Fecal metaproteomics enables functional characterization of remission in patients with inflammatory bowel disease

34

The gut microbiome is an important contributor to the development and the course of 35 inflammatory bowel disease (IBD). While changes in the gut microbiome composition were observed in 36 response to IBD therapy using biologics, studies elucidating human and microbial proteins and pathways 37 in dependence on therapy success are sparse. 38

Fecal samples of a cohort of IBD patients were collected before and after 14 weeks of treatment 39 with three different biologics. Clinical disease activity scores were used to determine the clinical response 40 and remission. Fecal metaproteomes of remitting patients (n=12) and of non-remitting patients (n=12) 41 were compared before treatment and changes within both groups were assessed over sampling time to 42 identify functional changes and potential human and microbial biomarkers. 43

The abundance of proteins associated with the intestinal barrier, neutrophilic granulocytes, and 44 immunoglobulins significantly decreased in remitting patients. In contrast, an increase of those proteins 45 was observed in non-remitting patients. There were significant changes in pathways of microbial 46 metabolism in samples from patients with remission after therapy . This included, for example, an 47 increased abundance of proteins from butyrat e fermentation. Finally, new potential biomarkers for the 48 prediction and monitoring of therapy success could be identified , e.g. human lysosome-associated 49 membrane glycoprotein 1, a cytotoxicity marker, or microbial anthranilate synthase component 2, a part 50 of the tryptophan metabolism. 51

Distinct changes of proteins related to gut inflammation and gut microbiome metabolism 52 showed whether IBD remission was achieved or not. This suggests that metaproteomics could be a useful 53 tool for monitoring remission in IBD therapies. 54

Clinical trials, Biomarkers, Microbiology 55 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 1. Introduction 56 Crohn's disease (CD) and ulcerative colitis (UC), summarized under the term inflammatory bowel diseases 57 (IBD), are chronic, recurrent inflammations of the intestinal tract that place enormous physical and 58 psychosocial burdens on the patient [1, 2]. IBD has been shown to be linked to the patient's genetics and 59 environmental factors . However, the patient's intestinal microbiome also plays a special role in the 60 etiology [3–6]. The permanent exposure of the intestinal epithelium to diverse microbial antigens requires 61 a fine orchestration of the immune response in the case of pathogen infection, while still exhibiting a 62 tolerance of commensal microorganisms [7]. This is ensured in the intestine of healthy individuals by an 63 intact epithelial barrier as well as a broad spectrum of immunologically active substances, such as 64 cytokines or adhesion molecules. In patients with IBD, the inflammatory response is inappropriate and the 65 epithelial barrier is damaged, which can lead to infiltration of the intestinal tissue with microorganisms 66 and consequent inflammatory cascades [8]. 67 Various therapeutic approaches such as drug, surgery, nutritional intervention, and fecal transplantation 68 are used to treat this inflammation and to achieve remission, a medical condition characterized by 69 decreasing severity of disease symptoms and mucosal healing [9]. Typical drugs are small-molecule drugs 70 and biologics. Small molecule drugs, such as thiopurines, aminosalicylates , or steroids, are inexpensive, 71 non-immunogenic molecules, but are characterized by a short half-life and low selectivity and efficacy [10, 72 11]. In contrast, biologics are large complex molecules, in the form of monoclonal antibodies or antibody-73 drug-conjugates, which specifically bind to certain molecules involved in IBD pathogenesis, particularly 74 those that interfere with immune cell communication or migration [12]. Although monoclonal antibodies 75 have a higher half -life, effectiveness , and selectivity, antibodies can be recognized as antigens by the 76 patient's immune system and thus their long-term effectiveness can be severely limited. In addition to the 77 immunogenicity of the antibodies, other factors contribute to widely varying efficacies of treatments in 78 different patients, such as genetic polymorphisms or metabolic potential of the patient's gut microbiome 79 [13–15]. 80 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint Despite increasing efforts and the introduction of novel targeted therapies in the treatment of IBD (anti-81 TNF, anti -a4b7 integrin, anti -IL12/23, anti -IL23, JAK inhibitors, S1P modulators) the overall long -term 82 disease control is still considerably low at approximately 40% [16]. 83 The early assessment of biomarkers that are associated with a favorable or unfavorable treatment 84 response to biologic therapies might therefore help to guide early treatment decision (e.g. swit ch 85 treatment class). Huge efforts have been undertaken to investigate the role of the intestinal microbiome 86 as a potential driver of treatment response in IBD. In this context most studies have initially focused on 87 analysis species diversity using 16S analysis, whilst lately metabolic properties (in -silico metabolic 88 modelling [17], fecal metabolomics [18]) have been the focus of investigation. 89 90 Although the metaproteome reflects the actual proteins expressed in the microbiome and the human 91 proteins in the stool sample and might show a stronger association with the inflammation state as gen-92 based methods [19], there are only very few publications on longitudinal studies of IBD patients using 93 metaproteomic methods compared to a multitude of metagenomic studies [20–24]. Therefore, this study 94 aimed to use metaproteomic approaches to investigate the host and microbial metabolism as well as the 95 abundance of proteins representing the host's immune response , to determine differences between 96 patients who achieved remission with biologics therapy and those who did not. 97 98 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 2. Material and Methods 99 Patients and sampling 100 In the course of this study, the 26 patients with IBD, including 13 patients diagnosed with CD and 13 101 patients diagnosed with UC, were treated with three different therapeutic agents (Infliximab: n=7, 102 Vedolizumab: n=9, Olamkizept n=10) at the University Hospital Schleswig Holstein , as described before 103 [13, 25] . At baseline and after 14 weeks of treatment, C-reactive protein (CRP) and Interleukin 6 (IL-6) in 104 the blood, calprotectin and blood leukocytes in the stool of the patients were measured by ELISA. Disease 105 intensity was characterized using the MAYO score for UC and the Harvey -Bradshaw index (HBI) for CD 106 patients. The participants provided paired stool samples at baseline and after 14 weeks of therapy. To 107 inactivate potential pathogenic material during sample processing, 200 µL of 1% Sodium docecyl sulfate 108 (SDS) solution were added to 200 µg of the stool samples and this solution was heated to 99°C for 10 109 minutes. 110 Metaproteomic analysis 111 Metaproteomic analysis was performed according to a previously established workflow [26, 27] . The 112 proteins of the solutions were extracted using phenol in a ball mill. The protein concentration was then 113 determined using the Amido Black assay. Subsequently, 10 µg protein of the samples were tryptically 114 digested (enzyme-substrate ratio 1:100) on a filter according to the filter-aided sample preparation (FASP) 115 protocol [28]. LC/MS -MS measurements were performed using an UltiMate 3000 RSLCnano 116 chromatography system (Thermo Fisher Scientific, Bremen, Germany) with a C18 reversed phase pre -117 column (Acclaim PepMap100 C18, Thermo Fisher Scientific, Bremen, Germany) and a C18 reversed-phase 118 separation column (Acclaim PepMap100 C18, Thermo Fisher Scientific, Bremen, Germany) coupled to a 119 timsTOF mass spectrometer (Bruker Daltonik GmbH, Bremen, Germany). Further details are summarised 120 in Supplementary Note 1. Compass DataAnalysis 5.1 software (Bruker Daltonik GmbH, Bremen, Germany) 121 was used to process and analyze the mass spectra. Using the search engine Mascot™ 2.6 (Matrix Science, 122 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint London, Great Britain) and the following parameters: enzyme trypsin, one missed cleavage, monoisotopic 123 mass, carbamidomethyl as fixed and oxidation as variable modification, ±0.02 Da precursor and ±0.02 Da 124 MS/MS fragment tolerance and a false discovery rate of 1 %, the peptides were annotated with the entries 125 of UniProtKB/SwissProt (16/01/2019) and a metagenome database published by Qin et al. [29], used in a 126 previous IBD study [30]. Identifications of redundant homologous proteins generated by these searches 127 were combined into protein groups (hereafter referred to as metaproteins) based on shared peptides 128 using MetaProteomAnalyzer, version 3.4. Metaproteins without functional or taxonomic annotation were 129 assigned using the BLAST and the UniProtKB/SwissProt database. Blast hits with an e -value <10 - 4 were 130 used for annotation according to the lowest common ancestor rule. Metaproteins with less than 10 spectra 131 were excluded from the following analysis. Subsequently, the spectral counts of the protein groups were 132 normalized to the total spectral count of the respective sample. Finally , a result matrix additionally 133 consisting of taxonomic annotation, enzyme commission (E.C.) number, Kyoto Encyclopedia of Genes and 134 Genomes (KEGG) orthologies, UniProtKB reference cluster, and UniProt keywords was generated, which 135 was used for the analysis. The MS files are available under accession number PXD053257 in the PRIDE 136 Archive. 137 Orthogonal validation of Lysosomal-associated membrane protein 1 (LAMP1) with independent validation 138 cohort using ELISA 139 For validation of biomarker discovery from metaproteomics experiments, fecale aspirates of n = 58 140 patients with UC aged 20-80 years were sampled between 2021 and 2023 at the First Medical Department 141 of University Hospital Schleswig-Holstein, Campus Kiel, Germany. All patients provided written informed 142 consent and biomaterial sampling from patients was approved by the local ethics committee of Kiel 143 University (Vote#: B231/98). Fecale aspirates were sampled during colonoscopy by flushing the distal 144 colon with water followed by aspiration of the fluids. Patients were recruited as a broad cross -sectional 145 cohort at different stages of disease activity and displayed an equal distribution of endoscopic Mayo 146 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint (eMayo) scores (0 – 3) and complete Mayo scores (0 – 11). Remission was defined as an endoscopic Mayo 147 Score of ≤ 1 and a complete Mayo score of ≤ 2. Samples were kept frozen at -80 degrees and measured by 148 ELISA according to the manufacturer (LAMP1: Abcam, ab277464) at the First Medical Department of 149 University Hospital Innsbruck, Austria. Measurements below the detection level were excluded from the 150 final analysis. 151 Statistical analysis 152 For statistical analysis, R-statistics (version 1.3.1093) was used. Unpaired samples (samples at baseline or 153 week 14 from patients with and without remission) were analyzed applying Wilcoxon rank sum test, paired 154 samples (samples at baseline and week 14 from patients with remission, samples at baseline and week 14 155 from patients without remission) were analyzed applying Wilcoxon signed rank test using the method 156 “wilcox.test”. The visualization with viol in plots was realized using the libraries “ggplot2” and 157 “ggstatsplot”. Volcano plots were created using Python (version 3.8.13) and the libraries “pandas”, 158 “numpy”, and “matplotlib”. 159 160 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 3. Results 161 Characteristics of the study cohort 162 163 In total, 26 patients provided fecal samples before and after 14 weeks of biologics therapy. The therapy 164 with one of three biologics (Infliximab: n=7, Vedolizumab: n=9, Olamkizept n=10) led to an overall 165 significant decrease in clinical disease scores (HBI: from 9.91 ± 7.32 to 4.18 ± 4.78, p-Value 0.042)(Mayo 166 score: from 7.15 ± 2.48 to 3.00 ± 2.99, p-value < 0.001)(Figure 1). As remission was characterized as a Mayo 167 score ≤2 respectively an HBI≤4 and a concentration of serum CRP ≤5 mg/L, 12 out of 26 patients reached 168 by definition a remitting state of the disease, 12 did not achieve remission and 2 patients were in remission 169 during the whole therapy. 19 patients showed an overall response to the therapy. In line with the MAYO 170 and HBI score, the concentration of Calprotectin decreased significantly by 77%, CRP by 73% and blood 171 leucocytes by 33% across all patients. The concentration of IL-6 was also reduced after therapy by 27%, 172 but this difference was not statistically significant (Figure 1). 173 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 174 Figure 1. Clinical and blood parameters of individuals with IBD at baseline and after 14 weeks of therapy with Infliximab, 175 Vedolizumab, or Olamkicept. Data are presented as violin plots, displaying the distribution of the data. Boxes depict the inner 176 quartile range, with the median (black line) and mean (red dot). The outer shape clarifies the kernel probability density of the 177 data. The dashed black lines indicate changes in the values for each patient . To analyze differences in paired samples Wilcoxon 178 signed-rank test was used (* p<0.05, ** p<0.01, *** p<0.001). Abbreviations: HBI (Harvey Bradshaw Index), CRP (C -reactive 179 protein), IL-6 (Interleukin-6) 180 181 Metaproteomics analysis 182 183 To more specifically characterize the developments of the host inflammation and metabolism as well as 184 the gut microbiome due to the induction of remission, we compared the fecal metaproteomes of remitting 185 patients (n=12) at baseline (BL), and after 14 weeks (W14) of biologics therapy. Furthermore, we analyzed 186 the changes in the fecal metaproteomes from patients (n=12) who did not achieve remission throughout 187 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint the therapy and, to check for possible metaprotein signatures predictive of therapy success, we compared 188 the metaproteomes of the two groups of patients before the therapy started. 189 On average over all measured samples, 10,268 (±2,149) metaproteins were identified and 31 % of spectra 190 were assigned to bacteria, 1 % to archaeal species, and 0.5 % were assigned to viral proteins. Moreover, 191 40 % of spectra were assigned to eukaryotic proteins (Supplementary Table 1). 192 193 194 195 Fecal metaproteome alterations through therapy in patients with remission 196 197 Numerous changes in the fecal metaproteomes of remitting patients (n=12) were observed. 198 The proportion of spectra assigned to human proteins decreased significantly from 15.7 % to 12.8% (p 199 <0.01) after the therapy in remitting patients (Figure 2A). Furthermore, 1,164 proteins were identified with 200 significantly altered abundance (p -value<0.05, Figure 2B). These included human proteins such as 201 immunoglobulin lambda variable 1-44 (p<0.001, fold change W14/BL: 0.52), lactotransferrin (p<0.001, fold 202 change W14/BL: 0.3), progranulin (p<0.001, fold change W14/BL: 0.43), stomatin-like protein 3 (p<0.001, 203 fold change W14/BL: 0.4) or peptidoglycan recognition protein 1 (PGLYRP1, p<0.001, fold change 204 W14/BL: 0.42), which were significantly reduced. In addition, the abundance of various microbial proteins 205 was also altered. The proteins with the lowest p -value included 50S ribosomal protein L7/L12 (p<0.001, 206 fold change W14/BL: 0.31), 3 -octaprenyl-4-hydroxybenzoate carboxy -lyase (p<0.001, fold change 207 W14/BL: 0.60), DNA-directed RNA polymerase subunit beta (p<0.001, fold change W14/BL: 0.41), PAN2-208 PAN3 deadenylation complex catalytic subunit (p<0.001, fold change W14/BL: 0.33) and small ribosomal 209 subunit biogenesis GTPase RsgA (p<0.001, fold change W14/BL: 0.32)(Table 1). 210 211 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 212 Figure 2. Proportion of spectra assigned to human proteins and differentially expressed metaproteins. A: Proportion of spectra 213 assigned to human proteins in samples from non -remitting and remitting patients at baseline and after 14 weeks of therapy 214 (Wilcoxon signed-rank test: ** p < 0.01. B: Volcano plot showing enrichment and depletion of metaproteins in baseline samples 215 from non-remitting patients compared to baseline samples from remitting patients. C: Volcano plot showing enrichment and 216 depletion of metaproteins in samples from remitting patients after 14 weeks of therapy compared to baseline. D: Volcano plot 217 showing enrichment and depletion of metaproteins in samples from non-remitting patients after 14 weeks of therapy compared 218 to baseline. In (B), (C), and (D), red dots describe sequences that have been enriched or depleted by factor 2 with a signifi cance 219 level of p < 0.05, calculated by Wilcoxon signed-rank test (B, C) and Wilcoxon rank sum test (D). The p -values and fold changes 220 depicted in (B), (C), and (D) are shown in Table S2. 221 222 Table 1. Significantly altered human and microbial metaproteins with lowest p -values (calculated with Wilcoxon signed -rank 223 test) in samples from patients with remission at week 14 compared to baseline. Unknown metaproteins were excluded. The 224 taxonomy column shows the lowest confirmed taxonomic rank. Metaproteins assigned to the kingdom Metazoa were considered 225 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint as host (human) metaprotein. Wilcoxon rank sum test was used to analyze differences between unpaired samples (n=24). The 226 function was inferred by UniProt keywords or GO term annotation. 227 Human metaproteins (ID) Description Function Taxonomy Average % abundance Fold change (W14/BL) p-Value 8343 Immunoglobulin lambda variable 1-44 Adaptive immunity Species: Homo Sapiens 3.93 x 10 -1 0.52 <0.001 10087 Lactotransferrin Antimicrobial Phylum: Chordata 1.34 x 10 -1 0.30 <0.001 12875 Granulins Cytokine Class: Mammalia 2.95 x 10 -2 0.43 <0.001 138 Stomatin-like protein 3 signal transduction Species: Homo Sapiens 8.96 x 10 -2 0.40 <0.001 11321 Peptidoglycan recognition protein 1 Innate immunity Species: Homo Sapiens 5.49 x 10 -2 0.42 <0.001 Microbial metaproteins (ID) Description Function Taxonomy Average % abundance Fold change (W14/BL) p-Value 15670 50S ribosomal protein L7/L12 Ribosomal protein Species: Persephonella marina 3.42 x 10 -2 0.31 <0.001 17989 3-octaprenyl-4-hydroxybenzoate carboxy-lyase Ubiquinone biosynthesis Species: Vibrio vulnificus 1.56 x 10 -2 0.60 <0.001 17075 DNA-directed RNA polymerase subunit beta Transcription Species: Clostridium novyi 1.12 x 10 -2 0.41 <0.001 15292 PAN2-PAN3 deadenylation complex catalytic subunit mRNA processing Species: Lodderomyces elongisporus 6.09 x 10 -3 0.33 <0.001 13358 Small ribosomal subunit biogenesis GTPase RsgA Ribosome biogenesis Genus: Bacillus 5.16 x 10 -3 0.32 <0.001 228 The comparison of functionally grouped proteins showed that intestinal barrier (p=0.034, fold change 229 W14/BL: 0.79), neutrophil granulocytes (p=0.009, fold change W14/BL: 0.56), "other immune cells" 230 (p=0.003, fold change W14/BL: 0.51) and immunoglobulin light chain proteins (p<0.001, fold change 231 W14/BL: 0.71) were significantly reduced in remission patients at week 14, compared to baseline (Figure 232 2 A, B). In contrast, we observed several microbial processes such as butyrate fermentation (p=0.042, fold 233 change W14/BL: 2.76), the pentose phosphate pathway ( PPW, p=0.007, fold change W14/BL: 1.79), 234 succinate synthesis (p=0.016, fold change W14/BL: 1.71) and sugar (p=0.003, fold change W14/BL: 1.97) 235 and peptide transport (p=0.029, fold change W14/BL: 2.83), that were significantly increased in patients 236 with clinical remissions (Figure 1C). Furthermore, we observed changes in the abundance of p roteins of 237 the blood (p=0.052, fold change W14/BL: 0.33), microbial lactate fermentation (p=0.569, fold change 238 W14/BL: 8.41) or the complement system (p=0.107, fold change W14/BL: 0.30) which however, were not 239 significant. 240 241 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 15 242 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint Figure 3. Overview of the functional assignment of identified metaproteins from patients at baseline and after 14 weeks of therapy. Data from patients who did not achieve remission 243 are displayed in blue, data from patients who achieved remission are displayed in orange. Significant changes are marked with *. Identified metaproteins were functionally assigned 244 to host metaproteins (A), microbial and human hydrolysis enzymes (B), microbial metabolism (C), and microbial transporters (D ) (Supplementary Table 3). Data are presented as 245 normalized average spectral abundance. All metaproteins assigned to the kingdom Metazoa were considered as host metaproteins. “Human trypsin” was shown as separated bar in 246 (B) since its abundance was potentially increased due to the trypsin used for tryptic digestion. The Wilcoxon rank sum test was used to analyze differences in independent samples (n 247 = 26, * p < 0.05, ** p < 0.01). The Wilcoxon signed rank test was used to analyze differences in dependent samples (n = 26, * p < 0.05, ** p < 0.01 for patients who achieved remission, 248 n = 26, * p < 0.05, ** p < 0.01 for patients who did not achieve remission ). Bar graphs show ± standard deviation.249 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 17 250 Fecal metaproteome alterations through therapy in patients without remission 251 252 In patients without remission (n=12), the proportion of spectra assigned to human proteins increased non-253 significantly from 14.8 % to 16.3 % and only 375 metaproteins changed significantly in abundance 254 (p<0.05). Human metaproteins which increased significantly, include LAMP1 (p=0.00 2, fold change 255 W14/BL: 3.0 5), polymeric immunoglobulin receptor ( PIGR, p=0.005, fold change W14/BL: 1.5 8), 256 pantetheinase (VNN1, p=0.005, fold change W14/BL: 1.70), calcium-activated chloride channel regulator 257 1 (CLCA1, p=0.007, fold change W14/BL: 1.84) and neutrophil defensin 3 ( DEF3A, p=0.007, fold change 258 W14/BL: 2.88). Microbial metaproteins which were significantly altered include e.g., ribosomal RNA large 259 subunit methyltransferase H (p<0.001, fold change W14/BL: 2.76) and anthranilate synthase component 260 2 (p=0.003, fold change W14/BL: 2.56) (Table 2). 261 Table 2. Significantly altered human and microbial metaproteins with lowest p -values (calculated with Wilcoxon signed -rank 262 test) in samples from patients without remission at week 14 compared to baseline. Unknown metaproteins were excluded. The 263 taxonomy column shows the lowest confirmed taxonomic rank. Metaproteins assigned to the kingdom Metazoa were considered 264 as host (human) metaprotein. The function was inferred by UniProt keywords or GO term annotation. 265 Human metaproteins (ID) Description Function Taxonomy Average % abundance Fold change (W14/BL) p-Value 12709 Lysosome-associated membrane glycoprotein 1 host-virus interaction Class: Mammalia 9.25 x 10 -3 3.05 0.002 8549 Polymeric immunoglobulin receptor polymeric Ig binding Class: Mammalia 5.75 x 10 -1 1.58 0.005 11174 Pantetheinase acute inflammatory response Class: Mammalia 4.69 x 10 -2 1.70 0.005 102 Calcium-activated chloride channel regulator 1 calcium transport Species: Homo sapiens 3.80 x 10 -1 1.84 0.007 368 Neutrophil defensin 3 Antimicrobial Order: Primates 1.29 x 10 -1 2.88 0.007 Microbial metaproteins (ID) Description Function Taxonomy Average % abundance Fold change (W14/BL) p-Value 14309 Ribosomal RNA large subunit methyltransferase rRNA processing Species: Clostridium difficile 7.77 x 10 -3 2.76 <0.001 15465 Anthranilate synthase component 2 tryptophan biosynthesis Species: Salmonella enterica 8.41 x 10 -3 2.56 0.003 11588 Uncharacterized protein HI_0392 acyltransferase activity Species: Bacillus subtilis 9.08 x 10 -3 2.01 0.005 11586 Alpha-monoglucosyldiacylglycerol synthase Carbohydrate metabolism Species: Streptococcus pneumoniae 1.03 x 10 -2 2.27 0.005 22126 Putative oligopeptide transport ATP-binding protein YkfD Translocase Species: Bacillus subtilis 8.44 x 10 -3 2.06 0.007 266 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 267 Even in patients who did not achieve remission, differences in the metabolic processes and host protein 268 abundance in stool metaproteome were observed after 14 weeks of therapy (Figure 3). Proteins of the 269 intestinal barrier (p=0.035, fold change W14/BL: 1.40), IG-J chains (p=0.017, fold change W14/BL: 2.22) 270 and human carbohydrate metabolism (p=0.011, fold change W14/BL: 1.49) were significantly increased in 271 abundance. An insignificant increase was observed for e.g., immunoglobulin light chain s (p=0.850, fold 272 change W14/BL: 1.07) and proteins from neutrophilic granulocytes (p=0.428, fold change W14/BL: 1.14). 273 In contrast, the microbial metabolic pathways of formate fermentation (p=0.013, fold change 274 W14/BL: 0.34) and pyruvate synthesis (p=0.004, fold change W14/BL: 0.50) were significantly decreased, 275 whereas vitamin B12 transport (p<0.001, fold change W14/BL: 1.60) was significantly increased. Human 276 blood (p=0.296, fold change W14/BL: 0.50) and complement system (p=0.266, fold change W14/BL: 0.65) 277 proteins were not significantly decreased. 278 279 280 Fecal metaproteome alterations between patients with and without remission at baseline 281 282 Whereas the proportion of spectra assigned to human proteins was not significantly different at baseline 283 between remitting and non-remitting patients (Figure 2A), a total of 550 metaproteins were significantly 284 altered between patients with and without remission at baseline (p-value < 0.05, Figure 2B). Human 285 proteins that were significantly increased in remission patients included DEF3A (p=0.003, ratio R/NR: 4.37), 286 HLA class II histocompatibility antigen, DR alpha chain (p=0.003, ratio R/NR: 5.27), or CLCA1 (p=0.004, ratio 287 R/NR: 2.74) (Table 3). In contrast, human metaproteins like centromere protein F were significantly 288 increased in non -remitting patients (p=0.002, R/NR: 0.17). Significantly changed microbial proteins 289 included riboflavin biosynthesis protein RibBA (p=0.002, ratio R/NR: 12.24), chloramphenicol 290 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint acetyltransferase 2 (p=0.002, R/NR: 0.17) and polyphosphate kinase (p=0.001, only in patients with 291 Remission). 292 293 Table 3. Significantly altered human and microbial metaproteins between patients with and without remission at baseline with 294 lowest p-Values. Unknown metaproteins were excluded. The taxonomy column shows the lowest confirmed taxonomic rank. 295 Metaproteins assigned to the kingdom Metazoa were considered as host (human) metaprotein. Wilcoxon rank sum test was used 296 to analyze differences between unpaired samples (n=24). The function was inferred by UniProt keywords or GO term annotation. 297 Human metaproteins (ID) Description Function Taxonomy Average % abundance Fold change (R/NR) p-Value 9938 Double homeobox protein 4-like protein 4 transcription regulation Species: Homo Sapiens 9.1 x 10 -2 7.65 0.001 23146 Centromere protein F cell cycle Species: Homo Sapiens 1.43 x 10 -3 0.17 0.002 368 Neutrophil defensin 3 antimicrobial Order: Primates 1.79 x 10 -1 4.37 0.003 17966 HLA class II histocompatibility antigen, DR alpha chain adaptive immunity Species: Homo Sapiens 8.36 x 10 -3 5.27 0.003 17342 Calcium-activated chloride channel regulator 1 calcium transport Class: Mammalia 1.29 x 10 -2 2.74 0.004 Microbial metaproteins (ID) Description Function Taxonomy Average % abundance Fold change (R/NR) p-Value 8869 Riboflavin biosynthesis protein RibBA riboflavin biosynthesis Superkingdom: Bacteria 1.65 x 10 -3 12.24 0.002 23145 Chloramphenicol acetyltransferase 2 antibiotic resistance Species: Escherichia coli 1.43 x 10 -3 0.17 0.002 23148 UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase cell wall biogenesis/degradation Species: Synechocystis sp. PCC 6803 1.04 x 10 -2 0.10 0.003 20508 Polyphosphate kinase kinase Class: Gammaproteobacteria 2.03 x 10 -1 only in patients with remission 0.003 3908 N-acetylneuraminate epimerase carbohydrate metabolism Family: Enterobacteriaceae 1.11 x 10 -3 only in patients without remission 0.003 298 299 To disentangle host and microbial processes , we independently assessed metaproteomes attributed to 300 humans or microbiota (Figure 3) . Concerning the functional processes of the gut microbiome, the 301 abundance of proteins involved in vitamin 12 transport was significantly higher (p=0.014, ratio R/NR: 1.86) 302 and the abundance of proteins of microbial methanogenesis (p =0.003, ratio R/NR: 0.34) was found 303 significantly lower in patients who achieved remission. (Figure 3C). Other non -significantly different 304 processes in terms of their protein abundance included bacteriocin transport (p=0.059, ratio R/NR: 3.88), 305 glycerol metabolism (p=0.159, ratio R/NR: 0.26) or succinate metabolism (p=0.143, ratio R/NR 0.74). 306 No processes to which human proteins were grouped, were significantly altered , except for proteins 307 assigned to carbohydrate hydrolysis (p=0.024, ratio R/NR: ). Proteins of neutrophil granulocytes (p=0.143, 308 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint ratio R/NR: 1.28) and proteins of mucosal functions (p=0.060, ratio R/NR: 1.42) were present in lower 309 abundance in patients who did not achieve remission. 310 311 312 Identification of disease-specific biomarker candidates for monitoring the success of IBD therapy 313 314 For the identification of disease-specific biomarker candidates for monitoring or predicting the success of 315 IBD therapy, only metaproteins with an average relative abundance of at least 0.01% in at least one patient 316 group (R -BL, R -W14, NR -BL, NR -W14) were considered. The most promising human and microbial 317 biomarkers, in our opinion, for which a pathogenesis contribution can be traced, are shown in Figure 4. 318 The first microbial biomarker shown, anthranilate synthase component 2 (Meta-Protein 15465, Salmonella 319 enterica) increases significantly during therapy in patients who did not achieve remission (NR-BL/NR-W14: 320 p<0.005). Furthermore, 3-octaprenyl-4-hydroxybenzoate carboxy -lyase (Meta -Protein 17989, Vibrio 321 vulnificus, R-BL/R-W14: p<0.001) was found significantly decreased and glutamate dehydrogenase (Meta-322 Protein 1031, Unknown Superkingdom , R-BL/R-W14: p=0.002 ) was found significantly increased in 323 remitting patients. 324 The abundance of the human proteins S100 -A8 (R-BL/R-W14: p=0.012) and CRP (R-BL/R-W14: p=0.036) 325 decreased in patients with remission in the course of therapy. Significant changes were also observed for 326 the human protein DEF3A (R-BL/NR-BL: p=0.001, R-BL/R-W14: p=0.021, NR-BL/NR-W14: p=0.007). Only in 327 patients without remission does the human protein LAMP1 (NR-BL/NR-W14: p=0.003) increase. In 328 contrast, the PGLYRP1 decreases significantly in patients with remission (R-BL/R-W14: p=0.001). Likewise, 329 the protein Granulins decreases significantly in remission patients (R-BL/R-W14: p<0.001). 330 In a first attempt to validate the finding from the metaproteomic biomarker discovery, LAMP1, as a 331 probable indicator for the failure of the therapy, was measured with ELISAs in fecal washes of UC patients 332 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint (Figure S1). A large but statistically insignificant (p=0.069) increase in the abundance of LAMP1 in was 333 found in patients, who were not in remission. 334 335 336 337 Figure 4. Panel of potential marker metaproteins as violin plots. Relative abundance of the marker metaproteins in fecal samples 338 from IBD patients who achieved remission (R) and IBD patients who did not achieve remission (NR) at baseline and after 14 weeks 339 of biologics therapy. The means are illustrated as red dots. P -values were calculated with Wilcoxon signed -rank test (between 340 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint timepoints of remitting or non-remitting patients) and Wilcoxon rank sum test (between remitting and non-remitting patients at 341 baseline). 342 343 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 4. Discussion 344 Fecal metaproteome alterations through therapy in patients with remission 345 With the help of the analysis of the fecal metaproteomes from patients who achieved remission, a large 346 number of changes in the course of therapy could be determined. These are obvious by the large number 347 of significantly altered proteins and the significantly decreased proportion spectra assigned to human 348 proteins in the course of the therapy , which was described previously to be indicative of disease activity 349 [31]. 350 The changes indicate on the one hand that the immune response of this group of patients has decreased 351 and on the other hand that proteins of the microbial metabolism are present in increased abundance. 352 The significantly decreased abundance of intestinal barrier proteins, as well as the decrease in blood 353 proteins, indicate an increased integrity of the epithelial layer of the intestine [32, 33]. Furthermore, the 354 decreased abundance of proteins from neutrophilic granulocytes is a sign of less translocation to the gut 355 of these immune cells [34, 35], which are an important driver of inflammation in IBD. The significantly 356 decreased human proteins, immunoglobulins, lactotransferrin , or progranulin have already been 357 associated with IBD [36–38]. Their decline in abundance suggests an ameliorated disease course. The 358 significant decrease in PGLYRP1 also indicates less dysbiosis with the gut microbiome, as this protein , 359 which is also part of neutrophil extracellular traps ( NETs), recognizes bacterial cell wall structures and 360 initiates the inflammatory response [39, 40]. 361 The reduced dysbiosis is also recognizable in the significantly increased abundance of microbial metabolic 362 pathways, such as the PPW, the tricarboxylic acid cycle /succinate metabolism, or butyrate fermentation. 363 The PPW is crucial for biomass increase and growth and an increased abundance of its enzymes may 364 indicate a prosperous intestinal microbiome [41], further supported by increased transport of sugars and 365 peptides. The increased metabolism of succinate, which promotes the release of pro -inflammatory 366 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint cytokines, could prevent the accumulation of it in the feces, which is a phenomenon observed in IBD 367 inflammation, and thereby further improve growth conditions [42]. Finally, butyrate fermentation in 368 particular is of great importance for the well-being of the patient, as butyrate serves as an energy source 369 for colonocytes and has great anti -inflammatory potential [43, 44]. The increased synthesis of butyrate 370 can also serve as an indicator of the stability of remission [45]. 371 372 Fecal metaproteome alterations through therapy in patients without remission 373 Based on the altered processes and proteins of the patients who did not achieve remission, a strongly 374 deviating reaction of the intestinal immune system and the gut microbiome to the therapy can be stated. 375 Immune-relevant processes and protein groups, like proteins from neutrophils or immunoglobulins, which 376 showed a clear decline in patients with remission, are not significantly decreased in this group of patients. 377 The increase of immunoglobulin light chain and immunoglobulin J -chain could also be a hint at the 378 formation of anti-drug antibodies, which are a major cause of therapy failure in biologics treatment [46], 379 The increased formation of microbiome targeting immunoglobulins was observed in active IBD [47], so 380 th95e target of those antibodies needs to be elucidated. Nevertheless, the secretion of immunoglobulins 381 into the gut lumen depends on the PIGR [48], which is one of the most significantly increased proteins in 382 non-remitting patients, supporting the hypothesis of an increased antibody response. 383 Other individual proteins, that have been described as the IBD susceptibility locus, such as VNN1 [49], or 384 that are associated with the immune response in IBD, like LAMP1 [50], are present in significantly 385 increased abundance after therapy. These findings indicate an ongoing inflammation whose character has 386 changed due to the induction of therapy. The reason for an ongoing inflammation could be the expansion 387 of apoptosis-resistant immune cells, which was observed in non -responding IBD patients in reaction to 388 biologics therapy. Immune modulatory therapies have also been shown to increase LAMP1 abundance 389 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [51], which is often used to measure degranulation and cytotoxic potential of cytotoxic lymphocytes [52]. 390 Increased cytotoxicity was observed in IBD patients in association with increased apoptosis in the intestinal 391 mucosa [53]. Identification of epithelial proteins like VNN1, PIGR, and CLCA1 among the most significantly 392 increased proteins and the significant increase in the abundance of intestinal barrier proteins in general, 393 would support this hypothesis. Increased abundance of proteins from neutrophilic granulocytes, especially 394 antimicrobial DEF3A, could be a reaction to increased microbial evasion due to barrier defects. As microbial 395 metabolite synthesis, like butyrate fermentation or formate synthesis, which is significantly decreased in 396 non-remitting patients, potentially limits the cytotoxic effects of immune cells [54, 55], the lack of a healthy 397 microbiome could contribute to therapy failure. 398 Nevertheless, there are some changes in the metaproteome of these patients that indicate at least a 399 slightly beneficial response to therapy. Proteins of microbial vitamin B12 transport were significantly 400 increased, which could indicate a reduced dysbiosis of the gut microbiome [56]. Furthermore, the 401 abundance of human blood proteins was halved. As fecal hemoglobin was previously identified as a 402 biomarker for intestinal inflammation [57], the decline of blood proteins could indicate at least a minor 403 amelioration of the disease. This is also supported by the reduction of proteins coupled to the complement 404 system. As this pathway is part of the innate immune response and the response to microbial -associated 405 molecular patterns this could also be a hint to less pronounced microbial dysbiosis. 406 407 Fecal metaproteome alterations between patients with and without remission at baseline 408 As there are no reliable biomarkers to predict the success of a therapy and clinicians often empirically try 409 different drugs to achieve remission [58], our final comparison aimed to identify metaprotein patterns, 410 that are likely to predict the success of biologics therapy before the therapy started. 411 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint In healthy individuals, the defense against pathogenic organisms in the human intestine takes place 412 through a balanced reaction of the innate and adaptive immune system [59]. Increased immune response 413 is seen at the onset of therapy, particularly in patients who subsequently achieve remission. Proteins of 414 the innate immune system, such as neutrophil DEF3A, and the adaptive immune system, such as HLA class 415 II histocompatibility antigen, are increased significantly in remitting patients . A high amount of alpha -416 defensins was indeed described as a prediction of therapy response in IBD therapy before [60]. 417 Furthermore, ROS are an important instrument of the intestinal immune response. But they represent a 418 double-edged sword in the pathogenesis of IBD, as a fine -tuned regulation is necessary to eliminate 419 pathogens, but at the same time avoid killing commensal microorganisms and damage tissue. The 420 increased abundance of microbial proteins that synthesize the antioxidants riboflavin and polyphosphate 421 in patients, who achieve remission suggests an active contribution of the gut microbiome to tissue 422 protection in these individuals [61]. In contrast, microbial processes that act as a sink for the anti -423 inflammatory molecule hydrogen [62], such as methanogenesis and glycerol metabolism, are more 424 prevalent in patients who do not achieve remission. These changes indicate an imbalance of the immune 425 response and the microbiome concerning oxidative stress, which possibly promotes the depletion of 426 beneficial microorganisms and tissue damage, which seems to extend during the therapy, as stated above. 427 This hypothesis is underlined by the increased abundance of blood proteins in patients without remission. 428 Furthermore, the increased abundance of chloramphenicol acetyltransferase 2 in non-remitting patients 429 hints at an involvement of the gut resistome, which is expanded in many diseases [63]. 430 For other significantly changed proteins like Double home obox protein 4-like protein 4 , a transcription 431 factor or centromere protein F, a marker for poor prognosis in cancer [64, 65], no connection to IBD or gut 432 immunity is described underlining the knowledge gaps still existing. 433 In general, predictive biomarkers are described as dependent on used biologic [66], which renders the 434 general assessment of predictive biomarkers hard in the present study. Nevertheless, as microbiome 435 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint fitness and human protein biomarkers are included in nearly all predictive models, the prediction of 436 therapy success using metaproteomics appears feasible. Taking into account other omics levels, like 437 metabolomics, and a personalized approach, thereby benefitting from knowledge about e.g., individual 438 polymorphisms of a patient, would further improve prediction potential [67]. 439 Identification of disease-specific biomarkers 440 To be able to predict the success of therapy before medication or to assess the effectiveness of ongoing 441 therapy, the search for biomarkers is a central topic of ongoing research. As we opted to identify markers, 442 which are present in easily detectable abundances, we only considered metaproteins with an average 443 relative abundance of at least 0.01% in at least one patient group (R-BL, R-W14, NR-BL, NR-W14). 444 In particular, fecal calprotectin (FC) and CRP have been investigated in a large number of studies using 445 antibody-based methods such as ELISA [68–70]. Although their use is often hindered by disadvantages 446 such as a lack of specificity [71], we considered them as biomarkers in our cohort. Both show a significant 447 decrease in patients with remission, but there is n o significant difference between patients with and 448 without remission at baseline to predict therapy success. Taken together with the observation of outlying 449 individuals, the need for more specific biomarkers or a panel of biomarkers is obvious. 450 The earlier-mentioned increased antimicrobial potential in patients without remission is supported by the 451 consideration of DEF3A. Furthermore, the data presented in Figure 4 regarding the abundance of the 452 proteins of neutrophil granulocytes are depicted reliably, so the protein seems to be suitable as a marker 453 protein for the invasion of this type of leukocyte. The violin plot (Figure 4) further implies that a high 454 amount of DEF3A, which has broad antimicrobial activity [72], in a patient's feces predicts therapy success 455 at baseline, whereas a low amount could be proof of a successful therapy after termination. Interestingly, 456 high DEF3A was also evaluated as a positive predictive marker in cancer immunotherapy [73]. 457 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint A marker for reduced dysbiosis in remission patients could also be the human PGLYRP1. Since this protein 458 is involved in antigen recognition [40], the decrease in this protein may indicate reduced invasion of 459 intestinal tissue by microorganisms. The significant decrease in granulins also indicates a decrease in 460 immune response and inflammation in remission patients. 461 In contrast, the significant increase in LAMP1, an essential protein of natural killer cell cytotoxicity [74], 462 may indicate a change in the nature of inflammation in patients without remission. This means physicians 463 may have to change the therapeutic strategy, for example to antibodies against NKG2D, which induces 464 cytolytic natural killer cells and T-cells upon interaction with stress-related molecules [75]. The validation 465 of LAMP1 with an ELISA also showed a higher abundance in non -remitting patients, but the increase was 466 not significant. Nevertheless, the strong elevation in distinct patients showed the relevance of the protein 467 for specific individuals supporting a precision medicine approach to the treatment of IBD patients. 468 As all human markers depicted in Figure 4 are also described in other malignancies [73, 76–80], their 469 specificity in IBD has to be evaluated further. Since microbial dysbiosis and associated functional and 470 taxonomic differences are hallmarks of IBD, there are also many attempts to discover microbial biomarkers 471 for disease monitoring or responsiveness to therapies . The significant differences in the microbial 472 biomarkers proposed in Figure 4 show, that they are suitable for disease monitoring. Markers for the 473 different development of microbial metabolism in patients with and without remission could be 474 anthranilate synthase component 2 (Meta-Protein 15465, Salmonella enterica), which is involved in of the 475 tryptophan metabolism and quorum sensing [81]. An increased tryptophan metabolism is under 476 investigation to aggravate disease activity in IBD patients [82]. Furthermore, anthranilate has antimicrobial 477 potential and by increasing in patients who did not achieve remission, the antimicrobial milieu in these 478 patients could be further worsened. Furthermore, the species the metaprotein is annotated to, Salmonella 479 enterica, is described to benefit from inflammation, to overcome colonization resistance and to play a role 480 in the onset of IBD [83, 84]. 481 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint In contrast a decreasing abundance of 3 -octaprenyl-4-hydroxybenzoate carboxylyase (Meta-Protein 482 17989, Vibrio vulnificus ), which catalyzes an intermediate step in the biosynthesis of ubiquinone, an 483 antioxidant and tested as supplement in inflammatory diseases, could serve as a marker of less oxidative 484 stress and a more hospitable environment for microorganisms. A similar sign of regrowth of a beneficial 485 gut flora could be glutamate dehydrogenase (Meta-Protein 1031, Unknown Superkingdom) , which has 486 great importance for microbes in gut colonization and is the most abundant metaprotein found in fecal 487 samples from healthy individuals [85, 86] . Nevertheless, as the gut microbiome underlies huge 488 interindividual and intraindividual changes, finding a single microbial marker protein remains a challenging 489 task, and a panel of marker proteins or the functional assessment of the microbiome might be more 490 promising. 491 492 Despite the ability of metaproteomics to characterize the induced changes in the remission of IBD patients 493 in contrast to non-remitting patients, the present study has limitations. Due to the small number of study 494 participants, the power of the statistical data analysis is limited. Furthermore, the field of participants is 495 quite heterogeneous with regard to their disease. In the evaluation, neither the type of disease (UC, CD) 496 or its local occurrence nor the type of therapeutic agent was considered. Due to the different mechanisms 497 of action of the drugs, different changes in the gastrointestinal microbiome are to be expected [87], 498 leading to specific metaproteome patterns especially between patients treated with different biologics. 499 This could minimalize the observations on the microbial aspects of the analysis. Larger clinical panels , 500 taking into account the different biologics and disease subtypes, to evaluate the differences identified here 501 would be useful. 502 Furthermore, the influence of environmental influences, e.g. changes in the patient’s diet in the course of 503 therapy, on the metaproteome of the study participants is difficult to assess, which could make the already 504 large interindividual differences even more significant. 505 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint Moreover, many identified meta-proteins, some of which changed significantly in the course of therapy, 506 could not be annotated. Analyzing different metaproteomic data sets may hold potential to help in 507 annotation of these potentially disease relevant features [88]. 508 509 Fecal metaproteomics enabled us to monitor the success of biological -based IBD therapy deduced from 510 proteins linked with inflammation and bacterial metabolisms, showing its potential for clinical application. 511 Remitting patients showed a decrease in immune functions, in particular, decreased abundance of 512 neutrophilic proteins and an increased abundance of several microbial functions, like butyrate synthesis. 513 The fecal metaproteomes of patients with an unsuccessful therapy were marked by increased abundance 514 of immunoglobulin proteins and signs of cytotoxicity and tissue damage. Furthermore, we proposed 515 several new potential biomarker candidates, which could also be of interest for further antibody -based 516 monitoring of IBD therapy. 517 518 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 5. Funding 519 This research was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft, 520 DFG) under Germany’s Excellence Strategy –EXC2167-Project ID 390884018 “Precision Medicine in Chronic 521 Inflammation” (K.A.) the RU5042-miTarget (to K.A.), the EKFS (Clinician Scientist Professorship, K.A.), the 522 BMBF (eMED Juniorverbund “Try-IBD” 01ZX1915A). 523 524 6. Acknoledgements 525 Not applicable. 526 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 7. Data Availability Statement 527 The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via 528 the PRIDE partner repository with the dataset identifier PXD053257. 529 Reviewer access details 530 Log in to the PRIDE website using the following details: 531 Project accession: PXD053257 532 Token: ZUI01Soinr2K 533 Alternatively, reviewer can access the dataset by logging in to the PRIDE website using the following 534 account details: 535 Username: [email protected] 536 Password: z6IsKunGzcvs 537 8. Conflicts of Interest: 538 Not applicable. 539 540 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 9. Author contributions: 541 ● Conceptualization: R.H., K.A. 542 ● Experiments metaproteomics: J.L., P.H. 543 ● ELISA experiments: L.W., S.N., F.T., S.T. 544 ● Data evaluation: J.L., M.W., L.W., S.N., F.T., S.T. 545 ● Bioinformatics: K.S., J.L., M.W. 546 ● Graphical abstract: K.S. 547 ● Supervision: R.H., K.A. 548 ● critical revision of the manuscript: D.B., M.G., R.H., K.A., K.S., U.R., P.R. 549 ● writing—original draft: M.W. 550 ● writing—review and editing: R.H., K.A. 551 552 All authors have read and agreed to the published version of the manuscript. 553 554 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint

555 [1] Yu YR, Rodriguez JR. Clinical presentation of Crohn's, ulcerative colitis, and indeterminate colitis: 556 Symptoms, extraintestinal manifestations, and disease phenotypes. Semin Pediatr Surg 2017; 26(6): 557 349–55 558 [https://doi.org/10.1053/j.sempedsurg.2017.10.003][PMID: 29126502] 559 [2] Bernabeu P, Belén-Galipienso O, van-der Hofstadt C, et al. Psychological burden and quality of life in 560 newly diagnosed inflammatory bowel disease patients. Front Psychol 2024; 15: 1334308 561 [https://doi.org/10.3389/fpsyg.2024.1334308][PMID: 38348263] 562 [3] Benoni C, Nilsson A. Smoking habits in patients with inflammatory bowel disease. A case -control 563 study. Scand J Gastroenterol 1987; 22(9): 1130 –6 564 [https://doi.org/10.3109/00365528708991970][PMID: 3423738] 565 [4] Glassner KL, Abraham BP, Quigley EMM. The microbiome and inflammatory bowel disease. J Allergy 566 Clin Immunol 2020; 145(1): 16–27 567 [https://doi.org/10.1016/j.jaci.2019.11.003][PMID: 31910984] 568 [5] Lange KM de, Moutsianas L, Lee JC , et al. Genome-wide association study implicates immune 569 activation of multiple integrin genes in inflammatory bowel disease. Nat Genet 2017; 49(2): 256 –61 570 [https://doi.org/10.1038/ng.3760][PMID: 28067908] 571 [6] Ye BD, McGovern DPB. Genetic variation in IBD: progress, clues to pathogenesis and possible clinical 572 utility. Expert Rev Clin Immunol 2016; 12(10): 1091 –107 573 [https://doi.org/10.1080/1744666X.2016.1184972][PMID: 27156530] 574 [7] Wu H-J, Wu E. The role of gut microbiota in immune homeostasis and autoimmunity. Gut Microbes 575 2012; 3(1): 4–14 576 [https://doi.org/10.4161/gmic.19320][PMID: 22356853] 577 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [8] Wei J, Feng J. Signaling pathways associated with inflammatory bowel disease. Recent Pat Inflamm 578 Allergy Drug Discov 2010; 4(2): 105–17 579 [https://doi.org/10.2174/187221310791163071][PMID: 20001899] 580 [9] Cai Z, Wang S, Li J. Treatment of Inflammatory Bowel Disease: A Comprehensive Review. Front Med 581 (Lausanne) 2021; 8: 765474 582 [https://doi.org/10.3389/fmed.2021.765474][PMID: 34988090] 583 [10] Liu J, Di B, Xu L-L. Recent advances in the treatment of IBD: Targets, mechanisms and related therapies. 584 Cytokine Growth Factor Rev 2023; 71 -72: 1 –12 585 [https://doi.org/10.1016/j.cytogfr.2023.07.001][PMID: 37455149] 586 [11] Na S-Y, Moon W. Perspectives on Current and Novel Treatments for Inflammatory Bowel Disease. Gut 587 Liver 2019; 13(6): 604–16 588 [https://doi.org/10.5009/gnl19019][PMID: 31195433] 589 [12] Yamamoto-Furusho JK, Parra -Holguín NN. Emerging therapeutic options in inflammatory bowel 590 disease. World J Gastroenterol 2021; 27(48): 8242 –61 591 [https://doi.org/10.3748/wjg.v27.i48.8242][PMID: 35068868] 592 [13] Aden K, Rehman A, Waschina S, et al. Metabolic Functions of Gut Microbes Associate With Efficacy of 593 Tumor Necrosis Factor Antagonists in Patients With Inflammatory Bowel Diseases. Gastroenterology 594 2019; 157(5): 1279-1292.e11 595 [https://doi.org/10.1053/j.gastro.2019.07.025][PMID: 31326413] 596 [14] Ben-Horin S, Chowers Y. Review article: loss of response to anti -TNF treatments in Crohn's disease. 597 Aliment Pharmacol Ther 2011; 33(9): 987 –95 598 [https://doi.org/10.1111/j.1365-2036.2011.04612.x][PMID: 21366636] 599 [15] Netz U, Carter JV, Eichenberger MR , et al. Genetic polymorphisms predict response to anti -tumor 600 necrosis factor treatment in Crohn's disease. World J Gastroenterol 2017; 23(27): 4958 –67 601 [https://doi.org/10.3748/wjg.v23.i27.4958][PMID: 28785150] 602 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [16] Mishra N, Aden K, Blase JI , et al. Longitudinal multi -omics analysis identifies early blood -based 603 predictors of anti -TNF therapy response in inflammatory bowel disease. Genome Med 2022; 14(1): 604 110 605 [https://doi.org/10.1186/s13073-022-01112-z][PMID: 36153599] 606 [17] Bauer E, Thiele I. From metagenomic data to personalized in silico microbiotas: predicting dietary 607 supplements for Crohn's disease. NPJ Syst Biol Appl 2018; 4: 27 608 [https://doi.org/10.1038/s41540-018-0063-2][PMID: 30083388] 609 [18] Vich Vila A, Hu S, Andreu -Sánchez S, et al. Faecal metabolome and its determinants in inflammatory 610 bowel disease. Gut 2023; 72(8): 1472 –85 611 [https://doi.org/10.1136/gutjnl-2022-328048][PMID: 36958817] 612 [19] Mills RH, Vázquez-Baeza Y, Zhu Q, et al. Evaluating Metagenomic Prediction of the Metaproteome in 613 a 4.5 -Year Study of a Patient with Crohn's Disease. mSystems 2019; 4(1) 614 [https://doi.org/10.1128/mSystems.00337-18][PMID: 30801026] 615 [20] Ding NS, McDonald JAK, Perdones -Montero A , et al. Metabonomics and the Gut Microbiome 616 Associated With Primary Response to Anti -TNF Therapy in Crohn's Disease. J Crohns Colitis 2020; 617 14(8): 1090–102 618 [https://doi.org/10.1093/ecco-jcc/jjaa039][PMID: 32119090] 619 [21] Effenberger M, Reider S, Waschina S, et al. Microbial Butyrate Synthesis Indicates Therapeutic Efficacy 620 of Azathioprine in IBD Patients. J Crohns Colitis 2021; 15(1): 88 –98 621 [https://doi.org/10.1093/ecco-jcc/jjaa152][PMID: 32687146] 622 [22] Lloyd-Price J, Arze C, Ananthakrishnan AN , et al. Multi-omics of the gut microbial ecosystem in 623 inflammatory bowel diseases. Nature 2019; 569(7758): 655 –62 624 [https://doi.org/10.1038/s41586-019-1237-9][PMID: 31142855] 625 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [23] Schierova D, Roubalova R, Kolar M , et al. Fecal Microbiome Changes and Specific Anti -Bacterial 626 Response in Patients with IBD during Anti -TNF Therapy. Cells 2021; 10(11) 627 [https://doi.org/10.3390/cells10113188][PMID: 34831411] 628 [24] Schirmer M, Franzosa EA, Lloyd -Price J , et al. Dynamics of metatranscription in the inflammatory 629 bowel disease gut microbiome. Nat Microbiol 2018; 3(3): 337 –46 630 [https://doi.org/10.1038/s41564-017-0089-z][PMID: 29311644] 631 [25] Schreiber S, Aden K, Bernardes JP , et al. Therapeutic Interleukin -6 Trans -signaling Inhibition by 632 Olamkicept (sgp130Fc) in Patients With Active Inflammatory Bowel Disease. Gastroenterology 2021; 633 160(7): 2354-2366.e11 634 [https://doi.org/10.1053/j.gastro.2021.02.062][PMID: 33667488] 635 [26] Sydor S, Dandyk C, Schwerdt J , et al. Discovering Biomarkers for Non -Alcoholic Steatohepatitis 636 Patients with and without Hepatocellular Carcinoma Using Fecal Metaproteomics. Int J Mol Sci 2022; 637 23(16) 638 [https://doi.org/10.3390/ijms23168841][PMID: 36012106] 639 [27] Biemann R, Buß E, Benndorf D , et al. Fecal Metaproteomics Reveals Reduced Gut Inflammation and 640 Changed Microbial Metabolism Following Lifestyle -Induced Weight Loss. Biomolecules 2021; 11(5) 641 [https://doi.org/10.3390/biom11050726][PMID: 34066026] 642 [28] Heyer R, Schallert K, Büdel A , et al. A Robust and Universal Metaproteomics Workflow for Research 643 Studies and Routine Diagnostics Within 24 h Using Phenol Extraction, FASP Digest, and the 644 MetaProteomeAnalyzer. Front Microbiol 2019; 10: 1883 645 [https://doi.org/10.3389/fmicb.2019.01883][PMID: 31474963] 646 [29] Qin J, Li R, Raes J , et al. A human gut microbial gene catalogue established by metagenomic 647 sequencing. Nature 2010; 464(7285): 59–65 648 [https://doi.org/10.1038/nature08821][PMID: 20203603] 649 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [30] Lehmann T, Schallert K, Vilchez -Vargas R, et al. Metaproteomics of fecal samples of Crohn's disease 650 and Ulcerative Colitis. J Proteomics 2019; 201: 93 –103 651 [https://doi.org/10.1016/j.jprot.2019.04.009][PMID: 31009805] 652 [31] Zhang X, Deeke SA, Ning Z , et al. Metaproteomics reveals associations between microbiome and 653 intestinal extracellular vesicle proteins in pediatric inflammatory bowel disease. Nat Commun 2018; 654 9(1): 2873 655 [https://doi.org/10.1038/s41467-018-05357-4][PMID: 30030445] 656 [32] Szymanska E, Wierzbicka A, Dadalski M, Kierkus J. Fecal Zonulin as a Noninvasive Biomarker of 657 Intestinal Permeability in Pediatric Patients with Inflammatory Bowel Diseases -Correlation with 658 Disease Activity and Fecal Calprotectin. J Clin Med 2021; 10(17) 659 [https://doi.org/10.3390/jcm10173905][PMID: 34501351] 660 [33] Fu Y, Wang L, Xie C, et al. Comparison of non-invasive biomarkers faecal BAFF, calprotectin and FOBT 661 in discriminating IBS from IBD and evaluation of intestinal inflammation. Sci Rep 2017; 7(1): 2669 662 [https://doi.org/10.1038/s41598-017-02835-5][PMID: 28572616] 663 [34] Guerrant RL, Araujo V, Soares E , et al. Measurement of fecal lactoferrin as a marker of fecal 664 leukocytes. J Clin Microbiol 1992; 30(5): 1238 –42 665 [https://doi.org/10.1128/jcm.30.5.1238-1242.1992][PMID: 1583125] 666 [35] Yadav SK, Ito N, Mindur JE , et al. Fecal Lcn-2 level is a sensitive biological indicator for gut dysbiosis 667 and intestinal inflammation in multiple sclerosis. Front Immunol 2022; 13: 1015372 668 [https://doi.org/10.3389/fimmu.2022.1015372][PMID: 36341389] 669 [36] Dai J, Liu W -Z, Zhao Y -P, Hu Y -B, Ge Z -Z. Relationship between fecal lactoferrin and inflammatory 670 bowel disease. Scand J Gastroenterol 2007; 42(12): 1440 –4 671 [https://doi.org/10.1080/00365520701427094][PMID: 17852860] 672 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [37] Wei F, Zhang Y, Jian J , et al. PGRN protects against colitis progression in mice in an IL -10 and TNFR2 673 dependent manner. Sci Rep 2014; 4: 7023 674 [https://doi.org/10.1038/srep07023][PMID: 25387791] 675 [38] Bourgonje AR, Roo -Brand G, Lisotto P , et al. Patients With Inflammatory Bowel Disease Show IgG 676 Immune Responses Towards Specific Intestinal Bacterial Genera. Front Immunol 2022; 13: 842911 677 [https://doi.org/10.3389/fimmu.2022.842911][PMID: 35693832] 678 [39] Cho JH, Fraser IP, Fukase K , et al. Human peptidoglycan recognition protein S is an effector of 679 neutrophil-mediated innate immunity. Blood 2005; 106(7): 2551 –8 680 [https://doi.org/10.1182/blood-2005-02-0530][PMID: 15956276] 681 [40] Dziarski R, Gupta D. The peptidoglycan recognition proteins (PGRPs). Genome Biol 2006; 7(8): 232 682 [https://doi.org/10.1186/gb-2006-7-8-232][PMID: 16930467] 683 [41] Rashida Z, Laxman S. The pentose phosphate pathway and organization of metabolic networks 684 enabling growth programs. Current Opinion in Systems Biology 2021; 28: 100390 685 [https://doi.org/10.1016/j.coisb.2021.100390] 686 [42] Wei Y-H, Ma X, Zhao J-C, Wang X-Q, Gao C-Q. Succinate metabolism and its regulation of host-microbe 687 interactions. Gut Microbes 2023; 15(1): 2190300 688 [https://doi.org/10.1080/19490976.2023.2190300][PMID: 36946592] 689 [43] Canani RB, Di Costanzo M, Leone L, Pedata M, Meli R, Calignano A. Potential beneficial effects of 690 butyrate in intestinal and extraintestinal diseases. World J Gastroenterol 2011; 17(12): 1519 –28 691 [https://doi.org/10.3748/wjg.v17.i12.1519][PMID: 21472114] 692 [44] Donohoe DR, Garge N, Zhang X, et al. The microbiome and butyrate regulate energy metabolism and 693 autophagy in the mammalian colon. Cell Metab 2011; 13(5): 517 –26 694 [https://doi.org/10.1016/j.cmet.2011.02.018][PMID: 21531334] 695 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [45] Buffet-Bataillon S, Bouguen G, Fleury F, Cattoir V, Le Cunff Y. Gut microbiota analysis for prediction of 696 clinical relapse in Crohn's disease. Sci Rep 2022; 12(1): 19929 697 [https://doi.org/10.1038/s41598-022-23757-x][PMID: 36402792] 698 [46] Marsal J, Barreiro-de Acosta M, Blumenstein I, Cappello M, Bazin T, Sebastian S. Management of Non-699 response and Loss of Response to Anti-tumor Necrosis Factor Therapy in Inflammatory Bowel Disease. 700 Front Med (Lausanne) 2022; 9: 897936 701 [https://doi.org/10.3389/fmed.2022.897936][PMID: 35783628] 702 [47] Masu Y, Kanazawa Y, Kakuta Y, et al. Immunoglobulin subtype-coated bacteria are correlated with the 703 disease activity of inflammatory bowel disease. Sci Rep 2021; 11(1): 16672 704 [https://doi.org/10.1038/s41598-021-96289-5][PMID: 34404881] 705 [48] Turula H, Wobus CE. The Role of the Polymeric Immunoglobulin Receptor and Secretory 706 Immunoglobulins during Mucosal Infection and Immunity. Viruses 2018; 10(5) 707 [https://doi.org/10.3390/v10050237][PMID: 29751532] 708 [49] Gensollen T, Bourges C, Rihet P, et al. Functional polymorphisms in the regulatory regions of the VNN1 709 gene are associated with susceptibility to inflammatory bowel diseases. Inflamm Bowel Dis 2013; 710 19(11): 2315–25 711 [https://doi.org/10.1097/MIB.0b013e3182a32b03][PMID: 23949622] 712 [50] Jin L, Li L, Hu C , et al. Integrative Analysis of Transcriptomic and Proteomic Profiling in Inflammatory 713 Bowel Disease Colon Biopsies. Inflamm Bowel Dis 2019; 25(12): 1906 –18 714 [https://doi.org/10.1093/ibd/izz111][PMID: 31173627] 715 [51] Muharam R, Febri RR, Prabowo KA, Bustami A, Mansur IG. Increased Levels of CD107a and 716 Intracellular Cytokines in IL-2 Stimulated PBMCs from Endometriosis Patients. Int J Inflam 2021; 2021: 717 5760959 718 [https://doi.org/10.1155/2021/5760959][PMID: 34616540] 719 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [52] Sudworth A, Dai K -Z, Vaage JT, Kveberg L. Degranulation Response in Cytotoxic Rat Lymphocytes 720 Measured with a Novel CD107a Antibody. Front Immunol 2016; 7: 572 721 [https://doi.org/10.3389/fimmu.2016.00572][PMID: 28003815] 722 [53] Souza HSP, Tortori CJA, Castelo-Branco MTL, et al. Apoptosis in the intestinal mucosa of patients with 723 inflammatory bowel disease: evidence of altered expression of FasL and perforin cytotoxic pathways. 724 Int J Colorectal Dis 2005; 20(3): 277 –86 725 [https://doi.org/10.1007/s00384-004-0639-8][PMID: 15503066] 726 [54] Zaiatz-Bittencourt V, Jones F, Tosetto M , et al. Butyrate limits human natural killer cell effector 727 function. Sci Rep 2023; 13(1): 2715 728 [https://doi.org/10.1038/s41598-023-29731-5][PMID: 36792800] 729 [55] Wenzel TJ, Gates EJ, Ranger AL, Klegeris A. Short -chain fatty acids (SCFAs) alone or in combination 730 regulate select immune functions of microglia -like cells. Mol Cell Neurosci 2020; 105: 103493 731 [https://doi.org/10.1016/j.mcn.2020.103493][PMID: 32333962] 732 [56] Kundra P, Geirnaert A, Pugin B, Morales Martinez P, Lacroix C, Greppi A. Healthy adult gut microbiota 733 sustains its own vitamin B12 requirement in an in vitro batch fermentation model. Front Nutr 2022; 734 9: 1070155 735 [https://doi.org/10.3389/fnut.2022.1070155][PMID: 36532531] 736 [57] Kato J, Hiraoka S, Nakarai A, Takashima S, Inokuchi T, Ichinose M. Fecal immunochemical test as a 737 biomarker for inflammatory bowel diseases: can it rival fecal calprotectin? Intest Res 2016; 14(1): 5–738 14 739 [https://doi.org/10.5217/ir.2016.14.1.5][PMID: 26884729] 740 [58] Moss AC. Approach to Treatment Failure in Inflammatory Bowel Disease. Gastroenterol Hepatol (N Y) 741 2022; 18(6): 360–3 742 [PMID: 36398141] 743 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [59] Round JL, Mazmanian SK. The gut microbiota shapes intestinal immune responses during health and 744 disease. Nat Rev Immunol 2009; 9(5): 313 –23 745 [https://doi.org/10.1038/nri2515][PMID: 19343057] 746 [60] Kanmura S, Uto H, Numata M, et al. Human neutrophil peptides 1-3 are useful biomarkers in patients 747 with active ulcerative colitis. Inflamm Bowel Dis 2009; 15(6): 909 –17 748 [https://doi.org/10.1002/ibd.20854][PMID: 19107772] 749 [61] Segawa S, Fujiya M, Konishi H, et al. Probiotic-derived polyphosphate enhances the epithelial barrier 750 function and maintains intestinal homeostasis through integrin-p38 MAPK pathway. PLoS One 2011; 751 6(8): e23278 752 [https://doi.org/10.1371/journal.pone.0023278][PMID: 21858054] 753 [62] Tian Y, Zhang Y, Wang Y , et al. Hydrogen, a Novel Therapeutic Molecule, Regulates Oxidative Stress, 754 Inflammation, and Apoptosis. Front Physiol 2021; 12: 789507 755 [https://doi.org/10.3389/fphys.2021.789507][PMID: 34987419] 756 [63] Fredriksen S, Warle S de, van Baarlen P, Boekhorst J, Wells JM. Resistome expansion in disease -757 associated human gut microbiomes. Microbiome 2023; 11(1): 166 758 [https://doi.org/10.1186/s40168-023-01610-1][PMID: 37507809] 759 [64] Landberg G, Erlanson M, Roos G, Tan EM, Casiano CA. Nuclear autoantigen p330d/CENP-F: A marker 760 for cell proliferation in human malignancies. Cytometry 1996; 25(1): 90 –8 761 [https://doi.org/10.1002/(SICI)1097-0320(19960901)25:13.0.CO;2-P] 762 [65] Zhuo Y -J, Xi M, Wan Y -P, et al. Enhanced expression of centromere protein F predicts clinical 763 progression and prognosis in patients with prostate cancer. Int J Mol Med 2015; 35(4): 966 –72 764 [https://doi.org/10.3892/ijmm.2015.2086][PMID: 25647485] 765 [66] Caenepeel C, Falony G, Machiels K, et al. Dysbiosis and Associated Stool Features Improve Prediction 766 of Response to Biological Therapy in Inflammatory Bowel Disease. Gastroenterology 2024; 166(3): 767 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint 483–95 768 [https://doi.org/10.1053/j.gastro.2023.11.304][PMID: 38096956] 769 [67] Stevens TW, Matheeuwsen M, Lönnkvist MH , et al. Systematic review: predictive biomarkers of 770 therapeutic response in inflammatory bowel disease -personalised medicine in its infancy. Aliment 771 Pharmacol Ther 2018; 48(11-12): 1213–31 772 [https://doi.org/10.1111/apt.15033][PMID: 30378142] 773 [68] Schoepfer AM, Beglinger C, Straumann A, Trummler M, Renzulli P, Seibold F. Ulcerative colitis: 774 correlation of the Rachmilewitz endoscopic activity index with fecal calprotectin, clinical activity, C-775 reactive protein, and blood leukocytes. Inflamm Bowel Dis 2009; 15(12): 1851 –8 776 [https://doi.org/10.1002/ibd.20986][PMID: 19462421] 777 [69] D'Haens G, Ferrante M, Vermeire S , et al. Fecal calprotectin is a surrogate marker for endoscopic 778 lesions in inflammatory bowel disease. Inflamm Bowel Dis 2012; 18(12): 2218 –24 779 [https://doi.org/10.1002/ibd.22917][PMID: 22344983] 780 [70] Anindita B, Sugihartono T, Miftahussurur M , et al. High levels of fecal calprotectin and C -reactive 781 protein in patients with colitis. J Med Life 2023; 16(1): 48 –51 782 [https://doi.org/10.25122/jml-2021-0311][PMID: 36873123] 783 [71] Nowak JK, Kalla R, Satsangi J. Current and emerging biomarkers for ulcerative colitis. Expert Rev Mol 784 Diagn 2023; 23(12): 1107–19 785 [https://doi.org/10.1080/14737159.2023.2279611][PMID: 37933807] 786 [72] Cunliffe RN. Alpha -defensins in the gastrointestinal tract. Mol Immunol 2003; 40(7): 463 –7 787 [https://doi.org/10.1016/s0161-5890(03)00157-3][PMID: 14568393] 788 [73] Berghmans E, Jacobs J, Deben C , et al. Mass Spectrometry Imaging Reveals Neutrophil Defensins as 789 Additional Biomarkers for Anti-PD-(L)1 Immunotherapy Response in NSCLC Patients. Cancers (Basel) 790 2020; 12(4) 791 [https://doi.org/10.3390/cancers12040863][PMID: 32252405] 792 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [74] Krzewski K, Gil-Krzewska A, Nguyen V, Peruzzi G, Coligan JE. LAMP1/CD107a is required for efficient 793 perforin delivery to lytic granules and NK -cell cytotoxicity. Blood 2013; 121(23): 4672 –83 794 [https://doi.org/10.1182/blood-2012-08-453738][PMID: 23632890] 795 [75] Vadstrup K, Bendtsen F. Anti-NKG2D mAb: A New Treatment for Crohn's Disease? Int J Mol Sci 2017; 796 18(9) 797 [https://doi.org/10.3390/ijms18091997][PMID: 28926962] 798 [76] Allin KH, Nordestgaard BG. Elevated C -reactive protein in the diagnosis, prognosis, and cause of 799 cancer. Crit Rev Clin Lab Sci 2011; 48(4): 155 –70 800 [https://doi.org/10.3109/10408363.2011.599831][PMID: 22035340] 801 [77] Dang Q, Zhou H, Qian J , et al. LAMP1 Overexpression Predicts for Poor Prognosis in Diffuse Large B -802 cell Lymphoma. Clin Lymphoma Myeloma Leuk 2018; 18(11): 749 –54 803 [https://doi.org/10.1016/j.clml.2018.07.288][PMID: 30082222] 804 [78] Klupp F, Kahlert C, Franz C, et al. Granulin: An Invasive and Survival-Determining Marker in Colorectal 805 Cancer Patients. Int J Mol Sci 2021; 22(12) 806 [https://doi.org/10.3390/ijms22126436][PMID: 34208547] 807 [79] Schnell A, Huang L, Regan BML , et al. Targeting PGLYRP1 promotes antitumor immunity while 808 inhibiting autoimmune neuroinflammation. Nat Immunol 2023; 24(11): 1908 –20 809 [https://doi.org/10.1038/s41590-023-01645-4][PMID: 37828379] 810 [80] Shabani F, Farasat A, Mahdavi M, Gheibi N. Calprotectin (S100A8/S100A9): a key protein between 811 inflammation and cancer. Inflamm Res 2018; 67(10): 801 –12 812 [https://doi.org/10.1007/s00011-018-1173-4][PMID: 30083975] 813 [81] Palmer GC, Jorth PA, Whiteley M. The role of two Pseudomonas aeruginosa anthranilate synthases in 814 tryptophan and quorum signal production. Microbiology (Reading) 2013; 159(Pt 5): 959 –69 815 [https://doi.org/10.1099/mic.0.063065-0][PMID: 23449919] 816 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint [82] Nikolaus S, Schulte B, Al-Massad N, et al. Increased Tryptophan Metabolism Is Associated With Activity 817 of Inflammatory Bowel Diseases. Gastroenterology 2017; 153(6): 1504 -1516.e2 818 [https://doi.org/10.1053/j.gastro.2017.08.028][PMID: 28827067] 819 [83] Schultz BM, Paduro CA, Salazar GA , et al. A Potential Role of Salmonella Infection in the Onset of 820 Inflammatory Bowel Diseases. Front Immunol 2017; 8: 191 821 [https://doi.org/10.3389/fimmu.2017.00191][PMID: 28293241] 822 [84] Stecher B, Robbiani R, Walker AW , et al. Salmonella enterica serovar typhimurium exploits 823 inflammation to compete with the intestinal microbiota. PLoS Biol 2007; 5(10): 2177 –89 824 [https://doi.org/10.1371/journal.pbio.0050244][PMID: 17760501] 825 [85] Kolmeder CA, Been M de, Nikkilä J, et al. Comparative metaproteomics and diversity analysis of human 826 intestinal microbiota testifies for its temporal stability and expression of core functions. PLoS One 827 2012; 7(1): e29913 828 [https://doi.org/10.1371/journal.pone.0029913][PMID: 22279554] 829 [86] Girinathan BP, Braun S, Sirigireddy AR, Espinola -Lopez J, Govind R. Correction: Importance of 830 Glutamate Dehydrogenase (GDH) in Clostridium difficile Colonization In Vivo. PLoS One 2016; 11(10): 831 e0165579 832 [https://doi.org/10.1371/journal.pone.0165579][PMID: 27768767] 833 [87] Ruszkowski J, Daca A, Szewczyk A, Dębska -Ślizień A, Witkowski JM. The influence of biologics on the 834 microbiome in immune-mediated inflammatory diseases: A systematic review. Biomed Pharmacother 835 2021; 141: 111904 836 [https://doi.org/10.1016/j.biopha.2021.111904][PMID: 34328113] 837 [88] Sun Z, Ning Z, Figeys D. The Landscape and Perspectives of the Human Gut Metaproteomics. Mol Cell 838 Proteomics 2024; 23(5): 100763 839 [https://doi.org/10.1016/j.mcpro.2024.100763][PMID: 38608842] 840 841 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 3, 2024. ; https://doi.org/10.1101/2024.07.02.24309587doi: medRxiv preprint