonic acid-induced colitis (n = 4) groups; B: Quantification of Ywhaz K3ac abundance. CD: Crohn’s disease.

DISCUSSION

Recent single-cell proteome and spatial proteome studies have revealed cell type specific transcripts or proteins quantification patterns[36-40], our segment level approach identified regionally dysregulated acetylation circuits, which is a necessary precursor to single-cell acetylproteome resolution in complex tissues.

Although histone modification remodeling was observed across all three colonic segments, the nature and degree of these changes differed substantially along the proximal-distal axis. Domain-centric clustering (Figure 4D) revealed that histone H2A-related domains exhibited distinct enrichment patterns among the three regions, while histone H2B domains were significantly enriched in M-colon and P-colon but not in D-colon. Consistently, molecular function enrichment (Figure 4B) showed that DNA binding and nucleic acid binding were highly significant in P-colon and D-colon but less pronounced in M-colon. These findings indicate that histone acetylation-driven chromatin remodeling occurs globally in the inflamed colon, yet the specific histone targets and the extent of remodeling vary by segment during TNBS-induced colitis.

Our data revealed distinct segment-specific histone acetylation patterns, particularly the prominent enrichment of histone H2A-associated domains and chromatin remodelers in M-colon (Figures 4C and D). These findings are consistent with the known role of histone acetylation as a bidirectional regulator of inflammation. On the pro-inflammatory side, Toll-like receptor (TLR) signaling promotes ATP-citrate lyase-dependent acetyl-CoA production to broadly enhance histone acetylation and drive pro-inflammatory gene expression[41]; LKB1 deficiency releases CRTC2-CREB signaling, enabling CBP/p300 to deposit H3K27ac at inflammatory gene loci[42]; the serine synthesis enzyme phosphoglycerate dehydrogenase sustains histone H3 lysine 9 and lysine 27 (H3K27) acetylation at the TLR4 promoter and acetylates NOD-like receptor protein 3 and apoptosis-associated speck-like protein containing a caspase recruitment domain to support interleukin-1β production[43]; and prostaglandin E2 induces H2A. Z acetylation at enhancers, increasing chromatin accessibility for rapid inflammatory gene induction[44]. On the anti-inflammatory side, lactic acid directly fuels H3K27 acetylation, inducing immunosuppressive genes such as Nr4a1 and establishing long-term lipopolysaccharide tolerance[45], while the gut microbial metabolite butyrate enhances H3 acetylation at the Foxp3 locus to promote regulatory T cell differentiation and suppress intestinal inflammation[46]. Thus, site-specific histone acetylation serves as a transcriptional rheostat that calibrates inflammatory responses. Given that M-colon showed the strongest NF-κB pathway activation alongside the most prominent histone domain clustering, it is plausible that region-specific histone acetylation contributes to the differential transcriptional activation of inflammatory mediators along the proximal-distal axis, although direct mechanistic evidence in the context of colitis remains to be established.

Cellular component enrichment (Figure 4A) further highlighted regional differences in gene regulatory mechanisms. The “protein-DNA complex” term was significantly enriched in M-colon and P-colon but not in D-colon, whereas “transcription factor binding” was highly enriched in D-colon while showing minimal changes in the other two segments. This suggests that although transcriptional regulation within the nucleus is altered throughout the colon, the underlying molecular strategies may differ between segments. For instance, acetylation levels of the basal transcription factors TATA-box binding protein associated factor 1 and Gtf2b were altered in all three regions. The transcription factor transcriptional factor 21 (TCF21) showed acetylation changes in both P-colon and D-colon, while metal-response element-binding transcription factor 2 was significantly decreased only in D-colon. Recent evidence indicates that high TCF21 expression is associated with stromal cell enrichment and promotes epithelial-mesenchymal transition through the LIM domain kinase 2-phospho-cofilin signaling axis[47]. Thus, it is plausible that acetylation-mediated changes in transcription factors such as TCF21 could influence nuclear localization or activity, potentially affect stromal cell function and contribute to segment-specific pathology; however, direct functional evidence is still lacking.

Analysis of kinase-associated molecular functions (Figure 4B) showed that “kinase regulator activity” was enriched at similar levels across the three segments, whereas “kinase inhibitor activity” displayed notable segment-specific differences. Given that kinases such as leucine-rich repeat kinase 2 (LRRK2) have been implicated in CD through effects on Paneth cell function, ROS release, autophagy, and cytokine secretion[48-50], it is not unexpected that acetylation changes in various kinases were detected. The observed differences in kinase inhibitor activity, however, have been rarely reported. Although our data do not establish a causal relationship between these acetylation changes and TNBS-induced colitis, they raise the possibility that modulation of kinase inhibitor pathways may represent a new avenue for therapeutic exploration in CD.

Beyond the segment-level patterns described above, our data also identified acetylation changes in specific proteins with established roles in CD pathogenesis. The elevated expression and acetylation of RIPK2, a key NOD2 effector and CD susceptibility gene product, alongside increased NOD2 expression, positions Kac as a potential modulator within the NOD2/NF-κB cascade[51-53]. Mass spectrometry further identified markedly increased acetylation at RIPK2-K508 in the colitis model, along with site-specific changes on MAP4K (K239), NF-κB (K353), Stat3 (K239), and Jak3 (K551). Furthermore, our data linked DAPs to cancer related microRNA pathways, both miR-146a and miR-155 known as CD regulators[54,55]. Notably, clinical cohort studies have established that CD is a risk factor for colorectal cancer[30], and our findings provide a potential molecular basis for exploring this association.

The region-specific acetylome remodeling described above, particularly the enrichment of chromatin modifiers and NF-κB pathway components, suggests that deacetylase dysfunction may contribute to sustained pathway activation. In CD inflammation, the function of histone deacetylases (HDACs) such as sirtuin 1 (SIRT1), which serve as negative regulators of NF-κB acetylation, is frequently suppressed[56,57], potentially contributing to the NF-κB hyperacetylation observed in M-colon. Therapeutic modulation of acetylation dynamics therefore holds promise, as HDAC inhibitors such as vorinostat have been shown to alleviate experimental colitis by restoring regulatory T-cell function, while SIRT1 activators attenuate fibrosis by deacetylating SMAD3[58]. Moreover, the regional specific acetylation patterns identified in this study suggest that the acetylation regulatory network in CD is far more complex than previously anticipated, and that site specific deacetylase targeting may offer advantages over pan-HDAC inhibition. The quantitative acetyl-proteomic data generated in this work may provide a useful resource for identifying potential therapeutic targets in CD drug development.

In the present study, we did not detect statistically significant differences in overall microbial community diversity or structure between the control and CD model groups. This lack of significance may be attributable to the small sample size (n = 3 per group) and the limited resolution of sequencing only the V6 region of 16S rRNA, and should not be interpreted as evidence of equivalence. At the phylum level, both groups were dominated by Bacteroidetes and Firmicutes, with the model group showing a trend toward increased Bacteroidetes and decreased Firmicutes (Figure 6A, D and E). This direction of change is consistent with the reduced Firmicutes and expanded Bacteroidetes reported in CD patients[59,60]. LEfSe analysis revealed trends of differential abundance at the genus level: Lactobacillus and Parabacteroides were relatively enriched in the CD model group, whereas Alistipes was relatively enriched in the control group. From the acetyl-proteomic perspective, we observed acetylation changes in several proteins with known roles in host-microbe interactions, including ATG16 L1 and NOD2. Previous studies have shown that NOD2 deficiency alters the gut microbiota and increases colitis susceptibility[61], while ATG16 L1 deficiency impairs bacterial clearance and promotes inflammation[62]. The enrichment of acetylated proteins in pathways related to bacterial infection (e.g., Salmonella infection) further suggests potential connections between microbial ecology and protein acetylation. In the absence of any interventional experiments, these associations await future investigation using approaches such as microbiota transfer.

To evaluate the translational relevance of our rat model findings to human disease, we compared our acetylated proteins with those curated in the compendium of protein lysine modifications (CPLM)[63,64]. Among the 604 human lysine-acetylated proteins in CPLM, 32 proteins overlapped with our dataset (Figure 2). Notably, several acetylated proteins implicated in human CD, including MAP4K, Stat3, Jak3, and NF-κB, were present in both our rat acetylome and the human CPLM database. Although the overall overlap is modest, these shared proteins, particularly those with established roles in CD, suggest that our rat model-based acetylomic analysis may capture acetylation events of potential relevance to human disease. However, the translational significance of individual acetylation events should be interpreted conservatively pending validation in human CD samples.

From a translational perspective, the region-specific acetylation signatures and hub proteins identified in this study may offer new directions for the clinical management of CD. The regional differences in processes such as histone modification re-modeling and kinase inhibitor activity (Figure 4) indicate that the molecular pathology of CD varies along the proximal-distal axis. This variation may inform future disease stratification strategies and could even help explain the segmental predilection of CD complications. Furthermore, several hub proteins occupying central network positions, including Ripk2, Stat1, and Nfkb2, have been previously reported to be involved in CD pathogenesis[29,31] and may represent candidate biomarkers for monitoring disease activity or predicting treatment response. The region-specific differences in kinase inhibitor activity, together with the prominent ubiquitin/proteasome pathway activation observed in M-colon, also raise the possibility of personalized therapeutic strategies, in which patients with lesions in different colonic segments might benefit from distinct inhibitory approaches targeting these segment-specific pathways. Although these hypotheses require validation in clinical cohorts, the quantitative acetyl-proteomic resource generated in this work provides a molecular foundation for future biomarker-driven and precision medicine approaches in CD.

While this study focused on mapping the acetylome landscape, deeper mechanistic insights into how the identified modifications influence transcriptional regulation, protein-protein interactions, and microbial ecology remain to be explored. Functional validation of specific acetylation sites using approaches such as site-directed mutagenesis or acetyl-mimetic mutants will be essential to establish causal relationships. Integrating our acetyl-proteomic data with transcriptomic, metagenomic, and metabolomic datasets from matched samples would further enable the construction of comprehensive multi-omics models of CD pathogenesis.

We acknowledge that this study has several limitations. First, the sample size is relatively small, which limits statistical power. Second, although the TNBS-induced colitis model reproduces key inflammatory features relevant to CD, it does not fully recapitulate the chronic, relapsing nature and the genetic and immunological heterogeneity of human CD; therefore, translational extrapolations of our findings should be made with appropriate caution. Third, the relationships between protein acetylation and gut microbiota composition reported here are entirely correlational, and establishing causality will require interventional approaches such as microbiota transfer. Finally, although we employed targeted PRM to validate the acetylation status of selected proteins, customized site-specific antibodies for comprehensive functional validation remains great value.

CONCLUSION

This study presents a comprehensive, segment-resolved acetylproteomic landscape in a rat model of chronic colitis, offering new insights into the potential roles of lysine acetylation in intestinal inflammation. Our data suggest that region-specific Kac patterns are associated with key inflammatory signaling pathways, including NF-κB, MAPK, and JAK-STAT3, and with CD-relevant proteins such as RIPK2, Ywhaz and LRRK2. By revealing how Kac relates to core pathological processes from inflammatory cascades to microbial ecology and chromatin dynamics, these findings, together with the integrative microbiome analysis, provide a multi-omics perspective for understanding CD progression. This work provides a resource for future mechanistic and translational investigations in CD.

ACKNOWLEDGEMENTS

The authors thank Meng-Li Wang from the First Affiliated Hospital of Henan University of Chinese Medicine for her assistance with rat husbandry and animal experiments, and Dr. Yi-Ming Wu from the Third Affiliated Hospital of Henan University of Chinese Medicine for his help with visualization. Their contributions allowed our research to proceed smoothly.