Bacterial composition comprising akkermansia species, for improving the response to car t-cell therapy

Abstract

The present invention is based on the demonstration that bacteria of the Akkermansia genus improve the response to CAR T-cells therapy. According to the invention, Akkermansia bacteria are administered to patients treated with CAR T-cells, especially if their gut microbiome was devoid of such bacteria before administering said CAR T-cells.

Description

[0001] Bacterial composition comprising Akkermansia species, for improving the response to CAR T-cell therapy

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to the field of anticancer therapy. More precisely, the invention is in the field of immunotherapy, especially using adoptive transfer of CAR T-cells. The invention provides compositions and methods useful for improving the response to CAR T-cells therapy by modulating the patient’s microbiota.

[0004] BACKGROUND OF THE INVENTION

[0005] Adoptive cell therapy, particularly chimeric antigen receptor (CAR) T cells, has revolutionized the treatment of B cell malignancies. Nowadays, CD19 CAR-T cells are the reference treatment for second-line therapy in the context of diffuse large B cell lymphoma (DLBCL), the most common subtype of B cell malignancies (Abramson et al., 2020; Bachy et al., 2022; Locke et al., 2022; Schuster et al., 2021). Recent reports have suggested that gut microbiota composition may be a predictive biomarker of CAR-T cell efficacy (Smith et al., 2022; Stein-Thoeringer et al., 2023). Indeed, the use of broadspectrum antibiotics (ATB) before CAR-T cell infusion was associated with worse progression-free survival and overall survival in a cohort of B cell malignancies (Smith et al., 2022). While the use of ATB before CAR-T cell infusion may be confounded with a higher tumor burden and poor performance status, it was also found that the prevalence of distinct intestinal taxa, such as Bacteroides eggerthii, Ruminococcus lactaris, Eubacterium spp., and Akkermansia muciniphila, were associated with CAR-T cell responsiveness in B cell malignancies.

[0006] Likewise, pioneering reports followed by broad metanalyses heralded the deleterious impact of ATB on the clinical outcome during immune checkpoint blockade with or without chemotherapy across various solid malignancies (Derosa et al., 2021 ; Elkrief et al., 2024; Routy et al., 2018). Moreover, preclinical studies demonstrated the cause effect relationship between fecal taxonomic composition or oral gavage with specific bacteria and response or resistance to immunotherapy (Chaput et al., 2017; Derosa et al., 2022, 2020; Frankel et al., 2017; Gopalakrishnan et al., 2018; Jiang et al., 2023; Matson et al., 2018; Pinato et al., 2019; Routy et al., 2018). Supporting these findings, the potential of harnessing the gut microbiota by fecal microbiota transplantation to circumvent primary resistance to or boosting the effects of ICB was unveiled in three phase I clinical trials (Baruch et al., 2021 ; Davar et al., 2021 ; Derosa et al., 2021 ; Qiu et al., 2020; Routy et al., 2023) in advanced melanoma patients. Among the key immunogenic species is Akkermansia muciniphila, an anaerobic Gram-negative bacterium from the phylum Verrucomicrobia specialized in the degradation of mucin (Derrien et al., 2004) and known to be a masterpiece of the intestinal homeostasis and host fitness.

[0007] Several studies have highlighted that the impact of Akkermansia on the clinical efficacy of anticancer therapeutics was not restricted to immune-checkpoint inhibitors. Indeed, the fecal presence of Akkermansia appears to be a promising predictor of responses to Chimeric Antigen Receptor (CAR) T-based therapies (Smith et al., 2022; Stein-Thoeringer et al., 2023). Akkermansia was the top enriched dominant taxon in patients with day 100 complete response (Smith et al., 2022). A parallel study corroborated the predictive role of Akkermansia in this scenario. Akkermansia has been identified as one the most important commensal in determining CAR-T responsiveness and was associated with pre-infusion peripheral T cell levels in cancer patients (Stein- Thoeringer et al., 2023). Conversely, in WO 2022 / 165255, over-representation of Akkermansia muciniphila in the intestinal microbiota of a patient was identified as indicating that the patient would likely not respond to an adoptive T-cell therapy.

[0008] While CAR-T cells represent a breakthrough immunotherapy in hematological malignancies and also hold promise in solid tumors (Bufalo et al., 2023; Heczey et al., 2023; Hegde et al., 2024; Sterner and Sterner, 2021 ; Vitanza et al., 2021), about 50% of diffuse large B cell lymphoma patients exhibit primary resistance to CAR- T cells, illustrated by treatment failure within the first 3 to 6 months following infusion (Bachy et al., 2022; Sterner and Sterner, 2021). Therefore, there is a crucial need to improve this response rate, potentially through combinatorial approaches. By modulating the host epithelial, immunological, and metabolic functions, gut microbiota-centered interventions may represent a valuable strategy independent and possibly synergistic with cell intrinsic options currently investigated. However, preclinical studies using syngeneic (as opposed to humanized) tumor models are mandatory given the impact of CAR-T cell conditioning (lymphodepletive chemotherapy) on homeostasis.

[0009] SUMMARY OF THE INVENTION

[0010] The inventors investigated the gut microbiota alterations associated with CAR-T cell failure in patients diagnosed with B cell malignancies, following infusions of commercial CAR-T cells, by performing longitudinal stool metagenomics coupled to plasma metabolomics. CAR-T cell conditioning (lymphodepletive chemotherapy), broadspectrum ATB and / or CAR-T cell infusion induced a profound gut dysbiosis characterized by a markedly reduced taxonomic richness and a loss of Akkermansia species. While alpha and beta diversity did not correlate with treatment outcomes, a distinct gut microbiota containing Akkermansia muciniphila was found in CAR-T cell responders accompanied by CAR-T cell accumulation and a Th1 cytokine pattern in their bone marrow (BM). Taking advantage of an immunocompetent mouse model of highgrade B cell lymphoma, they brought up the first proof-of-concept of a causal relationship between Akkermansia prevalence in the intestines and increased potency of CAR-T cells.

[0011] According to a first aspect, the present invention pertains to a bacterial composition comprising bacteria of the Akkermansia genus, for use in the treatment of a cancer amenable to CAR T cell therapy, wherein said composition is used in combination with CAR T cells targeting said cancer.

[0012] The invention also pertains to a combination of CAR T cells and bacteria of the Akkermansia genus, for use in the treatment of a cancer amenable to CAR T cell therapy.

[0013] Another aspect of the invention is a CAR T-cell for use in the treatment of a cancer amenable to CAR T cell therapy, wherein said CAR T-cell is used in combination with a bacterial composition comprising bacteria of the Akkermansia genus.

[0014] According to particular embodiments of the above aspects, the bacteria are of the Akkermansia massiliensis species.

[0015] BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1. PIONEER protocol clinical setting. Fecal material was prospectively and longitudinally collected before lymphodepletive chemotherapy, between 7 and 15 days, and 3 months after CAR-T cell therapy (A) in B-cell malignancies patients - mostly diffuse large B-cell lymphoma receiving commercial CD19 CAR-T cells, mostly axi-cel - observing a low abundance of Akkermansia muciniphila in the cohort, when compared to solid tumors (B-C).

[0017] Figure 2. Immunocompetent B-cell lymphoma-bearing mice. Subcutaneous injection of a c-myc driven B-cell lymphoma cell line expressing CD19, demonstrated by flow cytometry and fluorescence microscopy (A), resulted in the rapid development of an aggressive tumor in vivo (B) and tumor dissemination to all organs explored (C).

[0018] Figure 3. Murine mCD19 CAR-T cells. A CAR-T cell receptor targeting CD19 (A) demonstrated higher cytotoxicity in vitro at different effector: tumor cell ratios (B) and higher production of IFNy and GM-CSF, when compared to activated nontransduced T cells co-cultured with a lymphoma cell line (C-D). Figure 4. Pre-clinical CAR-T cells murine model in B-cell lymphomabearing mice. In vivo subcutaneous injection of murine CAR-T cells targeting CD19 resulted in an initial tumor regression, although relapse was observed 2 months after CAR-T cell infusion (A). CAR-T cells were found to be circulating in the blood of the treated mice 10 days after CAR-T cell infusion (B) and infiltrating all organs explored after relapse (C), including the tumor, which still expressed the target antigen CD19 at such time point (D).

[0019] Figure 5. Combination treatment of CAR-T cells + Akkermansia Eve#01. A significant increase in anti-tumor efficacy (A-C) was observed in mice receiving repetitive oral gavage with Akkermansia massiliensis (p=0.032; Mann-Whitney statistical test performed at day 10 after CAR-T cell intravenous injection between both treatment arms), as well as an increase in overall survival (D) (p<0.001 ; Log-rank test).

[0020] Figure 6. Tolerance of the treatment with CAR-T cells + Akkermansia Eve#01. Mice weight was stable over time and between treatment groups, during the course of the observation (A) and at the time of sacrifice (B), which was established according to tumor growth and behavioral appearance (C-D).

[0021] Figure 7. CAR-T cells expansion in the blood. No significant differences were observed in the proportion of circulating CAR-T cells in the blood of the mice receiving CAR-T cells + Akkermansia Eve#01 compared to the mice receiving CAR- T cells alone (A-B) at day 7 and 14 after CAR-T cell injection, although a significant increase in cytotoxicity was observed in mice receiving CAR-T cells + Akkermansia Eve#01 14 days after CAR-T cell injection, when compared to CAR-T cells alone (C).

[0022] Figure 8. 2 oxovaleric acid levels in responder (R) vs. non responder (NR) lymphoma patients at 7 days before and 3 months after CAR T cells infusion.

[0023] Mass spectrometry-based MS-MS plasma metabolomics has been performed in all lymphoma patients 7 days prior to CAR-T transfer and 3 months after CAR-T transfer in responders (R) and non responders (NR). Each dot represents one patient value. * statistical analyses using paired Student t’test and Wilcoxon.

[0024] Figure 9. Akkermansia massiliensis induced propionate, aKG and oxovalerate.

[0025] In vitro culturomics of two strains of Akkermansia, one A. muciniphila (Akk#5, SGB9226) and one A. massiliensis (Akk#1 , Akkermansia Eve#01) in appropriate culture medium in anaerobic conditions. Supernatants of Akkermansia have been harvested at day 2 and targeted MS-MS metabolomics has been performed comparing control (no Akkermansia spp) and both Akkermansia culture media at 48 hrs. Violin plots showing the results of three wells per condition. Box plots representing normalized values of propionic, 2 oxoglutaric and 2 oxovaleric acid in the culture supernatant of Akk#1 versus Akk#5. (B) Acetate, propionate and butyrate production (mM) during the treatment period of a SHIME simulating the impact of daily administration of Akk#1 on the gut microbiota from human cancer-patient. The negative control arm (CTRL) did not receive any test product. (C-D). Means ± SEM of tumor sizes of SPF versus FMT treated mice after Iso Ctrl mAb or after PD1 mAb in MCA-205-bearing supplemented with different Akk strains (of Akk#1= Akkermansia Eve#01 versus Akk#5= SGB9226 strain) (C) or propionate in drinking water (D).

[0026] Figure 10. Lymphoma patients exhibit a defect in propionate after CAR-T cell infusion.

[0027] Mass spectrometry-based MS-MS plasma metabolomics has been performed in all lymphoma patients 7 days prior to CAR-T transfer and 15 days as well as 3 months after CAR-T transfer. Each dot represents one patient value. * statistical analyses using paired Student t’test and Wilcoxon.

[0028] Figure 11. Association between gut bacterial composition at baseline and ORR.

[0029] (A) Schematic representation of the prospective longitudinal collection of fecal material, blood, tumor tissues, and bone marrow (BM) aspirations from 33 patients with B cell malignancies treated with commercial anti-CD19 CAR-T cells. Visit 1 occurred prior to lymphodepletive conditioning chemotherapy, visit 2 between days 7 and 15 post CAR-T cell infusion, and visit 3 at 3 months after CAR-T cell infusion. (B) Alpha diversity as determined by the Shannon index (left) and species richness index (right) assessed at visits 1 (color code grey), 2 (color code blue), and 3 (color code red). Data are represented as means ± standard error of the mean; statistical significance was evaluated using Wilcoxon signed-rank tests. (C) Comparison of bacterial alpha diversity between responders (ORR+) and non-responders (ORR-) at visits 1 , 2, and 3, calculated using the Shannon index (left) and species richness index (right). Results are shown as means ± standard error of the mean, with Wilcoxon signed-rank tests applied. (D) Beta diversity was analyzed using the Bray-Curtis metric to measure the distance between microbial communities, with PERMANOVA tests performed at visits 1 , 2, and 3. (E) Comparison of bacterial beta diversity between responders (ORR+) and non-responders (ORR-) at visits 1 (color code grey), 2 (color code blue), and 3 (color code red), using PERMANOVA tests. (F) A Linear Discriminant Analysis Effect Size (LEfSe) graph was generated using metagenomic data from baseline (visit 1) samples, highlighting the most discriminant species between ORR+ and ORR- patients based on LDA scores. (G) Kaplan-Meier curves, along with log-rank test analyses, depicting patient progression- free survival (PFS) stratified on the gut prevalence of Akkermansia muciniphila SGB9226 (Akk. spp.) at baseline. Figure 12. Akkermansia spp. supplementation improved the efficacy of CD19 / CD28- CAR-T cells against B cell lymphoma

[0030] (A) Schematic representation of the experimental design of the CD19 / CD28- CAR-T cell mouse model. (B-C) Tumor growth kinetics and animal survival following infusion of CAR-T with or without oral live biotherapeutics. Data are presented as means ± standard error of the mean of tumor sizes monitored by a caliper (B). Results from a representative experiment are depicted in (B). Three independent experiments were conducted, with 15 mice per treatment group. Tumor sizes were analyzed across all time points using one-way ANOVA, while unpaired nonparametric t-tests (Mann- Whitney) were used for statistical comparisons on specific days (day 10 and day 34). Survival analysis between the treatment arms was conducted using the Kaplan-Meier estimator and the log-rank test (C). (D) Flow cytometry analyses of the blood compartment on day 14 post-CAR-T cell IV injection. Evaluation of CAR-T cell expansion rate (left; % CD45.1 EGFR+cells) and percentage of persistent cells (right; % CD19+cells) in the blood. Data are presented as means ± standard error of the mean, and Mann-Whitney tests were performed for statistical analyses. (E-F) Flow cytometry analyses of the tumor compartment on day 24. Representative contour plots (E, left panel) and calculation of the absolyte numbers of CAR-T cell in the TME across treatment groups (E, right panel). Phenotype of tumor-infiltrating CAR-T cells, including quantification of the CD4 / CD8 ratio, and percentages of CD8+PD1+CAR-T cells, and IFNy+CD8+CAR-T cells (F). Data are shown as means ± standard error of the mean, with Mann-Whitney tests applied for statistical analyses.

[0031] Figure 13. Prevalence of intestinal Akkermansia spp. and lymphoma TME.

[0032] (A) CAR-T cell expansion in the blood of patients receiving anti-CD19 CAR-T cells, stratified by the presence of Akkermansia spp. based on available metagenomic sequencing data (n=6 Akk. spp.+; n=17 Akk. spp.-). CAR-T cell expansion is depicted as a percentage (left) and in absolute numbers per mL (right). Data are presented as means ± standard error of the mean, and Mann-Whitney tests were used for statistical significance. (B) Schematic representation of the experimental design. Tumor biopsies, conducted shortly after CAR-T cell infusion, were analyzed using spectral flow cytometry for patients with available metagenomic sequencing data (n=3 Akk. spp.+; n=3 Akk. spp.-). The secretome of the biopsied tumors was further analyzed in the tumor supernatant using MesoScale Discovery (MSD)-derived multiplex assays. (C) Spectral flow cytometry analysis of tumor biopsies. Unsupervised analyses were performed for all patients with available metagenomic sequencing data (n=6). Cell populations were visualized using Uniform Manifold Approximation and Projection (LIMAP) (not shown), and were categorized based on the presence of Akk. spp. as a group (not shown) and individually (not shown). Quantification of the percentage of CAR- T cells, CD8+ T cells, NK cells, and myeloid cells within the TME is shown. Data are expressed as means ± standard error of the mean, with Mann-Whitney tests applied for statistical comparisons. (D) Analysis of the tumor secretome in patient samples with available metagenomic data (n=6) focusing on levels of GM-CSF, IFNy, and LAG3. Results are shown as means ± standard error of the mean, with Mann-Whitney tests performed for statistical significance.

[0033] Figure 14. Akkermansia spp. promotes CAR-T cell homing in the bone marrow (BM)

[0034] (A-B) Flow cytometry analysis of the BM on day 7 post-treatment assessing CAR-T cell infiltration across treatment groups. Representative flow cytometry gating strategy for one sample per group (A, left panel), and absolute numbers of infiltrating CAR-T cells. (A, right panel). Data are presented as means ± standard error of the mean, with statistical analysis performed using Mann-Whitney tests. Quantification of the CD4 / CD8 CAR-T cell ratio, and percentages of effector memory CAR-T cells, CD8+PD1+CAR-T cells, and CD8+IFNy+CAR-T cells (B). Results are shown as means ± standard error of the mean, with Mann- Whitney tests applied for statistical significance. (C-D) Immunofluorescence analysis of the BM on day 7 post-treatment. Quantification of CAR-T cell infiltration (EGFR+cells / mm2tissue) (C), and regulatory T cells (CD4+ / FoxP3+cells / mm2tissue) (D). Absolute cell counts (right) are provided, with data presented as means ± standard error of the mean. Mann-Whitney tests were used for statistical comparisons. (E). Schematic representation of the experimental design in patients. BM aspirations performed after CAR-T cell infusion were analyzed using spectral flow cytometry and multiplex ELISA for BM plasma from patients with available metagenomic sequencing data (n=7). (F) Comparison of the percentage of total CAR-

[0035] T cells and CD8+CAR-T cells between Akk. spp.- and Akk. spp.+ patients (n=8), using supervised flow cytometry analyses. Data are displayed as means ± standard error of the mean, with Mann-Whitney tests performed for statistical analysis. (G) Analysis of BM secretome in patients with available metagenomic data (n=8) using multiplex kits, focusing on the levels of IL-2 and TNFa. Results are presented as means ± standard error of the mean, with Mann- Whitney tests applied for statistical significance.

[0036] Figure 15. Akkermansia spp. failed to translocate to distal organs, but coincided with Lactobacilli spp. accumulation in BM

[0037] (A) Schematic representation of the experimental design in tumor bearers. Single-cell suspensions of tumor and BM of animals treated with CAR-T cells alone or with CAR-T cells + Akk. spp. were divided into duplicates to perform culturomics and qRT-PCR experiments in parallel. (B) Detection of Akk. spp. bacterial infiltration via qRT- PCR in stool, tumor, and bone marrow samples 7 days post-intravenous CAR-T cell injection using specific probe sets. Data are presented as means ± standard error of the mean; Mann-Whitney tests were used for analysis. (C) MALDI-TOF mass spectrometry -based identification of bacterial colonies outgrowing after in vitro anaerobic cultivation of stools, tumor and BM cell suspensions in mice receiving various treatment modalities. No CAR T cells is shown in grey; CAR T cells shown in blue and CAR-T cells + Akk. p2261 is shown in pink (D) Comparison of bacterial alpha diversity between Akk. spp.+ and Akk. spp.- patients at visits 1 , 2, and 3, calculated by the Shannon index (left) and species richness index (right). Wilcoxon signed-rank tests were conducted.

[0038] Figure 16. Prevalence of Akkermansia spp. and systemic release of immunogenic indole metabolites

[0039] (A) Quantification of indole-3-propionic acid (I PA) at all three visits (n = 33 for visit 1 ; n = 33 for visit 2; n = 20 for visit 3). Data are presented as means ± standard error of the mean; Mann-Whitney tests were performed. (B) Correlation between levels of indole-3-propionic acid (I PA) and the prevalence of Akk. spp. in patients within available metagenomic sequencing data (n = 26 Akk. spp.-, n = 7 Akk. spp.+). Data are shown as means ± standard error of the mean; Mann- Whitney tests were performed. (C) Targeted mass spectrometry metabolomic analysis of supernatants from anaerobic cultures of Akkermansia spp. supplemented with 0.4 g / L of tryptophan, harvested after 72 hours. Three technical replicates were conducted. Data are presented as means ± standard error of the mean; Mann-Whitney tests were performed. (D-E) Levels of indole pathway metabolites — tryptophan, indole-3-propionic acid (I PA), and indole-3-aldehyde acid (I3A) — determined via targeted mass spectrometry analysis of mouse stool 7 days post-CAR-T cell IV injection (D) or BM supernatant 24 days after CAR-T cell IV injection (n = 10 mice per treatment group) (E). Data are shown as mean ± standard error of the mean; Mann- Whitney tests were performed. A representative experiment is shown.

[0040] Figure 17. Aryl-hydrocarbon receptor activation in CAR-T cells is required for the clinical benefit of Akkermansia spp. supplementation

[0041] (A) AhR activity was assessed using HepG2-Lucia™ AhR reporter cells after 24 hours of in vitro stimulation with supernatants from anaerobic cultures of Akkermansia spp., supplemented with 0.4 g / L of tryptophan for 72 hours. Three technical replicates were conducted. Data are presented as mean ± standard error of the mean; Mann-Whitney tests were performed. (B) AhR expression was measured by RT-qPCR (normalized to PPIA expression) in bone marrow (BM) cell suspensions (left) and whole tumors (right) from animals treated with CAR-T cells alone or CAR-T cells + Akk. spp. (n = 2-5, depending on mRNA yield). Data are shown as mean ± standard error of the mean; Mann-Whitney tests were performed. (C) Schematic representation of the CAR-T experimental design comparing WT and AhR deficient CAR-T cells. (D-E) Idem as in Fig. 2B showing tumor growth kinetics (D) and cross-sectional tumor size comparisons between treatment arms at day 13 (E). Two technical replicates were performed with 5 to 10 mice per treatment group. One-way ANOVA was used for tumor growth curve analyses across all time points. Data are presented as mean ± standard error of the mean; Mann- Whitney tests were performed.

[0042] Figure 18. Akkermansia muciniphila is accompanied by a healthy microbiome in lymphoma patients

[0043] (A) Fecal microbiota taxonomic differences observed at visit 3 between responder (ORR+) and non-responder (ORR-) patients. A Linear Discriminant Analysis Effect Size (LEfSe) graph was generated using metagenomic data from visit 3 samples (3 month post-CAR-T, n = 20), highlighting the most discriminant species in responders versus non-responders based on the LDA score. (B-C) Relative abundance (B) and prevalence (C) of Akkermansia muciniphila SGB9926 in the cohort of patients receiving anti-CD19 CAR-T cell therapy over time. Data are presented as means ± standard error of the mean; Wilcoxon-Mann-Whitney tests were performed. (D) Fecal microbiota taxonomic differences observed at baseline between Akkermansia spp.- and Akkermansia spp.+ patients. A Linear Discriminant Analysis Effect Size (LEfSe) graph was created using metagenomic data from visit 1 samples (baseline, n = 33), identifying the most discriminant species in Akkermansia spp.- versus Akkermansia spp.+ patients based on the LDA score.

[0044] Figure 19. Bioactivity of the CD19 / CD28- CAR-T cells in our mouse model

[0045] (A) In vitro cytotoxicity assay of CD19 / CD28- CAR-T cells against the tumor cells (refer to A) at various E:T ratios after 24 hours of co-culture, measured by flow cytometry (percentage of CD19+cells). (B) Concentrations of IFNy and GM-CSF in the supernatant from CAR-T cell-tumor cell co-cultures, assessed by flow cytometry multiplex assays. A representative experiment is depicted. (C) Relapse-free survival differences among mice receiving various treatments. Three independent experiments were conducted with 15 mice per treatment group, using Kaplan-Meier estimator and log-rank test for survival analysis. (D) Monitoring of body weight every two days across treatment groups. Data are presented as means ± standard error of the mean (15 mice per treatment group). (E) Qualitative toxicity score based on tumor and mouse behavioral appearance (score 0-3) between treatment arms. (F) Differences in blood cell populations between treatment arms, as determined by whole blood cell count. Data are shown as means ± standard error of the mean for the absolute number of granulocytes (left) and the percentage of lymphocytes (right). Mann-Whitney tests were performed between CAR-T cells alone and CAR-T cells + Akk. spp. (G) Comparison of anti-tumor efficacy in mice receiving repetitive oral gavages with Akkermansia spp. or Eggerthella lenta together with CAR-T cells. Data are presented as mean ± standard error of the mean (15 mice per treatment group). One-way ANOVA was used for tumor growth curve analyses across all time points. (H) Tumor volume at day 27 across three treatment groups. (I) Overall survival differences when Akkermansia spp. or E. lenta are added to the CAR-T cell therapeutic strategy. Survival analysis was performed using the Kaplan- Meier estimator and log-rank test. Three independent experiments were conducted with 15 mice per treatment group, and all technical replicates are represented. (J) Flow cytometry analysis of the TME at day 7. Evaluation of CAR-T cell infiltration between treatment arms and quantification of effector CAR-T cells (CD44highCD62L|OW). For all panels unless differently specified, data are presented as mean ± standard error of the mean (15 mice per treatment group); Mann-Whitney tests were performed.

[0046] Figure 20. Comprehensive phenotypic characterization of CAR-T cells infiltrating organs in our immunocompetent mouse model.

[0047] (A) Flow cytometry analyses of the spleen and tumor-draining lymph node (tdLN) compartments at day 7 (left) and day 24 (right). Evaluation of CAR-T cell infiltration across treatment groups. A representative experiment is depicted (comprising n = 15 mice per treatment group). (B) Immunofluorescence analysis of the bone marrow compartment at day 7, focusing on the myeloid cell population: macrophages (F4 / 80+cells / mm2tissue, top) and neutrophils (Ly6G+cells / mm2tissue, bottom). Absolute numbers are shown. (C-D) Flow cytometry analysis of the bone marrow (BM) (C) and spleen (D). Phenotypic characterization of the myeloid compartment: macrophages (MHCII+Ly6C'CD64+CD11 b+F480+among CD45+cells); neutrophils (MHClh Ly6C+CD11b+Ly6G+among CD45+cells); monocytes (MHCIF Ly6C+CD11 b+Ly6G' among CD45+cells) and dendritic cells (DCs) (MHCII+CD64' CD11 b'CD11c+CD103+among CD45+ cells). For all panels unless differently specified, data are presented as mean ± standard error of the mean (15 mice per treatment group); Mann-Whitney tests were performed.

[0048] Figure 21. Modulations of tryptophan-derived indoles in mice and patients during CAR-T therapy

[0049] (A) Relative abundance of Akkermansia 9226 (left) and Akkermansia 9228 (right) bacterial strains in murine stools 10 days after intravenous CAR-T cell injection, assessed by RT-qPCR (normalized to total bacterial expression). Data are presented as means ± standard error of the mean (n = 10 per treatment group). Mann- Whitney tests were performed. (B) Bacterial beta diversities at visits 1 , 2, and 3 comparing Akkermansia spp.+ and Akkermansia spp.- patients, using PERMANOVA tests (n = 33 for visit 1 ; n = 33 for visit 2; n = 20 for visit 3). (C-E) Levels of various polyamine metabolites at all three visits (n = 33 for visit 1 ; n = 33 for visit 2; n = 20 for visit 3) (C), or comparing ORR+ vs ORR- at visit 2 (n=33) (D-E). (F) Associations between levels of indole-3-acetic acid (left) and indole-3- aldehyde acid (right) and ORR at 6 months. (G) Targeted mass spectrometry-based metabolomic analyses of mice stools or BM at 7 days post-CAR-T cell IV injection. Graphs depict raw values in a representative experiment, each dot representing one mouse. For all panels unless differently specified, data are presented as mean ± standard error of the mean (15 mice per treatment group); Mann- Whitney tests were performed.

[0050] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0051] In the experimental part which follows, the inventors demonstrated, in a preclinical model of cancer, that modulating the intestinal microbiota through administration of bacteria of the Akkermansia genus led to a significant enhancement in anti-tumoral efficacy of a CAR-T cells therapy (compared to the same CAR-T cells therapy alone). This increased efficacy led to a heightened tumor control and a notable improvement in overall survival. Although the presence of Akkermansia muciniphila had been reported as a marker of CAR-T cell responsiveness, no causality was established. Since Akkermansia abundances were also associated with significantly higher baseline peripheral T cell levels, the increased response to CAR T-cells therapy could result from an improved quality and performance of the manufactured CAR-T cell product. Prior to the present invention, the skilled person could thus not know whether MCI with Akkermansia species would increase the efficacy of the response to CAR T-cells therapy.

[0052] The inventors also evidenced a causal relationship between Akkermansia prevalence in the intestines, bone marrow niching of CD8+CAR-T and increased potency of murine CD19 / CD28- CAR-T cells. Mechanistically, as shown in the experimental part below, the anticancer effects of the combinatorial regimen of CAR T-cells and Akkermansia spp. imply microbial indoles such as indole-3-propionate (I PA). I PA is an AhR ligand which is mandatory to account for the increased tumoricidal activity of CAR- T cells and which is depleted from patients’ serum 7-15 days post-CAR-T infusion in absence of Akkermansia spp.. IPA and other microbial indoles enhancing CAR T-cells cytotoxic activity can be produced by Akkermansia spp. within the gastrointestinal tract. They can also be produced by Lactobacilli spp., which were found to be enriched in the bone marrow (BM) and tumor microenvironment (TME) of mice treated with CAR T-cells and Akkermansia spp..

[0053] According to a first aspect, the present invention thus pertains to a composition for use in a method of treating cancer in a human subject who is receiving an adoptive immune cell therapy, the method comprising administering said composition to the subject and sequentially or concurrently administering the adoptive T-cells, wherein the composition is a bacterial composition comprising bacteria of the Akkermansia genus.

[0054] In the above, the “adoptive immune cell therapy” designates any cellular immunotherapy, such as Tumor-Infiltrating Lymphocyte (TIL) Therapy, Engineered T Cell Receptor (TCR) Therapy, Chimeric Antigen Receptor (CAR) T Cell Therapy, CAR Macrophage therapy and Natural Killer (NK) Cell Therapy.

[0055] According to a specific embodiment, the adoptive immune cell therapy is a CAR T-cell therapy, in which said CAR T-cells can be autologous or allogenic T-cells endowed with a chimeric antigen receptor (CAR) targeting the cancer to be treated.

[0056] When allogenic T-cells are used, they are preferably genetically engineered to inactivate the endogenous TCR (to prevent Graft versus Host Disease). They can also be genetically engineered to inactivate HLA molecules to prevent GvH reactions.

[0057] According to particular embodiments, the CAR T-cells have a CD28 costimulatory domain or a 4-1 BB costimulatory domain. The CAR T-cells specifically target antigens expressed by the tumor to be treated. Examples of antigens that can be targeted are CD19 and BCMA.

[0058] As illustrated in the experimental part below, the bacterial composition increases the response to CAR T-cell therapy, leading to an improvement in overall survival, compared to individuals treated only with CAR T-cells.

[0059] Akkermansia are anaerobic Gram-negative bacteria from the phylum Verrucomicrobia specialized in the degradation of mucin. The Akkermansia genus contains two validated species, i.e., A. muciniphila and A. glycaniphila (Derrien et al., 2004; Ouwerkerk et al., 2016). In 2021 , two putative new Akkermansia species, “Candidatus A. intesti navi urn" and “Candidatus A. intestinigallinarum” were described (Gilroy et al., 2021). Kim et al. observed a great genomic diversity within the species A. muciniphila (Kim et al., 2022). Recently, Ndongo et al. proposed to reclassify eight Akkermansia muciniphila strains and described two additional putative new species, “Akkermansia massiliensis sp. nov.” and “Candidatus Akkermansia timonensis”, isolated from human feces (Ndongo et al., 2022). In the experimental part below, the inventors used a live biotherapeutic product (LBP) currently used in clinics for the management of NSCLC and RCC cancer patients treated with ICI-based regimen (NCT05865730), Akkermansia Eve#01. The investigated the impact of administration of this LBP on the antitumoral efficacy of CAR-T cells-based therapy in a preclinical model of DLBCL. The strain present in the LBP Akkermansia Eve#01 was deposited at the Collection de souches de I’Unite des Rickettsies (CSLIR) under the reference CSLIR P2261 and was deposited on April 18, 2024 at the CNCM under the reference I-6073. This strain was initially identified as Akkermansia muciniphila and was then reclassified in the Akkermansia SGB9228 candidate species (WO2022 / 157207A1). Recently, it was again reclassified in the Akermansia massiliensis candidate species described by Ndongo et al. in 2022.

[0060] According to a particular embodiment, the bacterial composition comprises bacteria of the of Akkermansia massiliensis and / or Akkermansia muciniphila species. For example, the bacterial composition can comprise Akkermansia p2261 deposited on April 18, 2024 at the CNCM under the reference I-6073.

[0061] According to a particular embodiment of the invention, the bacterial composition comprises a population of bacteria of the Akkermansia genus which have been cultured in vitro (e.g. a purified population of bacteria of the Akkermansia genus). Other possible components of such a bacterial composition include salts (e.g., sodium propionate or another propionate salt), buffers, preservatives and any pharmaceutically acceptable recipients.

[0062] According to another particular embodiment, the bacterial composition comprises, in addition to bacteria of the Akkermansia genus, other beneficial bacteria. In particular, in Example 10, the inventors observed an accumulation of Lactobacilli spp. in the BM and TME of mice treated with CAR T-cells and receiving oral Akkermansia spp. Given that these compartments harbored detectable amounts of indoles and that such microbial indoles are reknown to induce AhR-dependent CD8+T cell fitness and reinvigoration, these data suggest that the combinatorial regimen of CAR T-cells and oral Akkermansia may mediate its anticancer effects by promoting the translocation of indole producing bacterial species, especially Lactobacillaceae family members such as Lactobacilli spp., including Lacticaseibacillus paracasei, Lactobacillus johnsonii and Limosilactobacillus reuteri. In another study, the inventors identified several Lactobacillaceae family members associated with an increased response to immunotherapy, possibly linked to their ability to produce ligands or agonists of the arylhydrocarbon receptor (AhR). These species include Limosilactobacillus reuteri, Lactobacillus gasseri, Limosilactobacillus fermentum, Lentilactobacillus parabuchneri, Limosilactobacillus vaginalis and Limosilactobacillus oris. Co-administration of Akkermansia spp. and such indole-producing Lactobacilli spp. could thus lead to an increase of the effects observed in the experimental part, and / or ensure that these effects are obtained in more patients. According to a particular embodiment, the bacterial composition hence comprises, in addition to Akkermansia, bacteria of one or several Lactobacillaceae family members, for example bacteria selected from Lacticaseibacillus paracasei, Lactobacillus johnsonii, Limosilactobacillus reuteri, Lactobacillus gasseri, Limosilactobacillus fermentum, Lentilactobacillus parabuchneri, Limosilactobacillus vaginalis and Limosilactobacillus oris.

[0063] According to another particular embodiment, the bacterial composition comprises a synthetic consortium of beneficial bacteria, i.e., several bacterial species obtained from in vitro cultures. The beneficial bacteria combined to Akkermansia in such a consortium can include Lactobacillaceae family members as described above and / or other bacteria known to be beneficial in a context of immunotherapy. Examples of bacteria which are considered as “beneficial bacteria” in the above are bacterial species identified as “SIG2” in WO2024 / 094817: Dorea formicigenerans, Dorea longicatena, Eubacterium ventriosum, Eubacterium rectale, Anaerostipes hadrus, Blautia wexlerae, Faecalibacterium prausnitzii, Roseburia hominis, Roseburia intestinalis, Roseburia inulinivorans, Ruminococcus bicirculans, Ruminococcus lactaris, Candidatus Cibiobacter qucibialis, Clostridiales bacterium KLE1615, Faecalibacillus intestinalis, Lachnospira eligens, Agathobaculum butyriciproducens, Anaerobutyricum hallii, Blautia massiliensis, Clostridiaceae bacterium, Clostridium sp AF34 10BH, Lachnospira pectinoschiza, Anaerotignum faecicola, Clostridiaceae bacterium OM08 6BH, Clostridiaceae unclassified SGB4769, Clostridiales unclassified SGB15145, Clostridium fessum, Clostridium sp AM22 11 AC, Clostridium sp AM333, Clostridium sp AM494BH, Coprobacter fastidiosus, Coprococcus comes, Coprococcus eutactus, Eubacterium ramulus, Faecalibacterium SGB15346, Firmicutes bacterium AF16 15, Gemmiger formicilis, Lachnospira sp NSJ 43, Lachnospiraceae bacterium OM04 12BH, Lachnospiraceae bacterium WCA3 601 WT 6H, Lacrimispora amygdalina, Mediterraneibacter butyricigenes, Oscillibacter sp ER4, Phocaeicola massiliensis and Roseburia sp AF02 12.

[0064] According to another specific embodiment, the bacterial composition comprises fecal microbiota. In such a case, the Akkermansia bacteria can originate from the fecal material and / or can have been added thereto from a purified culture. Such a composition for fecal microbiota transplantation (FMT) can also be enriched with other beneficial bacterial species, as described above. Whether comprising only Akkermansia spp, or Akkermansia spp plus other bacteria as recited above, the bacterial composition can be in the form of a live biotherapeutic product (LBP), or in the form of a food supplement.

[0065] According to another particular embodiment, the bacterial composition produces and / or releases at least one AhR ligand or AhR agonist. This can be tested, for example, in an in vitro culture of the bacteria comprised in said composition in an appropriate medium, or in an animal model (such as described below or in Sinha et al., Nat. Microbiol, 2024). More particularly, the bacterial composition can produce and / or release a molecule selected from microbial indoles such as indole-3-propionate, indole- 3-aldehyde, indole-3-acetate, indole lactate, indoxylsulfate, 5-hydroxyindole, 5- hydroxyindole-3-acetic acid, indole-3-butyrate, indole-3-pyruvate, indole-3- carboxaldehyde, diet derived indoles (e.g. derivatives of indole-3-carbinol) and salts therof.

[0066] The increase in CAR T-cells cytotoxic activity may also be favored or accelerated by administering microbial indoles with AhR agonist activity, in combination with the bacterial composition. These components may be added to the bacterial composition or administered separately. According to yet another particular embodiment, the bacterial composition thus comprises at least one AhR ligand or AhR agonist selected from the group consisting of microbial indoles such as indole-3-propionate, indole-3-aldehyde, indole-3-acetate, indole lactate, indoxylsulfate, 5-hydroxyindole, 5- hydroxyindole-3-acetic acid, indole-3-butyrate, indole-3-pyruvate, indole-3- carboxaldehyde, diet-derived indoles and salts therof.

[0067] As mentioned above, microbial indoles with AhR agonist activity can be administered separately from the bacterial composition comprising bacteria of the Akkermansia genus (and possibly other comonents as recited above). According to this aspect of the invention, the cancer patient receives sequentially or concurrently the adoptive T-cells, the bacterial composition and an AhR ligand and / or AhR agonist agonist selected from the group consisting of microbial indoles such as indole-3-propionate, indole-3-aldehyde, indole-3-acetate, indole lactate, indoxylsulfate, 5-hydroxyindole, 5- hydroxyindole-3-acetic acid, indole-3-butyrate, indole-3-pyruvate, indole-3- carboxaldehyde, diet-derived indoles and salts therof.

[0068] As shown in the experimental part below, at least in Example 9, intestinal residence of Akkermansia spp. can influence CAR-T cell potency within the bone marrow. Another aspect of the present invention is thus the use of a bacterial composition comprising Akkermansia spp., as described above, to increase the cytotoxic activity of CAR T-cells administered to a cancer patient. According to another aspect, the bacterial composition according to the invention decreases or prevents CAR T cell therapy-associated gut dysbiosis, as shown at least in Example 6 below.

[0069] According to a preferred embodiment, the bacterial composition comprises live bacteria of the Akkermansia genus. According to a specific embodiment, the bacterial composition is formulated so that live bacteria can engraft in the intestinal tract. The skilled person can use any appropriate route of administration for the bacterial composition. For example, the bacterial composition can be formulated for administration by oral route or by enema. According to a particular embodiment, encapsulated lyophilized bacteria (e.g., a lyophilized encapsulated strain of Akkermansia massiliensis, such as that present in the LBP Akkermansia Eve#01) are administered per os. The formulation of the bacterial composition preferably enables delivery of live bacteria to the ileum.

[0070] As shown in Example 4 below, metabolomics comparisons between responders (R) and non-reponders (NR) showed that 2 oxovalerate, a metabolite which level is increased by Akkermansia, is higher in NR at baseline, followed by a lower level 3 months after CAR T-cells infusion. According to the invention, the composition comprising bacteria of the Akkermansia genus as above-described is advantageously not administered before CAR T-cells administration, or only a few days (< 8 days) before.

[0071] As shown in Fig. 10, there is a drop of propionate levels is patients after CAR T-cells administration. According to a specific embodiment, propionate is administered together with the bacterial composition (or as part of this composition).

[0072] According to a specific embodiment, administration of the bacterial composition starts a few days the same week, preferably the same day as CAR T-cells infusion.

[0073] After the beginning of the treatment, the bacterial composition of the invention can be administered repeatedly, for example every day, every other day, two times a week, weekly or every two weeks.

[0074] According to a particular embodiment, the bacterial composition and the CAR T-cells are administered in combination with one or several other antineoplastic treatment(s) such as other immunotherapy treatments (e.g., immune checkpoint blockers), chemotherapy, radiotherapy, sugery, etc.

[0075] In what precedes, a “cancer amenable to CAR T-cell therapy” designates any cancer that can be treated bay CAR T-cell therapy, possibly combined with other antineoplastic treatment(s). As used herein, the terms “treat”, “treatment” and “treating” refer to any delay of the progression, reduction of severity, and / or duration of cancer. Examples of cancers that can be treated according to the present invention include leukemias (such as acute lymphatic leukemia (ALL), chronic lymphatic leukemia (CLL), acute myeloid leukemia (AML), chronic myeloid leukemia (CML)), lymphomas (such as diffuse large B-cell lymphoma (DLBCL), mantle cell lymphoma, follicular lymphoma, marginal zone lymphoma, Burkitt lymphoma or Hodgkin lymphoma), multiple myelomas, plasmocytomas, brain tumors (such as glioblastoma, neuroblastoma, astrocytoma, meningioma, medulloblastoma and peripheral neuroectodermal tumors), carcinomas (such as intestine carcinomas, esophageal carcinoma, labial carcinoma, larynx carcinoma, hypopharynx carcinoma, tongue carcinoma, salivary gland carcinoma, gastric carcinoma, adenocarcinoma, medullary thyroid carcinoma, papillary thyroid carcinoma, renal carcinoma, kidney parenchymal carcinoma, ovarian carcinoma, cervix carcinoma, uterine corpus carcinoma, endometrium carcinoma, chorion carcinoma, pancreatic carcinoma, prostate carcinoma, testis carcinoma, breast carcinoma, urinary carcinoma, hepatocellular carcinoma, gall bladder carcinoma, bronchial carcinoma, small cell lung carcinoma, non-small cell lung carcinoma), sarcomas (such as osteosarcoma, chondrosarcoma, myosarcoma, liposarcoma, fibrosarcoma and Ewing sarcoma) and melanomas.

[0076] According to a particular embodiment, the bacterial composition is administered to a cancer patient identified as lacking bacteria of the Akkermansia genus in his / her intestinal microbiome.

[0077] According to another aspect, the present invention relates to a combination of CAR T-cells and bacteria of the Akkermansia genus, for use in the treatment of a cancer amenable to CAR T cell therapy. According to this aspect, the CAR T-cells and the bacteria of the Akkermansia genus are administered through different routes (typically, infusion of CAR T-cells and administration of encapsulated lyophilized bacteria per os with ileal release). As mentioned above, combination regiment according to the invention can also comprise microbial indoles with AhR agonist activity.

[0078] The features described above regarding the bacterial composition, the CAR T-cells, the cancer to be treated, the patient’s microbiome, the sequence of administration of bacteria and CAR T-cells etc., all apply to the combination according to the invention.

[0079] According to another aspect, the present invention pertains to a CAR T- cell for use in the treatment of a cancer amenable to CAR T cell therapy, wherein said CAR T-cell is used in combination with a bacterial composition as described above.

[0080] The features described above regarding the bacterial composition, the CAR T-cells, the cancer to be treated, the patient’s microbiome, the sequence of administration of bacteria and CAR T-cells etc., all apply to the use of CAR T-cells according to the invention. According to another aspect, the present invention pertains to a theranostic method for assessing whether an individual having a cancer amenable to CAR T cell therapy needs a microbiota centered intervention (MCI) to respond to a CAR T cell therapy, comprising assessing the presence, in a sample from said individual (e.g., a stool sample), of bacteria of the Akkermansia genus, wherein the absence of bacteria of the Akkermansia genus is indicative that the patient needs a MCI with a bacterial composition as recited in any one of claim 1 to 7 to better respond to the CAR T cell therapy.

[0081] Another aspect of the invention is a method of treating a cancer amenable to CAR T-cell therapy, comprising administering to a patient CAR T-cells in combination with a bacterial composition as described above, and optionally at least one other anticancer treatment.

[0082] According to a particular embodiment, the method comprises:

[0083] (i) assessing, in a stool sample from a patient having a tumor, the relative abundance of bacteria of the Akkermansia genus;

[0084] (ii) administering CAR T-cells targeting the patient’s tumor; and

[0085] (iii) when the level assessed in step (i) is lower than a predetermined threshold (e.g., as defined in WO 2022 / 157207 A1), administering bacteria of the Akkermansia genus to the patient.

[0086] The features described above regarding the bacterial composition, the CAR T-cells, the cancer to be treated, the patient’s microbiome, the sequence of administration of bacteria and CAR T-cells, the possible additional antineoplastic treatment etc., all apply to the treatment methods according to the invention.

[0087] When performing the above method, the relative abundances of species of the Akkermansia genus can be measured by quantitative PCR using the following primers:

[0088] Table 1 Other characteristics of the invention will also become apparent in the course of the description which follows of the biological assays which have been performed in the framework of the invention and which provide it with the required experimental support, without limiting its scope.

[0089] EXAMPLES

[0090] Given the recent observations suggesting the predictive role of Akkermansia in CAR-T cell therapy response, we sought to investigate the impact of repetitive oral administration of Akkermansia Eve#01 (Everlmmune), currently used as a live biotherapeutic product (LBP) in clinics for the management of NSCLC and RCC cancer patients treated with ICI-based regimen (NCT05865730), on the antitumoral efficacy of CAR-T cells-based therapy in a preclinical model (DLBCL bearing mice).

[0091] Example 1 : Akkermansia muciniphila is underrepresented in the gut microbiota from CD19 CAR-T cells treated patients

[0092] We developed a clinical protocol, “PIONEER” (sponsor code-CSET n°3404 N°ID-RCB: 2022-A00472-41), in which we prospectively and longitudinally collected fecal material, from patients receiving commercial CD19 CAR-T cells, at different time-points: at baseline before lymphodepletive chemotherapy, between 7 and 15 days after CAR-T cell infusion, and 3 months after CAR-T cell infusion (Figure 1A). Shotgun metagenomic analyses were performed on the patient’s fecal samples, aiming at correlating the composition of the gut microbiota with response to CAR-T cells therapy.

[0093] Patients characteristics are described in Table 2. As expected, most of the patients had diffuse large B cell lymphoma and received axi-cel (CD28 co-stimulatory domain). The overall response rate observed (48.7%) was concordant with the ones observed in clinical trials.

[0094]

[0095] Table 2: Patients characteristics

[0096] Metagenomic sequencing were performed for 41 patients (data are still being collected) and analysis were obtained so far for 22 patients (Figure 1B). Strikingly, we observed an absence of Akkermansia muciniphila in the fecal material from most of our B cell lymphoma cohort (90.9 %). This was specific to this CAR T cell cohort as the presence of Akkermansia muciniphila was found in the feces material from 40% of lung cancer patients receiving immunotherapy (Figure 1C). There are various elements which could act as confounding factors in this low relative abundance, such as the age of the patients, much younger in the B-cell lymphoma cohort of patients, as well as potentially the prophylaxis treatment received in all patients (e.g. Bactrim), to which Akkermansia muciniphila is found to be sensitive. These factors are under investigation.

[0097] Thus, we cannot currently correlate the abundance of Akkermansia muciniphila to response to treatment, due to the lack of positivity in the cohort of patients receiving CAR-T cells amongst the PIONEER clinical protocol. Validation and expansion of the cohort, up to the inclusion of at least 60 patients, and further metagenomic analyses, are currently undergoing. This would allow to determine the predictive value of the gut microbiome - and Akkermansia muciniphila.

[0098] Example 2: Development of an immunocompetent mouse model of CAR-T cells therapy in B-cell Ivmphoma-bearinq mice

[0099] Most current available CAR-T cells mouse models rely on immunocompromised mice infused with human CAR-T cells, which limits the study of the interaction of CAR-T cells with the tumor immune-microenvironment. Therefore, we developed a fully immunocompetent mouse model of CAR-T cells in a B-cell lymphoma bearing animal. Briefly, a c-myc driven B-cell lymphoma cell line expressing mCD19 (Figure 2A) was injected subcutaneously in the right flank of C57BL / 6 immunocompetent female mice at a concentration of 500.000 cells. These mice developed a very rapid and aggressive subcutaneous tumor (Figure 2B), followed by a dissemination of the tumoral cells to all organs examined (lymph nodes, spleen, liver, and lung) (Figure 2C).

[0100] Murine CAR-T cells targeting CD19 were generated using a retroviral construct with a truncated human EGFR polypeptide co-expressed with the CAR receptor, and CD28 as co-stimulatory domain (Figure 3A). The generated CAR-T cells demonstrated a higher cytotoxicity in vitro (Figure 3B) and significant higher in vitro production of IFNy and GMCSF, when compared to non-transduced activated T cells, in co-culture against lymphoma cells (Figure 3C-D).

[0101] The generated CAR-T cells were injected in vivo via the tail vein in B-cell lymphoma bearing mice at different concentrations. Intravenous injection of 6x105murine anti-CD19 CAR-T cells in B-cell lymphoma-bearing mice having previously received a 4 grays lymphodepletive total body irradiation, resulted in an initial total regression of the tumor in half of the mice. However, ultimately, we observed a relapse, in the mice that had showed tumor control, almost 2 months after CAR-T cell injection. Conversely, the intravenous injection of a higher dose of 6x106murine anti-CD19 CAR- T cells in the same conditions, resulted in tumor regression in all mice, showing similarly a late-relapse two months after CAR-T cell injection (Figure 4A).

[0102] We were able to demonstrate CAR-T cell circulation in the blood at day 10 after CAR-T cells infusion via flow cytometry (Figure 4B) and CAR-T infiltration in all organs explored after relapse, notably in the bone marrow, which appears to be an important reservoir of CAR-T cells in vivo (Figure 4C). Crucially, we established that the failure to achieve long-term tumor control is not attributable to the loss of the tumor antigen, as the target antigen CD19 continued to be expressed in all tumors (Figure 4D).

[0103] Therefore, the administration of 6x105murine anti-CD19 CAR-T cells in B-cell lymphoma bearing mice appeared to be a relevant mouse model as tumor control was obtained in about 50% in the mice, mimicking the results observed in the clinic.

[0104] Example 3: Oral administration of Akkermansia (Akkermansia Eve #01) in parallel to CAR-T cells injection enables a better tumor control

[0105] Given the suggested predictive role of Akkermansia muciniphila in CAR- T cell therapy, and its near absence in the PIONEER clinical cohort of patients receiving this treatment, we investigated the effects of supplementing CAR-T cell treatment with bacteria of the Akkermansia genus. Specifically, we assessed the impact of administration of Akkermansia massiliensis on anti-tumor efficacy in our pre-clinical mouse model of CAR-T cell therapy in DLBCL-bearing mice.

[0106] We administered Akkermansia Eve#01 repeatedly through oral gavage to mice bearing lymphoma, starting on the day of CAR-T cell injection. We observed a significant enhancement in anti-tumoral efficacy in mice receiving the combination of CAR-T cells (at the established dose of 6x105cells) along with Akkermansia Eve#01 compared to those receiving CAR-T cells alone (p=0.032) (Figure 5A-C). This heightened tumor control was associated with a notable improvement in overall survival ((Figure 5D, p<0.001 ; assessed by Log-rank test, Mantel Cox).

[0107] We established the tolerance of supplementing Akkermansia Eve#01 to CAR-T cell therapy, as evidenced by the stable weight of mice (Figure 6A-B) in both treatment groups throughout all recorded time points, including at the time of sacrifice. Additionally, we devised an in-house toxicity scoring system to determine an humane end-point, taking into account factors such as tumor spread to other organs and behavioral indicators (Figure 6C-D), among others. The mechanism driving the heightened anti-tumor efficacy and prolonged survival in mice receiving CAR-T cells + Akkermansia Eve#01 is currently under investigation. However, we did not observe a difference in CAR-T cell expansion in the blood between the two treatment groups (CAR- T cells alone vs. CAR-T cells + Akkermansia Eve#01) on day 7 and day 14 post CAR-T cell infusion (Figure 7A-B). Interestingly, on day 14 after CAR-T cell injection, we did observe a significantly increased CAR-T cell cytotoxic activity in mice receiving treatment with CAR-T cells + Akkermansia Eve#01 (p=0.044) (Figure 7C). Nevertheless, the phenotypic characterization of CAR-T cells in the blood at both time points (day 7 and day 14 post CAR-T cell injection) via flow cytometry did not reveal significant differences between the treatment groups.

[0108] Further analyses are necessary to elucidate the mechanism underlying the enhanced anti-tumor efficacy when Akkermansia Eve#01 is combined with CAR-T cell therapy. Ongoing experimental designs are focused on assessing differences in cell populations at various time points, comparing mice receiving CAR-T cells with those receiving CAR-T cells + Akkermansia Eve#01 , and with a no-treatment control group. Notably, distinctions in cell populations are anticipated to be explored within both the T cell and myeloid cell compartments, as well as in terms of cytokine production in the tumor and all other organs examined.

[0109] Example 4: Metabolomics in lymphoma patients treated with CAR-T cells

[0110] Metabolomics comparisons between responders (R) and non-reponders (NR) showed, inter alia, an excess of 2 oxovalerate in NR at baseline, followed by a lower level 3 months after CAR T-cells infusion (Figure 8). Since Akkermansia Eve#01 was shown to induce 2 oxovalerate in mice (Figure 9), it is preferable not to administer Akkermansia before CAR T-cell infusion, but rather to start MCI with Akkermansia upon CAR T-cell infusion or later.

[0111] Moreover, there is a transient defect of propionate metabolites (namely 3 phenyl- and indole 3, propionic acids, Figure 10) at day 15 post-CAR-T infusion in vivo in lymphoma patients. Importantly, propionate is a key short chain fatty caid produced by Akkermansia Eve#01 (Figure 9). This provides a rationale to provide Akkermansia (e.g., A. massiliensis) after CAR-T cell infusion to sustain T cell survival.

[0112] Example 5: CAR-T cell therapy affects gut commensalism diversity without clinical significance

[0113] We conducted prospective CAR-T cell clinical and ancillary studies at Gustave Roussy, enrolling 33 patients treated with commercial anti-CD19 CAR-T cells with a median follow-up of 16.13 months (9.03; 19.20). Most patients were treated for B cell lymphoma and received CD28- CAR-T cells (Table 3).

[0114] The median age of the cohort was 61 years, with a predominance of males (64%). A complete response at 6 months was observed in 18 patients (55% of the cohort), which is consistent with real-life literature observations (Bachy et al., 2022) (Table 3). A large proportion of patients (82%) received cotrimoxazole prophylaxis before CAR-T cell infusion. However, only 7 patients (21 %) required curative antibiotics (ATB) before the infusion, and among them, only one received a broad-spectrum ATB with meropenem (in accordance with European recommendations (Hayden et al., 2022)). The entire cohort received broad-spectrum antibiotics within the month following CAR-T cell infusion. This period corresponded to aplasia and concomitant cytokine release syndrome.

[0115] Table 3. Patients' characteristics according to response. PPI: proton pump inhibitor, BZD: benzodiazepines, CRS: cytokine release syndrome, ICANS: immune effector cell-associated neurotoxicity syndrome, ICAHT: immune effector cell-associated hematotoxicity

[0116] We did not identify baseline classical clinical or biological parameters distinguishing responders from non-responders. Notably, administration of antibiotics before CAR-T cell infusion was not associated with dismal outcome in this cohort. Hence, we investigated whether the gut taxonomic composition could be a prognostic parameter in this context. This routine clinical management allowed the collection of stools, blood, tumor biopsies, and bone marrow (BM) aspirations at different time points. Specifically, fecal materials were obtained shortly before lymphodepletive chemotherapy (visit 1 = baseline), after CAR-T cell infusion between day +7 and day +15 (visit 2), and at 3 months following CAR-T cell infusion (visit 3) (Figure 11 A). Shotgun metagenomic analyses of the fecal samples were performed longitudinally in 33 patients for a total of 86 samples (n = 33 for visit 1 ; n = 33 for visit 2; n = 20 for visit 3). First, the bacterial diversity was assessed via alpha diversity calculated by Shannon and species richness indexes. Longitudinal analyses of alpha diversity of the gut microbiome showed a significant decrease in bacterial diversity after CAR-T cell infusion (visit 2) (P= 0.001) (Figure 11 B). Importantly, visit 2 corresponded to broad-spectrum ATB intake but also occurred after lymphodepletive chemotherapy, suggesting that this reduction of gut bacterial diversity may be caused by these treatments. However, we did not find any difference in alpha diversity between responders (ORR+) and non-responders (ORR-) at these three time points (Figure 11C) (P= 0.506 at visit 1 ; P= 0.823 at visit 2 for Shannon index and P= 0.564 for species richness index; P= 0.824 at visit 3 for Shannon index and P= 0.882 for species richness index). We then assessed beta diversity using the Bray- Curtis metric to calculate the distance between microbial communities and the PERMANOVA test to assess their relationship with variables of interest. Similarly, we observed a significant change in beta diversity across time points (Figure 11 D), again suggesting an effect of broad-spectrum ATB and / or lymphodepletive chemotherapy on gut microbial composition. However, we did not find any difference in beta diversity between ORR+ and ORR- patients at these three time points (Figure 11 E). Consequently, our findings suggest that a gut dysbiosis develops after CAR-T cell injection but does not significantly impact response to CAR-T cell therapy.

[0117] Example 6: Akkermansia muciniphila is associated with CAR-T cell clinical benefit

[0118] We next examined whether a specific gut bacterial composition could correlate with the response to CAR-T cell therapy. We performed linear discriminant analysis of effect size (LEfSe) and observed a distinct gut bacterial composition between ORR+ and ORR- patients, with an over-representation of specific bacterial species in ORR+ patients at baseline (visit 1). Strikingly, Akkermansia muciniphila SGB9226 species (Akk. spp.) showed the highest correlation with response to CAR-T cell therapy in our cohort of B cell malignancies (Figure 11 F). The distinctive bacterial taxa associated with response to therapy differed at visit 1 and visit 3 (Figure 18A), but the most significant association was the presence of Akkermansia muciniphila at baseline (visit 1). Although the clinical characteristics of patients positive and negative for Akk. spp. at baseline were generally similar (Table 4), the clinical outcome was markedly different when patients were classified by the presence of Akkermansia muciniphila in their gut microbiota at baseline (visit 1). Indeed, Akk. spp. positive patients presented a 100% 6- month ORR, while it fell to 42% in Akk. spp. negative patients (Table 4). This difference in ORR at 6 months correlated with prolonged progression-free survival (PFS) in patients positive for Akk. spp. (P= 0.024, Figure 11G).

[0119] Importantly, the prevalence of Akkermansia muciniphila was abnormally low in this cohort at baseline (only 20% of patients compared with 50% across many metastatic cancer types and healthy volunteers (Derosa et al., 2024, 2022; Birebent et al. submitted) and the relative abundance of Akk. spp. in Akk. spp. positive patients remained stable over time (Figure 18B-C). Interestingly, LEfSe analyses performed between Akk. spp. positive and Akk. spp. negative patients unveiled a broader repertoire of beneficial bacterial taxa associated with clinical response to PD1 blockade (belonging to “SIG2” (Derosa et al., 2024) i.e Oscillospiraceae and Lachnospiraceae family members), enriched in Akk. spp. positive patients, suggesting that the presence of Akk. spp. is a proxy for a healthy microbiome (Figure 18D).

[0120] In sum, about 80% of B cell lymphoma patients harbor a gut dysbiosis characterized by the absence of Akkermansia spp. coinciding with a loss of a healthy microbiota.

[0121] Table 4. Patients' characteristics according to Akkermansia muciniphila prevalence. PPI: proton pump inhibitor, BZD: benzodiazepines, CRS: cytokine release syndrome, ICANS: immune effector cell-associated neurotoxicity syndrome, ICAHT: immune effector cell-associated hematotoxicity

[0122] Example 7: Akkermansia spp. supplementation improved the efficacy of CD19 / CD28-E CAR-T cells against B cell lymphoma

[0123] Since Akkermansia muciniphila (SGB9226) was over-represented in ORR+ patients and its prevalence was decreased in B cell lymphoma patients, we next examined whether modulating the gut microbiota via supplementation with an immunogenic strain from the Akkermansia species could enhance the efficacy of CAR- T cell therapy.

[0124] Akkermansia muciniphila is an anaerobic Gram-negative bacterium from the phylum Verrucomicrobia, specialized in mucin degradation and known to play a key role in intestinal homeostasis and host fitness (Iwaza et al., 2022; Segers and Vos, 2023; van Passel et al., 2011). The Akkermansia genus contains two validated species; A. muciniphila SGB9226 and A. glycaniphila (Derrien et al., 2004; Ouwerkerk et al., 2016). Recently, Ndongo et al. proposed reclassifying eight Akkermansia muciniphila strains and described two additional putative new species, “Akkermansia massiliensis spp. nov.” and “Candidates Akkermansia timonensis" isolated from human feces (Ndongo et al., 2022). We used the strain Akkermansia massiliensis p2261 , deposited at the Collection de souches de I’Unite des Rickettsies (CSLIR) under the reference CSLIR P2261 and at the Collection Nationale de Cultures de Microorganismes (CNCM, Institut Pasteur, Paris) under the reference CNCM I-6073. This strain was initially identified as an Akkermansia muciniphila SGB9226 but was very recently reclassified as an Akkermansia massiliensis SGB9228 (WO2022 / 157207A1). Our group showed that the Akkermansia p2261 strain synergized with CTLA4 and / or PD1 blockade across several solid tumor models by promoting IL-12-dependent Th1 and Tc1 responses (Routy et al., 2018), specifically in settings of intestinal dysbiosis lacking Akkermansia spp. (Derosa et al., 2022). To test the effect of oral administration with Akkermansia massiliensis onto systemic CAR-T cell therapy, we developed a fully immunocompetent murine model of CD45.1+CD19 / CD28- CAR-T cell therapy in B cell lymphoma-bearing CD45.2+C56BL / 6J mice. Briefly, we subcutaneously injected the A- MYC- driven mouse B lymphoma cell line expressing CD19 as target antigen at a dose of 0.5 x 106cells into immunocompetent syngeneic C57BL / 6J mice. Tumor inoculation was performed after a lymphodepletive conditioning regimen consisting in 4 Gray (4 Gy) total body irradiation (TBI), and treated by a single intravenous (IV) injection of 6 x 105CD45.1+CAR-T cells per mouse together with the daily oral administration of Akkermansia p2261 (Figure 12A). The A-MYC- driven mouse B cell lymphoma rapidly expanded locally before invading secondary lymphoid organs and lungs, while sparing bone marrow (not shown), hence mimicking the natural growth pattern of human diffuse-large B cell lymphoma (DLBCL). The generated CD19 / CD28- CAR-T cells carried a truncated human-EGFR cassette (Figure 3A), co-expressed with the CAR receptor at the cell surface, allowing ex-vivo tracking of the transduced CD45.1+T cells. These transduced CAR-T cells were able to selectively kill murine B cell lymphoma cells in vitro (Figure 19A) and to produce higher levels of IFNy and GM-CSF when compared to untransduced activated T cells, when co-cultured with the tumor cell line in a dose-dependent manner (Figure 19B). Importantly, we observed a better tumor control in mice treated with Akkermansia spp. together with CAR-T cells (Figure 12B), resulting in significantly longer overall survival (Figure 12C) and delayed time to relapse (Figure 19C), compared to mice treated with CAR-T cells alone. Additionally, the supplementation of Akkermansia spp. to CAR-T cells was well tolerated, as assessed by mouse weight, toxicity score (Figure 19D and E), and recirculating lymphocytes and granulocytes post- CAR-T cell injection (Figure 19F). Importantly, oral administration of another bacterial species such as Eggerthella lenta (E. lenta), an anaerobic Grampositive bacillus of the Coriobacteriaceae family, did not lead to any improvement in CAR-T cell efficacy over negative controls for tumor growth kinetics (Figure 19G and H) (P= 0.503 at day 27 after CAR-T cell injection between CAR-T cell and CAR-T cell + E. lenta) or overall survival (Figure 191).

[0125] We then sought to identify mechanisms underlying the superior activity of CAR-T adjuvanted with oral Akkermansia spp. Therefore, we extensively explored different organs at different time points to elucidate the CAR-T cell dynamics and functions with or without gut taxonomic modulation. The percentage of CAR-T cells circulating in the blood (CD45.1+EGFR+) did not differ between treatment arms after injection, demonstrating that the improved anti-tumor efficacy is not caused by an increase in CAR-T cell expansion (P= 0.878, Figure 12D, right panel). Despite this similar expansion rate, the percentage of CD19+cells was significantly reduced in mice supplemented with Akk. spp. after CAR-T cells, suggesting that CAR-T cells from mice supplemented with Akk. spp. developed a better cytolytic activity (P= 0.044, Figure 12D, left panel). Importantly, we observed that CAR-T cell infiltration in the tumor at day 24 following CAR-T cell injection tended to be higher in mice supplemented with Akk. spp. together with CAR-T cells (P= 0.058) (Figure 12E). Yet, such a difference in tumor infiltration between treatment arms was not observed at earlier time points (Figure 19J, left panel). We then phenotypically characterized the CAR-T cells infiltrating the lymphoma. Intra-tumoral CAR-T cells from Akk. spp. treated animals exhibited a lower CD4 / CD8 ratio (Figure 12F, left panel). However, we did not observe any significant difference in terms of PD1 expression, IFNy release (Figure 12F right panel), nor CD44 or CD62 expression (Figure 19J, right panel).

[0126] Altogether, a syngeneic CAR-T cell lymphoma preclinical model allowed to show a cause-effect relationship between supplementation with an Akkermansia strain and CAR-T cell potency.

[0127] Example 8: Prevalence of intestinal Akkermansia and lymphoma tumor microenvironment

[0128] To extend these murine observations to patients, we analyzed blood and tumor samples from B cell lymphoma patients for whom we had metagenomic sequencing data. First, the peak of CAR-T cell expansion in our cohort of Akk. spp. negative (Akk-) (n=17) versus positive (Akk+) (n=6) patients did not differ (Figure 13A, P= 0.177). Next, we obtained tumor biopsies performed shortly after CAR-T cell infusion (Figure 11 A) from 3 Akk+ and 3 Akk- patients. Spectral flow cytometry analysis and tumor secretome were performed on these six lymphoma (Figure 13B). Unsupervised and supervised analyses of the tumor microenvironment (TME) highlighted high interindividual variability and lack of significant differences between Akk+ and Akk- patients (Figure 13C). Of note, the percentage of CAR-T cells infiltrating the tumor was very low and did not differ significantly between the Akk+ and Akk- patients (Figure 13C). Similarly, we failed to observe any significant difference in the tumor secretome, according to the intestinal prevalence of Akkermansia spp. (Figure 13D). Therefore, our data suggest that Akk. spp. does not primarily affect the TME at early time points post- CAR-T infusion.

[0129] Example 9: Akkermansia spp. supplementation promotes CAR-T cell homing to the bone marrow

[0130] In addition to examining the blood and tumor compartments, we studied primary (bone marrow (BM)) and secondary (spleen, tumor-draining lymph node (tdLN)) lymphoid organs in animals treated with CAR-T cells or the combination therapy. While we did not observe any drastic differences in the percentage of CAR-T cells in the spleen and tumor-draining lymph node (Figure 20A), there was a significant increase in the percentage of BM-infiltrating CAR-T cells at an early time point in mice receiving the combination treatment (Figure 14A, P= 0.004). We then studied the phenotypic characteristics of the BM infiltrating CAR-T cells. Interestingly, BM CAR-T cells from Akk. spp. treated animals exhibited a more cytotoxic phenotype, characterized by a decreased CD4 / CD8 ratio (P= 0.035), a higher proportion of effector CD8+T cells (CD44hiCD62L10) (P= 0.007), increased IFNy production (P= 0.019), and decreased PD- 1 expression (P= 0.002) (Figure 14B). Using immunofluorescence imaging, we confirmed the accumulation of BM infiltrating CAR-T upon oral administration of the live biotherapeutics (Akk+) (Figure 14C) while the number of CAR-Tregs remained unchanged inbetween treatment arms (Figure 14D). The myeloid cell compartment in the BM (Figure 20B-C) and in the spleen (Figure 20D) did not show significant differences between CAR-T cell-treated animals and those treated with the combination therapy. Specifically, no differences were observed in the percentages of macrophages, dendritic cells, monocytes, or neutrophils. To extend these observations to patients, we obtained bone marrow aspirations from 8 of our 33 patients (Table 3) and performed spectral flow cytometry and multiplex ELISA on the BM plasma to study the BM secretome (Figure 14E). Unsupervised analysis of BM compartments showed an interpatient variability (not shown) and allowed for the definition of various cell subsets, including CAR-T cells, which did not emerge as an individual cluster but could be localized within CD4+T cells or CD8+T cell clusters (not shown). The differences in cell populations were also assessed between Akk+ and Akk- patients (not shown). Despite a limited number of BM aspirations from Akk+ patients (representing only 20% of the cohort), we observed higher CD8+CAR-T cell infiltrates in the BM of Akk+ patients (Figure 14F). Strikingly, multiplex ELISA of BM plasma revealed a significant increase of type 1 cytokine (IFNy, TN Fa and IL-2) concentrations in Akk+ patients compared with Akk- counterparts (Figure 14G).

[0131] Taken together, these results suggest that the bone marrow serves as a crucial reservoir for CAR-T cells and that intestinal residence of Akkermansia spp. can influence CAR-T cell potency within the bone marrow.

[0132] Example 10: Akkermansia spp. failed to translocate to BM but released systemic immunogenic indoles

[0133] Since we observed increased CAR-T cell activation in mice treated with the live biotherapeutics Akk. spp., culminating in the early and late infiltration of BM and tumors respectively, we investigated whether Akk. spp. exerts direct effects on T cells. Akk. spp. is administered via oral gavage after 4 Gy lymphodepletive radiotherapy, which may alter the gut barrier facilitating translocation of Akkermansia spp. towards distal organs including BM. Therefore, we carried out the culturomics of stools, tumors, and BM after CAR-T cell injection in aerobic and anaerobic conditions for over 30 days (Figure 15A). Each bacterial isolate was further identified using matrix-assisted laser desorption / ionization time-of-flight mass spectrometry (MALDI-TOF MS). In parallel, RT- PCR was also performed after DNA extraction of the samples to detect the endogenous Akkermansia muciniphila as well as the exogenous Akkermansia massiliensis (Akk. p2261) using distinct and specific primers. As expected, Akk. p2261 was overrepresented in the stools of mice supplemented by oral gavage with Akk. p2261 (P= 0.004, Figure 15B). Indeed, the bacterial strain administered orally (Akk. p2261) and the bacterial species to which it belongs, Akkermansia massiliensis (SGB9228), were found and tended to increase overtime in murine stool samples, at the expense of the local resident Akkermansia muciniphila SGB 9226 (Figure 21A). Despite its intestinal residency, the exogenous Akk. p2261 was not detectable in the tumor nor in BM, as assessed by RT-PCR or anaerobic culturomics (Figure 15B and C). Conversely, we did observe a higher bacterial diversity in the tumor and BM as well as an accumulation of Lactobacilli spp. in mice treated with oral Akk. spp. (Figure 15C). Notably, we observed tumor enrichment with Lactobacillaceae family members including species such as Lactobacillus johnsonii reportedly associated with response to immunotherapy (Bender et al., 2023; Jia et al., 2024; Luu et al., 2021) as well as BM residence of immunogenic Enterococcus faecium (Griffin et al., 2021) in Akk+ mice. Interestingly, the indoleproducing healthy Lactobacillus reuteri was found in the stool culturomics post-CAR-T infusions.

[0134] Strikingly, when categorizing our patient cohort based on gut prevalence of Akk. spp. in shotgun MGS, we failed to notice the dramatic loss of commensal richness in Akk+ patients at visit 2, in sharp contrast with Akk- patients (Figure 11 B, Figure 15D). Similar results were found when analyzing the beta diversity, as it did not change overtime in Akk+ patients, in contrast to Akk- patients (Figure 21 B). Given that intestinal residence of Akk. spp. prevents CAR-T cell therapy-associated gut dysbiosis without operating translocation at distal sites, we searched for the bacterial metabolomic pathways that could modulate CAR-T potency. To gain more insight on the perturbations generated following CAR-T cell infusions and intercepted by gut Akk. spp. in patients, we performed a dynamic targeted mass spectrometry (MS-MS) on plasma from our cohort of B cell lymphoma patients receiving anti-CD19 CAR-T cell therapy, and correlated metabolite profiles with the 6 month- ORR and / or the presence of Akk. spp. in shotgun MGS.

[0135] Examining 273 metabolites belonging to bile acid and fatty acid metabolism, as well as glycolysis, aminoacids among others, we observed a trend towards differences in the polyamine pathway (comprising hypusine, spermidine, spermine, and its acetylated derivatives) that tended to be reduced at visit 2 and / or visit 3 compared to visit 1 (Figure 21C) and in ORR- (compared with ORR+) patients (Figure 21 D), supporting a prior report showing the critical role of EIF5A hypusination for CAR- T cell stemcellness (Klysz et al., 2024). Accordingly, plasma levels of inosine tended to be higher in responders at visit 2 post CAR-T cell therapy (Figure 21 E).

[0136] Focusing on statistically significant changes in plasma concentrations of metabolites overtime following CAR-T cell therapy, we found pharmacodynamic differences in the indole metabolites belonging to the tryptophan pathway. We detected an overall decrease in indoles after CAR-T cell injection (visit 2), which was then restored later at visit 3. Importantly, tryptophan plasma levels were stable across visits, suggesting that gut commensalism might have been in play to modulate indole plasma levels (not shown). The most affected indole metabolite was the indole-3-propionic acid (I PA) at visit 2 (Figure 16A). There was no significant difference in the plasma levels of indoles between ORR+ and ORR- patients (Figure 21 F). However, the drop in IPA was more exacerbated in Akk- patients compared with Akk+ individuals post-CAR-T cells (Figure 16B) suggesting that Akk. spp. metabolism could potentially support indole production. To support this contention, we first performed computational analyses and confirmed the presence of the enzyme responsible for indole production (enzyme 4.1.99.1) in the individual sequence of Akkermansia spp. SGB9226 found in our cohort, via KEGG Orthology and Links Annotation (KOALA) alignments to reference genomes. Secondly, we performed mass spectrometry (MS-MS)-based metabolomics of the Akkermansia p2261 supernatants during its expansion in vitro. Indeed, we detected a clear production of IPA in the supernatant of Akk cocultured with L-tryptophan (Figure 16C). Thirdly, we also conducted mass spectrometry (MS-MS)-based metabolomics of plasma, feces and bone marrow of animals receiving CAR-T plus or minus the live biotherapeutics Akk. p2261. In stools harvested at day 7 from CAR-T+ Akk+ treated mice, we observed a rise in tryptophan and the parallel drop in IPA and to some extent indole-3-aldehyde compared with controls (Figure 16D). In BM harvested at day 7 and day 24 from CAR-T+ Akk+ treated mice, we confirmed the late (but not early) drop in tryptophan as well as in the toxic indoxylsufate metabolite compared with controls (Figure 16E, Figure 21G). In plasma harvested from CAR-T+ Akk+ treated mice, we corroborated human findings with a general trend towards higher concentrations of indole metabolites at early timepoints (mostly indole 3 butyrate) and the late drop in the toxic indoxylsulfate (no shown), contrasting with the rise in inosine as previously reported (Klysz et al., 2024). Other metabolites, such as bile acids and the short-chain fatty acids were analysed without conclusive differences at the three visits (not shown).

[0137] Taken together, our data suggests that Akkermansia spp. do not translocate to organs, and metabolize tryptophan into indoles, such as indole-3-propionic acid, within the gastrointestinal tract, ultimately leading to their release and possible recirculation in BM.

[0138] Example 11 : Aryl -hydrocarbon receptor engagement in CAR-T cells reguired for antitumor efficacy of Akk. spp.

[0139] Indoles are a major source of endogenous aryl-hydrocarbon receptor (AhR) ligands (Hubbard et al., 2015; Lamas et al., 2018; Sari et al., 2020). Since Akk. p2261 produces indole-3 propionate (I PA) in vitro and is associated with tryptophan consumption and indole recirculation in plasma and BM, we tested whether such potential ligands act as AhR agonists on CAR-T cells in tumor bearers. To decipher whether Akk. p2261 culture supernatants in vitro could activate the AhR pathway, we took advantage of a human intestinal epithelial cell line (AhR reporter human liver cell line (HepG2luc) stably expressing an AhR-dependent reporter system containing three copies of the DNA-binding domain motif recognized by AhR (xenobiotic-responsive element; XRE) driving the transcription of the luciferase reporter gene (Modoux et al., 2022)). Albeit not strictly significant, the addition of tryptophan to bacterial cultures tended to increase luciferase activity in this assay (Figure 17A). Next, we investigated AhR-specific transcriptional activity in lymphoma and BM. RT-PCR performed in bone marrows and tumors from CAR-T cell + Akk-treated animals showed increased expression of the Ahr gene (P= 0.136 and P= 0.037, respectively, Figure 17B). Finally, to determine the implication of AhR in the beneficial effect of Akk supplementation, we generated AhR loss of function CAR-T cells by transfecting splenocytes from AhR' / _animals (Figure 17C). Strikingly, disease control mediated by Akk. spp. supplementation was abrogated when CAR-T cells did not express AhR, demonstrating that the efficacy of Akk. supplementation relied upon AhR engagement in CAR-T cells (Figure 17D and E).

[0140] Discussion

[0141] Research aimed at improving CAR-T cell potency against the CD19 target has been prolific over the last decade, but focused almost exclusively on improving cell- autonomous features. Here we present original findings demonstrating causality links between gut commensalism and CAR-T cell efficacy in preclinical and clinical settings of advanced lymphoma.

[0142] Why should we care about gut microbiome for CAR-T cell therapy of hematological malignancies? First, previous reports heralded that newly diagnosed patients with diffuse large B-cell lymphoma exhibit altered microbiota defined by low alpha diversities, over-representation of Enterobacteriaceae and virulence genes culminating in high levels of systemic IL-6 and febrile neutropenia and resistance to immunochemotherapy (Bigenwald and Zitvogel, 2023; Yoon et al., 2023). Secondly, here we show that refractory lymphoma patients exhibit a markedly low prevalence of Akkermansia muciniphila, a health-related bacterium associated with intestinal barrier fitness (Cani et al., 2022), compared with gender, age and geography-matched healthy individuals and various metastatic cancer patient populations previously described (Derosa et al., 2024, 2022). When Akkermansia muciniphila is absent from fecal materials, fecal microbiota transplantation (FMT) compromises response to anti-PD1 or -CTLA4 Abs in murine recipients implanted with transplantable solid tumors. These effects were attributed to i) a different composition of the intestinal commensalism in Akk+ versus Akk- individuals with a relative dominance of tolerogenic Enterocloster and Clostridium spp. (Derosa et al., 2022; Yonekura et al., 2022), ii) the intrinsic immunogenicity of distinct Akkermansia strains (including A. massiliensis) capable of compensating Akk- FMT by rescuing sensitivity to immunotherapy in an IL-12-dependent manner (Derosa et al., 2020; Routy et al., 2018) and iii) by restoring the expression of MAdCAM-1 on ileal endothelial cells, preventing the exodus of enterotropic immunosuppressive Tr17 cells towards tumor draining lymph node (Fidelle et al., 2023). Our findings revealed that CAR-T cell therapy exacerbated this gut dysbiosis, now characterized by a drastic loss of bacterial richness and accumulation of immunosuppressive pathobionts and oral taxa (such as Enterocloster asparagiformis, Veillonella rogosae, Selenomonadales), associated with comedications (ATB and proton pump inhibitors (Derosa et al., 2020), CAR-T cell resistance (Smith et al., 2022) and lack of response to immunotherapy (Derosa et al., 2024), more specifically in Akkermansia negative and / or ORR- individuals. This might be explained by the conditioning regimen consisting in lymphodepletive chemotherapy associated with quasi-systematic administration of broad-spectrum antibiotics. Importantly, we brought up evidence that patients harboring intestinal residence of Akkermansia muciniphila did not experience a drop in alpha diversity despite CAR-T cell therapy, instead exhibiting a gut microbiota enriched in Lachnospiraceae and Oscillospiracease family members associated with overall survival in advanced lung and kidney cancer patients (Derosa et al., 2024, 2022). Moreover, our study highlighted that bacterial diversity in the intestines coincides with bacterial diversity in distal tissues such as the bone marrow and tumor beds. In fact, this notion was already reported during PD1 and CTLA4 co-blockade in various tumor types (Choi et al., 2023; Mager et al., 2020, p. 202) as well as CpG- or STING agonist-based therapies (Mager et al., 2020; Shi et al., 2020), CAR-T cell therapy did not induce the distal engraftment of Akk. spp. but rather facilitate the local and distal seeding of indoleproducing bacterial species (Lactobacillaceae family members, including L. reuteri (Cervantes- Barragan et al., 2017). In our cohort of refractory lymphoma, we confirmed that gut prevalence of Akkermansia muciniphila is associated with longer progression- free survival to CAR-T cells. Although our study was limited to 33 patients, these findings support a previous report showing associations between Akkermansia and CAR-T cell efficacy in two distinct German and US cohorts (Stein-Thoeringer et al., 2023). Finally, our preclinical mouse model offers a direct cause-effect relationship between passive transfer of Akk. spp. by oral gavage and improved therapeutic outcome after CAR-T cell infusion.

[0143] The mechanisms underlying the impact of the gut microbiota onto CAR-T cell clinical efficacy remain poorly understood. This is partly due to the currently used pre-clinical models of CAR-T cell therapy, which involve immunocompromised mice infused with human CAR-T cells (Olson et al., 2018). To address this, we developed a fully immunocompetent mouse model of anti-CD19 / CD28- CAR-T cells in B cell lymphoma-bearing mice to study the impact of bacterial supplementation on the syngeneic immune system and CAR-T cell functional activity. Hence, we could demonstrate a clear benefit of adding an oral adjuvant to CD19 / CD28- CAR-T cells consisting in a live and immunogenic strain of Akkermansia i.e Akkermansia massiliensis that increased tumor control and overall survival, with an acceptable safety profile. Notably, this oral adjuvant enhanced CAR-T cell cytotoxic activity, Th1 cytokine and IPA release and their capacity to niche in the bone marrow, to give rise to CD8+ T cells (at the expense of CD4+ Th1) and to infiltrate lymphoma at early and late time points respectively. These findings suggest that the bone marrow serves as an important reservoir for CAR-T cells and could be crucial for inducing a Tc1 phenotype. Indeed, the significance of the bone marrow as a niche for CAR-T cells has been poorly studied. CAR-T cells have been primarily investigated in leukemia and multiple myeloma, where cancer cells reside in the bone marrow. In our mouse model where lymphoma cell dissemination in the bone marrow is not observed, unlike in lymph nodes and spleen, the bone marrow may have played an essential role in the therapeutic response. This contention is supported by our findings in lymphoma patients treated with anti-CD19 CAR-T cells. While longitudinal follow up of blood or tumor compartments failed to show any difference in CAR-T cell potency between Akk+ and Akk- patients, bone marrow aspirations revealed a clear association between intestinal residence of Akkermansia spp. and CD8+CAR-T cell accumulation, and Th1 cytokine release paralleling IPA release in the plasma. Based on this observation, one might speculate that CAR-T cells residing in the bone marrow may constantly replenish peripheral CAR-T cells for better tumor control. Indeed, pioneering reports unveiled that bone marrow can represent a nest for migratory memory T cells (Di Rosa and Gebhardt, 2016; Di Rosa and Pabst, 2005), suggesting a potential relevance of this compartment in CAR-T cell maintenance or potency.

[0144] As a correlate to the essential role of bone marrow to ensure CAR-T cell switch to CD8+Tc1 cells, we propose a direct role of distinct gut metabolites in inducing CAR-T cell potency in both, preclinical and clinical settings. Our metabolomic analyses revealed a consumption of the polyamine pathway in patients after CAR-T cell therapy, specifically in ORR- individuals. The relevance of this pathway in maintaining CAR-T sternness was recently demonstrated by Mackall’s group using inosine to diminish glycolysis, and increase mitochondrial capacity, and polyamine synthesis in CAR-T cells (Klysz et al., 2024). As a second finding, we observed a marked decrease in plasma indole metabolites right after CAR-T cell infusion (more specifically indole-3-propionic acid (IPA)) in our lymphoma patients, with a trend towards higher plasma concentrations of IPA in ORR+ or Akk+ patients. As documented in the literature, tryptophan undergoes endogenous metabolism via distinct pathways, including the kynurenine and serotonin pathways. However, the indole metabolic pathway is the only one exclusively dependent on microbial metabolism, meaning indoles are released only when the bacteria responsible for their production are present in the gut (Jiang et al., 2022). Previous research has revealed that indoles function as ligands and agonists for the aryl hydrocarbon receptor (AhR) (Agus et al., 2018; Hezaveh et al., 2022), an effect that can be further accentuated with short chain fatty acids that play the role of HDAC inhibitors opening the chromatin for accessibility to AhR gene targets (Modoux et al., 2022). This receptor has been extensively studied in the context of gut mucosal immunity (Lamas et al., 2018; Natividad et al., 2018) although the impact of tryptophan-enriched diet and indoles in cancer prognosis and T cell fate is somewhat controversial (Gutierrez-Vazquez and Quintana, 2018; Rueda et al., 2024). In the context of immunotherapy, indole-3 aldehyde has been shown to improve immune checkpoint inhibitor efficacy in melanoma (Bender et al., 2023) by driving AhR-dependent CREB activity, leading to a Tc1 phenotype. The microbiota from patients with pancreatic ductal adenocarcinoma (PDAC) produces indole-3-acetate to sensitize PDAC cells to chemotherapy, through an increase in oxidative reactions secondary to the modulation of myeloid peroxidase and an inhibition of autophagy in cancer cells (Tintelnot et al., 2023). lndole-3-carboxyaldehyde and kynurenic acid, produced by intestinal Lachnospiraceae and Enterococcus spp. convey radioprotection in lethal models of TBI (Guo et al., 2020). lndole-3-propionate resulting from the metabolic cooperation between Lactobacillus johnsonii and Clostridium sporogenes allow better response to PD-1 blockade by increasing the differentiation of progenitor exhausted CD8+T cells (Jia et al., 2024). Here we show using MS-MS-based metabolomics performed longitudinally in three compartments (feces, BM, tumor) in lymphoma-bearing mice that oral gavage with Akk. spp. is associated with fecal rise and BM drop in tryptophan, coinciding with accumulation of various indole derivatives in the BM compared with controls. Lastly, adoptive transfer of AhR loss of function CAR-T no longer protect animals against lymphoma dissemination.

[0145] Microbiota-centered interventions may involve various treatment strategies such as fecal microbiota transplantation (FMT), specific diets, or encapsulated live (LBP) or pasteurized biotherapeutics (Derosa et al., 2021 ; Plovier et al., 2017). While FMT and guided nutrition may seem impractical for patients undergoing CAR-T cell therapy due to aplasia, hospitalization requirements, and potential side effects, the safe administration of LBP delivering polyamines and indoles could represent a promising avenue to boost CAR-T cell therapy, specifically in patients harboring a gut dysbiosis at diagnosis prior to cell therapy. In conclusion, our findings suggest that bacterial supplementation with Akkermansia massiliensis or other immunogenic Akkermansia spp. deserves to be explored in a clinical setting of intestinal defect of Akkermansia muciniphila to improve response to CAR-T cell therapy.

[0146] Materials and methods for examples 5 to 11

[0147] • Clinical protocol study design and patient characteristics

[0148] Patients treated with commercial anti-CD19 CAR-T cell therapy (axi-cel, brexu-cel or tisa-cel) were enrolled in two studies: 1) PIONEER (NCT: 2022-A00472-41), an interventional study aiming at collecting blood, tumor biopsies and BM aspirations from day -7 before, up to 3 months after CAR-T cell infusion as well as 2) ONCOBIOTICS (NCT: 2017-A02010-53), a multicentric prospective observational study designed to evaluate the impact of the microbiome composition in the clinical outcome of cancer patients. Together, both studies allowed the collection of fecal material, blood, tumor and BM biopsies at different time-points of treatment. In total, 38 patients harboring a B cell lymphoma or B acute lymphoblastic leukemia were included in the study. However, only 33 could benefit from the omics analyses (metagenomics, metabolomics). Indeed, 5 patients were excluded because fecal materials from visit 1 were only obtained after (instead of before) lymphodepletive chemotherapy for CAR-T cell infusion. All patients’ characteristics are further described in Table 1 , including prophylaxis treatment and intake of antibiotics, as well as toxicities associated with treatment. The two studies were conducted in accordance with Good Clinical Practice guidelines and the provisions of the Declaration of Helsinki. All patients provided written informed consent and general data protection procedures and pseudo-anonymisation rules were applied according to Oncobiome H2020 model in the ClinicoBiome, Gustave Roussy. All data and sample collection were performed in compliance with regulatory, ethical and European General Data Protection Regulation (GDPR).

[0149] • Sample collection

[0150] Samples were collected from patients treated with CAR-T cell therapy at Gustave Roussy from 30 / 03 / 2021 to 04 / 05 / 2023, with a median follow-up of 16.13 months (9.03; 19.20). Fecal material and blood samples were collected at three different time-points: at baseline - shortly before lymphodepletive chemotherapy (visit 1), after CAR-T cell infusion - between 7 and 15 days after treatment (visit 2) and 3 months after CAR-T cell infusion (visit 3). Tumor biopsies were performed between day 7 and day 15 following CAR-T cell infusion (according to PIONEER protocol CT: 2022-A00472-41), while BM biopsies were collected for several patients at baseline and after CAR-T cell therapy (around 1 month after CAR-T cell infusion).

[0151] • Outcome measurements

[0152] The responder status of the cohort was defined by a complete persistent response at 6 months assessed by PET-CT, while the non-responder status corresponded to partial remission, stable disease or progressive disease. PET CT was also considered as the best strategy to assess disease staging for two acute B cell lymphoma patients, as they both presented extra-medullary disease. Responders and non-responders were defined as “overall response rate positive (ORR+)”, and “overall response rate negative” (ORR-) respectively henceforth.

[0153] • Shotgun metagenomic sequencing (MGS) on human fecal samples

[0154] Stools collected at baseline were processed for total DNA extraction and sequenced with Ion Proton technology following MetaGenoPolis (INRAE) France, as previously reported. Metagenomic analysis of fastq files was performed following previously published guidelines (Routy et al., 2018) for taxonomic (MetaPhlAn 4.0) and functional (HLIMAnN 3.0) (Jacomy et al., 2014) profiling of metagenomes. These two pipelines leverage a set of 99,200 high-quality and fully annotated reference microbial genomes spanning 16,800 species and the 87.3 million functional annotations available in UniProt. The taxonomic profiling and quantification of organisms’ relative abundances of all metagenomic samples were quantified using MetaPhlAn 4.0 with default parameters. The pipelines followed for the metagenomic sequencing included Akkermansia species, specifically the bacterial strain Akkermansia muciniphila SGB9226, hereby referred as Akkermansia spp.

[0155] • MGS bioinformatic analysis on human samples

[0156] Shotgun metagenomic analyses on the fecal samples were performed longitudinally for 33 patients at baseline. Taxonomic profiles were generated through MetaPhlAn v4 and used for downstream analysis. Alpha diversity was assessed in terms of Shannon index and Observed Richness. Statistically significant differences among timepoints were evaluated through the Wilcoxon signed-rank test, in which paired samples were considered for each comparison. The same test was also applied to compare Akk. spp. negative and Akk. spp. positive cohorts across timepoints. Differences between ORR- and ORR+ were instead computed using the Wilcoxon- Mann-Whitney test. Differentially abundant species between ORR- and ORR+ were identified through LEfSe (Linear discriminant analysis Effect Size) at baseline and later timepoint. The same algorithm was also considered to discriminate between Akk. spp. neg and Akk. spp. pos at baseline. Survival analysis at baseline was performed between Akk. spp. neg and Akk. spp. pos by considering the Kaplan-Meier estimator and the logrank test. Beta diversity was evaluated in terms of multidimensional scaling (MDS) using the R function cmdscale and by considering the Bray-curtis distance. Differences among timepoints and ORR cohorts were assessed through the PERMANOVA test (function adonis2 available in the R package vegan). The same methods were considered to discriminate between different timepoints in Akk. spp. pos and Akk. spp. neg cohorts.

[0157] • Metabolomic analyses on human plasma samples

[0158] Analyses of the metabolic pathways in patients receiving anti-CD19 CAR- T cells were performed on the plasma collected at all time-points (visit 1 , at baseline; visit 2, from 7 to 15 days after CAR-T cell injection and visit 3, 3 months after CAR-T cell infusion). Longitudinal analyses were performed, and the plasma metabolites levels were correlated with clinical variables (e.g. ORR at 6 months, presence of bacteria as determined by shotgun-metagenomic analyses). Samples were prepared and 273 metabolites were widely targeted for detection by liquid chromatography-mass spectrometry (LC-MS). 50 pl of collected human heparinized serum were mixed with 500 pl of ice-cold extraction mixture (methanol / water, 9 / 1 , -20 °C, with labeled internal standard). To facilitate endogenous metabolites extraction, samples were then completely homogenized (vortexed 5 min at 2500 rpm) and then centrifuged (10 min at 15,000 g, 4 °C). Biliary acids were analyzed by MRM analysis in negative polarity, gas temperature was set to 310°C with a gas flow of 9 L / min. Capillary voltage was set to 4.5 kV. Ten (10) pl of sample were injected on a Column Poroshell 120 EC-C8 1200bars (P / N 981758-902, 100 mm x 2.1 mm particle size 1.9 pm) from Agilent technologies, protected by a guard column XDB-C18 (5 mm x 2.1 mm particle size 1.8pm) and heated at 40°C by a Pelletier oven. Gradient mobile phase consisted of water with 0.2% of formic acid (A) and acetonitrile / isopropanol (1 / 1 ; v / v) (B) freshly made. Flow rate was set to 0.5 mL / min, and gradient as follow: initial condition was 70% phase A and 30% phase B, changing to 38% phase B over 2 minutes. Phases proportion was still over 2 minutes, then molecules were eluted using a gradient from 38% to 60% phase B over half a minute. Column was washed using 98% mobile phase B for 2 minutes and equilibrated using 30% mobile phase B for 1.5 min. Autosampler was kept at 4°C (Alves Costa Silva et al., 2024). Indoles and free fatty acids were analyzed by pseudo-targeted analysis with an UHPLC-HRAM (Ultra-High Performance Liquid Chromatography — High Resolution Accurate Mass), performed on a U3000 (Dionex) / Orbitrap q-Exactive (Thermo) coupling, previously described. All targeted treated data were merged and cleaned with a dedicated R (version 4.0) package (@Github / Kroemerlab / GRMeta).

[0159] • Organ dissociation in patient samples

[0160] As part of the PIONEER protocol, we were able to obtain tumor biopsies from 6 patients and BM aspirations from 8 patients, performed after CAR-T cell infusion. Tumor biopsies obtained with 14- to 16- gauge needles were immediately placed in an eppendorf tube filled with 1 mL RPMI. After weighing the tube containing the biopsies, approximately 500 uL of the supernatant was removed for storage at -80°C for subsequent secretome analysis.

[0161] After gentle mechanical dissociation, tumor samples were dissociated in RPMI with the human tumor dissociation kit (Cat. No. 130-095-929, Miltenyi) following the manufacturer’s instructions and an incubation at 37°C during 20 min with rotor movement. Bone marrow samples were received (1-2 mL) resuspended in EDTA. Samples were centrifuged and the supernatants were then frozen at -80°C for further analysis. The bone marrow mononuclear cells were then separated via Pancoll (Cat. No P04-60100, Dutscher), and further frozen at -80°C.

[0162] Flow cytometry analyses in patient samples We analysed patient’s tumor and BM biopsies after CAR-T cell infusion using spectral flow cytometry (Cytek Biosciences). Briefly, after obtaining single cell suspensions, tumors were stained for 30 minutes at 4°C with a mix of antibodies containing: anti-CAR-T CD19 (Cat. No. 130-127-342, clone REA1297, Miltenyi), anti- CCR6 (Cat. No. 566130, clone 11A9, BD), anti-CCR7 (Cat No. 566437, clone 3D12 RUO, BD), anti-CD11 b (Cat. No. 301325, clone ICRF44), anti-CD11c (Cat. No. 130-128- 247, clone REA618, Miltenyi), anti-CD123 (Cat. No. 130-115-272, clone REA918, Miltenyi), anti-CD127 (Cat. No. 351363, clone A019D5, Biolegend), anti-CD141 (Cat. No. 344113, clone M80, Biolegend), anti-CD16 (Cat. No. 130-113-395, clone REA423, Miltenyi), anti-CD163 (Cat. No. 333631 , clone GHI / 61 , Biolegend), anti-CD19 (Cat. No. 612916, clone SJ25C1 , BD), anti-CD1c (Cat. No. 331537, clone L161 , Biolegend), anti- CD20 (Cat. No. 612906, clone 2H7, BD), anti-CD206 (Cat. No. 130-126-623, clone DCN228, Miltenyi), anti-CD25 (Cat. No. 356145, clone M-A251 , Biolegend), anti-CD3 (Cat. No. 130-113-142, clone REA613, Miltenyi), anti-CD38 (Cat. No. 303549, clone HIT2, Biolegend), anti-CD39 (Cat. No. 749967, clone TU66, BD), anti-CD4 (Cat. No. 344669, clone SK3, Biolegend), anti-CD45 (Cat. No. 130-113-120, clone 5B1 , Miltenyi), CD45RA (Cat. No. 740298, clone H1100, BD), anti-CD56 (Cat. No. 362513, clone 5.1 H11 , Biolegend), anti-CD66b (Cat. No. 305111 , clone G10F5, Biolegend), CD79b (Cat. No. 612814, clone CB3-1 , BD), anti-CD8 (Cat. No. 344759, clone SK1 , Biolegend), anti-CXCR3 (Cat. No. 353755, clone G025H7, Biolegend), anti-CXCR5 (Cat. No. 356941 , clone J252D4, Biolegend), anti-FOLR2 (Cat. No. 391705, clone 94b / FOLR2, Biolegend), anti-HLA-DR (Cat. No. 307683, clone L243, Biolegend), anti-PD1 (Cat. No. 329927, clone EH12.2H7, Biolegend), anti-TCF1 (Cat. No. 905115, Cell Signaling), anti- Tim-3 (Cat. No. 345027, clone F38-2E2, Biolegend), anti-TREM2 (Cat. No. FAB17291G, clone 237920, RD), Viability (Cat. No. 423105, Biolegend). After obtaining single cell suspensions, bone marrow mononuclear cells were stained with a mix of antibodies for 30 minutes at 4°C, comprising: CD223 Antibody, anti-human, REAfinity™ (Cat. No. 130- 118-549, Miltenyi), BUV395 Mouse Anti-Human CD41a (Cat. No. 740295, BD), BUV496 Mouse Anti-Human CD90 (Cat. No. 741160, BD), BLIV661 Mouse Anti-Human CD39 (Cat. No. 749967, BD), BUV737 Mouse Anti-Human CD7 (Cat. No. 753774, BD), Brilliant Violet 421 ™ anti-human CD34 Antibody (Cat. No. 343610, Biolegend), Brilliant Violet 570™ anti-human CD3 Antibody (Cat. No. 300436, Biolegend), BV605 Mouse AntiHuman CD56 (Cat. No. 562779, BD), Brilliant Violet 650™ anti-human CD16 Antibody (302042, Biolegend), BV786 Mouse Anti-Human CD33 (Cat. No. 740974, BD), PerCP anti-human CD45 Antibody (Cat. No. 368506, Biolegend), PE / Fire™ 640 anti-human CD19 Antibody (Biolegend), CD19 CAR FMC63 Idiotype Antibody, REAfinity™ (Cat. No. 130-127-342, Miltenyi), BUV615 Rat Anti-Human CD115 (CSF-1 R) (Cat. No. 751279, BD), BB515 Armenian Hamster Anti-ICOS (CD278) (Cat No. 565880, BD), BUV805 Mouse Anti-Human CD38 (Cat. No.742074, BD), BV510 Mouse Anti-Human CD279 (PD-1) (Cat. No. 563076, BD).

[0163] • Immunoassays for cytokine quantification in patient samples

[0164] The supernatant recovered after tumor and BM processing was isolated from biopsies performed after CAR-T cell infusion and snap frozen at -80°C until analysis. Multiplex assay on human BM plasma was performed using the Human Immune Monitoring Panel, 65plex (Cat. No. EPX650- 10065-901 , ThermoFisher). The protocol was performed following manufacturer’s instructions, standards were reconstituted with Universal Assay Buffer diluted at 1 :2 in assay diluent and for analyte capture, the plate was incubated overnight at 4°C under agitation on an orbital shaker. Analyses were performed using a Bio-Plex 200 system equipped with Bio-Plex Manager software (Bio-Rad). Multiplex assays were also run on the tumor supernatant to study the tumor secretome via MSD-derived U-plex kits: GM-CSF I IFNy Uplex human Biomarker Group 1 (Cat. No. K15076M-1); PD-1 Uplex human Immuno-oncology Group 1 (Cat. No. K151AEM-1), LAG3 Uplex human Immuno-oncology Group 2 Cat. No. K151AEM-1).

[0165] • Production of murine anti-CD19 CD28- CAR-T cells

[0166] Murine CAR-T cells targeting CD19 were generated using retroviral transduction of T cells with an anti-CD19 plasmid construct (S087 pMP71-mCD19-29-p- tEGFR-PRE; 8066 base pairs, kindly provided by Dr Simon, S., Fred Hutchinson Cancer Center, Seattle, USA). The plasmid construct includes a cassette with a truncated human EGFR polypeptide co-expressed with the anti-mCD19 CAR receptor, and CD28 as a costimulatory domain, as well as a cassette for the resistance to ampicillin and the Gly4 linker. Constructs expressing the anti-mCD19 CAR receptor in retroviral vectors were transfected into the Platinum E cells (Plat-E) (Cell Biolabs, Cat No. RV101), in DMEM culture medium (ThermoFisher Scientific, Cat No. 10569010) and no antibiotics, with 15 g of DNA after a pre-incubation with Opti-MEM reduced serum medium (ThermoFisher Scientific, Cat No. 11058021) and TransIT-VirusGEN Transfection Reagent (Mirus, Cat No. MIR 6703). The supernatant was harvested after 36h. All supernatants harvested were pooled together, filtered at 0.45 pm and concentrated 100 times in PBS with the Retro-X-Concentrator (Takara Bio, Cat. No. 631455). The resulting virus was immediately stored at - 80 °C in single-use aliquots (250 pL). T cells were isolated from harvested secondary lymphoid organs (lymph nodes and spleen) from CD45.1 Ly5.1 mice using the Pan-T cell negative selection kit (Miltenyi, Cat. No 130-095-130). After isolation, T cells were activated for 24 hours with 1 :200 anti-CD3 cl.17A (Biolegend, Cat. No. 100201), anti-CD28 cl.37.51 (Biolegend, Cat. No 102101), and cytokines IL-15 at 5 ng / mL and IL-7 at 10 ng / mL. T cells were transduced with 3X concentrated virus and the polycationic solutions protamine sulfate and DEAE-dextran in a 2h spinfection at 32 °C. T cells were then incubated in a medium with RPMI and renovated cytokines for 24h. A second 2h spinfection was performed the following day and cells were then incubated for 48h before transduction efficiency assessment via flow cytometry and further in vivo injection into syngeneic mice of T cells that were at least 80% transduced (CD45.1 EGFR+).

[0167] • Mice

[0168] Six to eight-week-old C57BL / 6J female mice were obtained from Envigo (France) and kept in our animal facility under specific pathogen-free conditions and due to the received pre-conditioning irradiation regimen were always manipulated in sterile hoods. Similarly, six-to eight-week-old C57BL / 6J Ly5.1 mice from Charles River Laboratories, expressing the isoform CD45.1 encoded by the allelic variant Ptprca on T cells, were used for harvesting T cells which were further transduced for the CAR-T cell production. AhR mice (kindly provided by H. Sokol, Saint Antoine Research Center) allowed to harvest splenocytes for transduction with the construct. Animals were weighed weekly, and food consumption was monitored weekly in each cage. Animals were euthanized by CO2, and appropriate tissues were harvested. All experiments were performed in accordance with the Comite d’Ethique en Experimentation Animale (Ministere de I’Enseignement Superieur et de la Recherche, APAFIS #38264- 2022082408404336 v2).

[0169] • Cell culture and tumor cell lines for in vitro / in vivo assays

[0170] The A-MYC B cell lymphoma tumor cell line (# 123617 subclone # 7) was established from a primary A-MYC lymphoma on the C57BL6 / J genetic background where mice also carried also an Ezh2 conditional knock-out allele (Ezh2flox) (Casola, Petrocelli et al. unpublished results). The A-MYC lymphoma cell line was used for in vitro and in vivo assays. Tumor cells were cultured in complete DM EM medium (DM EM + Glutamax high glucose DMEM + Sodium Pyruvate 1 mM + 4 mM Glutamine) medium supplemented with 10% heat / fetal bovine serum (FBS), 1 % penicillin / streptomycin (P / S), 1% MEM Non-Essential Amino Acids and 0.1% p-mercaptoethanol (50 pM). Cells were maintained in vitro at low concentrations of 25.000 cells / mL for the optimization of cell viability and expanded prior to in vivo injection at early passages, and maintained at 37°C and 5% CO2. Cell lines were routinely tested for the absence of mycoplasma contamination (Venor-GeM Advance mycoplasma detection kit, Minerva Biolabs). The expression of the target antigen CD19 on the cell surface was demonstrated via flow cytometry and immunofluorescence. For in vivo assays, tumor cells were injected subcutaneously at a concentration of 500.000 cells per mouse. To assess the growth kinetics and in vivo progression, tumor growth was measured by a caliper, and tumor dissemination was assessed via organ examination after euthanasia (macroscopy, and hematoxylin and eosin staining).

[0171] • In vitro cytotoxicity assays

[0172] The cytotoxicity of the genetically engineered murine anti-CD19 CAR-T cells was demonstrated in in vitro co-cultures against the target tumor cell line A-MYC. Tumor cells were cultured for 24 hours in 96 well plates in complete DMEM complete medium, prior to the addition of T cells to the culture with complete RPMI medium, at different effector to target cell ratios (0.5:1 , 1 :1 , 5:1 and 10:1). Activated untransduced T cells were subjected to two serial spinfections and were used as control in co-cultures with the target cell line. The specific killing capacity was measured after 24 hours of coculture via the percentage of CD19+ alive cells at the different effector to target cell ratios (0.5:1 , 1 :1 , 5:1 and 10:1) by flow cytometry. The measurements were performed in triplicates for each condition. The supernatant of the co-culture was harvested, and the cytokine levels (I FNy, GM-CSF) were measured using MACSPlex multiplex assays (Cat. No. 130-101-740, Miltenyi) and the CytoFLEX (Beckman Coulter) flow cytometer.

[0173] • Bacterial culture of Akkermansia strains

[0174] We used the strain Akkermansia Eve#01 (Akkermansia massiliensis p2261 , Everlmmune proprietary), deposited at the Collection de souches de I’Unite des Rickettsies (CSLIR) under the reference CSLIR P2261 . To prepare mice for gavage with Akkermansia p2267, the bacterium was cultured on COS medium from bioMerieux, enriched with 5% sheep blood, under anaerobic conditions using a Whitley anaerobic chamber (Biolab Diagnostics). The bacteria were incubated for 48 hours. A suspension was then prepared, achieving a concentration of 1 optical density (OD) in 1 mL, equivalent to 109bacteria, in a sodium chloride (NaCI) solution for the gavage process.

[0175] • Pre-clinical immunocompetent anti-mCD19 CAR-T cell mouse model

[0176] To assess the interaction and modulation of the gut microbiota and the immune system, a fully immunocompetent murine model of anti-CD19 / CD28- CAR-T cell therapy in B cell lymphoma-bearing mice was developed. We subcutaneously injected the tumoral A-MYC driven mouse B lymphoma cell at a dose of 500.000 cells into the flank of immunocompetent syngeneic C57BI6 mice, after a pre-conditioning lymphodepletive regimen with 4 Gray (Gy) total body irradiation, followed by an intravenous (IV) injection of 6x105CAR-T cells per mouse. We assessed the effects of the irradiation via whole blood cell counting of the mice prior to and after irradiation, 7 days after CAR-T cell injection, by measuring the levels of granulocytes and lymphocytes in all treatment groups. For the generation of the model prior to the exploration of the effect of the bacterial supplementation, the optimal dose of infused 6x105CAR-T cells was validated after testing different numbers of CAR-T cells injected. For the exploration of the effect of the daily oral gavage with Akkermansia spp., we maintained the solid culture of Akkermansia p2261 bacterial strain. We performed oral gavage of the 1x109bacteria established by optical densitometry (equivalent of 1 DO) and after prior verification of the bacterial identity via matrix-assisted laser desorption / ionization time- of-flight mass spectrometry (MALDI-TOF MS). The anti-tumor efficacy was assessed by a caliper every two days (in mm2) and after euthanasia (metastases).

[0177] • CAR-T cell toxicity assessment

[0178] To assess the tolerance of the treatments administered to the animals, we weighed the mice every two days and developed an internal toxicity score comprising sudden increase or decrease of body weight, presence of necrosis, outgrowth of more than one tumoral mass or tumor dissemination, general appearance including piloerection and cachexia, as well as behavioral patterns. Based on these criteria, a score of 0 was assigned to mice with stable weight, active and interactive well-groomed animals and healthy appearance (silky well-kept hair); a score of 1 was assigned to mice with tumor dissemination and more than one tumoral mass, notably lymph nodes, affected motility specifically in the rear legs and the general absence of the first criteria; finally a score of 2 was assigned to mice with a rapid tumor growth and / or sudden signs of toxicity, such as the presence of more than one tumoral mass, weight loss or gain superior to 15%, severe behavioral (e.g. isolation) or appearance (e.g. loss of hair, piloerection, hunched back, etc.) abnormalities.

[0179] • Organ dissociation from murine samples

[0180] Tumor, bone marrow (BM), tumor-draining lymph nodes (tdLN) and spleen were harvested at different time points. Tumors and tumor-draining lymph nodes were mechanically digested via crashing the samples through 70 pm cell strainers (Cat. No. 352350, Falcon), followed by an enzymatic dissociation with DNAse I (150 U / rnL) and liberase (10pg / mL), and incubated for 30 min at 37°C. Then, the solution was filtered two times until the recovering of cell suspensions. Spleens were processed via manually crashing the samples through 70 pm cell strainers (Cat. No. 352350, Falcon), and incubated during 2 min at room temperature with red blood cell lysis buffer (Cat. No. 00- 4333-57, ThermoFisher Scientific) until the recovering of cell suspension. Similarly, BM single cell suspensions were obtained after getting femurs and tibias (4 bones per mice), flushing these bones with 23G 50mm needles and resuspended in PBS.

[0181] • Flow cytometry analyses for murine samples

[0182] All organs (tumor, tdLN, spleen and bone marrow) were stained for 30 minutes at 4°C with two distinct flow cytometry panels (T cell panel and cytokine panel). Additionally, the spleen and BM cell suspensions were stained for 30 minutes 4°C with a myeloid compartment panel. All panels comprised the following common surface antibodies: Brilliant Violet 421 ™ anti-human EGFR (Cat. No. 352911 , clone AY13), PerCP / Cyanine5.5 anti-mouse CD45.2 (Cat. No. 109828, clone 104), PE anti-mouse CD45.1 (Cat. No. 110708, clone A20) and Zombie NIR™ Fixable Viability Kit (Cat. No. 423106). The T cell and cytokine panel shared the following surface antibodies: Brilliant Violet 510™ anti-mouse CD4 (Cat. No. 100559, clone RM4-5), Alexa Fluor® 700 antimouse CD8a (Cat No. 100730, clone 53-6.7), PE / Cyanine7 anti-mouse CD19 (Cat. No. 115520, clone 6D5), Brilliant Violet 650™ anti-mouse TCR chain (Cat. No. 109251 , clone H57-597). The T cell panel exclusively included the following extracellular antibodies: Brilliant Violet 605™ anti-mouse / human CD44 (Cat. No. 103047, clone IM7), FITC anti-mouse CD25 (Cat. No. 101908, clone 3C7), APC anti-mouse CD279 (PD-1) (Cat. No. 109112, clone RMP1-30), PE / Dazzle™ 594 anti-mouse CD62L (Cat. No. 104448, clone MEL-14). For the cytokine panel, after a first incubation with the surface antibodies, cell suspensions were permeabilized (Cat. No. 00-5523-00, ThermoFisher Scientific) and further incubated for 30 minutes at 4°C with the following antibodies: APC anti-mouse IL-17A (Cat. No. 506916, clone TC11-18H10.1), Brilliant Violet 605™ antimouse TNFa (Cat. No. 506329, clone MP6-XT22), FITC anti-human / mouse Granzyme B (Cat. No. 515403, clone GB11), PE / Dazzle™ 594 anti-mouse IFNy (Cat. No. 505846, clone XMG1.2). Finally, with regards to the myeloid compartment cytometry panel, the spleen and bone marrow cell suspensions were additionally incubated with the following extracellular antibodies: Alexa Fluor® 700 anti-mouse F4 / 80 (Cat. No. 123130, clone BM8), FITC anti-mouse / human CD11 b (Cat. No. 101206, clone 132 CHF), Brilliant Violet 650™ anti-mouse l-A / l-E (Cat. No. 107641 , clone M5 / 114.15.2), Brilliant Violet 510™ anti-mouse Ly-6C Antibody (Cat. No. 128033, clone HK1.4), Brilliant Violet 605™ antimouse CD64 (FcyRI) Antibody (Cat. No. 139323, clone X54- 5 / 7.1), FITC anti-mouse Ly-6G Antibody (Cat. No. 127605, clone 1A8), APC anti- mouse CD103 Antibody (Cat. No. 121413, clone 2E7), PE / Cyanine7 anti-mouse CD11c Antibody (Cat. No. 117317, clone N418). All antibodies were purchased from Biolegend.

[0183] • Histology and immunohistochemistry

[0184] Femur, tdLN, spleen and lung were fixed in 4% paraformaldehyde and embedded in paraffin; 4pm sections were stained with hematoxylin-eosin-safranin for morphology. Paraffin sections were processed for heat-induced antigen retrieval (ER2 corresponding EDTA buffer pH9) for 20 min at 100°C. The signal was revealed with the Klear mouse kit (GBI labs). Single representative whole tissue section from each animal was digitalized using a slide scanner Olympus VS200.

[0185] • Immunofluorescence staining

[0186] The immunofluorescence multiplex panel was performed using sequential indirect immunolabeling with tyramid signal amplification technique. After tissue dewaxing at 72 °C during 30 sec, an antigen retrieval step at pH 9, at 100 °C for 20 min was done with an epitope retrieval solution (ER2). Next, endogenous peroxydases and non-specific binding sites were blocked with hydrogen peroxide (10 min at room temperature (RT)) and PKI blocking solution (5 min at RT, without wash) prior to the application of the primary antibody for 30 min at RT. This one was then recognized by an HRP polymere anti-mouse (GBI Labs, Klear mouse) for EGFR and an anti-rabbit HRP-polymer conjugated secondary antibodies for the others, incubated for 30 min at RT. The formation of highly reactive fluorochrome-tyramid complex (Akoya Opal®), able to bind the tyrosin of tissue in the vicinity of the antigen-antibody complex, was then catalyzed by HRP during an incubation step of 20 min at RT. The antibodies complex (but not the tyramids) were thereafter removed during a stripping step by incubation in ER solution at pH 6, at 97 °C during 20 min, before the next immuno-labeling cycle. To perform a 6-plex, we need to use, in last position, Opal® (TSA-DIG) Polaris 780, which is composed of a tyramid conjugated to digoxygenin, revealed after stripping by an antibody anti-DIG conjugated to Opal® 780 in an incubation step of 60 min at 37 °C. This TSA has to be added in the last cycle, without stripping. Slides were counterstained spectral DAPI (Akoya Biosciences) was added to the slides for 5 min and coverslips were mounted with mounting medium ProLongTM Diamond Antifade Mountant (I nvitrogenTM, ThermoFisher Scientific). Multiplex labeling was performed on an automated staining system Leica BOND RX (Leica Biosystems, Nussloch, Germany). The antibodies used were as follows: anti EGFR (Cat No. # ab264540, clone Emab-134), anti-CD4 (Cat. No. #25229, clone D7D2Z, Cell Signaling); anti-F4 / 80 (Cat No. #70076, clone D2S9R, Cell Signaling); anti-Ly6G (Cat No. #87048, clone E6Z1T, Cell Signaling); anti-FoxP3 (Cat No. #12653, clone D608R, Cell Signaling) and anti-CD8 (Cat No. #ab228965, clone EPR22331-54, Abeam) Several mandatory setup experiments were carried out before performing the multiplex labellings: validation of monoplex labelings, validation of the antigens resistance to multiple stripping cycles, and efficiency of stripping for each primary antibody. For the multiplex labeling, tissue sections were sequentially incubated with the following reagents and conditions (dilutions, OPAL) in this order: anti-EGFR 1 :50, Opal® 520 1 :400; anti-CD4 1 :50, Opal® 570 1 :1 / 200; anti-F4 / 80 1 :500, Opal® 690 1 :600; anti-Ly6G 1 :400, Opal 480 1 :300; anti-FoxP3, Opal 650 ;anti-CD8 1 :500; TSA DIG 1 :50 and Opal® Polaris 780 1 :25. All DAB stained slides were scanned at 20X using VS200 scanner (Olympus, Japan). Fluorescent staining slides were scanned at 20X using Polaris Scanner (Akoya Biosciences, Marlborough, MA, United States). Acquired field of view were color unmixed in Akoya's Inform software and then stitched together within QuPath vO.4.3 software (Bankhead et al., 2017). Quantifications were done in QuPath vO.4.3 software inside bone marrow tissue. A classifier was used to automatically detect these areas. For DAB-stained slides, color deconvolution (Hematoxylin, DAB and residual) was then applied to the images. The number of stained cells by square millimeters of analyzed tissue was enumerated and depicted. For fluorescent staining, cells were detected on the DAPI channel. Phenotyping of each cell was then assessed based on the amount of each fluorophore inside each cell. Results are expressed as cell numbers of each phenotype by square millimeters of analyzed tissue.

[0187] • Culturomics

[0188] For the identification of infiltrating bacteria, aerobic and anaerobic culturomics assays were performed on different mouse samples. Tumor and bone marrow samples from both treatment arms were harvested after CAR-T cell intravenous injection in sterile anaerobic conditions and processed into cell suspensions. Tumor samples were previously dissociated manually via the crashing through 70 pm cell strainers, and via enzymatic dissociation after an incubation with DNAse I (150 U / mL) and Liberase (10pg / mL). Similarly, bone marrow was harvested from femurs and tibias, flushed with PBS, up to a cell suspension. Fresh fecal samples were collected in sterile anaerobic conditions until inoculation or culture. All samples were pre-incubated in aerobic and anaerobic blood culture bottles enriched with 10% sheep blood, supplement mix, and 10% rumen fluid. The culturomics protocol was performed for over 30 days for the identification of bacterial species, every third day, from day 0 to day 30, 1.5 ml of the pre-cultures were extracted and cultivated (Lagier et al., 2015). After 24 hours of aerobic incubation and 48 hours of anaerobic incubation, each isolated bacterial colony was identified using matrix-assisted desorption-ionization-time-of-flight (MALDI-TOF) mass spectrometry with a Microflex spectrometer (Bruker Daltonics, Bremen, Germany), as previously described (Costa et al., 2018).

[0189] • Quantitative RT-PCR on murine samples

[0190] Polymerase Chain Reaction (PCR) was performed to detect bacterial presence in DNA extracted from BM, tumor and stool samples of mice receiving CAR-T cells and CAR-T cells supplemented with Akk. spp. Tumor, BM and stool were processed as previously described and we isolated the DNA from cell suspension using the QIAamp Fast DNA Stool Mini Kit (Cat. No. 74104, Qiagen), following the manufacturer's protocol provided in the kit. The samples were tested for the presence of Akkermansia p2261 as the bacterial strain administered orally with internal and negative controls (all bacteria and murine Akkermansia muciniphila 9226 bacterial strain), as well as positive controls (murine Akkermansia massiliensis 9228 bacterial strain), using specific primers. PCR experiments were similarly performed to detect the activity of the AhR pathway in mRNA extracted from tumor and BM cell suspensions after organ dissociation in mice from both treatment arms. RNA was extracted from cell suspensions using the Quick-RNA MicroPrep Kit (Cat. No. ZR1051 , Ozyme) following the manufacturer’s instructions. The samples were tested for the presence of AhR (Assay ID. Mm00478932_m1 , Cat. No. 4331182 Applied Biosystems TaqMan, Thermofisher Scientific).

[0191] • Metabolomics analyses on murine samples and bacterial culture supernatants in vitro

[0192] Analyses of the metabolic pathways were assessed in the pre-clinical murine model of CAR-T cell in B cell lymphoma bearing mice, in the plasma and stool supernatants at different time-points (day 7 and 24 after CAR-T cell injection). Targeted analyses of the indole metabolic pathway were performed on stools and bacterial culture supernatant after 72 hrs of anaerobic blood culture of Akkermansia p2261 . Culture medium was prepared as previously described (Iwaza et al., 2022). We used culture media alone, culture media together with Akk. p2261 , and culture media together with Akk. p2261 supplemented with L-Tryptophan at a dose of 0.4g / L. Supernatants were collected, and several fractions were split to be analyzed by different liquid chromatography coupled with mass spectrometers (LC / MS). Heparinized plasma and bacterial culture supernatants (25 pL and 100 pL respectively) were mixed with extraction mixture (250 pL, and 750 pL respectively), and then processed in the same way as previously described. Indoles were analyzed by pseudo-targeted analysis with an UHPLC-HRAM (Ultra-High Performance Liquid Chromatography — High Resolution Accurate Mass), performed on a U3000 (Dionex) / Orbitrap q-Exactive (Thermo) coupling, previously described. All targeted treated data were merged and cleaned with a dedicated R (version 4.0) package (@Github / Kroemerlab / GRMeta).

[0193] • AhR activation via HepG2 reporter cell line

[0194] The AhR activity was measured using HepG2-Lucia™ AhR reporter cells (InvivoGen, France) as previously described (Modoux et al., 2022). Cells were seeded into a 96-well plate and stimulated with for 24 hours according to the manufacturer’s protocol. Luciferase activity was measured using a luminometer and Quanti-Luc reagent (InvivoGen). The results were normalized based on the negative luciferase activity of the control and cytotoxicity measurement (LDH). The AhR activity was measured in the supernatant of the bacterial culture in different conditions (culture media alone, culture media together with Akk. p2261 , and culture media together with Akk. p2261 supplemented with L-Tryptophan at a dose of 0.4g / L).

[0195] • Statistics

[0196] Statistical significance was determined using different comparisons in a one-way or two-way analyses of variance (ANOVA) tests for tumor growth curve analyses, or Mann- Whitney tests for unpaired nonparametric t-test when only two groups were compared, using GraphPad Prism v10 built-in statistical analysis (GraphPad Software Inc., La Jolla, CA). For metabolomic and shotgun metagenomic statistically significant differences among timepoints were evaluated through the Wilcoxon signed- rank test, in which paired samples were considered for each comparison. The same test was also applied to compare Akk. spp. neg and Akk. spp. pos cohorts across timepoints. Differences between ORR- and ORR+ were instead computed using the Wilcoxon- Mann-Whitney test. Survival analysis at baseline was performed between Akk. spp. neg and Akk. spp. pos by considering the Kaplan-Meier estimator and the log-rank test. Differences among timepoints and ORR cohorts were assessed through the PERMANOVA test (function adonis2 available in the R package vegan). A P-value lower than P < 0.05 was considered significant in all statistical tests performed.

[0197] Example 12: Effects of microbial indoles on dendritic cells cross presentation

[0198] To dive into the potential mechanisms accounting for CAR-T or host T cell fitness, we used a DC-T cell co-culture system that was previously described (Zhao et al. 2021 ; Zhao et al. 2023), with non-genetically modified dendritic cells. In this assay, we used chicken ovalbumin (OVA), which can be presented by C57BI / 6 mouse-derived DCs in form of an octapeptide OVA257-264 abbreviated as SL8 (Rotzschke et al., 1991). The H2-Kb-SL8 complex is recognized by the TCR of B3Z hybridoma cell line, which produce interleukin-2 (IL-2) upon its stimulation. Hence, the measurement of the amount of IL-2 produced by B3Z cells cocultured with OVA-pulsed DCs can be used as a proxy for antigen presentation.

[0199] We examined the dose effects of numerous compounds onto the crosspresentation capacity of DCs loaded with ovalbumin (OVA) protein, monitored by the reactivity (IL-2 release) of an OVA257-264-specific, H-2Kb-restricted cytotoxic T lymphocyte (CTL) hybridoma (B3Z). Several indoles, including 5-hydroxyindole and 5- hydroxyindole-3-acetic acid and to a lesser extent indole-3-butyrate, indole-3-pyruvate and indole-3-carboxaldehyde, significantly augmented IL-2 production in this system (Table 5).

[0200] Table 5: mean fold change of IL-2 production in the presence of different concentrations of different indole compounds

[0201] Of note, doses that appeared optimal in DC / T cell cross talk with indoles (5-50pM) significantly augmented IL-2 production in this system measuring DC function. Altogether, these in vitro findings suggest that indoles could mediate its CD8+T cell effector- (CAR-T CD19 and / or host T cells) stimulatory effects via direct and indirect effects on DCs cross presentation.

[0202] REFERENCES

[0203] Abramson, J.S., Palomba, M.L., Gordon, L.L, Lunning, M.A., Wang, M., Arnason, J., Mehta, A., Purev, E., Maloney, D.G., Andreadis, C., Sehgal, A., Solomon, S.R., Ghosh, N., Albertson, T.M., Garcia, J., Kostic, A., Mallaney, M., Ogasawara, K., Newhall, K., Kim, Y., Li, D., Siddiqi, T., 2020. Lisocabtagene maraleucel for patients with relapsed or refractory large B-cell lymphomas (TRANSCEND NHL 001): a multicentre seamless design study. The Lancet 396, 839-852. https: / / doi.Org / 10.1016 / S0140-6736(20)31366-0

[0204] Agus, A., Planchais, J., Sokol, H., 2018. Gut Microbiota Regulation of Tryptophan Metabolism in Health and Disease. Cell Host Microbe 23, 716-724. https: / / d0i.0rg / l 0.1016 / j.chom.2018.05.003

[0205] Alves Costa Silva, C., Piccinno, G., Suissa, D., Bourgin, M., Schreibelt, G., Durand, S., Birebent, R., Fidelle, M., Sow, C., Aprahamian, F., Manghi, P., Puncochaf, M., Asnicar, F., Pinto, F., Armanini, F., Terrisse, S., Routy, B., Drubay, D., Eggermont, A.M.M., Kroemer, G., Segata, N., Zitvogel, L., Derosa, L., Bol, K.F., de Vries, LJ.M., 2024. Influence of microbiota-associated metabolic reprogramming on clinical outcome in patients with melanoma from the randomized adjuvant dendritic cell-based MIND-DC trial. Nat. Commun. 15, 1633. https: / / d0i.0rg / l 0.1038 / S41467-024-45357-1

[0206] Bachy, E., Le Gouill, S., Di Blasi, R., Sesques, P., Manson, G., Cartron, G., Beauvais, D., Roulin, L., Gros, F.X., Rubio, M.T., Bories, P., Bay, J.O., Llorente, C.C., Choquet, S., Casasnovas, R.-O., Mohty, M., Guidez, S., Joris, M., Loschi, M., Carras, S., Abraham, J., Chauchet, A., Drieu La Rochelle, L., Deau-Fischer, B., Hermine, O., Gastinne, T., Tudesq, J. J., Gat, E., Broussais, F., Thieblemont, C., Houot, R., Morschhauser, F., 2022. A real-world comparison of tisagenlecleucel and axicabtagene ciloleucel CAR T cells in relapsed or refractory diffuse large B cell lymphoma. Nat. Med. 28, 2145-2154. https: / / doi.org / 10.1038 / s41591-022- 01969-y

[0207] Bankhead, P., Loughrey, M.B., Fernandez, J.A., Dombrowski, Y., McArt, D.G., Dunne, P.D., McQuaid, S., Gray, R.T., Murray, L.J., Coleman, H.G., James, J.A., Salto-Tellez, M., Hamilton, P.W., 2017. QuPath: Open source software for digital pathology image analysis. Sci. Rep. 7, 16878. https: / / doi.org / 10.1038 / s41598-017-17204-5

[0208] Baruch, E.N., Youngster, I., Ben-Betzalel, G., Ortenberg, R., Lahat, A., Katz, L., Adler, K., Dick-Necula, D., Raskin, S., Bloch, N., Rotin, D., Anafi, L., Avivi, C., Melnichenko, J., Steinberg- Silman, Y., Mamtani, R., Harati, H., Asher, N., Shapira-Frommer, R., Brosh-Nissimov, T., Eshet, Y., Ben-Simon, S., Ziv, O., Khan, M.A.W., Amit, M., Ajami, N.J., Barshack, L, Schachter, J., Wargo, J.A., Koren, O., Markel, G., Boursi, B., 2021. Fecal microbiota transplant promotes response in immunotherapy-refractory melanoma patients. Science 371 , 602-609. https: / / doi.Org / 10.1126 / science.abb5920

[0209] Bender, M.J., McPherson, A.C., Phelps, C.M., Pandey, S.P., Laughlin, C.R., Shapira, J.H., Medina Sanchez, L., Rana, M., Richie, T.G., Mims, T.S., Gocher-Demske, A.M., Cervantes- Barragan, L., Mullett, S.J., Gelhaus, S.L., Bruno, T.C., Cannon, N., McCulloch, J.A., Vignali, D.A.A., Hinterleitner, R., Joglekar, A.V., Pierre, J.F., Lee, S.T.M., Davar, D., Zarour, H.M., Meisel, M., 2023. Dietary tryptophan metabolite released by intratumoral Lactobacillus reuteri facilitates immune checkpoint inhibitor treatment. Cell 186, 1846-1862. e26. https: / / doi.Org / 10.1016 / j.cell.2023.03.01 1

[0210] Bigenwald, C., Zitvogel, L., 2023. Light shed from the gut in large B-cell lymphoma. Blood 141 , 2165-2166. https: / / doi.org / 10.1182 / blood.2023019858

[0211] Bufalo, F.D., Angelis, B.D., Caruana, L, Baldo, G.D., loris, M.A.D., Serra, A., Mastronuzzi, A., Cefalo, M.G., Pagliara, D., Amicucci, M., Pira, G.L., Leone, G., Bertaina, V., Sinibaldi, M., Cecca, S.D., Guercio, M., Abbaszadeh, Z., laffaldano, L., Gunetti, M., lacovelli, S., Bugianesi, R., Macchia, S., Algeri, M., Merli, P., Galaverna, F., Abbas, R., Garganese, M.C., Villani, M.F., Colafati, G.S., Bonetti, F., Rabusin, M., Perruccio, K., Folsi, V., Quintarelli, C., Locatelli, F., 2023. GD2-CART01 for Relapsed or Refractory High-Risk Neuroblastoma. N. Engl. J. Med. 388, 1284- 1295. https: / / d0i.0rg / l 0.1056 / NEJMoa2210859 Cani, P.D., Depommier, C., Derrien, M., Everard, A., de Vos, W.M., 2022. Akkermansia muciniphila: paradigm for next-generation beneficial microorganisms. Nat. Rev. Gastroenterol. Hepatol. 19, 625-637. https: / / doi.org / 10.1038 / s41575-022-00631-9

[0212] Cervantes-Barragan, L., Chai, J.N., Tianero, Ma.D., Di Luccia, B., Ahern, P.P., Merriman, J., Cortez, V.S., Caparon, M.G., Donia, M.S., Gilfillan, S., Celia, M., Gordon, J.I., Hsieh, C.-S., Colonna, M., 2017. Lactobacillus reuteri induces gut intraepithelial CD4+CD8aa+ T cells. Science 357, 806-810. https: / / doi.org / 10.1126 / science.aah5825

[0213] Chaput, N., Lepage, P., Coutzac, C., Soularue, E., Le Roux, K., Monot, C., Boselli, L., Routier, E., Cassard, L., Collins, M., Vaysse, T., Marthey, L., Eggermont, A., Asvatourian, V., Lanoy, E., Mateus, C., Robert, C., Carbonnel, F., 2017. Baseline gut microbiota predicts clinical response and colitis in metastatic melanoma patients treated with ipilimumab. Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 28, 1368-1379. https: / / doi.org / 10.1093 / annonc / mdx108

[0214] Choi, Y., Lichterman, J.N., Coughlin, L.A., Poulides, N., Li, W., Del Valle, P., Palmer, S.N., Gan, S., Kim, J., Zhan, X., Gao, Y., Evers, B.M., Hooper, L.V., Pasare, C., Koh, A.Y., 2023. Immune checkpoint blockade induces gut microbiota translocation that augments extraintestinal antitumor immunity. Sci. Immunol. 8, eabo2003. https: / / doi.org / 10.1126 / sciimmunol.abo2003

[0215] Costa, O.Y.A., Raaijmakers, J.M., Kuramae, E.E., 2018. Microbial Extracellular Polymeric Substances: Ecological Function and Impact on Soil Aggregation. Front. Microbiol. 9, 1636. https: / / doi.Org / 10.3389 / fmicb.2018.01636

[0216] Davar, D., Dzutsev, A.K., McCulloch, J.A., Rodrigues, R.R., Chauvin, J.-M., Morrison, R.M., Deblasio, R.N., Menna, C., Ding, Q., Pagliano, O., Zidi, B., Zhang, S., Badger, J.H., Vetizou, M., Cole, A.M., Fernandes, M.R., Prescott, S., Costa, R.G.F., Balaji, A.K., Morgun, A., Vujkovic-Cvijin,

[0217] I., Wang, H., Borhani, A.A., Schwartz, M.B., Dubner, H.M., Ernst, S.J., Rose, A., Najjar, Y.G., Belkaid, Y., Kirkwood, J.M., Trinchieri, G., Zarour, H.M., 2021. Fecal microbiota transplant overcomes resistance to anti-PD-1 therapy in melanoma patients. Science 371 , 595-602. https: / / doi.Org / 10.1126 / science.abf3363

[0218] Derosa, L., lebba, V., Silva, C.A.C., Piccinno, G., Wu, G., Lordello, L., Routy, B., Zhao, N., Thelemaque, C., Birebent, R., Marmorino, F., Fidelle, M., Messaoudene, M., Thomas, A.M., Zalcman, G., Friard, S., Mazieres, J., Audigier-Valette, C., Sibilot, D.M.-, Goldwasser, F., Scherpereel, A., Pegliasco, H., Ghiringhelli, F., Bouchard, N., Sow, C., Darik, I., Zoppi, S., Ly, P., Reni, A., Daillere, R., Deutsch, E., Lee, K.A., Bolte, L.A., Bjork, J.R., Weersma, R.K., Barlesi, F., Padilha, L., Finzel, A., Isaksen, M.L., Escudier, B., Albiges, L., Planchard, D., Andre, F., Cremolini, C., Martinez, S., Besse, B., Zhao, L., Segata, N., Wojcik, J., Kroemer, G., Zitvogel, L., 2024. Custom scoring based on ecological topology of gut microbiota associated with cancer immunotherapy outcome. Cell 187, 3373-3389. e16. https: / / doi.Org / 10.1016 / j.cell.2024.05.029

[0219] Derosa, L., Routy, B., Desilets, A., Daillere, R., Terrisse, S., Kroemer, G., Zitvogel, L., 2021 . Microbiota-Centered Interventions: The Next Breakthrough in Immuno-Oncology? Cancer Discov. 11 , 2396-2412. https: / / doi.org / 10.1158 / 2159-8290.CD-21-0236

[0220] Derosa, L., Routy, B., Fidelle, M., lebba, V., Alla, L., Pasolli, E., Segata, N., Desnoyer, A., Pietrantonio, F., Ferrere, G., Fahrner, J.-E., Le Chatellier, E., Pons, N., Galleron, N., Roume, H., Duong, C.P.M., Mondragon, L., Iribarren, K., Bonvalet, M., Terrisse, S., Rauber, C., Goubet, A.- G., Daillere, R., Lemaitre, F., Reni, A., Casu, B., Alou, M.T., Alves Costa Silva, C., Raoult, D., Fizazi, K., Escudier, B., Kroemer, G., Albiges, L., Zitvogel, L., 2020. Gut Bacteria Composition Drives Primary Resistance to Cancer Immunotherapy in Renal Cell Carcinoma Patients. Eur. Urol. 78, 195-206. https: / / doi.Org / 10.1016 / j.eururo.2020.04.044

[0221] Derosa, L., Routy, B., Thomas, A.M., lebba, V., Zalcman, G., Friard, S., Mazieres, J., Audigier-Valette, C., Moro-Sibilot, D., Goldwasser, F., Silva, C.A.C., Terrisse, S., Bonvalet, M., Scherpereel, A., Pegliasco, H., Richard, C., Ghiringhelli, F., Elkrief, A., Desilets, A., Blanc-Durand, F., Cumbo, F., Blanco, A., Boidot, R., Chevrier, S., Daillere, R., Kroemer, G., Alla, L., Pons, N., Le Chatelier, E., Galleron, N., Roume, H., Dubuisson, A., Bouchard, N., Messaoudene, M., Drubay, D., Deutsch, E., Barlesi, F., Planchard, D., Segata, N., Martinez, S., Zitvogel, L., Soria,

[0222] J.-C., Besse, B., 2022. Intestinal Akkermansia muciniphila predicts clinical response to PD-1 blockade in patients with advanced non-small-cell lung cancer. Nat. Med. 28, 315-324. https: / / d0i.0rg / l 0.1038 / S41591-021-01655-5 Derrien, M., Vaughan, E.E., Plugge, C.M., de Vos, W.M., 2004. Akkermansia muciniphila gen. nov., sp. nov., a human intestinal mucin-degrading bacterium. Int. J. Syst. Evol. Microbiol. 54, 1469-1476. https: / / doi.Org / 10.1099 / ijs.0.02873-0

[0223] Di Rosa, F., Gebhardt, T., 2016. Bone Marrow T Cells and the Integrated Functions of Recirculating and Tissue-Resident Memory T Cells. Front. Immunol. 7, 51. https: / / doi.org / 10.3389 / fimmu.2016.00051

[0224] Di Rosa, F., Pabst, R., 2005. The bone marrow: a nest for migratory memory T cells. Trends Immunol. 26, 360-366. https: / / doi.Org / 10.1016 / j.it.2005.04.011

[0225] Elkrief, A., Mendez-Salazar, E.O., Maillou, J., Vanderbilt, C.M., Gogia, P., Desilets, A., Messaoudene, M., Kelly, D., Ladanyi, M., Hellmann, M.D., Zitvogel, L., Rudin, C.M., Routy, B., Derosa, L., Schoenfeld, A.J., 2024. Antibiotics are associated with worse outcomes in lung cancer patients treated with chemotherapy and immunotherapy. NPJ Precis. Oncol. 8, 143. https: / / doi.Org / 10.1038 / S41698-024-00630-w

[0226] Fidelle, M., Rauber, C., Alves Costa Silva, C., Tian, A.-L., Lahmar, L, de La Varende, A.-

[0227] L.M., Zhao, L., Thelemaque, C., Lebhar, I., Messaoudene, M., Pizzato, E., Birebent, R., Mbogning

[0228] Fonkou, M.D., Zoppi, S., Reni, A., Dalban, C., Leduc, M., Ferrere, G., Durand, S., Ly, P., Silvin, A., Mulder, K., Dutertre, C.-A., Ginhoux, F., Yonekura, S., Roberti, M.P., Tidjani-Alou, M., Terrisse, S., Chen, J., Kepp, O., Schippers, A., Wagner, N., Suarez-Gosalvez, J., Kobold, S., Fahrner, J.-E., Richard, C., Bosq, J., Lordello, L., Vitali, G., Galleron, N., Quinquis, B., Le Chatelier, E., Blanchard, L., Girard, J.-P., Jarry, A., Gervois, N., Godefroy, E., Labarriere, N., Koschny, R., Daillere, R., Besse, B., Truntzer, C., Ghiringhelli, F., Coatnoan, N., Mhanna, V., Klatzmann, D., Drubay, D., Albiges, L., Thomas, A.M., Segata, N., Danlos, F.-X., Marabelle, A., Routy, B., Derosa, L., Kroemer, G., Zitvogel, L., 2023. A microbiota-modulated checkpoint directs immunosuppressive intestinal T cells into cancers. Science 380, eabo2296. https: / / doi.Org / 10.1126 / science.abo2296

[0229] Frankel, A.E., Coughlin, L.A., Kim, J., Froehlich, T.W., Xie, Y., Frenkel, E.P., Koh, A.Y., 2017. Metagenomic Shotgun Sequencing and Unbiased Metabolomic Profiling Identify Specific Human Gut Microbiota and Metabolites Associated with Immune Checkpoint Therapy Efficacy in Melanoma Patients. Neoplasia N. Y. N 19, 848-855. https: / / doi.Org / 10.1016 / j.neo.2017.08.004

[0230] Gilroy R, et al. Extensive microbial diversity within the chicken gut microbiome revealed by metagenomics and culture. PeerJ. 2021 ;9:e10941 . doi: 10.7717 / peerj.10941.

[0231] Gopalakrishnan, V., Spencer, C.N., Nezi, L., Reuben, A., Andrews, M.C., Karpinets, T.V., Prieto, P.A., Vicente, D., Hoffman, K., Wei, S.C., Cogdill, A.P., Zhao, L., Hudgens, C.W., Hutchinson, D.S., Manzo, T., Petaccia de Macedo, M., Cotechini, T., Kumar, T., Chen, W.S., Reddy, S.M., Szczepaniak Sloane, R., Galloway-Pena, J., Jiang, H., Chen, P.L., Shpall, E.J., Rezvani, K., Alousi, A.M., Chemaly, R.F., Shelburne, S., Vence, L.M., Okhuysen, P.C., Jensen, V.B., Swennes, A.G., McAllister, F., Marcelo Riquelme Sanchez, E., Zhang, Y., Le Chatelier, E., Zitvogel, L., Pons, N., Austin-Breneman, J.L., Haydu, L.E., Burton, E.M., Gardner, J.M., Sirmans, E., Hu, J., Lazar, A.J., Tsujikawa, T., Diab, A., Tawbi, H., Glitza, LC., Hwu, W.J., Patel, S.P., Woodman, S.E., Amaria, R.N., Davies, M.A., Gershenwald, J.E., Hwu, P., Lee, J.E., Zhang, J., Coussens, L.M., Cooper, Z.A., Futreal, P.A., Daniel, C.R., Ajami, N.J., Petrosino, J.F., Tetzlaff,

[0232] M.T., Sharma, P., Allison, J.P., Jenq, R.R., Wargo, J.A., 2018. Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients. Science 359, 97-103. https: / / doi.Org / 10.1126 / science.aan4236

[0233] Griffin, M.E., Espinosa, J., Becker, J.L., Luo, J.-D., Carroll, T.S., Jha, J.K., Fanger, G.R., Hang, H.C., 2021 . Enterococcus peptidoglycan remodeling promotes checkpoint inhibitor cancer immunotherapy. Science 373, 1040-1046. https: / / doi.org / 10.1126 / science.abc91 13

[0234] Guo, H., Chou, W.-C., Lai, Y., Liang, K., Tam, J.W., Brickey, W.J., Chen, L., Montgomery,

[0235] N.D., Li, X., Bohannon, L.M., Sung, A.D., Chao, N.J., Peled, J.U., Gomes, A.L.C., van den Brink, M.R.M., French, M.J., Macintyre, A.N., Sempowski, G.D., Tan, X., Sartor, R.B., Lu, K., Ting, J.P.Y., 2020. Multi-omics analyses of radiation survivors identify radioprotective microbes and metabolites. Science 370, eaay9097. https: / / doi.org / 10.1126 / science.aay9097

[0236] Gutierrez-Vazquez, C., Quintana, F.J., 2018. Regulation of the Immune Response by the Aryl Hydrocarbon Receptor. Immunity 48, 19-33. https: / / doi.org / 10.1016 / jJmmuni.2017.12.012 Hayden, P.J., Roddie, C., Bader, P., Basak, G.W., Bonig, H., Bonini, C., Chabannon, C., Ciceri, F., Corbacioglu, S., Ellard, R., Sanchez-Guijo, F., Jager, U., Hildebrandt, M., Hudecek, M., Kersten, M.J., Kohl, U., Kuball, J., Mielke, S., Mohty, M., Murray, J., Nagler, A., Rees, J., Rioufol,

[0237] C., Saccardi, R., Snowden, J.A., Styczynski, J., Subklewe, M., Thieblemont, C., Topp, M., Ispizua, A.U., Chen, D., Vrhovac, R., Gribben, J.G., Kroger, N., Einsele, H., Yakoub-Agha, L, 2022. Management of adults and children receiving CAR T-cell therapy: 2021 best practice recommendations of the European Society for Blood and Marrow Transplantation (EBMT) and the Joint Accreditation Committee of ISCT and EBMT (JACIE) and the European Haematology Association (EHA). Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 33, 259-275. https: / / doi.Org / 10.1016 / j.annonc.2021 .12.003

[0238] Heczey, A., Xu, X., Courtney, A.N., Tian, G., Barragan, G.A., Guo, L., Amador, C.M., Ghatwai, N., Rathi, P., Wood, M.S., Li, Y., Zhang, C., Demberg, T., Di Pierro, E.J., Sher, A.C., Zhang, H., Mehta, B., Thakkar, S.G., Grilley, B., Wang, T., Weiss, B.D., Montalbano, A., Subramaniam, M., Xu, C., Sachar, C., Wells, D.K., Dotti, G., Metelitsa, L.S., 2023. Anti-GD2 CAR- NKT cells in relapsed or refractory neuroblastoma: updated phase 1 trial interim results. Nat. Med. 29, 1379-1388. https: / / doi.org / 10.1038 / s41591-023-02363-y

[0239] Hegde, M., Navai, S., DeRenzo, C., Joseph, S.K., Sanber, K., Wu, M., Gad, A.Z., Janeway, K.A., Campbell, M., Mullikin, D., Nawas, Z., Robertson, C., Mathew, P.R., Zhang, H., Mehta, B., Bhat, R.R., Major, A., Shree, A., Gerken, C., Kalra, M., Chakraborty, R., Thakkar, S.G., Dakhova, O., Salsman, V.S., Grilley, B., Lapteva, N., Gee, A., Dotti, G., Bao, R., Salem, A.H., Wang, T., Brenner, M.K., Heslop, H.E., Weis, W.S., Hicks, M.J., Gottschalk, S., Ahmed, N., 2024. Autologous HER2-specific CAR T cells after lymphodepletion for advanced sarcoma: a phase 1 trial. Nat. Cancer 5, 880-894. https: / / doi.org / 10.1038 / s43018-024-00749-6

[0240] Hezaveh, K., Shinde, R.S., Klotgen, A., Halaby, M.J., Lamorte, S., Ciudad, M.T., Quevedo, R., Neufeld, L., Liu, Z.Q., Jin, R., Grunwald, B.T., Foerster, E.G., Chaharlangi, D., Guo, M., Makhijani, P., Zhang, X., Pugh, T.J., Pinto, D.M., Co, LL., McGuigan, A.P., Jang, G.H., Khokha, R., Ohashi, P.S., O’Kane, G.M., Gallinger, S., Navarre, W.W., Maughan, H., Philpott, D.J., Brooks,

[0241] D.G., McGaha, T.L., 2022. Tryptophan-derived microbial metabolites activate the aryl hydrocarbon receptor in tumor-associated macrophages to suppress anti-tumor immunity. Immunity 55, 324-340. e8. https: / / doi.org / 10.1016 / jJmmuni.2022.01.006

[0242] Hubbard, T.D., Murray, LA., Perdew, G.H., 2015. Indole and Tryptophan Metabolism: Endogenous and Dietary Routes to Ah Receptor Activation. Drug Metab. Dispos. Biol. Fate Chem. 43, 1522-1535. https: / / doi.org / 10.1124 / dmd.115.064246

[0243] Iwaza, R., Wasfy, R.M., Dubourg, G., Raoult, D., Lagier, J.-C., 2022. Akkermansia muciniphila: The state of the art, 18 years after its first discovery. Front. Gastroenterol. 1. https: / / doi.org / 10.3389 / fgstr.2022.1024393

[0244] Jacomy, M., Venturini, T., Heymann, S., Bastian, M., 2014. ForceAtlas2, a Continuous Graph Layout Algorithm for Handy Network Visualization Designed for the Gephi Software. PLOS ONE 9, e98679. https: / / doi.org / 10.1371 / journal.pone.0098679

[0245] Jia, D., Wang, Q., Qi, Y., Jiang, Y., He, J., Lin, Y., Sun, Y., Xu, J., Chen, W., Fan, L., Yan, R., Zhang, W., Ren, G., Xu, C., Ge, Q., Wang, Lan, Liu, W., Xu, F., Wu, P., Wang, Y., Chen, S., Wang, Liangjing, 2024. Microbial metabolite enhances immunotherapy efficacy by modulating T cell sternness in pan-cancer. Cell 187, 1651-1665. e21 . https: / / doi.Org / 10.1016 / j.cell.2024.02.022

[0246] Jiang, H., Chen, C., Gao, J., 2022. Extensive Summary of the Important Roles of Indole Propionic Acid, a Gut Microbial Metabolite in Host Health and Disease. Nutrients 15, 151. https: / / doi.Org / 10.3390 / nu15010151

[0247] Jiang, Y., Jia, D., Sun, Y., Ding, N., Wang, L., 2023. Microbiota: A key factor affecting and regulating the efficacy of immunotherapy. Clin. Transl. Med. 13, e1508. https: / / d0i.0rg / l 0.1002 / ctm2.1508

[0248] Kim J-S, Kang SW, Lee JH, Park S-H, Lee J-S. The evolution and competitive strategies of Akkermansia muciniphila in gut. Gut Microbes. 2022;14:2025017. doi: 10.1080 / 19490976.2021 .2025017.

[0249] Klysz, D.D., Fowler, C., Malipatlolla, M., Stuani, L., Freitas, K.A., Chen, Y., Meier, S., Daniel, B., Sandor, K., Xu, P., Huang, J., Labanieh, L., Keerthi, V., Leruste, A., Bashti, M., Mata- Alcazar, J., Gkitsas, N., Guerrero, J.A., Fisher, C., Patel, Sunny, Asano, K., Patel, Shabnum, Davis, K.L., Satpathy, A.T., Feldman, S.A., Sotillo, E., Mackall, C.L., 2024. Inosine induces sternness features in CAR-T cells and enhances potency. Cancer Cell S1535610824000084. https: / / doi.Org / 10.1016 / j.ccell.2024.01.002

[0250] Lagier, J.-C., Hugon, P., Khelaifia, S., Fournier, P.-E., La Scola, B., Raoult, D., 2015. The rebirth of culture in microbiology through the example of culturomics to study human gut microbiota. Clin. Microbiol. Rev. 28, 237-264. https: / / doi.org / 10.1128 / CMR.00014-14

[0251] Lamas, B., Natividad, J.M., Sokol, H., 2018. Aryl hydrocarbon receptor and intestinal immunity. Mucosal Immunol. 1 1 , 1024-1038. https: / / doi.org / 10.1038 / s41385-018-0019-2

[0252] Locke, F.L., Miklos, D.B., Jacobson, C.A., Perales, M.-A., Kersten, M.-J., Oluwole, O.O., Ghobadi, A., Rapoport, A.P., McGuirk, J., Pagel, J.M., Munoz, J., Farooq, U., Meerten, T. van, Reagan, P.M., Sureda, A., Flinn, LW., Vandenberghe, P., Song, K.W., Dickinson, M., Minnema, M.C., Riedell, P.A., Leslie, L.A., Chaganti, S., Yang, Y., Filosto, S., Shah, J., Schupp, M., To, C., Cheng, P., Gordon, L.I., Westin, J.R., 2022. Axicabtagene Ciloleucel as Second-Line Therapy for Large B-Cell Lymphoma. N. Engl. J. Med. 386, 640-654. https: / / doi.Org / 10.1056 / NEJMoa2116133

[0253] Luu, M., Riester, Z., Baldrich, A., Reichardt, N., Yuille, S., Busetti, A., Klein, M., Wempe, A., Leister, H., Raifer, H., Picard, F., Muhammad, K., Ohl, K., Romero, R., Fischer, F., Bauer, C.A., Huber, M., Gress, T.M., Lauth, M., Danhof, S., Bopp, T., Nerreter, T., Mulder, I.E., Steinhoff, U., Hudecek, M., Visekruna, A., 2021. Microbial short-chain fatty acids modulate CD8+ T cell responses and improve adoptive immunotherapy for cancer. Nat. Commun. 12, 4077. https: / / doi.org / 10.1038 / s41467-021-24331-1

[0254] Mager, L.F., Burkhard, R., Pett, N., Cooke, N.C.A., Brown, K., Ramay, H., Paik, S., Stagg, J., Groves, R.A., Gallo, M., Lewis, LA., Geuking, M.B., McCoy, K.D., 2020. Microbiome-derived inosine modulates response to checkpoint inhibitor immunotherapy. Science 369, 1481-1489. https: / / doi.Org / 10.1126 / science.abc3421

[0255] Matson, V., Fessler, J., Bao, R., Chongsuwat, T., Zha, Y., Alegre, M.-L., Luke, J.J., Gajewski, T.F., 2018. The commensal microbiome is associated with anti-PD-1 efficacy in metastatic melanoma patients. Science 359, 104-108. https: / / doi.org / 10.1 126 / science.aao3290

[0256] Modoux, M., Rolhion, N., Lefevre, J.H., Oeuvray, C., Nadvornik, P., Illes, P., Emond, P., Parc, Y., Mani, S., Dvorak, Z., Sokol, H., 2022. Butyrate acts through HDAC inhibition to enhance aryl hydrocarbon receptor activation by gut microbiota-derived ligands. Gut Microbes 14, 2105637. https: / / d0i.0rg / l 0.1080 / 19490976.2022.2105637

[0257] Natividad, J.M., Agus, A., Planchais, J., Lamas, B., Jarry, A.C., Martin, R., Michel, M.-L., Chong-Nguyen, C., Roussel, R., Straube, M., Jegou, S., McQuitty, C., Le Gall, M., da Costa, G., Lecornet, E., Michaudel, C., Modoux, M., Glodt, J., Bridonneau, C., Sovran, B., Dupraz, L., Bado, A., Richard, M.L., Langella, P., Hansel, B., Launay, J. -M., Xavier, R.J., Duboc, H., Sokol, H., 2018. Impaired Aryl Hydrocarbon Receptor Ligand Production by the Gut Microbiota Is a Key Factor in Metabolic Syndrome. Cell Metab. 28, 737-749. e4. https: / / doi.Org / 10.1016 / j.cmet.2018.07.001

[0258] Ndongo, S., Armstrong, N., Raoult, D., Fournier, P.-E., 2022. Reclassification of eight Akkermansia muciniphila strains and description of Akkermansia massiliensis sp. nov. and Candidatus Akkermansia timonensis, isolated from human feces. Sci. Rep. 12, 21747. https: / / doi.org / 10.1038 / s41598-022-25873-0

[0259] Olson, B., Li, Y., Lin, Y., Liu, E.T., Patnaik, A., 2018. Mouse Models for Cancer Immunotherapy Research. Cancer Discov. 8, 1358-1365. https: / / doi.Org / 10.1158 / 2159-8290. CD- 18-0044

[0260] Ouwerkerk, J.P., van der Ark, K.C.H., Davids, M., Claassens, N.J., Finestra, T.R., de Vos, W.M., Belzer, C., 2016. Adaptation of Akkermansia muciniphila to the Oxic-Anoxic Interface of the Mucus Layer. Appl. Environ. Microbiol. 82, 6983-6993. https: / / doi.org / 10.1128 / AEM.01641- 16

[0261] Pinato, D.J., Howlett, S., Ottaviani, D., Urus, H., Patel, A., Mineo, T., Brock, C., Power, D., Hatcher, O., Falconer, A., Ingle, M., Brown, A., Gujral, D., Partridge, S., Sarwar, N., Gonzalez, M., Bendle, M., Lewanski, C., Newsom-Davis, T., Allara, E., Bower, M., 2019. Association of Prior Antibiotic Treatment With Survival and Response to Immune Checkpoint Inhibitor Therapy in Patients With Cancer. JAMA Oncol. 5, 1774-1778. https: / / doi.org / 10.1001 / jamaoncol.2019.2785 Plovier, H., Everard, A., Druart, C., Depommier, C., Van Hui, M., Geurts, L., Chilloux, J., Ottman, N., Duparc, T., Lichtenstein, L., Myridakis, A., Delzenne, N.M., Klievink, J., Bhattacharjee, A., van der Ark, K.C.H., Aalvink, S., Martinez, L.O., Dumas, M.-E., Maiter, D., Loumaye, A., Hermans, M.P., Thissen, J.-P., Belzer, C., de Vos, W.M., Cani, P.D., 2017. A purified membrane protein from Akkermansia muciniphila or the pasteurized bacterium improves metabolism in obese and diabetic mice. Nat. Med. 23, 107-113. https: / / doi.org / 10.1038 / nm.4236 Qiu, Q., Lin, Y., Ma, Y., Li, X., Liang, J., Chen, Z., Liu, K., Huang, Y., Luo, H., Huang, R., Luo, L., 2020. Exploring the Emerging Role of the Gut Microbiota and Tumor Microenvironment in Cancer Immunotherapy. Front. Immunol. 11 , 612202. https: / / doi.org / 10.3389 / fimmu.2020.612202

[0262] Routy, B., Le Chatelier, E., Derosa, L., Duong, C.P.M., Alou, M.T., Daillere, R., Fluckiger, A., Messaoudene, M., Rauber, C., Roberti, M.P., Fidelle, M., Flament, C., Poirier-Colame, V., Opolon, P., Klein, C., Iribarren, K., Mondragon, L., Jacquelot, N., Qu, B., Ferrere, G., Clemenson, C., Mezquita, L., Masip, J.R., Naltet, C., Brosseau, S., Kaderbhai, C., Richard, C., Rizvi, H., Levenez, F., Galleron, N., Quinquis, B., Pons, N., Ryffel, B., Minard-Colin, V., Gonin, P., Soria, J.-C., Deutsch, E., Loriot, Y., Ghiringhelli, F., Zalcman, G., Goldwasser, F., Escudier, B., Hellmann, M.D., Eggermont, A., Raoult, D., Albiges, L., Kroemer, G., Zitvogel, L., 2018. Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science 359, 91-97. https: / / doi.org / 10.1126 / science.aan3706

[0263] Routy, B., Lenehan, J.G., Miller, W.H., Jamal, R., Messaoudene, M., Daisley, B.A., Hes, C., Al, K.F., Martinez-Gili, L., Puncochaf, M., Ernst, S., Logan, D., Belanger, K., Esfahani, K., Richard, C., Ninkov, M., Piccinno, G., Armanini, F., Pinto, F., Krishnamoorthy, M., Figueredo, R., Thebault, P., Takis, P., Magrill, J., Ramsay, L., Derosa, L., Marchesi, J.R., Parvathy, S.N., Elkrief, A., Watson, LR., Lapointe, R., Segata, N., Haeryfar, S.M.M., Mullish, B.H., Silverman, M.S., Burton, J.P., Maleki Vareki, S., 2023. Fecal microbiota transplantation plus anti-PD-1 immunotherapy in advanced melanoma: a phase I trial. Nat. Med. 29, 2121-2132. https: / / doi.org / 10.1038 / s41591-023-02453-x

[0264] Rueda, G.H., Causada-Calo, N., Borojevic, R., Nardelli, A., Pinto-Sanchez, M.I., Constante, M., Libertucci, J., Mohan, V., Langella, P., Loonen, L.M.P., Wells, J.M., Collins, S.M., Sokol, H., Verdu, E.F., Bercik, P., 2024. Oral tryptophan activates duodenal aryl hydrocarbon receptor in healthy subjects: a crossover randomized controlled trial. Am. J. Physiol. Gastrointest. Liver Physiol. 326, G687-G696. https: / / doi.org / 10.1152 / ajpgi.00306.2023

[0265] Sari, Z., Miko, E., Kovacs, T., Janko, L., Csonka, T., Lente, G., Sebo, E., Toth, J., Toth, D., Arkosy, P., Boratko, A., Ujlaki, G., Torok, M., Kovacs, L, Szabo, J., Kiss, B., Mehes, G., Goedert, J. J., Bai, P., 2020. Indolepropionic Acid, a Metabolite of the Microbiome, Has Cytostatic Properties in Breast Cancer by Activating AHR and PXR Receptors and Inducing Oxidative Stress. Cancers 12, 2411. https: / / doi.org / 10.3390 / cancers12092411

[0266] Schuster, S.J., Tam, C.S., Borchmann, P., Worel, N., McGuirk, J.P., Holte, H., Waller, E.K., Jaglowski, S., Bishop, M.R., Damon, L.E., Foley, S.R., Westin, J.R., Fleury, L, Ho, P.J., Mielke, S., Teshima, T., Janakiram, M., Hsu, J.-M., Izutsu, K., Kersten, M.J., Ghosh, M., Wagner- Johnston, N., Kato, K., Corradini, P., Martinez-Prieto, M., Han, X., Tiwari, R., Salles, G., Maziarz, R.T., 2021 . Long-term clinical outcomes of tisagenlecleucel in patients with relapsed or refractory aggressive B-cell lymphomas (JULIET): a multicentre, open-label, single-arm, phase 2 study. Lancet Oncol. 22, 1403-1415. https: / / doi.org / 10.1016 / 51470-2045(21)00375-2

[0267] Segers, A., Vos, W.M. de, 2023. Mode of action of Akkermansia muciniphila in the intestinal dialogue: role of extracellular proteins, metabolites and cell envelope components. Microbiome Res. Rep. 2, N / A-N / A. https: / / doi.org / 10.20517 / mrr.2023.05

[0268] Shi, Y., Zheng, W., Yang, K., Harris, K.G., Ni, K., Xue, L., Lin, W., Chang, E.B., Weichselbaum, R.R., Fu, Y.-X., 2020. Intratumoral accumulation of gut microbiota facilitates CD47-based immunotherapy via STING signaling. J. Exp. Med. 217, e20192282. https: / / doi.Org / 10.1084 / jem.20192282

[0269] Sinha, A.K., Laursen, M.F., Brinck, J.E. et al. Dietary fibre directs microbial tryptophan metabolism via metabolic interactions in the gut microbiota. Nat Microbiol 9, 1964-1978 (2024).

[0270] Smith, M., Dai, A., Ghilardi, G., Amelsberg, K.V., Devlin, S.M., Pajarillo, R., Slingerland, J.B., Beghi, S., Herrera, P.S., Giardina, P., Clurman, A., Dwomoh, E., Armijo, G., Gomes, A.L., Littmann, E.R., Schluter, J., Fontana, E., Taur, Y., Park, J.H., Palomba, M.L., Halton, E., Ruiz, J., Jain, T., Pennisi, M., Afuye, A.O., Perales, M.-A., Freyer, C.W., Garfall, A., Gier, S., Nasta, S., Landsburg, D., Gerson, J., Svoboda, J., Cross, J., Chong, E.A., Giralt, S., Gill, S.L, Riviere, L, Porter, D.L., Schuster, S.J., Sadelain, M., Frey, N., Brentjens, R.J., June, C.H., Pamer, E.G., Peled, J.U., Facciabene, A., van den Brink, M.R.M., Ruella, M., 2022. Gut microbiome correlates of response and toxicity following anti-CD19 CAR T cell therapy. Nat. Med. 28, 713-723. https: / / doi.Org / 10.1038 / S41591-022-01702-9

[0271] Stein-Thoeringer, C.K., Saini, N.Y., Zamir, E., Blumenberg, V., Schubert, M.-L., Mor, U., Fante, M.A., Schmidt, S., Hayase, E., Hayase, T., Rohrbach, R., Chang, C.-C., McDaniel, L., Flores, I., Gaiser, R., Edinger, M., Wolff, D., Heidenreich, M., Strati, P., Nair, R., Chihara, D., Fayad, L.E., Ahmed, S., Iyer, S.P., Steiner, R.E., Jain, P., Nastoupil, L.J., Westin, J., Arora, R., Wang, M.L., Turner, J., Menges, M., Hidalgo-Vargas, M., Reid, K., Dreger, P., Schmitt, A., Muller- Tidow, C., Locke, F.L., Davila, M.L., Champlin, R.E., Flowers, C.R., Shpall, E.J., Poeck, H., Neelapu, S.S., Schmitt, M., Subklewe, M., Jain, M.D., Jenq, R.R., Elinav, E., 2023. A non- antibiotic-disrupted gut microbiome is associated with clinical responses to CD19-CAR-T cell cancer immunotherapy. Nat. Med. 29, 906-916. https: / / doi.org / 10.1038 / s41591-023-02234-6

[0272] Sterner, R.C., Sterner, R.M., 2021. CAR-T cell therapy: current limitations and potential strategies. Blood Cancer J. 11 , 69. https: / / doi.org / 10.1038 / s41408-021-00459-7

[0273] Tintelnot, J., Xu, Y., Lesker, T.R., Schonlein, M., Konczalla, L., Giannou, A.D., Pelczar, P., Kylies, D., Puelles, V.G., Bielecka, A.A., Peschka, M., Cortesi, F., Riecken, K., Jung, M., Amend,

[0274] L., Broring, T.S., Trajkovic-Arsic, M., Siveke, J.T., Renne, T., Zhang, D., Boeck, S., Strowig, T., Uzunoglu, F.G., Gungor, C., Stein, A., Izbicki, J.R., Bokemeyer, C., Sinn, M., Kimmelman, A.C., Huber, S., Gagliani, N., 2023. Microbiota-derived 3-IAA influences chemotherapy efficacy in pancreatic cancer. Nature 615, 168-174. https: / / doi.org / 10.1038 / s41586-023-05728-y van Passel, M.W.J., Kant, R., Zoetendal, E.G., Plugge, C.M., Derrien, M., Malfatti, S.A., Chain, P.S.G., Woyke, T., Palva, A., de Vos, W.M., Smidt, H., 2011 . The genome of Akkermansia muciniphila, a dedicated intestinal mucin degrader, and its use in exploring intestinal metagenomes. PloS One 6, e16876. https: / / doi.org / 10.1371 / journal.pone.0016876

[0275] Vitanza, N.A., Johnson, A.J., Wilson, A.L., Brown, C., Yokoyama, J.K., Kunkele, A., Chang, C.A., Rawlings-Rhea, S., Huang, W., Seidel, K., Albert, C.M., Pinto, N., Gust, J., Finn, L.S., Ojemann, J.G., Wright, J., Orentas, R.J., Baldwin, M., Gardner, R.A., Jensen, M.C., Park, J.R., 2021 . Locoregional infusion of HER2-specific CAR T cells in children and young adults with recurrent or refractory CNS tumors: an interim analysis. Nat. Med. 27, 1544-1552. https: / / doi.org / 10.1038 / s41591-021-01404-8

[0276] Yonekura, S., Terrisse, S., Alves Costa Silva, C., Lafarge, A., lebba, V., Ferrere, G., Goubet, A.-G., Fahrner, J.-E., Lahmar, L, Ueda, K., Mansouri, G., Pizzato, E., Ly, P., Mazzenga,

[0277] M., Thelemaque, C., Fidelle, M., Jaulin, F., Cartry, J., Deloger, M., Aglave, M., Droin, N., Opolon, P., Puget, A., Mann, F., Neunlist, M., Bessard, A., Aymeric, L., Matysiak-Budnik, T., Bosq, J., Hofman, P., Duong, C.P.M., Ugolini, S., Quiniou, V., Bernard, S., Ryffel, B., Kepp, O., Kroemer,

[0278] G., Routy, B., Lordello, L., Bani, M.-A., Segata, N., Yengej, F.Y., Clevers, H., Scoazec, J.-Y., Pasolli, E., Derosa, L., Zitvogel, L., 2022. Cancer Induces a Stress lleopathy Depending on p- Adrenergic Receptors and Promoting Dysbiosis that Contributes to Carcinogenesis. Cancer Discov. 12, 1128-1 151. https: / / doi.org / 10.1158 / 2159-8290.CD-21-0999

[0279] Yoon, S.E., Kang, W., Choi, S., Park, Y., Chalita, M„ Kim, H., Lee, J.H., Hyun, D.-W., Ryu, K.J., Sung, H., Lee, J.-Y., Bae, J.-W., Kim, W.S., Kim, S.J., 2023. The influence of microbial dysbiosis on immunochemotherapy-related efficacy and safety in diffuse large B-cell lymphoma. Blood 141 , 2224-2238. https: / / doi.org / 10.1182 / blood.2022018831

[0280] Zhao, L., Liu, P., Xie, W., Zhang, S., Thieme, S., Zitvogel, L., Kroemer, G., and Kepp, O. (2021). A genotype-phenotype screening system using conditionally immortalized immature dendritic cells. STAR Protoc 2, 100732. 10.1016 / j.xpro.2021 .100732.

[0281] Zhao, L., Liu, P., Mao, M., Zhang, S., Bigenwald, C., Dutertre, C.A., Lehmann, C.H.K., Pan,

[0282] H., Paulhan, N., Amon, L., et al. (2023). BCL2 inhibition reveals a dendritic cell-specific immune checkpoint that controls tumor immunosurveillance. Cancer Discov. 10.1158 / 2159-8290. CD-22- 1338.

Claims

CLAIMS1. A bacterial composition comprising bacteria of the Akkermansia genus, for use in the treatment of a cancer amenable to CAR T cell therapy, wherein said composition is used in combination with CAR T cells targeting said cancer.

2. The bacterial composition of claim 1 , for the use of claim 1 , wherein said composition comprises bacteria of the of Akkermansia massiliensis and / or Akkermansia muciniphila species.

3. The bacterial composition of claim 1 or claim 2, for the use of claim 1 , wherein said composition comprises a purified population of bacteria of the Akkermansia genus.

4. The bacterial composition of any one of claims 1 to 3, for the use of claim 1 , wherein said bacterial composition is formulated so that live bacteria can engraft in the intestinal tract.

5. The bacterial composition of any one of claims 1 to 4, for the use of claim 1 , wherein said bacterial composition further comprises bacteria selected from Lacticaseibacillus paracasei, Lactobacillus johnsonii, Limosilactobacillus reuteri, Lactobacillus gasseri, Limosilactobacillus fermentum, Lentilactobacillus parabuchneri, Limosilactobacillus vaginalis, and Limosilactobacillus oris.

6. The bacterial composition of any one of claims 1 to 5, for the use of claim 1 , wherein said bacterial composition produces and / or releases at least one compound selected from the group consisting of microbial indoles such as indole-3- propionate, indole-3-aldehyde, indole-3-acetate, indole lactate, indoxylsulfate, 5- hydroxyindole, 5-hydroxyindole-3-acetic acid, indole-3-butyrate, indole-3-pyruvate, indole-3-carboxaldehyde, diet-derived indoles and salts therof.

7. The bacterial composition of any one of claims 1 to 6, for the use of claim 1 , wherein said bacterial composition comprises at least one compound selected from the group consisting of microbial indoles such as indole-3-propionate, indole-3- aldehyde, indole-3-acetate, indole lactate, indoxylsulfate, 5-hydroxyindole, 5- hydroxyindole-3-acetic acid, indole-3-butyrate, indole-3-pyruvate, indole-3- carboxaldehyde, diet-derived indoles and salts therof.

8. The bacterial composition of any one of claims 1 to 7, for the use of claim 1 , wherein said bacterial composition increases the cytotoxic activity of the CAR T cells.

9. The bacterial composition of any one of claims 1 to 7, for the use of claim 1 , wherein said bacterial composition decreases or prevents CAR T cell therapy- associated gut dysbiosis.

10. The bacterial composition of any one of claims 1 to 7, for the use of any one of claims 1 , 8 and 9 wherein said composition is used in combination with at least one compound selected from the group consisting of microbial indoles such as indole-3-propionate, indole-3-aldehyde, indole-3-acetate, indole lactate, indoxylsulfate, 5-hydroxyindole, 5-hydroxyindole-3-acetic acid, indole-3-butyrate, indole-3-pyruvate, indole-3-carboxaldehyde, diet-derived indoles and salts therof.11 . The bacterial composition of any one of claims 1 to 7, for the use of any one of claims 1 and 8 to 10, wherein said bacterial composition is not administered before CAR T-cells administration.

12. The bacterial composition of any one of claims 1 to 7, for the use of any one of claims 1 and 8 to 11 , wherein said bacterial composition is first administered the same week, preferably the same day as CAR T-cells.

13. The bacterial composition of any one of claims 1 to 7, for the use of any one of claims 1 and 8 to 12, wherein said bacterial composition is administered several times after CAR T-cell administration.

14. The bacterial composition of any one of claims 1 to 7, for the use of any one of claims 1 and 8 to 13, wherein said cancer is a leukemia (such as acute lymphatic leukemia (ALL), chronic lymphatic leukemia (CLL), acute myeloid leukemia (AML), chronic myeloid leukemia (CML)), a lymphoma (such as diffuse large B-cell lymphoma (DLBCL), Burkitt lymphoma or Hodgkin lymphoma), a multiple myeloma, a plasmocytoma, a brain tumor (such as glioblastoma, neuroblastoma, astrocytoma, meningioma, medulloblastoma and peripheral neuroectodermal tumors), a carcinoma (such as intestine carcinomas, esophageal carcinoma, labial carcinoma, larynx carcinoma, hypopharynx carcinoma, tongue carcinoma, salivary gland carcinoma, gastric carcinoma, adenocarcinoma, medullary thyroid carcinoma, papillary thyroid carcinoma, renal carcinoma, kidney parenchymal carcinoma, ovarian carcinoma, cervix carcinoma, uterine corpus carcinoma, endometrium carcinoma, chorion carcinoma, pancreatic carcinoma, prostate carcinoma, testis carcinoma, breast carcinoma, urinary carcinoma, hepatocellular carcinoma, gall bladder carcinoma, bronchial carcinoma, small cell lung carcinoma, non-small cell lung carcinoma), a sarcoma (such as osteosarcoma, chondrosarcoma, myosarcoma, liposarcoma, fibrosarcoma and Ewing sarcoma) or a melanoma.

15. A combination of CAR T cells and bacteria of the Akkermansia genus, for use in the treatment of a cancer amenable to CAR T cell therapy.

16. The combination of claim 15, for the use of claim 15, wherein the bacteria of the Akkermansia genus are in a composition as recited in any one of claims17. The combination of claim 15 or 16, for the use of claim 15, wherein the CAR T-cells are autologous or allogenic T-cells.

18. The combination of any one of claims 15 to 17, for the use of claim 15, wherein the CAR T-cells have a CD28 costimulatory domain or a 4-1 BB costimulatory domain.

19. The combination of any one of claims 15 to 18, for the use of claim 15, wherein the CAR T-cells are specific for CD19 or BCMA.

20. The combination of any one of claims 15 to 19, for the use of claim 15, wherein CAR T-cells are administered before administration of the bacteria.

21. The combination of any one of claims 15 to 19, for the use of claim 15, wherein CAR T-cells are administered after a first administration of said bacteria.

22. The combination of any one of claims 15 to 21 , for the use of any one of claims 15, 20 and 21 , wherein bacteria are administered several times, e.g., daily, every other day or weekly to the patient after CAR T-cell administration.

23. The combination of any one of claims 15 to 22, for the use of any one of claims 15 to 22, for treating a patient also receiving at least one additional anticancer treatment.

24. The combination of any one of claims 15 to 22, for the use of any one of claims 15 to 23, for treating a patient also receiving at least one compound selected from the group consisting of microbial indoles such as indole-3-propionate, indole-3-aldehyde, indole-3-acetate, indole lactate, indoxylsulfate, 5-hydroxyindole, 5- hydroxyindole-3-acetic acid, indole-3-butyrate, indole-3-pyruvate, indole-3- carboxaldehyde, diet-derived indoles and salts therof.

25. A CAR T-cell for use in the treatment of a cancer amenable to CAR T cell therapy, wherein said CAR T-cell is used in combination with a bacterial composition as recited in any one of claims 1 to 7.

26. The CAR T-cell for use according to claim 25, wherein said CAR T-cell is as recited in any one of claims 17-19.

27. The bacterial composition for use of any one of claims 1-14, the combination for use of any one of claims 15-24 or the CAR T-cell for use of claim 25 or claim 26, for treating a patient having a cancer, wherein Akkermansia species are lacking in said patient’s gut microbiota.

28. A theranostic method for assessing whether an individual having a cancer amenable to CAR T cell therapy needs a microbiota centered intervention (MCI) to respond to a CAR T cell therapy, comprising assessing the presence, in a sample from said individual, of bacteria of the Akkermansia genus, wherein the absence of bacteria of the Akkermansia genus is indicative that the patient needs a MCI with abacterial composition as recited in any one of claim 1 to 7 to better respond to the CAR T cell therapy.

29. A method of treating a cancer amenable to CAR T-cell therapy, comprising administering with a bacterial composition as recited in any one of claims 1 to 7, CAR T-cells and optionally at least one other anticancer treatment.

Images