Next Article in Journal
Niclosamide Suppresses Migration and Invasion of Human Osteosarcoma Cells by Repressing TGFBI Expression via the ERK Signaling Pathway
Previous Article in Journal
Downregulation of Methionine Cycle Genes MAT1A and GNMT Enriches Protein-Associated Translation Process and Worsens Hepatocellular Carcinoma Prognosis
Previous Article in Special Issue
Uremic Toxins and Frailty in Patients with Chronic Kidney Disease: A Molecular Insight
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Gut Microbiota as a Source of Uremic Toxins

1
A.N. Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University, 119992 Moscow, Russia
2
V.I. Kulakov National Medical Research Center of Obstetrics, Gynecology and Perinatology, 117997 Moscow, Russia
3
Faculty of Bioengineering and Bioinformatics, Lomonosov Moscow State University, 119992 Moscow, Russia
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Int. J. Mol. Sci. 2022, 23(1), 483; https://doi.org/10.3390/ijms23010483
Submission received: 25 October 2021 / Revised: 26 December 2021 / Accepted: 28 December 2021 / Published: 1 January 2022
(This article belongs to the Special Issue Uremic Toxins and Metabolic Complications of Chronic Kidney Disease)

Abstract

:
Uremic retention solutes are the compounds that accumulate in the blood when kidney excretory function is impaired. Some of these compounds are toxic at high concentrations and are usually known as “uremic toxins”. The cumulative detrimental effect of uremic toxins results in numerous health problems and eventually mortality during acute or chronic uremia, especially in end-stage renal disease. More than 100 different solutes increase during uremia; however, the exact origin for most of them is still debatable. There are three main sources for such compounds: exogenous ones are consumed with food, whereas endogenous ones are produced by the host metabolism or by symbiotic microbiota metabolism. In this article, we identify uremic retention solutes presumably of gut microbiota origin. We used database analysis to obtain data on the enzymatic reactions in bacteria and human organisms that potentially yield uremic retention solutes and hence to determine what toxins could be synthesized in bacteria residing in the human gut. We selected biochemical pathways resulting in uremic retention solutes synthesis related to specific bacterial strains and revealed links between toxin concentration in uremia and the proportion of different bacteria species which can synthesize the toxin. The detected bacterial species essential for the synthesis of uremic retention solutes were then verified using the Human Microbiome Project database. Moreover, we defined the relative abundance of human toxin-generating enzymes as well as the possibility of the synthesis of a particular toxin by the human metabolism. Our study presents a novel bioinformatics approach for the elucidation of the origin of both uremic retention solutes and uremic toxins and for searching for the most likely human microbiome producers of toxins that can be targeted and used for the therapy of adverse consequences of uremia.

1. Introduction

Chronic kidney disease (CKD) is a common health problem in adults defined as a gradual loss in kidney function. It affects about 10% of the human population worldwide [1]. CKD is characterized by reduced glomerular filtration rate, enhanced urinary albumin excretion [2], and accumulation of many metabolic waste products in the organism that are normally excreted, predominantly by the kidneys. These metabolites are called uremic retention solutes or uremic toxins in the case of toxicity at uremic concentrations [3]. Their accumulation causes a great number of pathologies which are collectively named uremic syndrome or uremia [4]. The complications of uremia include multi-organ dysfunctions such as bone diseases, serositis, insulin resistance, renal fibrosis, podocyte dysfunction, decreased mental acuity, and various cardiovascular problems [4,5]. Some of these ailments are present in the World Health Organization list of widespread causes of death. Despite the severity of the consequences of uremia and despite intense study of the topic, the cellular and molecular mechanisms underlying syndrome development mostly remain unclear. The main reason for this is the huge and continuously expanding list of uremic retention solutes [6], which complicates the analysis of the impact of each of them and the elucidation of the possible interplays between them. Nowadays, the European Uremic Solutes Database (EUTox-DB), which was created by The European Uremic Toxin Work Group (EUTox), contains 130 solutes [7,8,9,10]. The substances belong to different chemical classes and participate in a great diversity of biochemical pathways, making further classification difficult.
Considering that the human gut metagenome contains 150 times more genes than the host [11], it is not surprising that intestinal bacteria produce a huge variety of unique substances, some of which could be related to uremic toxins. Indeed, there are some observations that the gut microbiota contributes to the uremic toxins production that inevitably aggravates the health status of CKD patients. Thus, as early as 1966, Einheber and Carter removed kidneys from germ-free rats and rats with normal microflora, thus creating rats that could not excrete uremic toxins and hence died due to uremia. Remarkably, the germ-free animals remained alive significantly longer than those with the normal gut microbiome [12]. Comparison of mural plasma from germ-free and conventional-microbiota rats revealed the emergence of many uremic retention solutes including indole-3-propionic acid, indoxyl sulfate, and p-cresyl sulfate in the conventional-microbiota rats after nephrectomy, which were less abundant in the germ-free animals [13]. Similarly, the microbiota was found to be important in the health/disease balance under renal failure in humans also. It was shown that hemodialysis patients who underwent a colon resection demonstrated lower levels of some uremic retention solutes and that dialysis patients with intact colons accumulated more than 30 additional substances in plasma. Those substances were assigned to the group of gut-derived uremic solutes [14].
Today, the list of microbiota-derived toxins is growing. The bacterial origin of substances such as p-cresyl sulfate, indoxyl sulfate, indole-3-acetic acid, trimethylamine, trimethylamine-N-oxide, hippuric acid, phenol, and phenylacetic acid has been proved in several independent experiments [15,16,17,18]. However, most compounds in the EUTox-DB are underexplored. According to the Human Metabolome Database, 69 uremic retention solutes are classified as endogenous, and 56 have not yet been classified [14,19,20,21,22].
In this study, we analyzed the Human Metabolome, Kyoto Encyclopedia of Genes and Genomes (KEGG), MetaCyc, and Human Protein Atlas databases to find microbial biochemical pathways and enzymes strongly associated with uremic toxin synthesis.

2. Results

We used a list of 142 substances regarded as uremic retention solutes for the analysis. These were the 130 solutes described in the EUTox database and an additional 12 compounds referred to in publications [15,23,24] as potential uremic toxins. Among them, 54 compounds were found in the KEGG database as products of, or participants in, some biochemical reactions, and these uremic retention solutes were included in further analysis.
Using data from KEGG, we assigned each toxin to a specific enzymatic reaction as a product or substrate. Then, using the NIH Human Microbiome Project database, all the bacteria in the human microbiome were identified. Lastly, all the metabolic pathways for these bacteria were found in KEGG. As a result, we obtained a complete list of the toxins with enzymatic pathways found in the bacteria of the human microbiome that could produce these reactions.
Full data on bacterial synthetic and degrading enzymatic reactions for uremic retention solutes are available in the Supplementary Table S1, with detailed data on the bacterial strains and KEGG reactions involved, including 186,186 toxin-reaction–bacteria links. These data were then subjected to a more in-depth analysis, as presented below.

2.1. Toxins That Can Be Synthesized by the Microbiome

In Table 1, we summarize the data on toxins associated with the human gut microbiome: the number of bacteria from the gut microbiome that have reactions described in KEGG for the synthesis or metabolism of a given toxin, the number of different reactions for the given toxin in bacteria, and the number of KEGG-described enzymatic reactions for the given toxin in the human organism.
Based on the KEGG database, we showed that only eight uremic retention solutes had no attributed pathways in human metabolism but had ascribed enzymes in bacteria (Table 1, and extended version in Supplementary Table S2). These compounds are mannitol, phenol, trimethylamine, oxalate, creatinine, trimethylamine-N-oxide, pseudouridine, and 3-(3-hydroxyphenyl) propanoic acid. It should be noted that our analysis included only enzymatic reactions described in KEGG, and thus, compounds such as creatinine or oxalate that can be produced non-enzymatically in human organisms fall into this category. Other important exceptions in this table are trimethylamine and trimethylamine-N-oxide, since there is an enzyme (flavin monooxygenase) in the human liver that determines the conversion of trimethylamine into trimethylamine-N-oxide [25]. However, due to the limitations of the KEGG database, our analysis yielded zero enzymes for these uremic toxins.
We also tested whether the number of potential toxin-synthesizing bacteria correlated with the toxin concentrations in either healthy or uremic conditions. However, the correlation coefficients for both conditions were non-significant and near-zero, and thus no strong correlation was observed between toxin concentration and the number of bacteria producing it.

2.2. Toxins with the Least Abundant Synthesizing Human Enzymes

Besides uremic retention solutes, which have no annotated synthesizing enzymes in Homo sapiens, we found 33 toxins for which a few enzymes (from one to three enzymes) could be assigned as synthetic in the human organism (Table 1). To test whether these enzymes result in a meaningful production of given toxins in humans, we evaluated the amount of these enzymes in the human body.
From the Human Protein Atlas database, we extracted the mRNA abundance data in different tissues and normalized it with respect to average tissue weight. We obtained an approximate abundance for particular enzymes in the human organism using this approach. All human gene expression abundance in the whole organism demonstrates a bimodal distribution, with a large portion of genes being poorly represented and others showing an almost normal distribution. In Figure 1, green dashes indicate the position of uremic-retention-solutes-synthesizing enzymes. According to our analysis, the most abundant enzymes are responsible for synthesizing creatine, indole-3-acetic acid, nicotinamide, methylglyoxal, and S-adenosylhomocysteine. Thus, these toxins are expected to be predominantly produced by the human organism. The five least abundant enzymes are responsible for the synthesis of melatonin, hexanal, orotidine, a-keto-d-guanidinovaleric acid, and 3-hydroxyanthranilic acid. Based on these results, we suggest that the human metabolism might play an insignificant role in producing these five compounds and that they are mainly produced by the microbiome.

2.3. Bacteria with the Ability to Synthesize or Metabolize Uremic Retention Solutes

Data from KEGG also allowed us to address the questions about the various bacteria’s ability to synthesize uremic retention solutes. Using KEGG, it was possible to extract data on whether a toxin is “upstream” or “downstream” in the specific pathway in a reaction, which theoretically should correlate with the toxin being synthesized or metabolized in a given reaction. We summed the number of toxins that certain bacteria can potentially synthesize or metabolize, using the KEGG data. Complete data for 142 bacteria taxa are available in Supplementary Table S3. From this list, 70 bacteria can potentially synthesize more than 20 uremic retention solutes, while only 20 species can metabolize a similar number of solutes.
We also estimated how many toxins gut bacteria can synthesize without being able to metabolize the same toxin, and vice versa. Using this approach, we discovered that Brevundimonas sp., Campylobacter coli, Desulfovibrio sp., Oxalobacter formigenes, Campylobacter upsaliensis, Helicobacter pylori, Phascolarctobacterium faecium, and Desulfovibrio piger can synthesize more than 14 toxins without being able to metabolize the same toxins.
In contrast, the following bacteria can only metabolize more than seven toxins without being able to synthesize them: Pediococcus acidilactici, Listeria innocua, Listeria grayi, Lactobacillus ruminis, Klebsiella oxytoca, Clostridium sporogenes, Escherichia coli, Klebsiella sp., Enterococcus faecalis, Ruminococcaceae bacterium, and Parvimonas micra.
However, in general, the reaction direction of KEGG pathways is not strictly determined. Some KEGG reactions are included in separated “modules”, where the reaction direction is given. Sadly, only 46 of 336 reactions are included in KEGG modules. We added information from the MetaCyc database, which contains more detailed information on reaction directions. Combining these two databases, we were able to strictly determine the direction of 151 of 336 enzymatic reactions that are responsible for toxin synthesis/metabolism. Using this approach, we could more accurately identify the ability of bacteria to synthesize (without decomposing) or decompose (without synthesizing) individual toxins. The most significant taxons of synthesizing and decomposing bacteria are given in Table 2; a complete list is available in Supplementary Table S3. Note that several bacterial taxa were included both in the list of toxin producers and in the list of toxin consumers. This apparent contradiction is explained by the fact that some species (genera) of bacteria can synthesize certain toxins but at the same time consume (that is, remove) some other toxins. For example, Pediococcus acidilactici has metabolic pathways for the synthesis of homocysteine, indole-3-acetic acid, myoinositol, nicotinamide, and y-guanidinobutyric acid and at the same time is able to consume creatinine, cytidine, methylglyoxal, S-adenosylhomocysteine, sorbitol, and xanthosine. Specific uremic retention solutes synthesized/decomposed by a particular bacterium can be found in Supplementary Table S4.

2.4. Human Microbiome Project metadata analysis

To check the data obtained from KEGG and bacterial genomes, we used the Human Microbiome Project (HMP2) database, which contains transcriptomic and proteomic data for the microbiota of 735 people. Using this approach, we determined which bacteria are present in the microbiome of what percentage of HMP2 patients, what mRNA of the uremic-retention-solutes-synthesizing enzymes are detected in the metatranscriptome, and for which of them the protein was detected in the proteome. The complete dataset can be seen in Supplementary Table S5.
Among uremic retention solutes, 46 had the mRNA of at least one potentially synthesizing enzyme detected in at least one patient’s microbiome. Twenty-four of them had at least one enzyme’s mRNA detected in 75% of all patients (Table 3). Thirteen toxins had at least one enzyme detected in less than 10% of patients, and these were: trimethylamine-N-oxide, pseudouridine, p-cresyl sulfate, 3-(3-hydroxyphenyl) propanoic acid, a-keto-d-guanidinovaleric acid, indole-3-acetic acid, y-guanidinobutyric acid, hyaluronic acid (hyaluronan), gentisic acid, kinurenine, dimethylamine, kynurenic acid, asymmetric dimethylarginine (ADMA), and taurocyamine.
We used KEGG data to determine which reactions could potentially yield uremic retention solutes and, using the EC code for enzymes, we determined the enzymes in the metatranscriptome that might be involved in uremic retention solutes synthesis (Table 4). For each of 173 such uremic-retention-solutes-synthesizing enzymes detected in the metatranscriptome, there was a calculated ratio of patients who had mRNA or protein detected in the microbiome samples. We analyzed the mRNA count to determine the abundance of each mRNA type. We ascribed all enzymes to quartiles according to the distribution of all mRNAs, which roughly described the abundances of mRNA in the metatranscriptome. The mRNAs for 64 uremic-retention-solutes-synthesizing enzymes were detected in more than 50% of patients, but only 9 of them had a protein detected in more than 20% of patients. A total of 21 enzymes’ mRNAs were ascribed to more than 100 different bacteria species, while 96 of them were ascribed to fewer than 10. The most common enzymes were associated with the synthesis of arginic acid, orotidine, hypoxanthine, inosine, nicotinamide, xanthine, xanthosine, methylglyoxal, and homocysteine.
Finally, we checked in what percentage of patients each bacterium was present and for how many toxins they expressed the mRNAs of enzymes (Table 5). A total of 313 bacteria species were detected in this database. The top 20 bacteria by frequency of occurrence in patients are listed in Table 5, and the full list can be found in Supplementary Table S5. Of these, 27 bacteria were present in more than 50% of patients and 93 in less than 10%. It should be noted that the abundance of bacteria increases the possibility of mRNA detection. Thus, rare bacteria have fewer “uremic enzymes” detected, but this can be explained by the detection threshold. The most common bacteria were Ruminococcus torques, Faecalibacterium prausnitzii, Eubacterium rectale, Bacteroides uniformis, Bacteroides ovatus, Bacteroides vulgatus, and Ruminococcus obeum. Escherichia coli and Klebsiella pneumoniae could synthesize the highest number of different toxins, at 31 and 30, respectively.

3. Discussion

The main goal of this study was to undertake a bioinformatic analysis of the possible contribution of the intestinal microbiota to the synthesis of uremic toxins and the development of uremia in renal failure conditions. To date, several experimental studies have demonstrated the possibility of such a link. In a recent study by Gryp et al. [26], bacteria that might synthesize uremic toxins and/or their precursors (p-cresyl sulfate, indoxyl sulfate, indole-3-acetic acid) were isolated and analyzed. The study revealed that some bacteria can release uremic toxins and/or precursors to the culture medium. These results correlate with our findings. For example, Lachnospiraceae, found by Gryp et al. to be one of the uremic-toxins-synthesizing bacteria, is prominently present in our genomic and transcriptomic analysis. The excellent study mentioned here highlighted some points that indicated the relevance of our bioinformatic analysis. As noted by Gryp et al., there was no direct correlation between bacteria with known capacities to synthesize uremic solutes and CKD-provoked microbiota patterns (although there was some correlation for dialysis patients). Note that this type of research requires complex protocol optimization for specific bacteria to be cultivated. Among all microbiota bacteria species, only a small fraction of gut bacterial species can be cultivated, and thus, many bacteria will not be checked in vitro for uremic toxins production. Studies focusing on clinical data from CKD patients have one fundamental limitation: they focus only on bacteria that change (increase or decrease) during CKD. However, these bacteria can produce uremic toxins, use them as substrates, or benefit/suffer from CKD-associated conditions not directly related to kidney dysfunction. Indeed, toxin-producing bacteria do not score significantly higher in CKD patients, as Gryp et al. found. For this reason, we focused on the theoretical capacity of bacteria to produce uremic toxins without focusing on their association with CKD or their abundance, thus bypassing the indicated limitations. We believe that such an approach can help find promising new targets (both bacteria and enzymes) for analysis, which might otherwise be hidden “beneath” major CKD-affected bacteria species.
We collected data on the potential for production of uremic retention solutes for each bacterium that could be found in the microbiome. The goal of this analysis was to identify “bad” and “good” bacteria in terms of uremic toxin production, i.e., to present putative targets for therapy of uremia. Of course, the fact that gut bacteria can potentially synthesize the toxin does not mean that the toxin inevitably enters the bloodstream since bacteria (the same or others) can consume it. Therefore, we also included the ability to metabolize toxins in our analysis. Predictably, the bacteria were the same as those synthesizing the toxin in most cases since the identified solute was just an intermediate in specific metabolic pathways. Theoretically, it is also possible to determine what bacteria can decompose some toxins without being able to synthesize them and, vice versa, what bacteria can synthesize certain toxins but do not have downstream metabolizing enzymes. Available databases do not include enough information for a full-scale analysis of this type.
Nevertheless, our approach still gives some promising results. We defined Desulfovibrio piger as a potentially “bad” bacterium, synthesizing more toxins than it can degrade. Aronov and colleagues showed that Desulfovibrio piger is associated with a bad prognosis for uremia [14]. However, in the HMP2 database, Desulfovibrio piger was detected only in 5% of people, while it expresses enzymes for the potential synthesis of up to 18 toxins. Eggerthella lenta was also a “bad”, toxin-synthesizing bacterium in our analysis, and in parallel with our study it was shown that this bacteria is associated with increased production of uremic toxins [27]. Eggerthella lenta was even less abundant in HMP2 patients: it was present only in 1.2% of people. Another “bad” bacterium Campylobacter upsaliensis, which was one of the most prominent “toxin-synthesizing bacteria” in our analysis, has been associated with uremic syndrome [28]. In a case study, “bad” Aeromonas veronii was described as causing uremic syndrome [29]. Among the potentially beneficial “toxin-decomposing” bacteria we identified Lactobacillus casei, Lactobacillus rhamnosus and Lactobacillus acidophilus, which were earlier described as a probiotic treatment for kidney failure. Despite the fact that none of them were common in HMP2 patients, their use as probiotics demonstrated positive effects [30,31,32,33,34,35]. Helicobacter pylori, which is conventionally considered to be a malignant bacterium, was also among the bacteria that can synthesize more toxins than they consume; however, there are no direct links between it and kidney diseases [36].
Pediococcus acidilactici is an excellent example of the limitations of the described approach. In our analysis, it was among both the synthesizing bacteria (it is able to synthesize five toxins without metabolizing them) and the toxin-metabolizing bacteria (six toxins). As mentioned above in the Results section, these are not the same toxins. The bacterium can synthesize homocysteine, indole-3-acetic acid, myoinositol, nicotinamide, and y-guanidinobutyric acid and can metabolize creatinine and cytidine methylglyoxal, S-adenosylhomocysteine, sorbitol, and xanthosine. Moreover, it can both synthesize and decompose nine more uremic retention solutes, which means that these solutes are intermediates in some metabolic pathways. For Pediococcus acidilactici, there exists a clinical trial where it was included in probiotic treatment for uremic patients, and thus, it could be defined as a “good” bacterium [37]. However, from these examples, one can easily see that it is quite possible to miss valid targets due to their scarcity in databases or a very broad range of accessible biochemical reactions for certain bacteria.
Moreover, it was found that the total presence of Klebsiella and Escherichia coli was significantly increased in the intestine of uremic patients [38]. Indeed, we identified these bacteria among those “metabolizing but not synthesizing” toxins. Using HMP2 data we have found, that Escherichia coli and Klebsiella pneumoniae are top-2 bacteria by number of detected enzymes’ mRNA which catalyze reactions with uremic retention solutes: 92 and 72 reactions for 31 and 30 different uremic toxins, accordingly. These findings perfectly aligns with mentioned study [38]. We expect such bacteria will benefit from an increase in uremic retention solute concentrations since they can utilize them during the reverse transport to the intestine from the blood. However, it is hard to distinguish the causality, i.e., whether the toxin increases because a bacterium produces it (and thus this is a “bad” bacterium) or whether the amount increases because it can utilize the toxin (and thus it is a “beneficial” one).
In this regard, it should be noted that bacteria can use uremic retention solutes as nutrients and thus eliminate hazardous solutes. For instance, some bacteria express urease (Pseudomonas spp.), the enzyme that catalyzes the hydrolysis of urea, or urate oxidase (Clostridia spp.), oxidizing uric acid [39]. Toxic substances such as oxalate and creatinine, when released to the gut, can also be subsequently metabolized by microbiota [40,41]. Microbiota species from the genera Oxalobacter, Lactobacillus, Bifidobacterium, Enterococcus, and Eubacterium, which are present in normal microbiota, are capable of degrading oxalate, hence diminishing its uremic accumulation [42]. According to our data, 39 bacterial species from the human microbiome are involved in compound breakdown.
We primarily focused on the microbiome’s effects on uremia development during our analysis. However, gut bacteria are also affected by uremic toxins. Uremia and chronic kidney disease alter the biochemical environment of the gut mainly through the entrance of urea into the intestine [43]. Uremic toxins are considered as a trigger for the systemic inflammation accompanying uremia [44]. A variety of conditions, including decreased pH and increased oxygen concentration, are observed in the intestinal lumen in CKD patients. Since the normal intestinal microbiota is sensitive to the milieu and predominantly consists of obligate anaerobes [45], such conditions are inappropriate for it. Hence, the normal symbiotic relationship between the microbiota and the host breaks down and results in dysbiosis [46]. The data regarding the total number of bacteria in healthy individuals and people with severe kidney disease show that the number of gut aerobes (Enterobacteria, Enterococci) is about 100 times higher in sick patients than in healthy ones, whereas the numbers of some anaerobes (Bifidobacteria) are decreased [38].
A comparison of microbial DNA isolated from the stools of healthy people and persons with end-stage renal disease (ESRD) reveals differences in 190 operational taxonomic units (OTUs) from 3 phyla and 13 families [43]. In general, the gut microbiota of CKD patients shows an abundance of pathogenic microflora, including some Clostridia, Bacteroidia, and Gammaproteobacteris (families Enterobacteriaceae, Halomonadaceae, Moraxellaceae, Pseudomonadaceae, Thiotrichaceae) [16,43,47]. These bacteria are able to survive in a transformed gut environment that has been changed due to the penetration of some uremic toxins into the gut. Some bacteria use uremic toxins as nutrients. For instance, urea could be utilized by urease, the enzyme that catalyzes the hydrolysis of urea (Pseudomonas spp.), and uric acid could be oxidized by urate oxidase (Clostridia spp.) [48]. Toxic substances such as oxalate and creatinine, when released to the gut, can be subsequently metabolized by changed microbiota also [40,49]. Thus, the uremic gut milieu works as a selective environment for microbiota resulting in uremia-associated biocenosis and, vice versa, uremic toxins can also be directly toxic to certain bacteria. For example, methylglyoxal is harmful to a number of microorganisms, both Gram-positive and Gram-negative species, since it reacts with the nucleophilic centers of macromolecules such as RNA, DNA, and proteins [50,51]. It has been suggested that, subsequently, the compound inhibits DNA replication and protein synthesis by interacting with the guanine residues of nucleic acids and their precursors [52]. In addition, methylglyoxal alters the formation of bacterial appendages, flagella, and fimbriae, resulting in impaired cell adherence and motility [53]. Nevertheless, some pathogenic species possess detoxification systems for this substance [54]. Probably, this is one of the major uremic toxins leading to a change in microflora under uremic conditions.
We also obtained a complete list of uremic toxins whose production is associated with the activity of various types of gut bacteria (Table 1). The relevance of the associations found for each solute is discussed below.
Trimethylamine-N-oxide and the closely related trimethylamine are well-known water-soluble uremic toxins. They are both considered to be microbiome-related [55]. Vanholder et al. estimated that they were among the most critical toxins due to demonstrated experimental and clinical toxic effects [56]. We found 141 human gut microbial species that theoretically contribute to trimethylamine-N-oxide production via two enzymatic reactions. On the other hand, mRNA for trimethylamine-N-oxide reductase, which can produce trimethylamine-N-oxide, was detected in the microbiota metatranscriptome of only 10% of patients (Table 4). This finding highlights the fact that bacterial enzyme abundance in a human microbiome is not an ideal predictor of the contribution to uremic toxin production. Apart from this, trimethylamine-N-oxide could be produced in the liver from trimethylamine by flavin monooxygenases, but since trimethylamine has a microbial origin the production of this and trimethylamine-N-oxide by the liver will at least in part depend on the intestinal microbiota. As well as renal failure, trimethylamine-N-oxide induces cardiovascular problems and stimulates upregulation of a variety of macrophage scavenger receptors related to atherosclerosis development [25]. However, the exact mechanism by which trimethylamine-N-oxide accumulation leads to atherosclerosis is still unproven [57]. Moreover, a link between increased levels of trimethylamine-N-oxide and the risk of a heart attack has been demonstrated [58]. Additionally, this compound is a suggested candidate for mediating type-2 diabetes mellitus [59,60].
3-(3-Hydroxyphenyl) propanoic acid was earlier described as being produced by the microbiome [61] and associated with schizophrenia and autism [62]. In our study, 103 bacteria were identified as able to produce this compound. Interestingly, this was again a uremic toxin produced by an enzyme that was present in only 10% of people in the HMP2 study (Table 4). However, in patients where this mRNA was detected, its quantity was in Q1 of all mRNA types in the transcriptome. Clinical observations in hemodialysis patients showed that this compound was reduced more than 10-fold in colectomy patients [15].
Another uremic toxin defined in our study as bacteria-derived, i.e., mannitol, is produced via a simple reaction in bacterial fructose and mannose metabolism. D-mannitol-1-phosphate phosphohydrolase catalyzes the hydrolysis of D-mannitol 1-phosphate to D-mannitol and phosphate. We found 227 gut bacteria species that produce the reaction. In contrast to trimethylamine-N-oxide and pseudouridine, mannitol can be further metabolized by several bacterial enzymes. However, the ratio of its synthesis to its metabolism is unknown. At the organism level, mannitol causes over-diuresis, with consequent dehydration [63]. Moreover, it demonstrates cytotoxicity to renal tubular epithelial cells, destroying the cell cytoskeleton [64].
We identified five specific bacterial enzymes that catalyze the production of phenol (an uremic solute with well-documented adverse effects in humans) causing protein denaturation with subsequent spreading necrosis [65]. Altogether, 192 gut bacterial species possess at least one of these enzymes. Phenol could be produced from tyrosine by microbial tyrosine phenol-lyase or may be synthesized from 4-hydroxybenzoate by 4-hydroxybenzoate decarboxylase or from catechol by phenol 2-monooxygenase, as parts of the aminobenzoate degradation pathway. The mRNA for three of these enzymes was found in 90% of people, and mRNA for arylesterase was in Q1 of all mRNA types for most patients.
Oxalate is a toxin enzymatically produced by bacteria, while humans synthesize it non-enzymatically. Oxalate is formed during purine metabolism when bacterial oxamate amidohydrolase catalyzes the transformation of oxamate to oxalate. We detected this reaction in 76 prokaryotic species from the human microbiome. In the HMP2 data, mRNA for oxalate-synthesizing enzymes was detected in the metatranscriptomes of 23% of patients, and all of these mRNA types were in Q1 of all mRNA types. As already mentioned, humans can produce oxalate non-enzymatically: for instance, ascorbic acid is metabolized in the human body with oxalate as an output [66]. Furthermore, significant amounts of oxalate enter the organism with food. The substance is known as the main component of kidney stones [67], and it modulates the immune system through an induced synthesis of cytokines, chemoattractants, and other inflammatory signal molecules, causing degradation of IκBα in proximal tubular cells [68]. It also has an unfavorable impact on mitochondrial function [69]. Note that the enzymes identified by database-based approaches need additional validation: oxalate and creatinine were regarded in this approach as “non-human origin” toxins, but they are produced in human organisms non-enzymatically, like pseudouridine, which is one of the main RNA catabolites [70,71,72].
In addition to toxins for which no potential synthetic pathways have been found in human metabolism, we showed that for several toxins there are only between one and three human enzymes that can potentially synthesize them. However, the question remains regarding whether these enzymes contribute significantly to toxins synthesis, i.e., how many of these enzymes there are in the organism. We suggested that an optional way to evaluate enzyme abundance in the whole organism was to use the Human Protein Atlas database, normalizing data on the mRNA abundance with respect to tissue weight. While understanding all the limitations of this approach, we believe that it can provide helpful information. Thus, we estimated that some of these human toxin-producing enzymes presented at a reasonable level in humans compared to others. Therefore, we propose that these enzymes have a low contribution to uremic retention solutes, and the role of the microbiota predominates in the synthesis of corresponding substances. For example, the orotidine concentration in the blood of uremic patients is quite high (1.20 (+/−1.60) mg/L), while the enzyme orotate phosphoribosyltransferase does not abound in human organisms (Figure 1). It is important to note that many uremic retention solutes are metabolically connected substances. One of the most important uremic retention solutes synthesis pathways is tryptophan metabolism through indolic and kynurenine pathways [73]. The following toxins are generated through this pathway: indoxyl sulfate (IS), indoxyl-β-d-glucuronide, and indole-3-acetic acid (IAA). Predictably, bacteria that have enzymes responsible for synthesizing one of the connected toxins usually have a whole pathway. It can be seen in Figure 2 that there is a noticeable clusterization between solutes of a similar chemical nature. For example, xanthosine, xanthine, hypoxanthine, and inosine, which are interconnected via one enzymatic reaction in the purine metabolism pathway, are clustered together via common bacteria species. This illustrates exactly the indicated phenomenon: toxins of a similar chemical nature are synthesized by the same types of bacteria.

Limitations

The crucial limitation of this approach is database completeness. Not all uremic retention solutes were found in the KEGG database, enzymatic reactions were not always listed, and not all reaction directions (synthesis or degradation) were given. However, most of these databases are updated annually. We hope that it will be possible to include even more important details in the analysis, such as enzymatic reaction constants, bacteria abundance in the microbiome, and others.

4. Materials and Methods

4.1. Database Usage

The uremic toxins list was taken from the EUTox database (https://database.uremic-toxins.org/home.php, last accessed 1 December 2020) and contained 134 compounds. Additionally, 12 compounds were added after the literature analysis.
The Human Metabolome Database (https://hmdb.ca/, last accessed 1 December 2020) [22] was used to check the origins of toxins.
The human microbiome bacteria list was obtained from the NIH Human Microbiome Project (https://commonfund.nih.gov/hmp, last accessed 1 December 2020) [74]. Different strains were collapsed into a single data point for a bacterial species if possible. A list of 500 bacteria was obtained.
The Human Protein Atlas (https://www.proteinatlas.org/, last accessed 1 December 2020) was used to roughly estimate the abundance of enzymes in the human organism, using data on mRNA levels and average tissue weights.
The MetaCyc database (https://metacyc.org/, last accessed 1 December 2020) [75] was used to manually check the directions of the reactions if this information was not available in KEGG.
The Human Microbiome Project (HMP2) (https://ibdmdb.org/, last accessed 1 August 2021) [76] contains multi-omics data on the human microbiome in healthy and ill patients.

4.2. Analysis

We analyzed 130 uremic toxins from the European Uremic Solutes Database (EUTox-DB) plus 12 from the literature analysis. For 97 toxins, we found identifiers in the KEGG database (release: 1 July 2020) [77].
The receipt and processing of information obtained from the KEGG database were carried out using the R package KEGGREST (version 1.26.1) [78].
A total of 69 toxins were involved in any KEGG reaction, and 54 toxins were involved in enzymatic reactions for which the metabolic pathway is known.
A total of 482 bacteria were identified in KEGG for which the genus, species, and strain (or only genus and species) corresponded to the bacteria from the HMP list. All enzymes involved in all metabolic pathways described for each bacterium from the list and separately for humans were obtained from the KEGG database.
We selected 67 enzymes that can metabolize uremic toxins and have a KEGG reaction ID and a KEGG EC ID. For each of the 67 enzymes, we received information about its identification at the protein (Proteomics section) and transcriptome (Metatranscriptomes section) levels by the donor or by bacteria, using HMP2 data.
All data manipulations and visualizations were performed using R version 3.6.3 and the packages tidyverse (version 1.3.1) and data.table (version 1.14.2).

4.3. Code Availability

All code used in the pipeline is available in GitHub at https://github.com/Zharikova/toxins, last accessed 24 October 2021.

5. Conclusions

We strongly recommend that our data and approach are not considered to be solid evidence of some bacteria being beneficial or malignant in conditions of uremia, but rather that they represent a tool for new insights during experimental analysis. We believe that even in limited form these approaches can contribute to the study of uremia-related bacteria and might contribute to the development of uremia treatment, suggesting new target bacteria and critical enzymatic reactions.

Supplementary Materials

The following are available online at https://www.mdpi.com/article/10.3390/ijms23010483/s1. Figure S1: Expanded cluster heatmap, which represents a clustering of toxins by bacteria that can potentially synthesize them. Table S1: All found toxin-reaction–bacteria connections. Table S2: Aggregated data on toxin-reaction–bacteria connections. Table S3: Full aggregated data on bacteria’s ability to synthesize or metabolize toxins. Table S4: List of toxins synthesized or metabolized by each bacterium. Table S5: Analysis of the presence of mRNA and proteins of potential uremic-toxins-synthesizing enzymes in metatranscriptomic and proteomic data from Human Microbiome Project database.

Author Contributions

Conceptualization E.Y.P.; methodology, A.A.Z.; software, A.A.Z.; validation, E.Y.P. and E.A.D.; formal analysis, V.A.P., E.A.D., and N.V.A.; writing—original draft preparation, V.A.P. and E.A.D.; writing—review and editing V.A.P.; supervision, D.B.Z.; project administration, E.Y.P. All authors have read and agreed to the published version of the manuscript.

Funding

The study was supported by the Russian Science Foundation (grant #21-75-30009).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Acknowledgments

The computational resources of the Makarich HPC cluster were provided by the A.N.Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University, Moscow, Russia.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Levin, A.; Tonelli, M.; Bonventre, J.; Coresh, J.; Donner, J.-A.; Fogo, A.B.; Fox, C.S.; Gansevoort, R.T.; Heerspink, H.J.L.; Jardine, M.; et al. Global kidney health 2017 and beyond: A roadmap for closing gaps in care, research, and policy. Lancet 2017, 390, 1888–1917. [Google Scholar] [CrossRef]
  2. Jha, V.; Garcia-Garcia, G.; Iseki, K.; Li, Z.; Naicker, S.; Plattner, B.; Saran, R.; Wang, A.Y.-M.; Yang, C.-W. Chronic kidney disease: Global dimension and perspectives. Lancet 2013, 382, 260–272. [Google Scholar] [CrossRef]
  3. Vanholder, R.; De Smet, R.; Glorieux, G.; Argilés, A.; Baurmeister, U.; Brunet, P.; Clark, W.; Cohen, G.; De Deyn, P.P.; Deppisch, R.; et al. Review on uremic toxins: Classification, concentration, and interindividual variability. Kidney Int. 2003, 63, 1934–1943. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  4. Meyer, T.W.; Hostetter, T.H. Uremia. N. Engl. J. Med. 2007, 357, 1316–1325. [Google Scholar] [CrossRef]
  5. Lau, W.L.; Savoj, J.; Nakata, M.B.; Vaziri, N.D. Altered microbiome in chronic kidney disease: Systemic effects of gut-derived uremic toxins. Clin. Sci. 2018, 132, 509–522. [Google Scholar] [CrossRef] [Green Version]
  6. Meijers, B.; Glorieux, G.; Poesen, R.; Bakker, S.J.L. Nonextracorporeal methods for decreasing uremic solute concentration: A future way to go? Semin. Nephrol. 2014, 34, 228–243. [Google Scholar] [CrossRef]
  7. Vanholder, R.; Baurmeister, U.; Brunet, P.; Cohen, G.; Glorieux, G.; Jankowski, J. European uremic toxin work group a bench to bedside view of uremic toxins. J. Am. Soc. Nephrol. 2008, 19, 863–870. [Google Scholar] [CrossRef] [Green Version]
  8. Meert, N.; Schepers, E.; De Smet, R.; Argiles, A.; Cohen, G.; Deppisch, R.; Drüeke, T.; Massy, Z.; Spasovski, G.; Stegmayr, B.; et al. Inconsistency of reported uremic toxin concentrations. Artif. Organs 2007, 31, 600–611. [Google Scholar] [CrossRef]
  9. Duranton, F.; Cohen, G.; De Smet, R.; Rodriguez, M.; Jankowski, J.; Vanholder, R.; Argiles, A. European uremic toxin work group normal and pathologic concentrations of uremic toxins. J. Am. Soc. Nephrol. 2012, 23, 1258–1270. [Google Scholar] [CrossRef] [Green Version]
  10. Almeras, C.; Argilés, A. The general picture of uremia. Semin. Dial. 2009, 22, 329–333. [Google Scholar] [CrossRef]
  11. Qin, J.; Li, R.; Raes, J.; Arumugam, M.; Burgdorf, K.S.; Manichanh, C.; Nielsen, T.; Pons, N.; Levenez, F.; Yamada, T.; et al. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 2010, 464, 59–65. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  12. Einheber, A.; Carter, D. The role of the microbial flora in uremia. I. Survival times of germfree, limited-flora, and conventionalized rats after bilateral nephrectomy and fasting. J. Exp. Med. 1966, 123, 239–250. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  13. Wikoff, W.R.; Anfora, A.T.; Liu, J.; Schultz, P.G.; Lesley, S.A.; Peters, E.C.; Siuzdak, G. Metabolomics analysis reveals large effects of gut microflora on mammalian blood metabolites. Proc. Natl. Acad. Sci. USA 2009, 106, 3698–3703. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  14. Aronov, P.A.; Luo, F.J.-G.; Plummer, N.S.; Quan, Z.; Holmes, S.; Hostetter, T.H.; Meyer, T.W. Colonic contribution to uremic solutes. J. Am. Soc. Nephrol. 2011, 22, 1769–1776. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  15. Mair, R.D.; Sirich, T.L.; Plummer, N.S.; Meyer, T.W. Characteristics of colon-derived uremic solutes. Clin. J. Am. Soc. Nephrol. 2018, 13, 1398–1404. [Google Scholar] [CrossRef] [Green Version]
  16. Kikuchi, M.; Ueno, M.; Itoh, Y.; Suda, W.; Hattori, M. Uremic toxin-producing gut microbiota in rats with chronic kidney disease. Nephron 2017, 135, 51–60. [Google Scholar] [CrossRef]
  17. Ramezani, A.; Massy, Z.A.; Meijers, B.; Evenepoel, P.; Vanholder, R.; Raj, D.S. Role of the gut microbiome in uremia: A potential therapeutic target. Am. J. Kidney Dis. 2016, 67, 483–498. [Google Scholar] [CrossRef] [Green Version]
  18. Gryp, T.; De Paepe, K.; Vanholder, R.; Kerckhof, F.-M.; Van Biesen, W.; Van de Wiele, T.; Verbeke, F.; Speeckaert, M.; Joossens, M.; Couttenye, M.M.; et al. Gut microbiota generation of protein-bound uremic toxins and related metabolites is not altered at different stages of chronic kidney disease. Kidney Int. 2020, 97, 1230–1242. [Google Scholar] [CrossRef] [Green Version]
  19. Wishart, D.S.; Tzur, D.; Knox, C.; Eisner, R.; Guo, A.C.; Young, N.; Cheng, D.; Jewell, K.; Arndt, D.; Sawhney, S.; et al. HMDB: The human metabolome database. Nucleic Acids Res. 2007, 35, D521–6. [Google Scholar] [CrossRef] [PubMed]
  20. Wishart, D.S.; Knox, C.; Guo, A.C.; Eisner, R.; Young, N.; Gautam, B.; Hau, D.D.; Psychogios, N.; Dong, E.; Bouatra, S.; et al. HMDB: A knowledgebase for the human metabolome. Nucleic Acids Res. 2009, 37, D603–10. [Google Scholar] [CrossRef]
  21. Wishart, D.S.; Jewison, T.; Guo, A.C.; Wilson, M.; Knox, C.; Liu, Y.; Djoumbou, Y.; Mandal, R.; Aziat, F.; Dong, E.; et al. HMDB 3.0—The human metabolome database in 2013. Nucleic Acids Res. 2013, 41, D801–7. [Google Scholar] [CrossRef] [PubMed]
  22. Wishart, D.S.; Feunang, Y.D.; Marcu, A.; Guo, A.C.; Liang, K.; Vázquez-Fresno, R.; Sajed, T.; Johnson, D.; Li, C.; Karu, N.; et al. HMDB 4.0: The human metabolome database for 2018. Nucleic Acids Res. 2018, 46, D608–D617. [Google Scholar] [CrossRef] [PubMed]
  23. Addi, T.; Dou, L.; Burtey, S. Tryptophan-derived uremic toxins and thrombosis in chronic kidney disease. Toxins 2018, 10, 412. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  24. Tanaka, H.; Sirich, T.L.; Meyer, T.W. Uremic solutes produced by colon microbes. Blood Purif. 2015, 40, 306–311. [Google Scholar] [CrossRef]
  25. Wang, Z.; Klipfell, E.; Bennett, B.J.; Koeth, R.; Levison, B.S.; Dugar, B.; Feldstein, A.E.; Britt, E.B.; Fu, X.; Chung, Y.-M.; et al. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease. Nature 2011, 472, 57–63. [Google Scholar] [CrossRef] [Green Version]
  26. Gryp, T.; Huys, G.R.B.; Joossens, M.; Van Biesen, W.; Glorieux, G.; Vaneechoutte, M. Isolation and quantification of uremic toxin precursor-generating gut bacteria in chronic kidney disease patients. Int. J. Mol. Sci. 2020, 21, 1986. [Google Scholar] [CrossRef] [Green Version]
  27. Wang, X.; Yang, S.; Li, S.; Zhao, L.; Hao, Y.; Qin, J.; Zhang, L.; Zhang, C.; Bian, W.; Zuo, L.; et al. Aberrant gut microbiota alters host metabolome and impacts renal failure in humans and rodents. Gut 2020, 69, 2131–2142. [Google Scholar] [CrossRef]
  28. Carter, J.E.; Cimolai, N. Hemolytic-uremic syndrome associated with acute campylobacter upsaliensis gastroenteritis. Nephron 1996, 74, 489. [Google Scholar] [CrossRef]
  29. Figueras, M.J.; Aldea, M.J.; Fernández, N.; Aspíroz, C.; Alperi, A.; Guarro, J. Aeromonas hemolytic uremic syndrome. A case and a review of the literature. Diagn. Microbiol. Infect. Dis. 2007, 58, 231–234. [Google Scholar] [CrossRef]
  30. Firouzi, S.; Mohd-Yusof, B.-N.; Majid, H.-A.; Ismail, A.; Kamaruddin, N.-A. Effect of microbial cell preparation on renal profile and liver function among type 2 diabetics: A randomized controlled trial. BMC Complement. Altern. Med. 2015, 15, 433. [Google Scholar] [CrossRef] [Green Version]
  31. Cruz-Mora, J.; Martínez-Hernández, N.E.; del Campo-López, F.M.; Viramontes-Hörner, D.; Vizmanos-Lamotte, B.; Muñoz-Valle, J.F.; García-García, G.; Parra-Rojas, I.; Castro-Alarcón, N. Effects of a symbiotic on gut microbiota in mexican patients with end-stage renal disease. J. Ren. Nutr. 2014, 24, 330–335. [Google Scholar] [CrossRef]
  32. Guida, B.; Germanò, R.; Trio, R.; Russo, D.; Memoli, B.; Grumetto, L.; Barbato, F.; Cataldi, M. Effect of short-term synbiotic treatment on plasma p-cresol levels in patients with chronic renal failure: A randomized clinical trial. Nutr. Metab. Cardiovasc. Dis. 2014, 24, 1043–1049. [Google Scholar] [CrossRef]
  33. Fagundes, R.A.B.; Soder, T.F.; Grokoski, K.C.; Benetti, F.; Mendes, R.H. Probiotics in the treatment of chronic kidney disease: A systematic review. J. Bras. Nefrol. 2018, 40, 278–286. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  34. Alatriste, P.V.M.; Arronte, R.U.; Espinosa, C.O.G.; Cuevas, M.d.L.Á.E. Effect of probiotics on human blood urea levels in patients with chronic renal failure. Nutr. Hosp. 2014, 29, 582–590. [Google Scholar]
  35. Dehghani, H.; Heidari, F.; Mozaffari-Khosravi, H.; Nouri-Majelan, N.; Dehghani, A. Synbiotic supplementations for azotemia in patients with chronic kidney disease: A randomized controlled trial. Iran. J. Kidney Dis. 2016, 10, 351–357. [Google Scholar]
  36. Sugimoto, M.; Yamaoka, Y. Review of Helicobacter pylori infection and chronic renal failure. Ther. Apher. Dial. 2011, 15, 1–9. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  37. Effect of Prebiotics and/or Probiotics on Uremic Toxins and Inflammation Markers in Peritoneal Dialysis Patients. Available online: https://clinicaltrials.gov/ct2/show/NCT03770611 (accessed on 12 July 2021).
  38. Hida, M.; Aiba, Y.; Sawamura, S.; Suzuki, N.; Satoh, T.; Koga, Y. Inhibition of the accumulation of uremic toxins in the blood and their precursors in the feces after oral administration of lebenin, a lactic acid bacteria preparation, to uremic patients undergoing hemodialysis. Nephron 1996, 74, 349–355. [Google Scholar] [CrossRef] [PubMed]
  39. Wong, J.; Piceno, Y.M.; DeSantis, T.Z.; Pahl, M.; Andersen, G.L.; Vaziri, N.D. Expansion of urease- and uricase-containing, indole- and p-cresol-forming and contraction of short-chain fatty acid-producing intestinal microbiota in ESRD. Am. J. Nephrol. 2014, 39, 230–237. [Google Scholar] [CrossRef] [Green Version]
  40. Hatch, M.; Freel, R.W.; Vaziri, N.D. Intestinal excretion of oxalate in chronic renal failure. J. Am. Soc. Nephrol. 1994, 5, 1339–1343. [Google Scholar] [CrossRef] [PubMed]
  41. Dunn, S.R.; Gabuzda, G.M.; Superdock, K.R.; Kolecki, R.S.; Schaedler, R.W.; Simenhoff, M.L. Induction of creatininase activity in chronic renal failure: Timing of creatinine degradation and effect of antibiotics. Am. J. Kidney Dis. 1997, 29, 72–77. [Google Scholar] [CrossRef]
  42. Miller, A.W.; Dearing, D. The metabolic and ecological interactions of oxalate-degrading bacteria in the mammalian gut. Pathogens 2013, 2, 636–652. [Google Scholar] [CrossRef] [Green Version]
  43. Vaziri, N.D.; Wong, J.; Pahl, M.; Piceno, Y.M.; Yuan, J.; DeSantis, T.Z.; Ni, Z.; Nguyen, T.-H.; Andersen, G.L. Chronic kidney disease alters intestinal microbial flora. Kidney Int. 2013, 83, 308–315. [Google Scholar] [CrossRef] [Green Version]
  44. Anders, H.-J.; Andersen, K.; Stecher, B. The intestinal microbiota, a leaky gut, and abnormal immunity in kidney disease. Kidney Int. 2013, 83, 1010–1016. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  45. Savage, D.C. Microbial ecology of the gastrointestinal tract. Annu. Rev. Microbiol. 1977, 31, 107–133. [Google Scholar] [CrossRef]
  46. Nigam, S.K.; Bush, K.T. Uraemic syndrome of chronic kidney disease: Altered remote sensing and signalling. Nat. Rev. Nephrol. 2019, 15, 301–316. [Google Scholar] [CrossRef]
  47. Levy, M.; Kolodziejczyk, A.A.; Thaiss, C.A.; Elinav, E. Dysbiosis and the immune system. Nat. Rev. Immunol. 2017, 17, 219–232. [Google Scholar] [CrossRef]
  48. Termén, S.; Tollin, M.; Rodriguez, E.; Sveinsdóttir, S.H.; Jóhannesson, B.; Cederlund, A.; Sjövall, J.; Agerberth, B.; Gudmundsson, G.H. PU.1 and bacterial metabolites regulate the human gene CAMP encoding antimicrobial peptide LL-37 in colon epithelial cells. Mol. Immunol. 2008, 45, 3947–3955. [Google Scholar] [CrossRef]
  49. Mitch, W.E.; Collier, V.U.; Walser, M. Creatinine metabolism in chronic renal failure. Clin. Sci. 1980, 58, 327–335. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  50. Papoulis, A.; Al-Abed, Y.; Bucala, R. Identification of N2-(1-carboxyethyl)guanine (CEG) as a guanine advanced glycosylation end product. Biochemistry 1995, 34, 648–655. [Google Scholar] [CrossRef] [PubMed]
  51. Lo, T.W.; Westwood, M.E.; McLellan, A.C.; Selwood, T. Binding and modification of proteins by methylglyoxal under physiological conditions. A kinetic and mechanistic study with N alpha-acetylarginine, N alpha-acetylcysteine, and N alpha-acetyllysine, and bovine serum albumin. J. Biol. 1994, 269, 32299–32305. Available online: https://www.jbc.org/content/269/51/32299.short (accessed on 24 October 2021). [CrossRef]
  52. Krymkiewicz, N.; Diéguez, E.; Rekarte, U.D. Properties and mode of action of a bactericidal compound (= methylglyoxal) produced by a mutant of Escherichia coli. J. Bacteriol. 1971, 108, 1338–1347. Available online: https://jb.asm.org/content/108/3/1338.short (accessed on 24 October 2021). [CrossRef] [PubMed] [Green Version]
  53. Rabie, E.; Serem, J.C.; Oberholzer, H.M.; Gaspar, A.R.M.; Bester, M.J. How methylglyoxal kills bacteria: An ultrastructural study. Ultrastruct. Pathol. 2016, 40, 107–111. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  54. Booth, I.R.; Ferguson, G.P.; Miller, S.; Li, C.; Gunasekera, B.; Kinghorn, S. Bacterial production of methylglyoxal: A survival strategy or death by misadventure? Biochem. Soc. Trans. 2003, 31, 1406–1408. [Google Scholar] [CrossRef] [PubMed]
  55. Mei, Z.; Chen, G.-C.; Wang, Z.; Usyk, M.; Yu, B.; Baeza, Y.V.; Humphrey, G.; Benitez, R.S.; Li, J.; Williams-Nguyen, J.S.; et al. Dietary factors, gut microbiota, and serum trimethylamine-N-oxide associated with cardiovascular disease in the hispanic community health study/study of latinos. Am. J. Clin. Nutr. 2021, 113, 1503–1514. [Google Scholar] [CrossRef]
  56. Vanholder, R.; Pletinck, A.; Schepers, E.; Glorieux, G. Biochemical and clinical impact of organic uremic retention solutes: A comprehensive update. Toxins 2018, 10, 33. [Google Scholar] [CrossRef] [Green Version]
  57. Velasquez, M.T.; Ramezani, A.; Manal, A.; Raj, D.S. Trimethylamine N-oxide: The good, the bad and the unknown. Toxins 2016, 8, 326. [Google Scholar] [CrossRef] [Green Version]
  58. Zhu, W.; Gregory, J.C.; Org, E.; Buffa, J.A.; Gupta, N.; Wang, Z.; Li, L.; Fu, X.; Wu, Y.; Mehrabian, M.; et al. Gut microbial metabolite TMAO enhances platelet hyperreactivity and thrombosis risk. Cell 2016, 165, 111–124. [Google Scholar] [CrossRef] [Green Version]
  59. Kim, Y.; Keogh, J.; Clifton, P. A Review of potential metabolic etiologies of the observed association between red meat consumption and development of type 2 diabetes mellitus. Metabolism 2015, 64, 768–779. [Google Scholar] [CrossRef]
  60. Nowiński, A.; Ufnal, M. Trimethylamine N-oxide: A harmful, protective or diagnostic marker in lifestyle diseases? Nutrition 2018, 46, 7–12. [Google Scholar] [CrossRef]
  61. Wang, D.; Ho, L.; Faith, J.; Ono, K.; Janle, E.M.; Lachcik, P.J.; Cooper, B.R.; Jannasch, A.H.; D’Arcy, B.R.; Williams, B.A.; et al. Role of intestinal microbiota in the generation of polyphenol-derived phenolic acid mediated attenuation of Alzheimer’s disease β-amyloid oligomerization. Mol. Nutr. Food Res. 2015, 59, 1025–1040. [Google Scholar] [CrossRef]
  62. Shaw, W. Increased urinary excretion of a 3-(3-hydroxyphenyl)-3-hydroxypropionic acid (HPHPA), an abnormal phenylalanine metabolite of Clostridia spp. in the gastrointestinal tract, in urine samples from patients with autism and schizophrenia. Nutr. Neurosci. 2010, 13, 135–143. [Google Scholar] [CrossRef] [PubMed]
  63. Volarevic, V.; Djokovic, B.; Jankovic, M.G.; Harrell, C.R.; Fellabaum, C.; Djonov, V.; Arsenijevic, N. Molecular mechanisms of cisplatin-induced nephrotoxicity: A balance on the knife edge between renoprotection and tumor toxicity. J. Biomed. Sci. 2019, 26, 25. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  64. Shi, J.; Qian, J.; Li, H.; Luo, H.; Luo, W.; Lin, Z. Renal tubular epithelial cells injury induced by mannitol and its potential mechanism. Ren. Fail. 2018, 40, 85–91. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  65. Downs, J.W.; Wills, B.K. Phenol toxicity. In StatPearls; StatPearls Publishing: Treasure Island, FL, USA, 2020. [Google Scholar]
  66. Knight, J.; Madduma-Liyanage, K.; Mobley, J.A.; Assimos, D.G.; Holmes, R.P. Ascorbic acid intake and oxalate synthesis. Urolithiasis 2016, 44, 289–297. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  67. Brzica, H.; Breljak, D.; Burckhardt, B.C.; Burckhardt, G.; Sabolić, I. Oxalate: From the environment to kidney stones. Arh. Hig. Rada Toksikol. 2013, 64, 609–630. [Google Scholar] [CrossRef] [Green Version]
  68. Jonassen, J.A.; Kohjimoto, Y.; Scheid, C.R.; Schmidt, M. Oxalate toxicity in renal cells. Urol. Res. 2005, 33, 329–339. [Google Scholar] [CrossRef]
  69. Cao, L.-C.; Honeyman, T.W.; Cooney, R.; Kennington, L.; Scheid, C.R.; Jonassen, J.A. Mitochondrial dysfunction is a primary event in renal cell oxalate toxicity. Kidney Int. 2004, 66, 1890–1900. [Google Scholar] [CrossRef] [Green Version]
  70. Feher, J. 7.4—Tubular reabsorption and secretion. In Quantitative Human Physiology, 2nd ed.; Feher, J., Ed.; Academic Press: Boston, MA, USA, 2017; pp. 719–729. ISBN 9780128008836. [Google Scholar]
  71. Fargue, S.; Milliner, D.S.; Knight, J.; Olson, J.B.; Lowther, W.T.; Holmes, R.P. Hydroxyproline metabolism and oxalate synthesis in primary hyperoxaluria. J. Am. Soc. Nephrol. 2018, 29, 1615–1623. [Google Scholar] [CrossRef] [Green Version]
  72. Penzo, M.; Guerrieri, A.N.; Zacchini, F.; Treré, D.; Montanaro, L. RNA pseudouridylation in physiology and medicine: For better and for worse. Genes 2017, 8, 301. [Google Scholar] [CrossRef] [Green Version]
  73. Sallée, M.; Dou, L.; Cerini, C.; Poitevin, S.; Brunet, P.; Burtey, S. The aryl hydrocarbon receptor-activating effect of uremic toxins from tryptophan metabolism: A new concept to understand cardiovascular complications of chronic kidney disease. Toxins 2014, 6, 934–949. [Google Scholar] [CrossRef]
  74. Consortium, T.H.M.P. The human microbiome project consortium structure, function and diversity of the healthy human microbiome. Nature 2012, 486, 207–214. [Google Scholar]
  75. Caspi, R.; Altman, T.; Billington, R.; Dreher, K.; Foerster, H.; Fulcher, C.A.; Holland, T.A.; Keseler, I.M.; Kothari, A.; Kubo, A.; et al. The MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of pathway/genome databases. Nucleic Acids Res. 2014, 42, D459–D471. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  76. Integrative HMP (iHMP) research network consortium the integrative human microbiome project. Nature 2019, 569, 641–648. [CrossRef] [PubMed] [Green Version]
  77. Kanehisa, M.; Sato, Y.; Furumichi, M.; Morishima, K.; Tanabe, M. New approach for understanding genome variations in KEGG. Nucleic Acids Res. 2019, 47, D590–D595. [Google Scholar] [CrossRef] [Green Version]
  78. KEGGREST. Available online: https://bioconductor.org/packages/release/bioc/html/KEGGREST.html (accessed on 20 May 2021).
Figure 1. Abundance of human mRNA expression in the whole organism: green lines—genes of uremic-retention-solutes-synthesizing enzymes.
Figure 1. Abundance of human mRNA expression in the whole organism: green lines—genes of uremic-retention-solutes-synthesizing enzymes.
Ijms 23 00483 g001
Figure 2. Cluster heatmap representing a clustering of toxins according to bacteria that can potentially synthesize them. Color represents the number of enzymatic reactions in given bacteria that can potentially lead to the synthesis of a toxin (see Supplementary Figure S1 for a high-resolution version with all bacteria names expanded).
Figure 2. Cluster heatmap representing a clustering of toxins according to bacteria that can potentially synthesize them. Color represents the number of enzymatic reactions in given bacteria that can potentially lead to the synthesis of a toxin (see Supplementary Figure S1 for a high-resolution version with all bacteria names expanded).
Ijms 23 00483 g002
Table 1. Potential synthesis and metabolism of uremic retention solutes by microbiome bacteria species and human organism.
Table 1. Potential synthesis and metabolism of uremic retention solutes by microbiome bacteria species and human organism.
ToxinNumber of Synthesizing or Metabolizing Bacteria SpeciesNumber of Different Reactions in KEGG Number of Enzymes in Human
Mannitol5640
Phenol3740
Trimethylamine1940
Oxalate2030
Creatinine7420
Trimethylamine-N-oxide1620
Pseudouridine2010
3-(3-Hydroxyphenyl) propanoic acid1310
S-Adenosylhomocysteine143351
Homocysteine141121
Argininic Acid143111
Putrescine119111
Methylglyoxal138101
Hypoxanthine13791
Urea6991
Xanthine14071
Nicotinamide14071
Cytidine13871
Uridine13871
anthranilic acid10871
Inosine14261
Indole-3-acetic acid11761
3-hydroxyanthranilic acid7061
a-keto-d-Guanidinovaleric Acid2761
Myoinositol11651
Phenylacetic acid10651
Sorbitol9451
Dimethylamine1551
Orotic Acid14341
Xanthosine14241
y-guanidinobutyric Acid5941
Monomethylamine2241
p-Cresyl sulfate8131
Uric Acid3431
Kinurenine2031
Orotidine14321
Quinolinic Acid8921
Creatine4421
Gentisic acid1221
N-Acetylhistamine8211
Hippuric acid1211
Taurocyamine911
Melatonin211
Ethylamine111
Table 2. Bacteria determined as potential producers or consumers of some uremic retention solutes using analysis of KEGG modules with the MetaCyc database. Bacteria presented have more than five toxin-producing reactions (without decomposing the same toxin) or more than four toxin-decomposing reactions (without synthesizing the same toxin). A complete list is available in Supplementary Table S3. Bacteria identified as both synthesizing and decomposing are presented in bold.
Table 2. Bacteria determined as potential producers or consumers of some uremic retention solutes using analysis of KEGG modules with the MetaCyc database. Bacteria presented have more than five toxin-producing reactions (without decomposing the same toxin) or more than four toxin-decomposing reactions (without synthesizing the same toxin). A complete list is available in Supplementary Table S3. Bacteria identified as both synthesizing and decomposing are presented in bold.
KEGG + MetaCyc
Synthesis
(of More Than Five Toxins)
Decomposition
(of More Than Four Toxins)
Oxalobacter formigenes
Stenotrophomonas sp.
Campylobacter coli
Brevundimonas sp.
Campylobacter upsaliensis
Geobacter sp.
Escherichia sp.
Rhizobium sp.
Desulfovibrio sp.
Methylobacterium sp.
Dialister pneumosintes
Helicobacter pylori
Desulfovibrio piger
Edwardsiella tarda
Morganella morganii
Gordonibacter pamelaeae
Clostridium sp.
Eggerthella lenta
Coprococcus catus
Pediococcus acidilactici
Lactobacillus fermentum
Corynebacterium ammoniagenes
Desulfitobacterium hafniense
Paenisporosarcina sp.
Paenibacillus sp.
Christensenella minuta
Aeromonas veronii
Propionibacterium sp.
Helicobacter bilis
Phascolarctobacterium faecium
Lachnospiraceae bacterium
Helicobacter cinaedi
Escherichia sp.
Rhizobium sp.
Pediococcus acidilactici
Lactobacillus ruminis
Stenotrophomonas sp.
Lactobacillus fermentum
Lactobacillus paracasei
Lactobacillus casei
Lactobacillus rhamnosus
Flavobacteriaceae bacterium
Klebsiella sp.
Lactobacillus amylolyticus
Corynebacterium ammoniagenes
Clostridium sporogenes
Sphingomonas sp.
Hafnia alvei
Escherichia coli
Providencia alcalifaciens
Providencia rustigianii
Enterobacter cloacae
Listeria grayi
Kocuria sp.
Lactobacillus acidophilus
Pseudomonas sp.
Klebsiella oxytoca
Lactobacillus helveticus
Table 3. Uremic retention solutes and their synthesizing enzymes’ mRNA abundance in HMP2. Presented uremic retention solutes had at least one enzyme’s mRNA represented in more than 90% of patients. The following information is given in parentheses for each enzyme: mRNA - percentage of patients who have mRNA for this enzyme; Q - quartile for this mRNA abundance; prot - percent of patients who have protein for this enzyme detected in proteome. The complete list is available in Supplementary Table S5.
Table 3. Uremic retention solutes and their synthesizing enzymes’ mRNA abundance in HMP2. Presented uremic retention solutes had at least one enzyme’s mRNA represented in more than 90% of patients. The following information is given in parentheses for each enzyme: mRNA - percentage of patients who have mRNA for this enzyme; Q - quartile for this mRNA abundance; prot - percent of patients who have protein for this enzyme detected in proteome. The complete list is available in Supplementary Table S5.
Uremic Retention SolutesNumber of Different Enzymes Where mRNA DetectedPercentage of Patients Who Have at Least One Enzyme’s mRNA for This Uremic Retention SoluteEnzymes
Argininic Acid1199.73%Nitric-oxide synthase (NADPH) (mRNA: 2.59%, Q1, prot: NA); Glycine amidinotransferase (mRNA: 0.68%, Q1, prot: NA); Arginine N-succinyltransferase (mRNA: 5.17%, Q1, prot: NA); Arginine--pyruvate transaminase (mRNA: 0.14%, Q1, prot: NA); Arginine kinase (mRNA: 9.12%, Q2, prot: NA); [Protein ADP-ribosylarginine] hydrolase (mRNA: 2.86%, Q2, prot: NA); Arginase (mRNA: 82.31%, Q2, prot: 0.22%); Arginine deiminase (mRNA: 83.27%, Q2, prot: 8.22%); Arginine decarboxylase (mRNA: 91.29%, Q2, prot: 32%); Argininosuccinate lyase (mRNA: 99.18%, Q4, prot: 14%); Arginine--tRNA ligase (mRNA: 99.59%, Q4, prot: 42%)
S-Adenosylhomocysteine6499.59%Homocysteine S-methyltransferase (mRNA: 4.49%, Q1, prot: NA); Protein-S-isoprenylcysteine O-methyltransferase (mRNA: 0%, QNA, prot: NA); Caffeoyl-CoA O-methyltransferase (mRNA: 8.98%, Q1, prot: NA); Uroporphyrinogen-III C-methyltransferase (mRNA: 52.65%, Q1, prot: 1.33%); Site-specific DNA-methyltransferase (cytosine-N(4)-specific) (mRNA: 2.45%, Q1, prot: NA); Precorrin-2 C(20)-methyltransferase (mRNA: 57.96%, Q1, prot: NA); Precorrin-3B C(17)-methyltransferase (mRNA: 78.91%, Q2, prot: NA); Precorrin-6B C(5,15)-methyltransferase (decarboxylating) (mRNA: 86.94%, Q2, prot: 0.22%); Precorrin-4 C(11)-methyltransferase (mRNA: 71.56%, Q1, prot: NA); Trans-aconitate 2-methyltransferase (mRNA: 8.98%, Q1, prot: 0.22%)... (see all 64 enzymes in ST 4)
Orotidine299.59%Orotate phosphoribosyltransferase (mRNA: 99.05%, Q4, prot: 35.78%); Orotidine-5’-phosphate decarboxylase (mRNA: 99.32%, Q4, prot: 63.78%)
Uridine799.46%Pyrimidine-nucleoside phosphorylase (mRNA: 50.88%, Q1, prot: 0.22%); Uridine phosphorylase (mRNA: 91.84%, Q2, prot: 4.67%); Uridine kinase (mRNA: 99.05%, Q4, prot: 2.89%); 5’-nucleotidase (mRNA: 98.64%, Q4, prot: 1.56%); 3’-nucleotidase (mRNA: 19.46%, Q1, prot: 0.89%); Uridine nucleosidase (mRNA: 0.54%, Q1, prot: NA); Cytidine deaminase (mRNA: 87.62%, Q2, prot: 11.56%)
Hypoxanthine899.32%Xanthine dehydrogenase (mRNA: 80%, Q2, prot: 0.89%); Purine-nucleoside phosphorylase (mRNA: 99.18%, Q4, prot: 62.67%); Thymidine phosphorylase (mRNA: 58.5%, Q1, prot: 0.89%); S-methyl-5’-thioinosine phosphorylase (mRNA: 0.41%, Q1, prot: NA); Hypoxanthine phosphoribosyltransferase (mRNA: 97.69%, Q3, prot: 6.89%); Purine nucleosidase (mRNA: 28.44%, Q1, prot: NA); Ribosylpyrimidine nucleosidase (mRNA: 3.67%, Q1, prot: NA); Adenine deaminase (mRNA: 96.46%, Q2, prot: 0.22%)
Cytidine699.32%Pyrimidine-nucleoside phosphorylase (mRNA: 50.88%, Q1, prot: 0.22%); Uridine kinase (mRNA: 99.05%, Q4, prot: 2.89%); 5’-nucleotidase (mRNA: 98.64%, Q4, prot: 1.56%); 3’-nucleotidase (mRNA: 19.46%, Q1, prot: 0.89%); Ribosylpyrimidine nucleosidase (mRNA: 3.67%, Q1, prot: NA); Cytidine deaminase (mRNA: 87.62%, Q2, prot: 11.56%)
Orotic Acid499.32%Dihydroorotate dehydrogenase (NAD(+)) (mRNA: 91.43%, Q3, prot: 14.44%); Dihydroorotate dehydrogenase (quinone) (mRNA: 93.47%, Q3, prot: 0.44%); Dihydroorotate oxidase (fumarate) (mRNA: 2.86%, Q1, prot: NA); Orotate phosphoribosyltransferase (mRNA: 99.05%, Q4, prot: 35.78%)
Nicotinamide799.18%Purine-nucleoside phosphorylase (mRNA: 99.18%, Q4, prot: 62.67%); Nicotinamide phosphoribosyltransferase (mRNA: 0.41%, Q1, prot: NA); NAD(+) ADP-ribosyltransferase (mRNA: 0.41%, Q2, prot: 0%); Purine nucleosidase (mRNA: 28.44%, Q1, prot: NA); Uridine nucleosidase (mRNA: 0.54%, Q1, prot: NA); ADP-ribosyl cyclase/cyclic ADP-ribose hydrolase (mRNA: 0%, QNA, prot: NA); Nicotinamidase (mRNA: 19.32%, Q1, prot: NA)
Xanthine699.18%Xanthine dehydrogenase (mRNA: 80%, Q2, prot: 0.89%); Purine-nucleoside phosphorylase (mRNA: 99.18%, Q4, prot: 62.67%); Xanthine phosphoribosyltransferase (mRNA: 98.37%, Q4, prot: 16.22%); Hypoxanthine phosphoribosyltransferase (mRNA: 97.69%, Q3, prot: 6.89%); Purine nucleosidase (mRNA: 28.44%, Q1, prot: NA); Guanine deaminase (mRNA: 42.04%, Q1, prot: 0%)
Inosine699.18%Purine-nucleoside phosphorylase (mRNA: 99.18%, Q4, prot: 62.67%); Inosine kinase (mRNA: 17.69%, Q1, prot: NA); 5’-nucleotidase (mRNA: 98.64%, Q4, prot: 1.56%); Purine nucleosidase (mRNA: 28.44%, Q1, prot: NA); Ribosylpyrimidine nucleosidase (mRNA: 3.67%, Q1, prot: NA); Adenosine deaminase (mRNA: 36.46%, Q1, prot: 0.22%)
Myoinositol599.18%Inositol 2-dehydrogenase (mRNA: 16.33%, Q1, prot: 0.89%); Galactinol--raffinose galactosyltransferase (mRNA: 0.27%, Q2, prot: NA); CDP-diacylglycerol--inositol 3-phosphatidyltransferase (mRNA: 3.54%, Q2, prot: NA); Inositol-phosphate phosphatase (mRNA: 87.48%, Q2, prot: 0.22%); Alpha-galactosidase (mRNA: 99.05%, Q3, prot: 11.56%)
Xanthosine399.18%Purine-nucleoside phosphorylase (mRNA: 99.18%, Q4, prot: 62.67%); 5’-nucleotidase (mRNA: 98.64%, Q4, prot: 1.56%); Purine nucleosidase (mRNA: 28.44%, Q1, prot: NA)
Methylglyoxal899.05%Aldehyde reductase (mRNA: 11.43%, Q1, prot: NA); Methylglyoxal reductase (NADPH) (mRNA: 5.71%, Q2, prot: NA); Glycerol dehydrogenase (mRNA: 54.01%, Q1, prot: 7.11%); Glyoxylate reductase (NADP(+)) (mRNA: 17.28%, Q1, prot: NA); Lactaldehyde dehydrogenase (mRNA: 24.76%, Q1, prot: 0.89%); D-lactate dehydratase (mRNA: 15.78%, Q1, prot: 0.22%); Methylglyoxal synthase (mRNA: 98.91%, Q4, prot: 58.22%); Lactoylglutathione lyase (mRNA: 66.8%, Q2, prot: 0.67%)
Homocysteine899.05%Homocysteine S-methyltransferase (mRNA: 4.49%, Q1, prot: NA); Methionine synthase (mRNA: 82.45%, Q2, prot: 0.44%); 5-methyltetrahydropteroyltriglutamate--homocysteine S-methyltransferase (mRNA: 38.23%, Q1, prot: 18.67%); Cystathionine gamma-synthase (mRNA: 29.52%, Q1, prot: 0.67%); O-acetylhomoserine aminocarboxypropyltransferase (mRNA: 94.01%, Q2, prot: 21.56%); Adenosylhomocysteinase (mRNA: 97.82%, Q3, prot: 16.22%); Cystathionine beta-synthase (mRNA: 4.35%, Q1, prot: NA); S-ribosylhomocysteine lyase (mRNA: 97.82%, Q3, prot: 42.89%)
Sorbitol499.05%L-iditol 2-dehydrogenase (mRNA: 63.27%, Q1, prot: 4.89%); Aldehyde reductase (mRNA: 11.43%, Q1, prot: NA); Hexokinase (mRNA: 33.88%, Q1, prot: NA); Alpha-galactosidase (mRNA: 99.05%, Q3, prot: 11.56%)
Phenylacetic acid698.37%Phenylacetaldehyde dehydrogenase (mRNA: 4.08%, Q1, prot: NA); Aldehyde dehydrogenase (NAD(P)(+)) (mRNA: 5.85%, Q1, prot: NA); Penicillin amidase (mRNA: 62.99%, Q1, prot: 0.44%); Amidase (mRNA: 1.09%, Q1, prot: NA); Nitrilase (mRNA: 0.27%, Q1, prot: NA); Phenylacetate--CoA ligase (mRNA: 98.23%, Q4, prot: 12.44%)
Putrescine1198.1%Non-specific polyamine oxidase (mRNA: 0.41%, Q2, prot: NA); Putrescine carbamoyltransferase (mRNA: 1.5%, Q1, prot: NA); Diamine N-acetyltransferase (mRNA: 83.4%, Q2, prot: NA); Spermidine synthase (mRNA: 76.87%, Q2, prot: 0.44%); Homospermidine synthase (mRNA: 0.54%, Q1, prot: NA); Diamine transaminase (mRNA: 0.14%, Q1, prot: NA); Putrescine aminotransferase (mRNA: 16.19%, Q1, prot: NA); N-carbamoylputrescine amidase (mRNA: 28.84%, Q1, prot: NA); Agmatinase (mRNA: 74.29%, Q2, prot: 1.33%); Ornithine decarboxylase (mRNA: 48.71%, Q2, prot: 2.89%); Glutamate--putrescine ligase (mRNA: 6.8%, Q1, prot: NA)
Quinolinic Acid298.1%Nicotinate-nucleotide diphosphorylase (carboxylating) (mRNA: 77.55%, Q2, prot: 8.89%); Quinolinate synthase (mRNA: 97.82%, Q3, prot: 4.67%)
3-hydroxyanthranilic acid497.82%Catalase peroxidase (mRNA: 4.9%, Q2, prot: 5.11%); Catalase (mRNA: 97.69%, Q3, prot: 22.22%); 3-hydroxyanthranilate 3,4-dioxygenase (mRNA: 0.27%, Q2, prot: NA); Kynureninase (mRNA: 0.27%, Q2, prot: NA)
Urea896.87%Urease (mRNA: 58.78%, Q2, prot: 0.67%); Arginase (mRNA: 82.31%, Q2, prot: 0.22%); Agmatinase (mRNA: 74.29%, Q2, prot: 1.33%); Creatinase (mRNA: 0.95%, Q2, prot: NA); Allantoicase (mRNA: 0.68%, Q2, prot: NA); Guanidinobutyrase (mRNA: 0.27%, Q1, prot: NA); Ureidoglycolate lyase (mRNA: 5.44%, Q1, prot: 0.22%); Urea carboxylase (mRNA: 27.76%, Q1, prot: NA)
Anthranilic acid596.33%Anthranilate phosphoribosyltransferase (mRNA: 94.29%, Q2, prot: 2%); Arylformamidase (mRNA: 0.41%, Q1, prot: NA); Kynureninase (mRNA: 0.27%, Q2, prot: NA); Anthranilate synthase (mRNA: 55.1%, Q1, prot: 0.89%); Anthranilate--CoA ligase (mRNA: 0.27%, Q1, prot: NA)
Phenol389.66%Arylesterase (mRNA: 16.73%, Q1, prot: NA); 4-hydroxybenzoate decarboxylase (mRNA: 3.95%, Q2, prot: NA); Tyrosine phenol-lyase (mRNA: 86.67%, Q3, prot: NA)
Uric Acid380.00%FAD-dependent urate hydroxylase (mRNA: 0.68%, Q1, prot: NA); Xanthine dehydrogenase (mRNA: 80%, Q2, prot: 0.89%); 8-oxoguanine deaminase (mRNA: 0%, QNA, prot: NA)
Trimethylamine378.23%Betaine reductase (mRNA: 75.65%, Q3, prot: 0.44%); Trimethylamine dehydrogenase (mRNA: 0.27%, Q1, prot: NA); Trimethylamine-N-oxide reductase (cytochrome c) (mRNA: 10.2%, Q1, prot: 0.44%)
Hippuric acid 144.63%Hippurate hydrolase (mRNA: 44.63%, Q1, prot: NA)
Monomethylamine142.99%Sarcosine reductase (mRNA: 42.99%, Q1, prot: 0.44%)
Creatine328.44%Creatine kinase (mRNA: 1.5%, Q1, prot: 0%); Creatininase (mRNA: 26.39%, Q1, prot: 0.67%); Creatinase (mRNA: 0.95%, Q2, prot: NA)
Creatinine126.39%Creatininase (mRNA: 26.39%, Q1, prot: 0.67%)
Oxalate322.72%Formyl-CoA transferase (mRNA: 20.54%, Q1, prot: 1.33%); CoA:oxalate CoA-transferase (mRNA: 3.27%, Q1, prot: 0.22%); Oxalate--CoA ligase (mRNA: 0.14%, Q1, prot: NA)
Mannitol321.9%D-arabinitol 4-dehydrogenase (mRNA: 0.68%, Q1, prot: NA); Mannitol dehydrogenase (mRNA: 0.41%, Q1, prot: NA); Mannitol 2-dehydrogenase (mRNA: 21.22%, Q1, prot: NA)
Phenol sulfate220.27%Aryl-sulfate sulfotransferase (mRNA: 5.99%, Q1, prot: NA); Arylsulfatase (mRNA: 15.37%, Q1, prot: NA)
Arab(in)itol111.43%Aldehyde reductase (mRNA: 11.43%, Q1, prot: NA)
Trimethylamine-N-oxide110.2%Trimethylamine-N-oxide reductase (cytochrome c) (mRNA: 10.2%, Q1, prot: 0.44%)
Pseudouridine19.66%Pseudouridine kinase (mRNA: 9.66%, Q1, prot: NA)
p-Cresyl sulfate29.12%4-hydroxyphenylacetate decarboxylase (mRNA: 0.95%, Q2, prot: NA); 2-iminoacetate synthase (mRNA: 8.71%, Q1, prot: NA)
3-(3-Hydroxyphenyl) propanoic acid19.12%3-(3-hydroxy-phenyl)propanoic acid hydroxylase (mRNA: 9.12%, Q1, prot: NA)
a-keto-d-Guanidinovaleric Acid25.85%D-amino-acid transaminase (mRNA: 5.71%, Q1, prot: NA); Arginine--pyruvate transaminase (mRNA: 0.14%, Q1, prot: NA)
Indole-3-acetic acid33.95%Aldehyde dehydrogenase (NAD(+)) (mRNA: 2.72%, Q1, prot: NA); Amidase (mRNA: 1.09%, Q1, prot: NA); Nitrilase (mRNA: 0.27%, Q1, prot: NA)
y-guanidinobutyric Acid31.9%Glycine amidinotransferase (mRNA: 0.68%, Q1, prot: NA); Amidase (mRNA: 1.09%, Q1, prot: NA); Guanidinobutyrase (mRNA: 0.27%, Q1, prot: NA)
Hyaluronic acid (Hyaluronan)11.9%Hyaluronan synthase (mRNA: 1.9%, Q1, prot: 0.22%)
Gentisic acid11.22%3-hydroxybenzoate 6-monooxygenase (mRNA: 1.22%, Q1, prot: 0.22%)
Kinurenine40.95%Kynurenine 3-monooxygenase (mRNA: 0%, QNA, prot: NA); Kynurenine--oxoglutarate transaminase (mRNA: 0.82%, Q2, prot: 0%); Arylformamidase (mRNA: 0.41%, Q1, prot: NA); Kynureninase (mRNA: 0.27%, Q2, prot: NA)
Dimethylamine20.95%Trimethylamine dehydrogenase (mRNA: 0.27%, Q1, prot: NA); Dimethylargininase (mRNA: 0.68%, Q2, prot: NA)
Kynurenic Acid10.82%Kynurenine--oxoglutarate transaminase (mRNA: 0.82%, Q2, prot: 0%)
Taurocyamine10.68%Glycine amidinotransferase (mRNA: 0.68%, Q1, prot: NA)
Asymmetric Dimethylarginine (ADMA)10.68%Dimethylargininase (mRNA: 0.68%, Q2, prot: NA)
Table 4. Occurrence and abundance of uremic-retention-solutes-synthesizing enzymes in patients. Presented enzymes were detected in more than 10% of patients. The full list is available in Supplementary Table S4.
Table 4. Occurrence and abundance of uremic-retention-solutes-synthesizing enzymes in patients. Presented enzymes were detected in more than 10% of patients. The full list is available in Supplementary Table S4.
Enzyme NameNumber of Bacteria That Express mRNA for This EnzymePercentage of Patients with Detected mRNA for This EnzymeQuartile of mRNA AbundancePercentage of Patients with Indicated Enzyme Detected in the ProteomeA Toxin That Could Be Potentially Synthesised by the Indicated Enzyme
Orotidine-5’-phosphate decarboxylase20599.32%263.78%Orotidine
Purine-nucleoside phosphorylase18799.18%262.67%Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine
Methylglyoxal synthase12498.91%358.22%Methylglyoxal
S-ribosylhomocysteine lyase12997.82%342.89%Homocysteine
Arginine--tRNA ligase22199.59%242%Argininic Acid
Orotate phosphoribosyltransferase19999.05%335.78%Orotic Acid; Orotidine
Arginine decarboxylase3991.29%232%Argininic Acid
Catalase5697.69%322.22%3-hydroxyanthranilic acid
O-acetylhomoserine aminocarboxypropyl-transferase4694.01%221.56%Homocysteine
5-methyltetrahydropteroyltriglutamate-homocysteine S-methyltransferase3138.23%118.67%Homocysteine
Xanthine phosphoribosyltransferase16198.37%216.22%Xanthine
Adenosylhomocysteinase6897.82%216.22%Homocysteine; S-Adenosylhomocysteine
Dihydroorotate dehydrogenase (NAD(+))2891.43%214.44%Orotic Acid
Argininosuccinate lyase19299.18%214%Argininic Acid
Phenylacetate--CoA ligase8498.23%212.44%Phenylacetic acid
tRNA-2-methylthio-N(6)-dimethylallyladenosine synthase18599.05%211.56%S-Adenosylhomocysteine
Alpha-galactosidase7199.05%211.56%Myoinositol; Sorbitol
Cytidine deaminase4887.62%311.56%Cytidine; Uridine
Table 5. Top 20 (in terms of occurrence) bacteria that metabolize uremic retention solutes.
Table 5. Top 20 (in terms of occurrence) bacteria that metabolize uremic retention solutes.
BacteriaPercentage of Patients with Detected mRNA for Indicated Bacteria Number of Uremic Retention Solutes for Which mRNA Was Detected of at Least One of the Toxin-Producing EnzymesNumber of Different Enzymes Detected for Indicated BacteriaUremic Retention Solutes That Could Be Potentially Synthesized by Indicated Bacteria
Ruminococcus torques93.61%2245Sorbitol; Arab(in)itol; Methylglyoxal; Hypoxanthine; Uric Acid; Xanthine; S-Adenosylhomocysteine; Homocysteine; Inosine; Nicotinamide; Xanthosine; Orotic Acid; Orotidine; anthranilic acid; Uridine; Putrescine; Quinolinic Acid; Cytidine; Myoinositol; Urea; Argininic Acid; Phenylacetic acid
Faecalibacterium prausnitzii93.33%1734Sorbitol; Homocysteine; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; Putrescine; Cytidine; Uridine; Myoinositol; Urea; Argininic Acid; Methylglyoxal
Eubacterium rectale81.77%1727Sorbitol; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; anthranilic acid; Uridine; Putrescine; Homocysteine; Quinolinic Acid; Cytidine; Methylglyoxal; Argininic Acid
Bacteroides uniformis79.05%1820S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; anthranilic acid; Quinolinic Acid; Cytidine; Uridine; Myoinositol; Sorbitol; Methylglyoxal; Argininic Acid; Homocysteine; Phenylacetic acid
Bacteroides ovatus78.37%2134Sorbitol; Mannitol; 3-hydroxyanthranilic acid; Orotic Acid; S-Adenosylhomocysteine; Putrescine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotidine; anthranilic acid; Homocysteine; Quinolinic Acid; Cytidine; Uridine; Myoinositol; Phenylacetic acid; Methylglyoxal; Argininic Acid
Bacteroides vulgatus77.14%1724Orotic Acid; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotidine; anthranilic acid; Cytidine; Uridine; Myoinositol; Sorbitol; Methylglyoxal; Argininic Acid; Homocysteine; Phenylacetic acid
Ruminococcus obeum75.78%2341Sorbitol; Arab(in)itol; Methylglyoxal; Hypoxanthine; Uric Acid; Xanthine; S-Adenosylhomocysteine; Inosine; Nicotinamide; Xanthosine; Orotic Acid; Orotidine; anthranilic acid; Quinolinic Acid; Uridine; Putrescine; Homocysteine; Cytidine; Phenol sulfate; Phenol; Urea; Argininic Acid; Phenylacetic acid
Lachnospiraceae bacterium68.98%1316S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; Cytidine; Uridine; Homocysteine; Argininic Acid; Urea
Roseburia inulinivorans62.04%1521S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; Uridine; Putrescine; Quinolinic Acid; Cytidine; Methylglyoxal; Argininic Acid; Homocysteine
Bacteroides thetaiotaomicron61.77%19303-hydroxyanthranilic acid; Orotic Acid; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotidine; anthranilic acid; Homocysteine; Quinolinic Acid; Cytidine; Uridine; Myoinositol; Sorbitol; Methylglyoxal; Argininic Acid; Phenylacetic acid
Clostridium bolteae61.5%1622Sorbitol; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; anthranilic acid; Uridine; Quinolinic Acid; Methylglyoxal; Argininic Acid; Homocysteine; Phenylacetic acid
Bacteroides caccae60.95%1826Orotic Acid; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotidine; anthranilic acid; Homocysteine; Quinolinic Acid; Cytidine; Uridine; Myoinositol; Sorbitol; Methylglyoxal; Argininic Acid; Phenylacetic acid
Bacteroides xylanisolvens60.14%2137Sorbitol; Mannitol; Orotic Acid; S-Adenosylhomocysteine; Putrescine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotidine; anthranilic acid; Homocysteine; Quinolinic Acid; Cytidine; Uridine; Myoinositol; Phenol sulfate; Phenylacetic acid; Methylglyoxal; Argininic Acid
Dorea longicatena59.32%1625Sorbitol; Methylglyoxal; Homocysteine; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; Uridine; Quinolinic Acid; Myoinositol; Argininic Acid; Phenylacetic acid
Roseburia hominis59.32%1621S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; anthranilic acid; Putrescine; Quinolinic Acid; Cytidine; Uridine; Methylglyoxal; Argininic Acid; Homocysteine
Roseburia intestinalis57.14%2042Sorbitol; Hypoxanthine; Uric Acid; Xanthine; S-Adenosylhomocysteine; Inosine; Nicotinamide; Xanthosine; Orotic Acid; Orotidine; anthranilic acid; Quinolinic Acid; Cytidine; Uridine; Putrescine; Homocysteine; Phenol; Myoinositol; Argininic Acid; Methylglyoxal
Parabacteroides merdae56.05%1723Orotic Acid; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotidine; anthranilic acid; Quinolinic Acid; Cytidine; Uridine; Homocysteine; Argininic Acid; Urea; Methylglyoxal; Phenylacetic acid
Bifidobacterium longum55.1%1621S-Adenosylhomocysteine; Orotic Acid; Orotidine; anthranilic acid; Xanthine; Hypoxanthine; Putrescine; Homocysteine; Cytidine; Inosine; Uridine; Xanthosine; Myoinositol; Sorbitol; Nicotinamide; Argininic Acid
Flavonifractor plautii54.83%1317S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; Cytidine; Uridine; Homocysteine; Argininic Acid; Urea
Alistipes putredinis54.56%15193-hydroxyanthranilic acid; Homocysteine; S-Adenosylhomocysteine; Hypoxanthine; Inosine; Nicotinamide; Xanthine; Xanthosine; Orotic Acid; Orotidine; Quinolinic Acid; Cytidine; Uridine; Methylglyoxal; Argininic Acid
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Popkov, V.A.; Zharikova, A.A.; Demchenko, E.A.; Andrianova, N.V.; Zorov, D.B.; Plotnikov, E.Y. Gut Microbiota as a Source of Uremic Toxins. Int. J. Mol. Sci. 2022, 23, 483. https://doi.org/10.3390/ijms23010483

AMA Style

Popkov VA, Zharikova AA, Demchenko EA, Andrianova NV, Zorov DB, Plotnikov EY. Gut Microbiota as a Source of Uremic Toxins. International Journal of Molecular Sciences. 2022; 23(1):483. https://doi.org/10.3390/ijms23010483

Chicago/Turabian Style

Popkov, Vasily A., Anastasia A. Zharikova, Evgenia A. Demchenko, Nadezda V. Andrianova, Dmitry B. Zorov, and Egor Y. Plotnikov. 2022. "Gut Microbiota as a Source of Uremic Toxins" International Journal of Molecular Sciences 23, no. 1: 483. https://doi.org/10.3390/ijms23010483

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop