"Toxic" microbiome and progression of Chronic Kidney Disease. A prospective study – CKD_Microbiome

The CKD_Microbiome project is an ancillary study to the national prospective CKD-REIN study, which included patients with CKD stages 2 to 5, nationally representative, recruited from 40 nephrology centers across France. A total of 240 patients provided stool samples, along with blood and urine samples (T0). The patients also completed general questionnaires and a food frequency questionnaire. Among them, 103 patients provided a second set of samples three years later (T1).

From the stool samples, bacterial DNA was extracted and sequenced using shotgun metagenomic sequencing. The sequences were compared to the "Integrated Gut Catalogue 2 (IGC2)" and the Human Oral Microbiome Catalogue. Metagenomic species pangenomes (MSPs) were extracted from these catalogs.

From the blood samples, 10 uremic toxins and 3 precursors were measured at T0 and T1 and quantified via UHPLC-MS/MS.

Statistical analyses were then conducted to address the defined objectives. The comparison of CKD patients with healthy individuals was carried out using the Milieu Intérieur cohort. Diversity indices (alpha and beta), as well as the Gut Microbiome Health Index (GMHI), were compared in the first three objectives, along with the microbiota composition. CKD severity was defined as: moderate (n=130, eGFR ≥ 30 mL/min/1.73 m²) and severe (n=110, eGFR < 30 mL/min/1.73 m²). CKD progression was defined based on a decline of more than 30%, initiation of renal replacement therapy, or death: 68 fast progressors and 172 slow progressors were identified. Cliff’s delta analyses were performed to determine the species most discriminative in each group. Integrated multi-omics analyses were performed using the "Data Integration Analysis for Biomarker discovery using Latent cOmponents (DIABLO)" model.

Regarding the experimental part, 10 CKD patients from the CKD_Microbiome study and 10 healthy volunteers were selected as donors and provided fresh stool samples. The stools were pooled and used for fecal microbiota transplantation (FMT) as a mixed fecal suspension, administered via oral gavage to each antibiotic-treated mouse for three consecutive days. The gavage process was repeated biweekly. The mice were sacrificed two weeks after the start of the FMT. Stool, blood, and tissue samples were collected, urea concentrations were measured, and bacterial 16S rRNA sequencing in the cecum was performed to analyze the intestinal microbiota.

Our results show that the gut microbiota evolves progressively during the course of chronic kidney disease (CKD):

The Gut Microbiome Health Index (GHMI) tends to be lower in CKD patients compared to healthy controls. We identified around 10% of the total microbiome as being significantly different between these two populations. The species enriched in CKD patients mainly belonged to the Enterocloster and Hungatella genera (Lachnospiraceae family), with a tendency for a higher abundance of uremic toxin (UT)-producing species in CKD patients.

Regarding disease severity, although diversity indices and GHMI did not differ significantly between groups, the microbiota composition differed, with a marked increase in toxin-producing species in patients with severe CKD compared to those with moderate CKD. Correlation analyses showed that these species enriched in severe patients were strongly correlated with blood concentrations of UTs.

Through integrated multi-omics models, we identified markers that differentiate patients according to CKD severity:

The species enriched in the severe group included D. fairfieldensis msp_0295, Bacteroides clarus msp_0063, and Blautia obeum msp_0722, which harbor genes essential for the production of uremic toxins. Additionally, severe CKD patients were characterized by specific dietary habits, such as relatively high consumption of alcohol and hot drinks, as well as clinical characteristics including inflammation and elevated blood levels of several UTs.

We also identified markers that differentiate slow progressors from fast progressors:

Several UT-producing species, such as D. bacterium_MSP 226 and Alistipes putredinis_MSP 0090, were significantly more abundant in fast progressors. Interestingly, F. prausnitzii 4_msp0389 was associated with slower CKD progression. Higher blood levels of uremic toxins and the presence of albuminuria were also correlated with faster disease progression.

After 3 years, the gut microbiota composition was altered, with a significant loss in bacterial richness and an increase in the number of UT-producing species, which was accompanied by a rise in blood UT concentrations. However, the ratio of UT-producing species to total species decreased in patients who consumed probiotics, increased their vegetable intake, or reduced their protein consumption.

Finally, the experimental part demonstrated that fecal microbiota transplantation (FMT) from CKD patients into mice induced early kidney fibrosis and an increase in UTs.

Our project offers new insights into the link between the gut microbiome, uremic toxins, and diet, Our study underscores the critical role of the microbiome and its metabolites in chronic kidney disease (CKD), suggesting that the microbiome represents a potential therapeutic target for slowing CKD progression. Specifically, we identified certain bacterial species involved in the production of uremic toxins (UTs), which could be targeted in future clinical trials. Reducing these species may offer a promising strategy for mitigating CKD progression. Additionally, our findings indicate that dietary modifications—such as reducing protein intake, increasing vegetable consumption, or taking probiotics—can influence gut microbiota composition. This opens new avenues for clinical trials, where diet changes or probiotic supplementation could be explored as strategies to improve gut microbiome health, limit UT production, and slow the progression of CKD.

Chronic kidney disease (CKD) affects 10 to 15% of the adult population and is associated with high risks of progression to end-stage kidney disease (ESKD), cardiovascular disease and mortality. Few treatments are currently available to slow CKD progression and reduce its complications. The composition of gut microbiota in CKD has attracted particular interest, because of its role in colon-derived uremic solutes. Food protein fermentation by the gut microbiota leads to the generation of waste metabolites, known as uremic toxins that are normally cleared by the kidneys. These toxins promote CKD progression, insulin resistance and inflammation. Nevertheless, the relations between features of the gut microbiota, uremic toxin production and CKD progression have not been explored prospectively in humans. The overall goal of the CKD_Microbiome project is to investigate the role and mechanisms of gut microbiota-related toxicity in the progression of CKD and the potential impact of diet on that relation. It includes an observational and an experimental component.
The observational component of the project is an ancillary study based on the on-going CKD-REIN study, a large prospective cohort that included 3033 adult patients with non-dialysis CKD stages 3-5 (estimated glomerular filtration rate [eGFR] < 60 mL/min/1.73m²), from 40 nationally representative nephrology outpatient facilities, recruited from 01/2014 to 03/2016. Patient follow-up is on-going and will last 5 years, with an annual clinical and biological data collection. The CKD_Microbiome study is based on a subsample of 240 patients with simultaneous blood and stool samples collected twice, 3 years apart, at the 2-year (completed) and 5-year (to be completed) follow-up examinations, together with data on food intake. Highly standardized protocols will be used throughout to extract total DNA from stools, shotgun sequence the DNA, and deduce the presence and abundance of each microbial species and functionalities for all samples. In parallel, protein-bound and unbound uremic toxins will be measured in blood. Availability of these unique samples and the expertise in metagenomic profiling of INRA-MetaGenoPolis will allow us to assess both cross-sectional and longitudinal associations of gut microbiome composition and functionalities with kidney function and its decline over 3 years, and to test the hypothesis (primary objective) that some microbiota compositions worsen CKD progression through production of uremic toxin precursors. This observational component will also evaluate the potential modulating effect of dietary patterns on these associations (secondary objective).
In the experimental component of the project, we will investigate the causality of the relation between microbial species, identified by metagenomics, and uremic toxin production. We will study the impact of faecal microbiota transplantation (FMT) from CKD patients with high vs low levels of uremic toxins, into antibiotic-treated mice and investigate their uremic toxin production.
The CKD_Microbiome consortium is composed of 5 complementary teams providing the expertise required in metagenomics and bioinformatics (SD Ehrlich, INRA U1367), uremic toxin analyses (JC Alvarez, ZA Massy, UVSQ), epidemiology and biostatistics (B Stengel, INSERM U1018, S Wagner, CIC1433, SD Ehrlich, INRA U1367), nutrition (S Wagner, CIC1433, D Fouque, INSERM U1060), nephrology and basic sciences (L Koppe, D Fouque, INSERM U1060) as well as operational capacity, which form the pillars of the project.
The CKD_Microbiome project will provide new conceptual and mechanistic leads, based on 'toxic' microbiome tests performed in animal models and will improve our understanding of gut microbiota-related toxicity in CKD. This may identify new prognostic biomarkers of CKD progression in humans and provide the basic data to design interventions by food modulation, to change the microbiota, in order to diminish its toxicity.