Skip to main content
NCBI home page
The Veterinary Quarterly logoLink to The Veterinary Quarterly
. 2025 Jan 2;45(1):1–15. doi: 10.1080/01652176.2024.2447601

Metabolomics reveals alterations in gut-derived uremic toxins and tryptophan metabolism in feline chronic kidney disease

Laurens Van Mulders a,b,✉, Ellen Vanden Broecke a,b, Ellen De Paepe a, Femke Mortier b, Lynn Vanhaecke a,c, Sylvie Daminet b
PMCID: PMC11703532  PMID: 39745207

Abstract

Chronic Kidney Disease (CKD) is one of the most common conditions affecting felines, yet the metabolic alterations underlying its pathophysiology remain poorly understood, hindering progress in identifying biomarkers and therapeutic targets. This study aimed to provide a comprehensive view of metabolic changes in feline CKD across conserved biochemical pathways and evaluate their progression throughout the disease continuum. Using a multi-biomatrix high-throughput metabolomics approach, serum and urine samples from CKD-affected cats (n = 94) and healthy controls (n = 84) were analyzed with ultra-high-performance liquid chromatography-high-resolution mass spectrometry. Significant disruptions were detected in tryptophan (indole, kynurenine, serotonin), tyrosine, and carnitine metabolism, as well as in the urea cycle. Circulating gut-derived uremic toxins, including indoxyl-sulfate, p-cresyl-sulfate, and trimethylamine-N-oxide, were markedly increased, primarily due to impaired renal excretion. However, alternative mechanisms, such as enhanced bacterial formation from dietary precursors like tryptophan, tyrosine, carnitine, and betaine, could not be ruled out. Overall, the findings suggest that metabolic disturbances in feline CKD are largely driven by the accumulation of gut-derived uremic toxins derived from precursors highly abundant in the feline diet. These insights may link the strict carnivorous nature of felines to CKD pathophysiology and highlight potential avenues for studying preventive or therapeutic interventions.

Keywords: Feline chronic kidney disease, metabolomics, pathophysiology, fundamental nephrology, biochemical pathway analysis, gut microbiota derived uremic toxins

Graphical abstract

graphic file with name TVEQ_A_2447601_UF0001_C.jpg

Open in a new tab

Introduction

Chronic kidney disease (CKD) is a widespread disease that affects >10% of the general human population worldwide (Kovesdy 2022). Surpassing the human scenario, CKD is one of the most common chronic conditions in felines, affecting at least 30% of cats aged over 12 years (Marino et al. 2014) and accounting for 14% of domesticated cat mortality (O’Neill et al. 2015). Feline CKD shows striking similarities with human CKD in terms of clinical presentation, diagnosis, and disease progression (Ghys et al. 2015; Sparkes et al. 2016; Ruiz-Ortega et al. 2020).

In humans, tubulointerstitial fibrosis represents the ultimate sequala of a diverse spectrum of underlying renal disorders (Kriz and Lehir 2005). Remarkably, tubulointerstitial fibrosis is already present in the early stages of feline CKD (Chakrabarti et al. 2013), consistent with the hypothesis that chronic overload of the metabolic active renal tubular cells is a critical initial driver for subsequent oxidative stress and cellular damage (Brown et al. 2016; Gyurászová et al. 2020). Although the etiopathogenesis is likely multifactorial, potential underlying causes of feline CKD and more specifically tubular inflammation, may be related to high protein intake (Stenvinkel et al. 2018). This theory is indirectly supported by the observation that not only the domestic cat but most, if not all felids (i.e. tigers, cheetahs, …), all obligate carnivores, are extremely prone to CKD (Newkirk et al. 2011). Furthermore, the distinct involvement of protein-derived uremic toxins in tubular injury (Vanholder et al. 2016) and the ensuing heightened risk of advancing to end-stage renal disease with increased daily protein intake has been clearly established in both human (Metzger et al. 2018) and feline populations (Elliott et al. 2000). Therefore, feline CKD can primarily be regarded as an evolutionary model of naturally occurring CKD, characterized by intrinsic tubulointerstitial dysfunction as a hallmark. Despite its biological and clinical significance, the metabolic alterations underpinning CKD in felines remain incompletely understood. Improved comprehension of alterations within conserved biochemical pathways may yield substantial insights into fundamental mechanisms that underlie feline CKD and may form the first step in the discovery of disease biomarkers and specific therapeutic targets.

In recent years, metabolomics has emerged as a powerful technology for unraveling the complex web of metabolic alterations associated with various disease states, including CKD (Hocher and Adamski 2017). Advances in analytical techniques have already shown potential for the investigation of CKD pathophysiology in felines (Nealon et al. 2024). Serum and urine represent ideal biomatrices that facilitate in-depth comprehension of renal function. While the serum metabolome offers a holistic snapshot of the endogenous biological processes within a system, alterations observed in the urinary metabolome may directly reflect the excretory function of the deteriorating kidney (Zhang et al. 2014).

Through mapping changes in the metabolic fingerprint of diseased versus healthy cats, we expect to reveal which pathways are markedly disrupted in the feline CKD state. Metabolomic fingerprinting is a qualitative (untargeted) approach that ensures maximal molecular coverage (Boelaert et al. 2014; Hocher and Adamski 2017). In parallel, by targeted profiling of a priori selected metabolites and thus pathways, we may capture comprehensive insights into specific underlying pathophysiologic alterations. Metabolic profiling methods have been developed and optimized to analyze specific metabolites of interest, providing higher sensitivity and selectivity than untargeted approaches (Johnson et al. 2016; Hocher and Adamski 2017). In the specific context of naturally occurring feline CKD, this signifies the detection of specific subsets of protein-derived uremic toxins and associated pathways potentially implicated in the development of tubulointerstitial fibrosis such as the tryptophan, carnitine and tyrosine metabolism (Rhee et al. 2013; Kimura et al. 2016; Hu et al. 2018; Chen et al. 2020; Cheng et al. 2020).

Materials and methods

Ethics statement

The study protocol was approved by the ethical committee of the faculties of Faculties of Veterinary Medicine and Bioscience Engineering, Ghent University, Belgium (Approval no. EC2020-081). All feline specimens in this study were obtained prospectively from privately owned domestic cats through a proactive recruitment process. Written informed consent was obtained prior to inclusion in the study.

Study design

The study population comprised two general cohorts: (I) A cohorts of cats diagnosed with CKD (II) A control cohort of healthy domestic senior cats (over the age of 10 years) (Figure 1).

Figure 1.

Figure 1.

Open in a new tab

The study design consists of a comparison between the serum and urine metabolome of a group of cats diagnosed with chronic kidney disease (CKD; n = 94) and a healthy control (n = 84) group. Serum and urine samples were collected, followed by chemical extraction according to feline-specific optimized and validated methods. Samples from both matrices were analyzed by Ultra-High-performance liquid chromatography coupled to high-Resolution Mass spectrometry (UHPLC-HRMS). Generated data was further analyzed using an untargeted approach to discover which pathways are altered in feline CKD. Specific metabolites in these pathways were identified in a targeted manner and statistically compared between the two health states, based on this the biological relevance of the disrupted metabolic pathways was further elucidated.

The source population consisted of 289 cats (i.e. either presented as healthy or diagnosed with CKD) that were actively recruited via the research facility of the Small Animal Clinic at the Faculty of Veterinary Medicine, Ghent University, between July 2021 and December 2022.

Owners were asked to complete a standardized questionnaire designed to assess the overall health status of the cats. The questionnaire included a detailed anamnesis covering past and present clinical symptoms, medical history, dietary habits, and any prescribed medications. Furthermore, physical examination, blood pressure measurement, complete blood count, serum biochemistry profiles (including total thyroxine [tT4], symmetric dimethylarginine [SDMA], and electrolytes), infectious serological testing (specifically for feline leukemia virus and feline immunodeficiency virus [FIV]), and complete urinalysis, including urinary protein:creatinine ratios (UPC) and bacterial culture were performed, adhering to established protocols (Paepe et al. 2013; Reppas and Foster, 2016a, 2016b; Acierno et al. 2018; Little et al. 2020).

Confirmation of CKD diagnosis was based on a compatible medical history, physical examination, and laboratory findings. Key laboratory criteria included persistently elevated serum creatinine (sCr) >1.83 mg/dl (>161.8 µmol/l) in combination with a decreased urinary specific gravity (<1.035), according to international guidelines (Sparkes et al. 2016). Cats suspected of acute kidney injury (AKI) were not included. The sCr cutoff in this study was based on a previously established lab-specific reference interval comprising 130 healthy cats (Ghys et al. 2015). Subsequently, the CKD population was further categorized into three distinct groups: (1) stage 2 CKD (n = 59), (2) stage 3 CKD (n = 29), and (3) stage 4 CKD (n = 6). The classification of feline CKD stages was determined based on sCr values, employing a four-tier staging system as defined by the International Renal Interest Society (IRIS) (IRIS 2023). Cats showing evidence of any other significant non-CKD related health abnormality, were not included.

The healthy control cohort included only animals that showed no significant abnormalities based on the questionnaire in combination with the listed examinations. Reasons for exclusion were recorded. Additionally, cats with potential renal function abnormalities (sCr >1.83 mg/dl, SDMA >14 ug/dl, UPC >0.4) were not eligible for inclusion in the control cohort.

A subpopulation of animals (validation cohort) was selected to address confounding factors and to further validate the observed findings in the specific context of the disease state. To achieve this, specific restrictions were applied to the CKD cohort, excluding cats if they were undergoing medical therapy (i.e. telmisartan, benazepril, amlodipine, mirtazapine, phosphate binders, and lactulose) and/or receiving specific renal dietary formulae. Inclusion within the healthy subgroup necessitated that the cats maintained a state of health during a 6-month longitudinal follow-up, in accordance with the aforementioned criteria. Moreover, healthy animals in our validation subgroup that showed an increase in sCr or SDMA by more than 25% or beyond the reference interval, persistent suboptimal concentrated urine (USG < 1.035), and/or developed proteinuria (UPC > 0.4) were excluded. Subsequently, a similar number of healthy cats, compared to the CKD subgroup, were randomly selected using simple random sampling following the longitudinal follow-up.

Sample collection

Blood samples were acquired through venipuncture, specifically from the jugular or cephalic veins (23-gauge needle) and collected in plain serum tubes (VACUETTE® Serum Tubes with Gel). The samples were allowed to clot for maximum 15 min at room temperature, followed by centrifugation (5 min; 1931xg) at 4 °C. Within a narrow time window (15 min), sterile urine samples were obtained through ultrasound-guided cystocentesis, utilizing a 22-gauge needle and transferred to plain tubes. The samples were then centrifuged (3 min; 355xg) at room temperature. After centrifugation, both the serum and urine supernatants were aliquoted into Eppendorf® tubes and preserved at a temperature of −80 °C until further analysis.

Reagents and chemicals

All 52 analytical standards (Table S1) and internal standards (ISTDs) (Table S2) were purchased from reputable sources, including Sigma-Aldrich (St. Louis, Missouri, USA), ICN Biomedicals Inc. (Ohio, USA), TLC Pharmchem (Vaughan, Ontario, Canada), and Cambridge Isotope Laboratories Inc. (Tewksbury, Massachusetts, USA).

All solvents utilized in this experiment, i.e. acetonitrile, acetone and methanol were of LC–MS grade for extraction purposes and sourced from Fisher Scientific (Loughborough, UK) and VWR International (Merck, Darmstadt, Germany). Ultrapure water was obtained through the utilization of a purified-water system, Arium® 611UV (Sartorius, Göttingen, Germany).

UHPLC-HRMS analysis

For both biomatrices (urine, serum), the same UHPLC and HRMS settings were applied. For chromatographic separation, a Vanquish Horizon UHPLC system (Thermo Fisher Scientific, San José, CA, USA) was used with an Acquity UPLC HSS C18 column T3 1.8 µm (150 mm x 2.1 mm) (Waters, Manchester, UK) at a column temperature of 45 °C. An injection volume of 10 µL at a maximum injection time of 150 ms was applied. Ultrapure water and acetonitrile, both acidified with 0.1% formic acid, were used as analytical solvents. A constant flow rate of 0.4 mL.min−1 was maintained throughout the analysis. An Exploris 120 stand-alone bench top orbitrap High Resolution Mass Spectrometer (Thermo Fisher Scientific, San José, CA, USA), equipped with a heated electrospray ionization source (HESI II), operating in polarity switching full scan mode at a scan range of 53–800 Da and with a mass resolution of 120,000 full width at half maximum (1 Hz), was used. Prior to mass spectrometric detection, initial instrument calibration was achieved by infusing ready-to-use calibration mixtures (Thermo Fisher Scientific, San José, CA, USA) (Vangeenderhuysen et al. 2023).

For each biomatrix, Quality Control samples (QCs) were generated by pooling sample extracts specific to the respective matrix. To adjust for instrumental fluctuations, QCs were run in duplicate after every 10 samples. Furthermore, at the start of the analyses, samples from the QC pool were injected to facilitate column conditioning, tailored to the specific biomatrix under investigation. An analytical standard mixture containing all a priori selected analytical standards (n = 52) (i.e. associated with the urea acid cycle, tryptophan, tyrosine, carnitine and purine metabolism) was injected before, in the middle and at the end of the analyses. This procedure enabled the assessment of the operational parameters of the device and ensured accurate annotation of the designated metabolites.

Urine and serum extraction protocols

We designed, optimized, and validated specific metabolomics extraction protocols for feline serum and urine samples targeting selected metabolic pathways (Vanden Broecke et al. 2024). Analytical optimization was conducted using a fractional factorial design (FFD) within JMP™ 16 software. Subsequently, validation of the analytical performance pertaining to the feline serum and urine metabolome was undertaken. This procedure adhered to the methodology outlined by De Paepe et al. (De Paepe et al. 2018), attuned to emphasize targeted compounds within the chosen metabolic pathways (Supplementary information: Extraction Protocols).

Untargeted data processing

In the context of untargeted data processing, the software package Compound Discoverer™ 3.3 (Thermo Fisher Scientific, San José, CA, USA) was employed to achieve automated peak extraction, peak alignment, deconvolution, and noise removal. Key parameters utilized for feature extraction included a minimum peak intensity of 500,000 arbitrary units, a retention time width of 0.3 min, and a mass window of 6 ppm. The primary feature descriptors encompassed retention time, m/z-value, and signal intensity.

Outputs of the serum untargeted data preprocessing were subjected to multivariate statistical analysis using SIMCA™ 17.0 software (Umetrics AB, Umea, Sweden). First, additional data cleaning strategies were applied including background subtraction and QC normalization (to correct for intra-analytical fluctuations). Subsequently, the data was pareto-scaled and log10 transformed to standardize the range of signal intensities and induce a normal data distribution. Principal Component Analysis (PCA-X) was conducted for data exploration, enabling assessment of natural sample patterns according to their inherent metabolic fingerprints and identification of outliers. Thereafter, Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) was employed to establish predictive models and define which fraction of the metabolome is responsible for the differentiation according to health status. OPLS-DA models were validated to ensure robustness, significance, and model fit by assessing various quality parameters, including R2(Y), Q2, permutation testing (n = 100), and cross-validated ANOVA (CV-ANOVA) (p-value < 0.05).

The discriminative quality of all features was investigated based on variance importance in projection (VIP) scores >1, S-plot correlations |p(corr)|>0.5, jackknifed confidence intervals not across zero, and p-values < 0.05. All features were ranked according to their VIP-scores (p-value < 0.05) and subjected to pathway analysis and network mapping in MetaboAnalyst 5.0 (Xia Lab, McGill University, Quebec, Canada) using the mummichog algorithm (Pang et al. 2021).

Targeted data processing

Targeted data processing was executed using Xcalibur 3.0™ (Thermo Fisher Scientific, San José, CA, USA). In this process, the identification of compounds relied on their m/z-value, carbon-isotope (C-isotope) profile, MS/MS fragmentation pattern and retention time in relation to that of the internal standard (Vanden Broecke et al. 2024). Compound abundance was determined based on the total area under the curve. Subsequently, the data underwent QC normalization. In the case of urine data, an additional normalization step was applied, where the data was adjusted by urinary creatinine levels. This adjustment was made to ascertain the kidney’s actual excretion capacity, independent of variations in urinary dilution, given that total daily urinary creatinine excretion remains relatively constant across a broad spectrum of glomerular filtration rate values (Ortiz et al. 2011; Ryan et al. 2011).

Group comparisons were conducted using QC normalized compound abundance, without absolute quantification of metabolites. To assess statistical differences between groups for the identified metabolites, we employed R (version 4.2.1), SPSS Statistics 28, and MetaboAnalyst 5.0. Initially, normality was evaluated using the Shapiro-Wilk test, and the data distribution was visualized through histograms and QQ-plots. For two-group comparisons, the independent Mann-Whitney U test was employed. To address the issue of multiple comparisons, a post-hoc correction procedure, specifically the Benjamini-Hochberg method, was applied. Heat maps displaying the abundance levels of all chosen serum metabolites were generated. Hierarchical clustering was conducted using the Pearson correlation coefficient and Ward linkage method. To assess the behavior of metabolites in the progression of CKD, Spearman correlation analysis was performed assessing the relationship with sCr, further clarified by visualization statistics.

Results

Cohort selection and exclusion of CKD and healthy cats

Out of our source population of 289 recruited cats, 143 were presented by their owners as CKD patients and 146 as healthy controls. Out of 143 cats initially presented as CKD patients, 49 were excluded for the following reasons: 17 cats did not have a confirmed CKD diagnosis, 8 had stage 1 CKD, 11 had concurrent hyperthyroidism, 3 were FIV-positive, 6 had other concurrent illnesses (e.g. chronic enteropathy, pancreatitis, elevated liver enzymes, hyperaldosteronism, and polycystic kidney disease), 1 had suspected AKI on CKD, and in 3 cats, there was insufficient sample available. This resulted in a CKD cohort of 94 cats that met the inclusion criteria. The percentages of cats prescribed renal dietary formulae and medications are presented in Table S3. After implementing specific restrictions on renal dietary therapy and prescribed medications, a CKD (validation) subgroup of 31 cats remained. Out of 146 cats initially presented as healthy, 62 were excluded for the following reasons: 14 had increased SDMA and/or sCr levels, 3 had a UPC > 0.4, 25 were diagnosed with hyperthyroidism, 8 were FIV-positive, 10 had other concurrent illnesses (e.g. chronic enteropathy, elevated liver enzymes, unexplained weight loss, chronic rhinitis/conjunctivitis, bacterial cystitis, hypercalcemia), and 2 had insufficient sample available. This resulted in a healthy control cohort of 84 cats that met the inclusion criteria. Thirty-four healthy cats, which remained healthy after a follow-up period of six months, were randomly selected to form the control (validation) subgroup.

Differences in clinical and laboratory parameters between the healthy and CKD cohort

The median age of cats in both cohorts was 13 years (Table S4). The CKD cohort (n = 94) included 43 males and 51 females, while the healthy cohort (n = 84) consisted of 42 males and 42 females, all of which were spayed/neutered regardless of their group. The CKD cohort had a lower body weight compared to the healthy cohort (p < 0.001), and there were no significant differences in systolic blood pressure between the two groups (p = 209). Renal values, including sCr (p < 0.001), SDMA (p < 0.001), and urea (p < 0.001), were significantly higher in the CKD cohort compared to the healthy cohort. Both groups showed similar albumin levels (p = 0.946). Also, potassium levels were comparable (p = 0.557). However, the CKD cohort showed significantly higher phosphate (p < 0.001) and total calcium levels (p < 0.001). Additionally, USG was significantly lower in the CKD cohort (p < 0.001), and the UPC was mildly higher (p = 0.044) (Table S4).

Metabolic fingerprinting uncovered the most prominently CKD impacted pathways

Untargeted data processing generated 7220 compounds. Robust analytical performance was noted, with well-defined QC clustering in PCA-X modeling (Figure S1). Unsupervised PCA-X score plots further demonstrated distinct clustering patterns based on inherent metabolic fingerprints, with clear trends related to CKD stage progression observed along the first principal component axis (Figure 2A). The supervised OPLS-DA model demonstrated a distinct separation between the two groups (CKD n = 94 and healthy n = 84), indicating strong discriminative power (Figure 2B). Building upon this robust model, a subsequent analysis was undertaken to precisely identify which portion of the metabolome plays a pivotal role in distinguishing between the two health states. A total of 183 compounds were retained as being significantly associated with CKD. Mapping the retained metabolites to their respective pathways revealed significant disruptions in the urea acid cycle, tryptophan metabolism, and tyrosine metabolism (Figure 2C).

Figure 2.

Figure 2.

Open in a new tab

A: PCA-X score plot visualizing natural variance in the serum metabolome among healthy controls and different CKD stages: Healthy (green, n = 84), stage 2 CKD (pink, n = 59), stage 3 CKD (red, n = 29), and stage 4 CKD (dark red, n = 6). Figure 2B: Score plot of the supervised OPLS-DA model showing distinct separation between CKD (dark red, n = 94) and healthy (green, n = 84) states according to the serum metabolome. The model exhibits excellent performance characteristics (R2Y:0.96, Q2:0.71). Figure 2C: Visual representation of the most dysregulated metabolic pathways in feline CKD, based on metabolic fingerprinting, using the mummichog algorithm. The X-axis represents the enrichment factor and is derived from the number of hits within a particular metabolic pathway (increasing range: small to large dots). The Y-axis represents the significance level (-log10(p)) of the perturbed pathway.

Targeted metabolite analysis confirmed distinct CKD-related metabolic changes

The untargeted hypothesis generating approach substantiated the validity of the chosen metabolites (i.e. belonging to the urea acid cycle, tryptophan, and tyrosine metabolism) for deeper exploration of the pathways impacted by CKD in the targeted experiments. In addition to these included pathways, carnitine and purine metabolism were also prioritized, based on in-depth review of the scientific literature (Rhee et al. 2013; Boelaert et al. 2014; Zhang et al. 2014; Johnson et al. 2016; Kimura et al. 2016; Hocher and Adamski 2017; Hu et al. 2018; Chen et al. 2020; Cheng et al. 2020; Nealon et al. 2024).

Visual representation of the comparative abundance of the selected metabolites revealed good natural clustering between the healthy controls and the CKD group (Figure 3). Furthermore, at the individual metabolite level, our findings also showed distinct clustering patterns among the up- or downregulated compounds originating from various metabolic pathways (Figure 3).

Figure 3.

Figure 3.

Open in a new tab

Heat map of all a priori selected serum metabolites, representing natural clustering between healthy cats (n = 84) and cats diagnosed with CKD (n = 94) at the group level and inherent clustering of up- or downregulated or unaltered metabolites at the compound level.

In the subpopulation (after applied restrictions) of healthy cats (n = 34) and cats diagnosed with CKD (n = 31), statistical two-group comparisons revealed specific up- or downregulation of metabolites within the selected pathways (Table S5). To elucidate the dynamic behavior of metabolites across the spectrum of CKD progression, Spearman correlation analysis enabled the evaluation of associations between each specific compound and sCr levels, utilizing serum (n = 178) and urine (n = 173) data (Table S6), as elaborated below.

Tryptophan metabolism experienced significant alterations

Tryptophan showed a notable and statistically significant reduction in serum and urinary abundance within the CKD group (p < 0.001) with a clear downward trend in serum as CKD progresses. In contrast, metabolites from all three tryptophan-derived pathways demonstrated substantial elevations in serum (Figure 4), resulting in a completely reversed tryptophan/catabolite ratio precisely at the onset of stage 2 CKD (Figure S2). Moreover, the intrinsic associations of the indole, kynurenine, and serotonin pathways showed a significant positive trend with further progression of CKD (Table S6 and Figure S2).

Figure 4.

Figure 4.

Open in a new tab

Box plot representation of the selected serum and urine metabolites from the tryptophan metabolism comparing healthy controls (green; n = 84) with specific CKD stages. Significance levels (*<0.05; **<0.001) for the comparison healthy vs. CKD (all stages) are provided. Clear up- or downward trends in the indole, serotonin and kynurenine pathways are portrayed in accordance with CKD progression from stage 2 (pink; n = 59) to stage 3 (red; n = 29) and 4 (dark red; n = 6). Urine metabolites are normalized to urine creatinine.

Specifically, within the indole pathway, both the bacterial tryptophan catabolite indole and its uremic toxin derivative, indoxyl-sulfate, showed significant upregulation in serum (p < 0.001) (Figure 4) but only mildly increased urinary excretion (indole: p = 0.031; indoxyl-sulfate: p = 0.002) (Figure 4). Urinary excretion of indole increased with further deteriorating renal function but remained unaltered over the progression spectrum for indoxyl-sulfate (Table S6). Serum indole-3-propionic acid, indole-3-acetic acid and its intermediate compound tryptamine remained unchanged between the two health states, although urinary excretion of indole-3-acetic acid was slightly reduced (p = 0.020).

Furthermore, while the abundance of serum kynurenine remained unchanged, there were significant increases in kynurenic acid, 3-hydroxykynurenine, and its derivative xanthurenic acid (p = 0.017) (Figure 4). Urinary kynurenine (p < 0.001) and kynurenic acid (p = 0.001) were significantly elevated, whilst 3-hydroxykynurenine (p < 0.001) was strongly decreased (Figure 4). Additionally, an upregulation of the catabolite quinolinic acid (p = 0.005) in serum was observed in combination with increased urinary excretion (p < 0.001), whereas the other catabolite, picolinic acid, did not show significant alterations in serum nor urine. Only excretion of 3-hydroxy-kynurenine and xanthurenic acid showed a downward trend, signifying increased retention of these compounds with the advancement of CKD.

Finally, serum serotonin remained unchanged, whereas its precursor, 5-hydroxytryptophan, and its catabolite, 5-hydroxyindole-3-acetic acid, exhibited significant increases (p < 0.001) (Figure 4). However, only urinary 5-hydroxytryptophan was elevated (p < 0.001) (Figure 4), and the urinary excretion of 5-hydroxytryptophan increased as renal function further deteriorated, while there was no significant change observed in the excretion of 5-hydroxyindole-acetic acid across the progression spectrum (Table S6).

Tyrosine catabolites were retained in feline CKD

Serum tyrosine abundance remained relatively consistent between the two groups (Figure S3), albeit showing a mild decrease, as demonstrated by the observed negative correlations (Table S6) with sCr and a subsequent reduction in urinary concentration (Figure S3). Conversely, its precursor, phenylalanine, displayed mild downregulation in serum at the significance threshold (p = 0.041), but the pattern lacks statistical significance over the CKD progression spectrum. The catabolites, hippuric acid and phenylacetylglutamine, did not demonstrate notable alterations in serum, despite the latter one showing upward trends with CKD advancement. None of the aforementioned metabolites were altered in urine. Interestingly, both the bacterial catabolite p-cresol (p = 0.006) and the uremic toxin p-cresyl-sulfate (p = 0.006) exhibited similar and significant increases in serum, clearly worsening with disease progression (Figure S3), without augmented urinary excretion (Figure S3). Furthermore, the tyrosine-derived compound 3-hydroxyphenylacetic acid remained unaltered in both serum and urine.

Carnitine metabolism was upregulated toward trimethylamine-N-oxide (TMAO)

Serum carnitine showed a pronounced and statistically significant increase in abundance within the feline CKD group (p < 0.001) with a clear upward trend as CKD progresses, while urinary abundance remained unaltered over all stages (Figure 5A, Tables S5 and S6). No discernible differences were observed in either serum or urinary deoxycarnitine and choline. Conversely, serum betaine (p = 0.016) decreases as renal function deteriorates (Figure 5A), while its catabolic product, dimethylglycine, showed significant upregulation in both serum (p < 0.001) and, to a lesser extent, in urine (p = 0.033), worsening with CKD progression. Additionally, urinary betaine only decreased in advanced stages. Remarkably, serum TMAO, which is also a uremic toxin, exhibited a substantial increase (p < 0.001), especially in advanced stages (Figure 5A), resulting only in a mildly increased urinary excretion (p = 0.020), remaining unaltered with increasing progression (Figure 5A and Table S6). Noteworthy, the ratio of betaine and its downstream metabolite TMAO reverses exactly at the point of stage 2 disease development (Figure 5B). In contrast, the intermediate compound, trimethylamine (TMA), remained unchanged in serum but increased in urine with disease advancement (p = 0.019).

Figure 5.

Figure 5.

Open in a new tab

A) Box plot representation of the selected serum and urine metabolites from the carnitine metabolism comparing healthy controls (green; n = 84) with specific disease stages. Significance levels (*<0.05; **<0.001) for the comparison healthy vs. CKD (all stages) are provided. Clear up- or downward trends are portrayed in accordance with CKD progression from stage 2 (pink; n = 59) to stage 3 (red; n = 29) and 4 (dark red; n = 6). Urine metabolites are normalized to urine creatinine. B) Trend lines representing the linear relationship between serum creatinine (X-axis) and specific serum metabolites (Y-axis) from the carnitine pathway. Red lines mean upward trends, blue lines mean downward trends and grey lines mean unaltered behavior in accordance with CKD progression from stage 2 to stage 3 and 4. Spearman correlation coefficients are provided with significance levels (*<0.05; **<0.001) for every metabolite. The positive ratio of betaine/carnitine and its catabolites is completely reversed precisely at the onset of stage 2 CKD.

Well-established molecular changes were noted in the urea cycle

Key metabolites of the urea cycle itself, namely L-citrulline, L-arginine, and L-ornithine, showed no significant alterations between the two health states in the serum metabolome (Figure S4). However, correlation analysis unveiled an upward trend in the case of serum and urinary L-citrulline, in advanced disease stages (Table S6). In contrast, serum L-arginine showed a mild decrease concomitant with heightened urinary excretion. Furthermore, the end products and byproducts of the amino acid metabolism, such as urea (p < 0.001) and N,N-dimethylarginine (p < 0.001), displayed a substantial increase in serum (Figure S4). With CKD progression serum dimethylarginine gradually rose, accompanied by reduced urine abundance. Notably, both L-aspartate (p < 0.001) and guanidinosuccinate (p < 0.001) exhibited similar and significant increases in their serum (Figure S4) in conjunction with elevated urinary abundance (p = 0.014 and p < 0.001, respectively) within the CKD group (Figure S4). Conversely, the precursor of creatinine, guanidinoacetate, showed a mild downregulation, just reaching statistical significance (p = 0.023). Creatine, an intermediate in this pathway, did not show discriminative changes in its serum. Furthermore, the polyamines spermidine and putrescine demonstrated only mild decreasing and increasing trends in serum, respectively, while their urinary abundance remained stable across the spectrum of CKD progression (Table S6).

Purine metabolism remained generally unaffected

Within the context of feline CKD, no significant serum alterations were observed in the purine metabolism, except for a substantial increase in allantoin (p < 0.001), worsening with CKD progression (Figure S5). Conversely, the urinary excretion of guanine, adenine, adenosine, and xanthosine showed a ­significant reduction across the spectrum of CKD progression, with no discernible alterations in serum (Table S6 and Figure S5).

Discussion

Our study employed state-of-the-art metabolomics to improve the understanding of specific pathological mechanisms underlying feline CKD. Untargeted examination of dynamic changes in metabolic signatures identified which portion of the metabolome is altered in the feline CKD state, while the targeted approach shed light on specific metabolic shifts in distinct pathways.

Most perturbed pathways in feline CKD point toward accumulation of gut-derived uremic toxins originating from the diet. In general, three different underlying mechanisms may theoretically explain the increase in circulating gut-derived uremic toxins: (I) enhanced density of dietary precursors; (II) increased microbial formation; and (III) reduced fractional renal excretion (Figure 6) (Meijers and Evenepoel 2011; Ramezani et al. 2016; Wang et al. 2019; Bush et al. 2020; Graboski and Redinbo 2020; Vanholder et al. 2023).

Figure 6.

Figure 6.

Open in a new tab

Nutritional Precursors tryptophan, tyrosine, carnitine, and betaine are highly abundant in the diet of strict carnivores, such as felines. These Precursors are partially degraded in the gut to the bacterial byproducts indole, p-cresol, and trimethylamine (TMA). These compounds are subsequently processed by the liver, resulting in the formation of the uremic toxins indoxyl-sulfate, p-cresyl-sulfate, and trimethylamine-N-oxide (TMAO). These solutes undergo renal excretion by the tubule cells, which is associated with oxidative stress promoting tubular inflammation and secondary fibrosis. This process is hypothesized to contribute to the progression and development of feline CKD.

(I) Influenced by an inherent carnivorous biology, specific gut-derived uremic toxin precursors, i.e. tryptophan, tyrosine, carnitine, and betaine, are highly abundant in the diet of felids (Schmidt et al. 2016). Recent evidence suggests that dietary macronutrient composition directly impacts circulating levels of gut-derived uremic toxins in healthy adult cats. For example, cats fed high-protein diets showed significantly increased circulating indoxyl-sulfate concentrations whilst tryptophan itself was not increased, conform with findings in humans and mice (Deng et al. 2014; Poesen et al. 2015).

(II) In the context of feline CKD, our data raises the possibility of an increase in microbial formation. This phenomenon may be theoretically attributed to the observed increased serum abundance of indole and p-cresol (but not TMA) in relation to the decreasing abundance of their dietary precursors: tryptophan, tyrosine, and betaine (Meijers and Evenepoel 2011; Wang et al. 2019; Graboski and Redinbo 2020; Nealon et al. 2024). However, urinary metabolite/creatinine ratios for indole and p-cresol did not increase, as opposed to serum metabolite abundance, ­suggesting that the latter results from impaired renal clearance rather than increased formation. Nevertheless, decreased fecal microbiome diversity and richness is commonly described in human patients with impaired renal function and more recently also in cats suffering from CKD (Vaziri et al. 2013; Summers et al. 2019; Nealon et al. 2024). In addition, intestinal dysbiosis and increased abundance of proteolytic bacteria may be potential drivers for augmented formation of uremic toxins (Wang et al. 2020).

(III) Importantly, we discovered that urinary clearances of these solutes reached a plafond over the feline CKD progression spectrum. In other words, renal excretion does not correspond to higher concentrations in circulating serum. The observed saturable excretion indicates that renal retention may be the main factor in the systemic rise of toxic solutes in cats with CKD. Indeed, the clearance of these solutes is partly reliant on specific transporters located in the renal tubules (Masereeuw et al. 2014; Bush et al. 2020). Therefore, elevated serum/urine ratios of either indoxyl-sulfate, p-cresyl-sulfate or TMAO could also serve as markers of tubular dysfunction.

Our results indicate that serum indole and indoxyl-sulfate accumulation are directly related to loss of renal function with serum abundance correlating to feline CKD severity. Similarly, our findings also suggest that feline CKD is associated with elevated circulating p-cresol and p-cresyl-sulfate, which tend to rise as the disease progresses. Lastly, we observed an increased accumulation of TMAO, also correlating with disease progression. This also raises the hypothesis that higher concentrations of these circulating uremic solutes may contribute to tubular overload and consequent induction of tubular damage (Mutsaers et al. 2015).

In the case of indoxyl-sulfate, it is well-established that decreasing renal function leads to its accumulation in proximal tubular cells, which may promote CKD progression through the induction of reactive oxidative stress, increased production of profibrotic cytokines, and inhibition of tubular proliferation (Miyazaki et al. 1997; Chen et al. 2020; Cheng et al. 2020). Additionally, Chen and colleagues found that plasma indoxyl-sulfate serves as a potential predictor of disease progression in feline CKD (Chen et al. 2018). Similarly, p-cresyl-sulfate has been associated with tubular inflammation (Poveda et al. 2014) and may contribute to the progression of CKD, as serum total p-cresyl-sulfate is linked to renal progression and all-cause mortality in human patients across different stages of CKD (Wu et al. 2011; Chen et al. 2020). Recent studies have also indicated a potential mechanistic link between TMAO and progressive renal dysfunction, highlighting an association between TMAO levels, renal tubulointerstitial fibrosis, and functional impairment (Bain et al. 2006; Gupta et al. 2020; Zhang et al. 2021). Furthermore, elevated baseline TMAO levels have been associated with an increased risk of developing CKD, as demonstrated through a combined human epidemiological and metabolomic approach (Rhee et al. 2013).

These observations are of particular interest as tubulointerstitial inflammation is the hallmark of feline CKD (Chakrabarti et al. 2013; Brown et al. 2016). Furthermore, tubular inflammation may set off a cascade of pathological metabolic reactions in the mammalian kidney (Vanholder et al. 2023). Interestingly, in human CKD patients, it is well-established that indoleamine 2,3-dioxygenase (IDO), an inducible tubular enzyme, is typically upregulated in response to a pro-inflammatory environment (Mohib et al. 2008; Schefold et al. 2009; Jensen et al. 2021). Increased IDO activity is responsible for the degradation of tryptophan to kynurenine, accounting for approximately 95% of dietary tryptophan catabolism (Bender 1983; Le Floc’h et al. 2011). This aligns with our findings, indicating significant alterations impacting the kynurenine pathway in feline CKD. In contrast to humans, serum kynurenine itself remained relatively stable, likely due to an increased excretion and its further conversion to kynurenic acid, 3-hydroxykynurenine, and catabolites such as xanthurenic and quinolinic acid. Interestingly, our findings indicate that urinary 3-hydroxykynurenine excretion shows a strong negative correlation with renal function, even as serum abundance increases with the progression of CKD. This suggests that tubular excretion is the primary mechanism for elimination of 3-hydroxykynurenine (Bahn et al. 2005). Importantly, given that tubulointerstitial lesions are the hallmark of feline CKD, biomarkers of tubular excretion can provide additional information on renal health beyond mere estimators of glomerular function, such as creatinine (Naud et al. 2011). Remarkably, in humans, three large-cohort studies recently demonstrated excellent predictive abilities of metabolites from the kynurenine pathway in the development of overt CKD (Goek et al. 2013; Rhee et al. 2013; Lee et al. 2020).

Likewise, tryptophan metabolism was also further upregulated along the serotonin pathway, normally accounting for 5% of tryptophan metabolization (Richard et al. 2009). Although serum serotonin remained stable, accumulation of the degradation product, 5-hydroxyindoleacetic acid, was strongly associated with declining renal function in the feline CKD state (Rhee et al. 2013). Increased formation of 5-hydroxyindoleacetic acid results from upregulated mitochondrial monoamine oxidase A (MAO-A) activity (Celada and Artigas 1993). Typically, MAO-A activity is also stimulated in pro-inflammatory conditions, and further contributes to free radical production and subsequent oxidative stress (Sturza et al. 2019). Indeed, the ratio of serotonin to 5-hydroxyindoleacetic acid has been outlined as a surrogate marker of tubular oxidative stress and inflammation (Lee et al. 2021).

Persistent tubular inflammation may eventually lead to irreversible fibrosis and loss of nephrons in the mammalian kidney (Chevalier 2016). When a certain percentage of nephrons is lost, the glomerular filtration rate (GFR) will decline, aggravating the retention of a scala of uremic solutes (López-Novoa et al. 2011). It is widely acknowledged that components of the urea cycle show significant deviations in association with a diminishing GFR (Weiss and Kim 2011). Especially sCr, urea, and symmetric dimethyl arginine (SDMA) have long been established as diagnostic biomarkers of a reduced GFR in both human and feline CKD (Sparkes et al. 2016; Bidin et al. 2019). Detection of these retained metabolites also further confirms the validity of the methodologies employed in this study.

Together, current evidence suggests that high protein consumption and the increased formation and retention of gut-derived uremic toxins may contribute to tubular inflammation (Vanholder et al. 2023). This process has the potential to trigger a series of metabolic reactions, which could play a role in the progression of CKD. However, the extent of this impact and the underlying mechanisms are not yet fully established and require further investigation. As such, to date it remains unclear whether diets high in protein can induce de novo renal dysfunction in healthy felines (Stenvinkel et al. 2018; Nealon et al. 2024).

The main challenge regarding study design was to obtain correct subject stratification (i.e. CKD/healthy), as the diagnosis of CKD was based on elevated sCr in combination with a decrease in USG. Therefore, one could expect that early renal deterioration was missed, and the risk to even confound the analyses by misclassifications in the healthy group existing. As validated markers of tubular damage/dysfunction are currently lacking in cats, we aimed to address this problem through removal of healthy animals in our validation subgroup that showed an increase in sCr or SDMA within or exceeding the reference-interval, a persistent suboptimal concentrated urine, and/or developed proteinuria of renal origin after a 6-month longitudinal follow-up (IRIS 2023). Another limitation regarding study design and stratification is the absence of imaging for diagnosing CKD. While idiopathic fibrosis is the primary cause of renal dysfunction in cats, less common causes, such as renal lymphoma or PKD, may have been erroneously classified within the CKD group (Brown et al. 2016).

An important bottleneck in untargeted metabolomics remains the accurate annotation of compounds of interest. Current algorithms only provide plausible compound matches with existing databases without definitive certainty. To overcome this obstacle, we selected and validated a panel of analytical standards based on a priori knowledge and pathway analysis, allowing correct annotation (Vanden Broecke et al. 2024). An important limitation regarding the presented metabolomic data is the lack of absolute quantification, therefore only relative changes between groups could be reported. Lastly, the performance of urine creatinine normalization can be dependent on other factors than urinary dilution (Vogl et al. 2016). Alternatives such as pre-analytical dilution of urine specimens to a fixed creatinine concentration are impractical because of the number of samples but could provide a more accurate representation of urinary compound concentrations (Vogl et al. 2016).

Conclusion

We demonstrated comprehensive fundamental insights in CKD-related metabolic shifts of conserved biochemical pathways. Significant disruptions in tryptophan metabolism (indole, kynurenine, and serotonin pathways), as well as alterations in carnitine and tyrosine metabolism, and the urea cycle were observed. The most affected pathways in feline CKD suggest an accumulation of gut-derived uremic toxins from dietary precursors. This is primarily due to impaired renal excretion but may potentially also involve changes in intrinsic or bacterial enzyme activity. These findings could link the strict carnivorous nature of felines to the pathophysiology of CKD and offer possible insights into future therapeutic approaches.

The acquired fundamental insights in metabolic shifts underlying feline CKD can be further exploited for: (I) Evaluating the capacity of specific biomarkers to reflect tubular dysfunction and/or inflammation, alongside existing diagnostic markers of glomerular filtration. (II) Delving deeper into the potential causal relationship between specific metabolic changes and the progression of CKD. (III) Investigating to what extent nutrient intake directly influences metabolic shifts toward uremic toxin production, while at the same time induces tubular inflammation/fibrosis and eventually development of overt CKD. Such approaches may further enhance the exploration of specific therapeutic or preventative targets.

Supplementary Material

Supplemental Material
TVEQ_A_2447601_SM9717.docx (2.4MB, docx)

Funding Statement

This research was funded by the Research Foundation – Flanders (FWO, SBP 2020 006001), the Special Research Fund of Ghent University (BOF, DOC 2020 004701), the Industrial Research Fund of Ghent University (IOF, F2021/IOF-Equip/014), and by the EveryCat Health Foundation (CaPK22-002).

Author contributions statement

Laurens Van Mulders: Conceptualization, Methodology, Formal analysis, Investigation, Writing - original draft, Visualization, Project administration, Funding acquisition. Ellen Vanden Broecke: Data curation, Investigation, Resources, Funding acquisition. Femke Mortier: Investigation; Resources. Ellen De Paepe: Methodology, Software, Formal analysis, Writing - review & editing. Lynn Vanhaecke: Conceptualization, Methodology, Software, Resources, Supervision, Writing - review & editing, Project administration, Funding acquisition. Sylvie Daminet: Conceptualization, Methodology, Resources, Supervision, Writing - review & editing, Project administration, Funding acquisition. The authors wish to thank the owners of the cats that donated a blood and urine sample, and everyone involved in the data collection process (Lisa Stammeleer, Marleen Brans, Sofie Marynissen, Dominique Paepe), and the analytical procedures (Beata Pomian, Dirk Stockx, Tom Cools). All authors have given approval to the final version of the manuscript. We acknowledge the use of ChatGPT (OpenAI) for language optimization in drafting and refining this manuscript. The tool was used to enhance clarity, coherence, and overall readability during the preparation of the text.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

References

  1. Acierno MJ, Brown S, Coleman AE, Jepson RE, Papich M, Stepien RL, Syme HM.. 2018. ACVIM consensus statement: guidelines for the identification, evaluation, and management of systemic hypertension in dogs and cats. J Vet Intern Med. 32(6):1803–1822. doi: 10.1111/jvim.15331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bahn A, Ljubojevic M, Lorenz H, Schultz C, Ghebremedhin E, Ugele B, Sabolic I, Burckhardt G, Hagos Y.. 2005. Murine renal organic anion transporters mOAT1 and mOAT3 facilitate the transport of neuroactive tryptophan metabolites. Am J Physiol Cell Physiol. 289(5):C1075–1084. doi: 10.1152/ajpcell.00619.2004. [DOI] [PubMed] [Google Scholar]
  3. Bain MA, Faull R, Fornasini G, Milne RW, Evans AM.. 2006. Accumulation of trimethylamine and trimethylamine-N-oxide in end-stage renal disease patients undergoing haemodialysis. Nephrol Dial Transplant. 21(5):1300–1304. doi: 10.1093/ndt/gfk056. [DOI] [PubMed] [Google Scholar]
  4. Bender DA. 1983. Biochemistry of tryptophan in health and disease. Mol Aspects Med. 6(2):101–197. doi: 10.1016/0098-2997(83)90005-5. [DOI] [PubMed] [Google Scholar]
  5. Bidin MZ, Shah AM, Stanslas J, Seong CLT.. 2019. Blood and urine biomarkers in chronic kidney disease: an update. Clin Chim Acta. 495(2019):239–250. doi: 10.1016/j.cca.2019.04.069. [DOI] [PubMed] [Google Scholar]
  6. Boelaert J, t’Kindt R, Schepers E, Jorge L, Glorieux G, Neirynck N, Lynen F, Sandra P, Vanholder R, Sandra K, et al. 2014. State-of-the-art non-targeted metabolomics in the study of chronic kidney disease. Metabolomics. 10(3):425–442. doi: 10.1007/s11306-013-0592-z. [DOI] [Google Scholar]
  7. Brown CA, Elliott J, Schmiedt CW, Brown SA.. 2016. Chronic kidney disease in aged cats: clinical features, morphology, and proposed pathogeneses. Vet Pathol. 53(2):309–326. doi: 10.1177/0300985815622975. [DOI] [PubMed] [Google Scholar]
  8. Bush KT, Singh P, Nigam SK.. 2020. Gut-derived uremic toxin handling in vivo requires OAT-mediated tubular secretion in chronic kidney disease. JCI Insight. 5(7):e13381. doi: 10.1172/jci.insight.133817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Celada P, Artigas F.. 1993. Plasma 5‐hydroxyindoleacetic acid as an indicator of monoamine oxidase‐A inhibition in rat brain and peripheral tissues. J Neurochem. 61(6):2191–2198. doi: 10.1111/j.1471-4159.1993.tb07459.x. [DOI] [PubMed] [Google Scholar]
  10. Chakrabarti S, Syme HM, Brown CA, Elliott J.. 2013. Histomorphometry of feline chronic kidney disease and correlation with markers of renal dysfunction. Vet Pathol. 50(1):147–155. doi: 10.1177/0300985812453176. [DOI] [PubMed] [Google Scholar]
  11. Chen CN, Chou CC, Tsai PSJ, Lee YJ.. 2018. Plasma indoxyl sulfate concentration predicts progression of chronic kidney disease in dogs and cats. Vet J. 232:33–39. doi: 10.1016/j.tvjl.2017.12.011. [DOI] [PubMed] [Google Scholar]
  12. Chen Y, Zelnick LR, Wang K, Hoofnagle AN, Becker JO, Hsu C-Y, Feldman HI, Mehta RC, Lash JP, Waikar SS, et al. 2020. Kidney clearance of secretory solutes is associated with progression of CKD: the CRIC study. J Am Soc Nephrol. 31(4):817–827. doi: 10.1681/ASN.2019080811. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cheng T-H, Ma M-C, Liao M-T, Zheng C-M, Lu K-C, Liao C-H, Hou Y-C, Liu W-C, Lu C-L.. 2020. Indoxyl sulfate, a tubular toxin, contributes to the development of chronic kidney disease. Toxins. 12(11):684. doi: 10.3390/toxins12110684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Chevalier RL. 2016. The proximal tubule is the primary target of injury and progression of kidney disease: role of the glomerulotubular junction. Am J Physiol Renal Physiol. 311(1):F145–F161. doi: 10.1152/ajprenal.00164.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. De Paepe E, Van Meulebroek L, Rombouts C, Huysman S, Verplanken K, Lapauw B, Wauters J, Hemeryck LY, Vanhaecke L.. 2018. A validated multi-matrix platform for metabolomic fingerprinting of human urine, feces and plasma using ultra-high-performance liquid-chromatography coupled to hybrid orbitrap high-resolution mass spectrometry. Anal Chim Acta. 1033(2018):108–118. doi: 10.1016/j.aca.2018.06.065. [DOI] [PubMed] [Google Scholar]
  16. Deng P, Jones JC, Swanson KS.. 2014. Effects of dietary macronutrient composition on the fasted plasma metabolome of healthy adult cats. Metabolomics. 10(4):638–650. doi: 10.1007/s11306-013-0617-7. [DOI] [Google Scholar]
  17. Elliott J, Rawlings JM, Markwell PJ, Barber PJ.. 2000. Survival of cats with naturally occurring chronic renal failure: effect of dietary management. J Small Anim Pract. 41(6):235–242. doi: 10.1111/j.1748-5827.2000.tb03932.x. [DOI] [PubMed] [Google Scholar]
  18. Ghys LFE, Paepe D, Duchateau L, Taffin ERL, Marynissen S, Delanghe J, Daminet S.. 2015. Biological validation of feline serum cystatin C: the effect of breed, age and sex and establishment of a reference interval. Vet J. 204(2):168–173. doi: 10.1016/j.tvjl.2015.02.018. [DOI] [PubMed] [Google Scholar]
  19. Goek O-N, Prehn C, Sekula P, Römisch-Margl W, Döring A, Gieger C, Heier M, Koenig W, Wang-Sattler R, Illig T, et al. 2013. Metabolites associate with kidney function decline and incident chronic kidney disease in the general population. Nephrol Dial Transplant. 28(8):2131–2138.   doi: 10.1093/ndt/gft217. [DOI] [PubMed] [Google Scholar]
  20. Graboski AL, Redinbo MR.. 2020. Gut-derived protein-bound uremic toxins. Toxins. 12(9):590. doi: 10.3390/toxins12090590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gupta N, Buffa JA, Roberts AB, Sangwan N, Skye SM, Li L, Ho KJ, Varga J, DiDonato JA, Tang WHW, et al. 2020. Targeted inhibition of gut microbial trimethylamine N-oxide production reduces renal tubulointerstitial fibrosis and functional impairment in a murine model of chronic kidney disease. Arterioscler Thromb Vasc Biol. 40(5):1239–1255. doi: 10.1161/ATVBAHA.120.314139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gyurászová M, Gurecká R, Bábíčková J, Tóthová L.. 2020. Oxidative stress in the pathophysiology of kidney disease: implications for noninvasive monitoring and identification of biomarkers. Oxid Med Cell Longev. 2020:5478708–5478711. doi: 10.1155/2020/5478708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hocher B, Adamski J.. 2017. Metabolomics for clinical use and research in chronic kidney disease. Nat Rev Nephrol. 13(5):269–284. doi: 10.1038/nrneph.2017.30. [DOI] [PubMed] [Google Scholar]
  24. Hu J-R, Coresh J, Inker LA, Levey AS, Zheng Z, Rebholz CM, Tin A, Appel LJ, Chen J, Sarnak MJ, et al. 2018. Serum metabolites are associated with all-cause mortality in chronic kidney disease. Kidney Int. 94(2):381–389. doi: 10.1016/j.kint.2018.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. [IRIS] International Renal Interest Society Ltd . 2023. www.iris-kidney.com. [cited 2024 Jan 16].
  26. Jensen CG, Jensen MS, Tingskov SJ, Olinga P, Nørregaard R, Mutsaers HAM.. 2021. Local inhibition of indoleamine 2,3-dioxygenase mitigates renal fibrosis. Biomedicines. 9(8):856. doi: 10.3390/biomedicines9080856. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Johnson CH, Ivanisevic J, Siuzdak G.. 2016. Metabolomics: beyond biomarkers and towards mechanisms. Nat Rev Mol Cell Biol. 17(7):451–459. doi: 10.1038/nrm.2016.25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kimura T, Yasuda K, Yamamoto R, Soga T, Rakugi H, Hayashi T, Isaka Y.. 2016. Identification of biomarkers for development of end-stage kidney disease in chronic kidney disease by metabolomic profiling. Sci Rep. 6(1):26138. doi: 10.1038/srep26138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kovesdy CP. 2022. Epidemiology of chronic kidney disease: an update 2022. Kidney Int Suppl. 12(1):7–11. doi: 10.1016/j.kisu.2021.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kriz W, Lehir M.. 2005. Pathways to nephron loss starting from glomerular diseases—insights from animal models. Kidney Int. 67(2):404–419. doi: 10.1111/j.1523-1755.2005.67097.x. [DOI] [PubMed] [Google Scholar]
  31. Le Floc’h N, Otten W, Merlot E.. 2011. Tryptophan metabolism, from nutrition to potential therapeutic applications. Amino Acids. 41(5):1195–1205. doi: 10.1007/s00726-010-0752-7. [DOI] [PubMed] [Google Scholar]
  32. Lee H, Jang HB, Yoo MG, Park SI, Lee HJ.. 2020. Amino acid metabolites associated with chronic kidney disease: an eight-year follow-up Korean epidemiology study. Biomedicines. 8(7):222. doi: 10.3390/biomedicines8070222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lee HS, Kim SM, Jang JH, Park HD, Lee SY.. 2021. Serum 5-hydroxyindoleacetic acid and ratio of 5-hydroxyindoleacetic acid to serotonin as metabolomics indicators for acute oxidative stress and inflammation in vancomycin-associated acute kidney injury. Antioxidants. 10(6):1330. doi: 10.3390/antiox10060895. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Little S, Levy J, Hartmann K, Hofmann-Lehmann R, Hosie M, Olah G, Denis KS.. 2020. 2020 AAFP feline retrovirus testing and management guidelines. J Feline Med Surg. 22(1):5–30. doi: 10.1177/1098612X19895940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. López-Novoa JM, Rodríguez-Peña AB, Ortiz A, Martínez-Salgado C, López Hernández FJ.. 2011. Etiopathology of chronic tubular, glomerular and renovascular nephropathies: clinical implications. J Transl Med. 9(1):13. doi: 10.1186/1479-5876-9-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Marino CL, Lascelles BDX, Vaden SL, Gruen ME, Marks SL.. 2014. Prevalence and classification of chronic kidney disease in cats randomly selected from four age groups and in cats recruited for degenerative joint disease studies. J Feline Med Surg. 16(6):465–472. doi: 10.1177/1098612X13511446. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Masereeuw R, Mutsaers HA, Toyohara T, Abe T, Jhawar S, Sweet DH, Lowenstein J.. 2014. The kidney and uremic toxin removal: glomerulus or tubule? Semin Nephrol. 34(2):191–208. doi: 10.1016/j.semnephrol.2014.02.010. [DOI] [PubMed] [Google Scholar]
  38. Meijers BK, Evenepoel P.. 2011. The gut–kidney axis: indoxyl sulfate, p-cresyl sulfate and CKD progression. Nephrol Dial Transplant. 26(3):759–761. doi: 10.1093/ndt/gfq818. [DOI] [PubMed] [Google Scholar]
  39. Metzger M, Yuan WL, Haymann J-P, Flamant M, Houillier P, Thervet E, Boffa J-J, Vrtovsnik F, Froissart M, Bankir L, et al. 2018. Association of a low-protein diet with slower progression of CKD. Kidney Int Rep. 3(1):105–114. doi: 10.1016/j.ekir.2017.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Miyazaki T, Ise M, Hirata M, Endo K, Ito Y, Niwa T, et al. 1997. Indoxyl sulfate stimulates renal synthesis of transforming growth factor-β1 and progression of renal failure. Kidney Int Suppl. 63:S-1. [PubMed] [Google Scholar]
  41. Mohib K, Wang S, Guan Q, Mellor AL, Sun H, Du C, Jevnikar AM.. 2008. Indoleamine 2,3-dioxygenase expression promotes renal ischemia-reperfusion injury. Am J Physiol Renal Physiol. 295(1):F226–234. doi: 10.1152/ajprenal.00567.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Mutsaers HA, Stribos EG, Glorieux G, Vanholder R, Olinga P.. 2015. Chronic kidney disease and fibrosis: the role of uremic retention solutes. Front Med. 2(2015):60. doi: 10.3389/fmed.2015.00060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Naud J, Michaud J, Beauchemin S, Hébert M-J, Roger M, Lefrancois S, Leblond FA, Pichette V.. 2011. Effects of chronic renal failure on kidney drug transporters and cytochrome P450 in rats. Drug Metab Dispos. 39(8):1363–1369. doi: 10.1124/dmd.111.039115. [DOI] [PubMed] [Google Scholar]
  44. Nealon NJ, Summers S, Quimby J, Winston JA.. 2024. Untargeted metabolomic profiling of serum from client-owned cats with early and late-stage chronic kidney disease. Sci Rep. 14(1):4755. doi: 10.1038/s41598-024-55249-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Newkirk KM, Newman SJ, White LA, Rohrbach BW, Ramsay EC.. 2011. Renal lesions of nondomestic felids. Vet Pathol. 48(3):698–705. doi: 10.1177/0300985810382089. [DOI] [PubMed] [Google Scholar]
  46. O’Neill DG, Church DB, McGreevy PD, Thomson PC, Brodbelt DC.. 2015. Longevity and mortality of cats attending primary care veterinary practices in England. J Feline Med Surg. 17(2):125–133. doi: 10.1177/1098612X14536176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Ortiz A, Sanchez-Niño MD, Sanz AB.. 2011. The meaning of urinary creatinine concentration. Kidney Int. 79(7):791. doi: 10.1038/ki.2011.1. [DOI] [PubMed] [Google Scholar]
  48. Paepe D, Verjans G, Duchateau L, Piron K, Ghys L, Daminet S.. 2013. Routine health screening: findings in apparently healthy middle-aged and old cats. J Feline Med Surg. 15(1):8–19. doi: 10.1177/1098612X12464628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Pang Z, Chong J, Zhou G, de Lima Morais DA, Chang L, Barrette M, Gauthier C, Jacques P-É, Li S, Xia J, et al. 2021. MetaboAnalyst 5.0: narrowing the gap between raw spectra and functional insights. Nucleic Acids Res. 49(W1):W388–W396. doi: 10.1093/nar/gkab382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Poesen R, Mutsaers HAM, Windey K, van den Broek PH, Verweij V, Augustijns P, Kuypers D, Jansen J, Evenepoel P, Verbeke K, et al. 2015. The influence of dietary protein intake on mammalian tryptophan and phenolic metabolites. PLoS One. 10(10):e0140820. (2015) doi: 10.1371/journal.pone.0140820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Poveda J, Sanchez-Niño MD, Glorieux G, Sanz AB, Egido J, Vanholder R, Ortiz A.. 2014. p-Cresyl sulfate has pro-inflammatory and cytotoxic actions on human proximal tubular epithelial cells. Nephrol Dial Transplant. 29(1):56–64. doi: 10.1093/ndt/gft367. [DOI] [PubMed] [Google Scholar]
  52. Ramezani A, Massy ZA, Meijers B, Evenepoel P, Vanholder R, Raj DS.. 2016. Role of the gut microbiome in uremia: a potential therapeutic target. Am J Kidney Dis. 67(3):483–498. doi: 10.1053/j.ajkd.2015.09.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Reppas G, Foster SF.. 2016a. Practical urinalysis in the cat: 1: urine macroscopic examination ‘tips and traps. J Feline Med Surg. 18(3):190–202. doi: 10.1177/1098612X16631228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Reppas G, Foster SF.. 2016b. Practical urinalysis in the cat: 2: urine microscopic examination ‘tips and traps. J Feline Med Surg. 18(5):373–385. doi: 10.1177/1098612X16643249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Rhee EP, Clish CB, Ghorbani A, Larson MG, Elmariah S, McCabe E, Yang Q, Cheng S, Pierce K, Deik A, et al. 2013. A combined epidemiologic and metabolomic approach improves CKD prediction. J Am Soc Nephrol. 24(8):1330–1338. doi: 10.1681/ASN.2012101006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Richard DM, Dawes MA, Mathias CW, Acheson A, Hill-Kapturczak N, Dougherty DM.. 2009. L-tryptophan: basic metabolic functions, behavioral research and therapeutic indications. Int J Tryptophan Res. 2:45–60. doi: 10.4137/ijtr.s2129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Ruiz-Ortega M, Rayego-Mateos S, Lamas S, Ortiz A, Rodrigues-Diez RR.. 2020. Targeting the progression of chronic kidney disease. Nat Rev Nephrol. 16(5):269–288. doi: 10.1038/s41581-019-0248-y. [DOI] [PubMed] [Google Scholar]
  58. Ryan D, Robards K, Prenzler PD, Kendall M.. 2011. Recent and potential developments in the analysis of urine: a review. Anal Chim Acta. 684(1–2):8–20. doi: 10.1016/j.aca.2010.10.035. [DOI] [PubMed] [Google Scholar]
  59. Schefold JC, Zeden J-P, Fotopoulou C, von Haehling S, Pschowski R, Hasper D, Volk H-D, Schuett C, Reinke P.. 2009. Increased indoleamine 2,3-dioxygenase (IDO) activity and elevated serum levels of tryptophan catabolites in patients with chronic kidney disease: a possible link between chronic inflammation and uraemic symptoms. Nephrol Dial Transplant. 24(6):1901–1908. doi: 10.1093/ndt/gfn739. [DOI] [PubMed] [Google Scholar]
  60. Schmidt JA, Rinaldi S, Scalbert A, Ferrari P, Achaintre D, Gunter MJ, Appleby PN, Key TJ, Travis RC.. 2016. Plasma concentrations and intakes of amino acids in male meat-eaters, fish-eaters, vegetarians and vegans: a cross-sectional analysis in the EPIC-Oxford cohort. Eur J Clin Nutr. 70(3):306–312. doi: 10.1038/ejcn.2015.144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Sparkes AH, Caney S, Chalhoub S, Elliott J, Finch N, Gajanayake I, Langston C, Lefebvre HP, White J, Quimby J, et al. 2016. ISFM consensus guidelines on the diagnosis and management of feline chronic kidney disease. J Feline Med Surg. 18(3):219–239. doi: 10.1177/1098612X16631234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Stenvinkel P, Painer J, Kuro-O M, Lanaspa M, Arnold W, Ruf T, Shiels PG, Johnson RJ.. 2018. Novel treatment strategies for chronic kidney disease: insights from the animal kingdom. Nat Rev Nephrol. 14(4):265–284. doi: 10.1038/nrneph.2017.169. [DOI] [PubMed] [Google Scholar]
  63. Sturza A, Popoiu CM, Ionică M, Duicu OM, Olariu S, Muntean DM, Boia ES.. 2019. Monoamine oxidase-related vascular oxidative stress in diseases associated with inflammatory burden. Oxid Med Cell Longev. 2019:8954201. doi: 10.1155/2019/8954201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Summers SC, Quimby JM, Isaiah A, Suchodolski JS, Lunghofer PJ, Gustafson DL.. 2019. The fecal microbiome and serum concentrations of indoxyl sulfate and p‐cresol sulfate in cats with chronic kidney disease. J Vet Intern Med. 33(2):662–669. doi: 10.1111/jvim.15389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Vanden Broecke E, Van Mulders L, De Paepe E, Daminet S, Vanhaecke L.. 2024. Optimization and validation of metabolomics methods for feline urine and serum towards application in veterinary medicine. Anal Chim Acta. 1310(2024):342694. doi: 10.1016/j.aca.2024.342694. [DOI] [PubMed] [Google Scholar]
  66. Vangeenderhuysen P, Van Arnhem J, Pomian B, De Graeve M, De Commer L, Falony G, Raes J, Zhernakova A, Fu J, Hemeryck LY, et al. 2023. Dual UHPLC-HRMS metabolomics and lipidomics and automated data processing workflow for comprehensive high-throughput gut phenotyping. Anal Chem. 95(22):8461–8468. doi: 10.1021/acs.analchem.2c05371. [DOI] [PubMed] [Google Scholar]
  67. Vanholder R, Glorieux G.. Uremic toxicity. In: Nissenson AR, and Fine RN, editors. Handbook of dialysis therapy. 6th ed. Amsterdam: Elsevier; 2023; p. 16–44. doi: 10.1016/B978-0-323-79135-9.00002-1. [DOI] [Google Scholar]
  68. Vanholder RC, Eloot S, Glorieux GL.. 2016. Future avenues to decrease uremic toxin concentration. Am J Kidney Dis. 67(4):664–676. doi: 10.1053/j.ajkd.2015.08.029. [DOI] [PubMed] [Google Scholar]
  69. Vaziri ND, Wong J, Pahl M, Piceno YM, Yuan J, DeSantis TZ, Ni Z, Nguyen T-H, Andersen GL.. 2013. chronic kidney disease alters intestinal microbial flora. Kidney Int. 83(2):308–315. doi: 10.1038/ki.2012.345. [DOI] [PubMed] [Google Scholar]
  70. Vogl FC, Mehrl S, Heizinger L, Schlecht I, Zacharias HU, Ellmann L, Nürnberger N, Gronwald W, Leitzmann MF, Rossert J, et al. 2016. Evaluation of dilution and normalization strategies to correct for urinary output in HPLC-HRTOFMS metabolomics. Anal Bioanal Chem. 408(29):8483–8493. doi: 10.1007/s00216-016-9974-1. [DOI] [PubMed] [Google Scholar]
  71. Wang X, Yang S, Li S, Zhao L, Hao Y, Qin J, Zhang L, Zhang C, Bian W, Zuo L, et al. 2020. Aberrant gut microbiota alters host metabolome and impacts renal failure in humans and rodents. Gut. 69(12):2131–2142. doi: 10.1136/gutjnl-2019-319766. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Wang Z, Bergeron N, Levison BS, Li XS, Chiu S, Jia X, Koeth RA, Li L, Wu Y, Tang WHW, et al. 2019. Impact of chronic dietary red meat, white meat, or non-meat protein on trimethylamine N-oxide metabolism and renal excretion in healthy men and women. Eur Heart J. 40(7):583–594. doi: 10.1093/eurheartj/ehy799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Weiss RH, Kim K.. 2011. Metabolomics in the study of kidney diseases. Nat Rev Nephrol. 8(1):22–33. doi: 10.1038/nrneph.2011.152. [DOI] [PubMed] [Google Scholar]
  74. Wu I-W, Hsu K-H, Lee C-C, Sun C-Y, Hsu H-J, Tsai C-J, Tzen C-Y, Wang Y-C, Lin C-Y, Wu M-S, et al. 2011. p-Cresyl sulfate and indoxyl sulfate predict progression of chronic kidney disease. Nephrol Dial Transplant. 26(3):938–947. doi: 10.1093/ndt/gfq580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Zhang A, Sun H, Qiu S, Wang X.. 2014. Metabolomics insights into pathophysiological mechanisms of nephrology. Int Urol Nephrol. 46(5):1025–1030. doi: 10.1007/s11255-013-0600-2. [DOI] [PubMed] [Google Scholar]
  76. Zhang W, Miikeda A, Zuckerman J, Jia X, Charugundla S, Zhou Z, Kaczor-Urbanowicz KE, Magyar C, Guo F, Wang Z, et al. 2021. Inhibition of microbiota dependent TMAO production attenuates chronic kidney disease in mice. Sci Rep. 11(1):518. doi: 10.1038/s41598-020-80063-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
TVEQ_A_2447601_SM9717.docx (2.4MB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Articles from The Veterinary Quarterly are provided here courtesy of Taylor & Francis

RESOURCES