As a pesticide, thiram is widely used in agriculture to eliminate pests, and it also causes great environmental pollution and poses a certain threat to animal health. It’s well known that the contamination of feed with thiram in the environment can lead to tibial chondrogenesis(TD) in chickens. The imbalance of intestinal flora and related metabolites is closely related to bone development. Unfortunately, the relationship between the intestinal flora of TD broilers and serum metabolites is unclear. Our results demonstrated that broilers exposed to thiram showed typical lameness and the white cartilage thrombus in the growth plate, accompanied by hepatotoxicity and intestinal injury. We found that the intestinal flora of TD group was out of balance, the diversity was significantly increased with Corynebacterium significantly enriched. Moreover, the metabolome results showed alterations in 10 serum metabolites, with Glucosylceramide being considerably up-regulated, resulting in sphingolipid metabolism problem, which is critical in the etiology of TD. The comprehensive correlation analysis showed the relationship between intestinal microflora and Sphingolipid metabolism in TD broilers. Thiram aggravates tibial chondrodysplasia by affecting the changes in the composition and structure of the intestinal microflora of broilers and the disorder of sphingomyelin metabolism. Collectively, these findings provide novel insight into the pathogenesis of TD from the perspective of thiram-induced gut microbiota and metabolic disorders.
Research Article
Integrated microbiome-metabolome reveals thiram aggravates tibial dyschondroplasia through disturbing sphingolipid metabolism
https://doi.org/10.21203/rs.3.rs-2674389/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Thiram
TD
gut microbiota
Serum metabolites
Chickens
-
TD is associated with toxic effect of thiram exposure in feed.
-
Thiram induce significant changes in the sphingolipid metabolism.
-
The intestinal flora disbalance associated with thiram induced TD.
-
Integrated microbiome-metabolome reveals thiram aggravates TD.
Pesticides are widely employed in aquatic environments to control illnesses, pests, and weeds, public health to combat vector-borne infectious diseases, and agricultural productivity to promote economic growth(Zhang and Yang 2021). Although the widespread use of these synthetic pesticides helps to advance agricultural progress, it also significantly damages the environment. Pesticide residues carried by crops infiltrate animals through the food chain through the majority of animal feeds, endangering their lives and well-being(Sharma et al. 2019). Thiram(IUPAC: tetramethylthiuram disulfide), a carbamate pesticide, enters the food chain as a result of abuse as an insecticide or fungicide in agriculture and environmental pollution in the ecosystem, causing tibial achondroplasia (TD) in broilers fed thiram-contaminated feed(Rath et al. 2005). Thiram damages vascular endothelial cells in the tibial growth plate, limits angiogenesis, and inhibits tibial growth plate angiogenesis(Zhang et al. 2020). Thiram is metabolized into carbon disulfide in animals, causing liver damage. Aside from the liver, thiram can harm a number of other tissue systems in animals, including the kidneys and intestines(Kong et al. 2020). The dosage necessary for thiram damage varies depending on the organ. In experimental animals, repeated daily oral doses of 50mg/kg can be expected to result in weakness, muscle incoordination and limb paralysis. There was no effect on dogs fed with 200ppm feed for one year (4mg/kg)(Liu et al. 2022a). After adding 50mg/kg thiram, the chickens began to develop typical TD symptoms and associated claudication(Rath et al. 2007; Huang et al. 2021; Liu et al. 2022b). The survival rate of zebra fish decreased when exposed to 1 µM and 10 µM, and the hatching was delayed at the dose of 0.1 µM and 1 µM(Chen et al. 2019). In addition, the residues of thiram can be passed on to consumers through the food chain and may have toxic effects, causing unpredictable life risks.
An ever increasing number of studies have shown that the piece and variety of host stomach microbiota are effortlessly impacted by different natural poisons, including food synthetics and pesticides, which demonstrates that the job of stomach microbiota is increasingly more significant in the investigation of the poisonousness of pesticides to non-target organic entities. The intestinal microbiome is involved in regulating a wide range of biological activities, involving intestinal physiology(Mangiola et al., 2018), production and absorption of nutrients(Rowland et al., 2018), host growth(Makki et al., 2018), energy balance(Jumpertz et al., 2011), metabolic function(Martinez et al., 2017), immune system function (Fung et al., 2017) and inflammatory processes(Zhang and Li, 2014). There is developing confirmation that intestinal microbiome assumes a paramount part in regulating important biological processes and the pathogenesis of many complex diseases(Gentile and Weir, 2018). It is worth mentioning that the function of gut bacteria in bone health and illness has garnered increasing attention in recent years, and the influence of intestinal microbiota on bone growth and development has also become a research hotspot(Chen et al., 2017; McHugh, 2019). According to research, bone-related diseases such as osteoarthritis and osteoporosis are linked to an imbalance of gut flora(Gill et al., 2018). At present, those exploration on the interaction between the intestinal flora and the skeletal system focuses on three key areas, 1) the influence of microorganisms on the immune system, also known as bone immunology; 2) hormonal pathways (such as steroid hormones, PTH and vitamin D); 3) The synthesis of bacterial metabolites, which can send signals to the bone cells (Charles et al., 2014; Zaiss et al., 2019). These studies supporting the influence of intestinal bacterial communities on skeletal development provide important reference values for our research on TD through 16S rRNA intestinal microbiome sequencing.
Comprehensive intestinal microbiome and metabonomic analysis is a popular way to diagnose diseases. Comprehensive analysis of intestinal microbiome and metabolic group data can not only provide insight into the effects of lifestyle and dietary factors on chronic and acute diseases, such as autoimmune diseases, inflammatory intestinal diseases(Lloyd-Price et al. 2019), cancer, type 2 diabetes and obesity(Zitvogel et al. 2018), cardiovascular diseases(Thingholm et al. 2019) and nonalcoholic fatty liver disease(Caussy et al. 2018) but it also helps in potential diagnostic and treatment targets.
The gut microbiota has been shown to play an indispensable role in the pathogenesis of bone-related diseases. At present, the relationship between the gut microbiome and metabolism of TD affected broilers has not yet been determined, and more in-depth research was urgently needed. Therefore, we induced TD with thiram, and observed the incidence of the disease in broilers on the 7th, 10th, 14th and 18th day. The illness was visible in broilers fed thiram-contaminated feed, and the characteristic signs of TD in broilers persisted until the 18th day, so we mainly carried out the experimental study on 18th day. We conducted 16S sequencing and metabolomics research on healthy broilers and broilers with TD. We found that the intestinal microbial balance of TD broilers was disturbed and Sphingolipid metabolism was altered. In addition, we created a map that depicts the relationship between the gut flora and Sphingolipids. These findings underscore the importance of gut microbiota in regulating host lipid metabolism and shed information on the pathogenic mechanism of TD.
2.1 Animals and experimental design
A total of 100 healthy 1-day-old broilers were adaptively raised for 2 days before being randomly assigned to one of two groups: control (regular diet) or thiram-induced TD (50 mg/kg thiram was added to the feed). The broiler chickens were reared with free feeding and drinking water, and the temperature, humidity and light of the chicken house were strictly controlled. The feed intake and weight of the broiler chickens and the health status of the flock are observed and recorded on daily basis. Broilers were sampled on 7th, 10th, 14th and 18th day, and 10 chickens were randomly picked from each group and were slaughtered to collect the intestines, liver, and tibia. A portion of the tissues were put in 4% paraformaldehyde, while the remainder were frozen in a -80℃ refrigerator for later use. Intestinal contents were collected for intestinal microbial determination. Blood was collected for serum biochemical and metabolomics experiment.
2.2 Pesticide
Thiram(C6H12N2S4, 98%purity): IUPAC, tetramethylthiuram disulfide; obtained from Hubei Maoerwo Biomedicine.
2.3 Correspondence coefficient evaluation
The broiler's body weight and feed consumption of each repetitive broiler was weighed on 7th, 10th, 14th and 18th day respectively, the average daily gain of each stage of broiler growth was calculated. In addition to this, tibial length, weight, diameter of middle shaft of tibia and growth plate width were measured. GraphPadPrism7.0 was used to analyze the data and to draw the relevant graphs. After counting and weighing the visceral organs, the equation below was used to calculate the organ index. organ index = organ weight / body weight of broilers.
2.4 H&E staining and pathological observation
Pathological sections were prepared, stained with HE, and pathological changes were examined using a light microscope (Leica, Germany).
2.5 Serum separation and biochemistry
Blood samples from all broilers were collected and the serum was separated by centrifuging the blood for 5 minutes @ 3000rpm at 4℃. The concentrations of serum total protein (TP), serum total cholesterol (TC), serum triglyceride (TG), serum urea (UREA), albumin (ALB-II), serum glucose (GLU), aspartate aminotransferase (AST) and alkaline phosphatase (ALP) were measured by Mindray BS-380 automatic biochemical analyzer.
2.6 16S rRNA gene sequence analysis
After extracting the total DNA from the sample, primers were designed based on the conservative area, and a sequencing connector was added at the end of the primer. After the PCR amplification, the products were purified, quantified, and homogenized to form a sequencing library. The newly constructed library was first subjected to a quality inspection. Qualified libraries were subjected to high-throughput sequencing using Illumina HiSeq 2500 to obtain the original sequence. Trimmomaticv0.33 and cutadapt1.9.1 were used for quality filtering to obtain Clean Reads. Using Usearchv10 software, the Clean Reads of each sample were spliced by overlap, and then the length of the stitched data was filtered according to the length range of different regions. Using UCHIMEv4.2 software, the chimera sequence was identified and removed, and the final effective data was obtained.
According to 97% similarity, the obtained sequences were divided and clustered by OTU, and the representative sequences were classified, identified and phylogenetic analysis was done. The intestinal microbial alpha diversity was determined based on the relative abundance distribution of OTU in each sample. To analyze the difference and resemblance between various samples and groups, βdiversity was evaluated using weighted and unweighted UniFrac distance.
2.7 Metabolomics analysis
Serum (150µL) was taken into a 1.5mL EP tube, 2µL L-2-chloro-phenylalanine (300µg/mL) was added as an internal standard, with 450µL of organic solvent (methanol: acetonitrile = 1:9, -80℃ precooling), later it was shaken and mixed well and was refrigerated at -20℃ for 30min. Then, after vortexing for two minutes, 300µl ultrapure water was added to each tube, and was centrifuged @ 12000r/min at 4℃ to collect supernatant. Finally, the supernatant was transferred to the sample vial for LC-MS analysis after being filtered via a 0.22m filter membrane.
As per our prior study technique(Yang et al. 2021), the data were entered into the SIMCA-P V14.1 program (Umetrics, Uppsala, Sweden) for statistical analysis. The significance of variables (VIP > 1) was predicted using principal component analysis (PCA), followed by an orthogonal partial least square discriminant analysis (OPLS-DA) model. Then, by integrating the VIP value of OPLS-DA multivariate statistical analysis with the t-test p value of univariate statistical analysis, the metabolites with significant differences were filtered. The key metabolic pathways were also identified using MetaboAnalyst (https://www.metaboanalyst.ca/).
2.8 Integrative analysis of the microbiome and metabolome
The data from significantly different metabolites and 16S sequencing were submitted to an automated microbiome and metabolome integrative analysis pipeline (M2IA) (http://m2ia.met-bioinformatics.cn) for metabolome and microbiome inquiry. The six primary phases of the whole data analysis workflow in M2IA are as follows: (1) data upload and preprocessing; (2) comprehensive similarity analysis of microbiome and metabolome data matrices; (3) metabolite microbial pair analysis; (4) integrated analysis of several regression-based data sets; (5) metabolic functional network analysis; and (6) automatically created HTML reports for overview.
2.9 Statistical analysis
GraphPad Prism 7.0 (GraphPad Software Inc., La Jolla, CA) was used to do statistical analysis on the data gathered. One-way analysis of variance was used to examine the data (ANOVA). The data is communicated as a mean and standard deviation (SD). *P < 0.05, **P < 0.01 and ***P < 0.001 were deemed statistically significant differences.
3.1 Thiram induced tibial dyschondroplasia (TD) in broilers
Tibial dyschondroplasia appeared in the broilers after the exposure to the thiram. Broilers in the TD group showed symptoms of difficulty in standing and lameness (Fig. 1a), and decreased daily weight gain (Fig. 1c). A thickened, avascular cartilage plug was formed at the proximal end of the tibia after TD (Fig. 1b). Tibia weight, length and diaphyseal width of the TD group was lower than that of the control group (Fig. 1d, e, f). On different days, the width of the growth plate of the control group was larger than that of the TD group (Fig. 1g), and the growth plate width coefficient of the TD group decreased significantly (Fig. 1i). However, we found that the tibial index in the TD group on different days was significantly higher than that of the control group (Fig. 1h). It is reasonable to infer that thiram induced TD to inhibit the growth performance of broilers, especially in the daily gain of broilers. The tibia index of the TD group increased significantly due to the large difference in body weight between the two groups.
3.2 The effect of thiram on cartilage development.
The observation of tissue sections showed that the blood vessels in the growth plate of TD group (Fig. 2e, h) were significantly less than those in the control group. In addition, the chondrocytes in different states in the proliferative zone and hypertrophic zone were further observed at a magnification of 400 x light microscope. It was found that the cells in the proliferative zone of TD group were irregularly arranged(Fig. 2b, c), and the chondrocytes in the hypertrophic zone showed nuclear condensation and dissolution, resulting in the increase of vacuolar cartilage sac(Fig. 2f, g). Abnormal apoptosis accumulation of chondrocytes in hypertrophic area is the direct cause of white cartilage plugs in growth plate.
3.3 The influences of thiram on broiler livers and intestines
Liver of the control group was normal purple-red, while that of the TD group was khaki-yellow (Fig. 3a). The liver index of TD group decreased significantly (Fig. 3b). These results showed that thiram induced hepatotoxicity in broilers. Further observation of the HE section of the liver showed that hepatocyte lysis, structural disorder of hepatic cord and inflammatory cell infiltrations were found in TD group (Fig. 3c). Intestinal histopathological examination showed that the columnar cells of intestinal villous epithelium in TD group were dissolved (Fig. 3d).
Serum biochemical results showed that serum glucose (GLU) was significantly down-regulated and aspartate aminotransferase (AST) was considerably up-regulated in TD group (p < 0.05), indicating pathological changes in liver. There were corresponding changes in serum total protein (TP), serum total cholesterol (TC), serum triglyceride (TG), serum urea (UREA), albumin (ALB-II) and alkaline phosphatase (ALP), but the difference was found to be not significant (Fig. 3e).
3.4 Data acquisition and analysis
We performed a research of the gut microbiota to evaluate the efficacy of thiram on the gut. In the result report, 8 samples were sequenced to obtain a total of 640,011 pairs of Reads, double-ended Reads quality control, after splicing, a total of 635,515 Clean Reads were generated, and each sample generated at least 79,057 Clean Reads, with an average of 79,439 Clean Reads (TABLE A2). After classification and identification, 356 OTUs (FIG. A1a, b, c) were obtained depending on 97% nucleotide sequence commonality. In addition, we also observed that there are 244 core OTUs in the bacterial community, responsible for nearly 68.54% of the entire OTU. In each sample, the species accumulation and rarefaction curves were fairly flat and had a disposition to saturate features, suggesting that virtually all of the bacterial species were recognized (FIG. A1d, e). Additionally, the rank abundance curves for all samples were broad and slowly falling, indicating adequate evenness and abundance.
3.5 Thiram altered the intestinal microbiota significantly.
In order to assess the distinction of bacterial diversity between the control group and the TD group, we compared the sequences to evaluate α-diversity and β-diversity. The results showed that Shannon index (0.9173 ± 0.1994) and Simpson index (0.2138 ± 0.06959) were increased significantly in TD group, while there was no significant difference between ACE index (3.029 ± 24.49) and Chao1 (10.23 ± 23.38) in control group and TD group (FIG .4a). To assess the variability and resemblance of individuals within and within groups, principal coordinate analysis (PCoA) and unweighted paired group coupled with arithmetic mean tree (UPGMA) analysis were performed. The unweighted PCoA map revealed that the TD group gathered closely, whereas the control group separated, which was compatible with the conclusions of the UPGMA analysis. These results indicated that the diversity of intestinal flora may be strongly affected by the thiram (FIG. A1g).
Through microbial classification and identification, the relative proportions of dominant groups at different classification levels were determined, and substantial differences in the abundance and composition of intestinal bacteria were found in the two groups. A total of 11 phyla were distinguished from 8 samples at the phylum level. According to the classification results (Fig. 4d), in the control group and TD group, Proteobacteria (76.55%, 67.91%), Firmicutes (22.37%, 28.04%), Actinobacteria (0.53%, 3.09%), Cyanobacteria (0.27%, 0.53%) are the dominant taxa, accounting for more than 99% of the total taxa. In addition, Bacteroidetes (0.10%, 0.22%), Fusobacteria (0.04%, 0.07%), Patescibacteria (0.00%, 0.05%) and Gemmatimonadetes (0.00%, 0.04%) were all low in abundance in the both groups. Even the abundance of Patescibacteria and Gemmatimonadetes in the control group was 0.00%. In order to further explore the influence of thiram on the taxonomic composition of the intestinal microbiota of broilers, a total of 182 genera were discovered from the intestinal flora of broilers. Among the identified genera, Escherichia-Shigella (74.24%, 63.35%) was the main one, followed by Candidatus_Arthromitus (12.01%, 12.29%), Enterococcus (1.59%, 2.77%) and Lactobacillus (0.77%, 4.87%). Staphylococcus (0.40%, 0.54%) and [Ruminococcus]_torques_group (0.55%, 0.86%) accounted for less in the total genera group (Fig. 4e). The cluster analysis represented by the heat map further analyzed the relative abundance of microorganisms in terms of phylum and genus level (Fig. 4b, c).
In order to further analyze the changes in the taxonomic composition of intestinal microorganisms in the control group and TD group, Metastats analysis (Table 1) was carried out on broilers with different categorization levels. The abundance of Bacteroidetes in the TD group was considerably greater than that of the control group at the phylum level while compared to the control group. At the genus level, there were substantial variations between the two groups in 12 genera. Among these differential taxa, the relative abundance of one genus (CHKCI001) decreased significantly under the influence of thiram, while the relative abundance of 11 genera increased significantly, including Bradyrhizobium, uncultured_bacterium_f_Beggiatoaceae, Aerococcus, Alloprevotella, Rothia, Pseudochrobactrum, Lotus_japonicus, uncultured_bacterium_o_Microtrichales, Desulfovibrio, Brevibacterium and Massilia. Given that this discriminant analysis cannot separate the predominant taxa, we employed a combination of linear discriminant analysis (LEFSE) and linear discriminant analysis (LDA) to detect specific bacteria relevant to thiram-induced TD (Fig. 4g, h). The results showed that the TD group was rich in o__Corynebacteriales, f__Corynebacteriaceae and g__Corynebacterium_1 at genus level, which demonstrated that the TD group had a substantially higher number of microorganisms than the control group, and the variations in these microflorae were significant enough to differentiate the microflora between the control and TD groups.
|
Taxonomic Rank |
Relative Abundance |
P value |
||
|---|---|---|---|---|
|
Con |
TD |
|||
|
phylum |
Bacteroidetes |
1.03E-03 |
2.19E-03 |
0.0343 |
|
genus |
Bradyrhizobium |
0.00E + 00 |
3.01E-04 |
0.00E + 00 |
|
genus |
uncultured_bacterium_f_Beggiatoaceae |
1.66E-05 |
8.35E-04 |
0.00151 |
|
genus |
Aerococcus |
5.95E-04 |
3.60E-03 |
0.00616 |
|
genus |
Alloprevotella |
0.00E + 00 |
4.36E-04 |
0.00813 |
|
genus |
Rothia |
1.99E-05 |
6.75E-04 |
0.0098 |
|
genus |
Pseudochrobactrum |
2.98E-05 |
8.97E-04 |
0.0121 |
|
genus |
Lotus_japonicus |
6.71E-06 |
4.83E-04 |
0.0145 |
|
genus |
uncultured_bacterium_o_Microtrichales |
9.77E-05 |
8.85E-04 |
0.0206 |
|
genus |
CHKCI001 |
1.12E-03 |
2.04E-05 |
0.0376 |
|
genus |
Desulfovibrio |
0.00E + 00 |
3.66E-04 |
0.0398 |
|
genus |
Brevibacterium |
1.33E-05 |
8.94E-05 |
0.047 |
|
genus |
Massilia |
5.24E-05 |
8.15E-04 |
0.0476 |
The network analysis was carried out by Python to clarify the relationship among different bacterial genera in the intestinal microflora. The results showed that there was positive correlation among 247 groups and negative correlation among 87 groups. The correlation was represented by a weight value, and the greater the value, the stronger the link. (Fig. 4i) shows the top 50 most relevant genera. The bacterial network consists of 79 nodes and 334 edges. Among them, Novosphingobium has a favorable correlation with Ohtaekwangia, Clostridium sensu stricto 1, and Bombella. Ohtaekwangia and Bombella are both positively linked with Clostridium sensu stricto 1. Ohtaekwangia and Bombella are positively correlated. All of their weight value is 1.00. Pantoea was negatively correlated with [Ruminococcus]_gauvreauii_group and Ruminiclostridium_9 (0.9762).
Through the analysis of the composition and difference of KEGG metabolic pathway, we can observe the differences and changes of microbial community functional genes in metabolic pathway among different groups of samples, which is an effective means to study the changes of metabolic function of community samples to adapt to environmental changes. The findings revealed that the alterations in the biological community in the control and TD groups were mostly engaged in Carbohydrate metabolism, Lipid metabolism, Amino acid metabolism, Energy metabolism, Glycan biosynthesis and metabolism, Metabolism of cofactors and vitamins and Membrane transport (FIG. 4h).
3.6 Alteration of serum metabolome in thiram-induced TD broilers
To investigate the particular effects of thiram on serum, the metabolic features of serum on the 18th day were analyzed using UHPLC/Orbitrap MS in positive and negative ionization modes. An unsupervised PCA approach was carried out to visualize general clustering and trend across control and TD groups in both positive and negative ion modes. As depicted in (FIG. A2a, b), the clustering of QC samples indicates that the analysis method is stable and reproducible. In order to differentiate the various metabolites between the two groups, OPLS-DA was used, and the permutation test was conducted through 200 permutations to improve the prediction capacity. As demonstrated in (FIG. A2c, d), distinctly segregated clusters were found between the two groups' metabolite profiles. All R2 and Q2 values in the validation of the OPLSDA model were lower than the initial values in the validation plot, and the regression line intercept of Q2 was negative, indicating that the developed discriminant model was not overfitting (FIG. A2e, f). Variables with VIP > 1.0 in the OPLSDA model were visualized in associated Splots and were considered for further analysis using the Student's t-test, and those with P < 0.05 appeared to be possible differential metabolites.
In this study, a total of 10 significantly different metabolites were identified in serum (VIP > 1, p < 0.05), including 3 positive poles and 7 negative poles (Fig. 5a, b, c). Nine metabolites such as N-Desmethylcitalopram, 33-Deoxy-33-hydroperoxyfurohyperforin, Imipramine, Corchorusoside B, Myricatomentoside I and Glucosylceramide in serum were significantly up-regulated after exposure to thiram, but only PE (DiMe (13prime5) / MonoMe (11mem3)) decreased. KEGG enrichment analysis of differential metabolites showed that two metabolic pathways, Drug metabolism-cytochrome P450 and Sphingolipid metabolism, had a significant impact in the evolution of TD in broilers (Fig. 5d). Among them, Imipramine and Glucosylceramide2 upregulated differential metabolites related to the regulation of these two metabolic pathways, suggesting that broilers exposed to thiram may regulate Sphingolipid metabolism by changing Glucosylceramide, and then affect the occurrence of TD. Interestingly, we also found that the up-regulated metabolites of Cavipetin B, N-Desmethylcitalopram and Myricatomentoside I are alkaline substances, among which N-Desmethylcitalopram is a very strong alkaline compound, which indicates that exposure to thiram can affect the acid-base environment of serum.
3.7 Correlation analysis of the gut microbiota and serum metabolites
In order to determine the correlation between significantly different serum metabolites and the changes of intestinal microflora in TD broilers, we carried out Spearman correlation analysis on intestinal microflora and 10 significant differences in serum metabolites. Glucosylceramide and PE (DiMe (13Magne5) / MonoMe (11pyr3)) were discovered to be strongly associated to alterations in intestinal flora(Fig. 6a). The Spearman correlation (Fig. 6b) between 6 genera and 2 metabolites associated with TD was further observed in the related network. In addition, we found that Glucosylceramide and PE (DiMe (13,5) / MonoMe (11,3)) are the hubs of the entire network. The results showed that Corynebacterium_1, Lotus_japonicus, Bradyrhizobium and uncultured_bacterium_f_Beggiatoaceae were substantially favorably connected with an increase in Glucosylceramide, but inversely linked with PE (DiMe (13,5) / MonoMe (11,3)). CHKCI001 has a high negative correlation with Glucosylceramide, and a positive correlation with PE (DiMe (13,5) / MonoMe (11,3)). In summary, these results indicate that thiram-induced TD can affect the structure and composition of the intestinal microflora. In turn, the dysbiosis of the gut microflora plays the most important role in TD.
Thiram, an organic sulfur molecule, which is widely used in agriculture for pest management due to its germicidal and insecticidal properties, but it is also a major polluter of the environment and a hazard to animal health(Sharma et al. 2003). Thiram is like a double-edged sword. There is no doubt that the role of thiram in agricultural protection cannot be ignored, but its exposure concentration in the environment is positively correlated with its threat to animal health. Many studies have reported the toxicity of thiram. An in vitro study showed thiram-induced cytotoxicity and oxidative stress in human erythrocytes(Salam et al. 2020). Thiram causes leg deformity disease in poultry - tibial cartilage dysplasia, which is also a cause of thyroid dysfunction in aquatic organisms(Belaid and Sbartai 2021). Thiram induces myocardial oxidative damage and apoptosis in broilers by interfering with cardiac metabolism. Our previous studies have found that residual Thiram in the feed inhibits immune system stress signals and disrupts sphingolipid metabolism in chickens(Liu et al. 2022b). In addition, thiram is also toxic to goats, fish, and birds(H. 2010). In our study, 50mg/kg thiram was mixed with feed to induce TD. Whether the pollution caused by thiram application can accumulate to this concentration in the biological chain has not been reported. The concentration of thiram exposed to the environment through pesticides remains to be further explored in the future. Here, we used 50mg/kg thiram-induced TD is basically consistent with the symptoms of natural onset, so it is used to explore the mechanism of this concentration of thiram on TD.
Consumption of thiram-contaminated feed or water by broilers can induce tibial achondroplasia (TD)(Zhang et al. 2018). Diseased broilers are often characterized by dyskinesia and limited feeding, resulting in hindered growth and development, slow body weight growth and inhibition of growth performance(Tong et al. 2018; Zhang et al. 2018). When TD occured, the Col II and ACAN of the growth plate significantly decreased with other cartilage development-related genes. In our study, thiram-induced TD broilers also showed obvious symptoms. Lameness and inability to stand caused difficulty in feeding. Therefore, the daily weight gain was lower than healthy broilers, and the tibia weight, length, width and growth plate width index were also indicating that the tibia development rate of diseased broilers is lower than that of healthy broilers. Interestingly, the tibia index of the TD group increased significantly. However, we reasonably inferred that the weight gains of TD broilers caused by thiram is much lower than that of the control group, and the tibia index increased significantly due to the excessive weight difference between the two groups. In addition, white avascular cartilage plugs of TD broilers can be clearly observed on the epiphyseal cut surface of the tibia. According to previous studies, fat solubility may be the reason for the formation of TD induced by thiram, which makes it easier for thiram to combine with the cell membrane and damage chondrocytes, due to this, chondrocytes gather and blood vessels cannot invade the hypertrophic area. As angiogenesis is blocked this results in the decrease vasculature in the TD affected growth plate, thus depriving the cells of essential nutrients and trace elements. This also inhibits the vascularization of tibial growth plate and hinders the removal of dead chondrocytes, thus forming TD lesion area in the growth plate region(Li et al. 2020b).
Prior studies indicate that thiram exposure can cause liver damage(Li et al. 2007). According to our findings, the liver index of thiram-induced TD broilers was vastly smaller than that of the control group. In order to further evaluate the degree of liver injury in broilers, biochemical and histological examinations were carried out on the liver of broilers. It is well-known that ALT and AST in serum are the biomarkers of hepatocyte injury(Zhou et al. 2018). Furthermore, several studies have indicated that pesticide exposure might result in a large rise in AST and ALP activities, suggesting liver physiological abnormalities(Kong et al. 2020). In our experiment, the levels of ALT and AST activity in the serum of chickens given access to thiram elevated, causing histological alterations in the liver in the form of inflammatory cell infiltration and hepatic cord dysfunction, which indicate that increased ALT and AST activity may cause hepatocyte structural integrity to deteriorate and result in the leaking of these enzymes into the blood(Wiest et al. 2017; Waheed et al. 2020). Our experimental results also proved that the intestinal tract was also damaged in thiram-induced TD broilers, and the pathological observation showed that the columnar cells of small intestinal villous epithelium dissolved and showed vacuoles. As we all know, there have a tight association between intestinal tract, intestinal flora and liver, and that enterohepatic axis plays an important role in the metabolism and physiological function(Tripathi et al. 2018; Albillos et al. 2020). In other words, the liver can affect the intestinal tract, especially the intestinal flora, which in turn affects the liver through the intestinal flora(Wang et al. 2020). Therefore, we further explored the changes of intestinal microorganisms in TD by 16S rRNA sequencing.
Some studies have shown that exposure to thiram can significantly change the richness and diversity of fecal microbiota in broilers. Our results showed that the Shannon index and Simpson index of TD affected broilers increased significantly in TD group, indicating the increase in α diversity of intestinal flora. Proteobacteria (76.55%, 67.91%), Firmicutes (22.37%, 28.04%) and Actinobacteria (0.53%, 3.09%) were the dominant intestinal microflora in the two groups, which is the classical structure of intestinal microorganisms in broilers(Mazidi et al. 2016; Clavijo and Flórez 2018). In order to further analyze the alterations in the taxonomic composition of intestinal bacteria between the control and TD groups, we carried out Metastats analysis. The findings indicate that the abundance of Bacteroidetes in the TD group grew considerably at the phylum level, which was accompanied by a rise in the abundance of 11 bacterial genera at the genus level and a drop in the abundance of 1 bacterial species. That is, in addition to CHKCI001, Bradyrhizobium, uncultured_bacterium_f_Beggiatoaceae, Aerococcus, Alloprevotella, Rothia, Pseudochrobactrum, Lotus_japonicus, uncultured_bacterium_o_Microtrichales, Desulfovibrio, Brevibacterium and Massilia all increased significantly in the TD group. As we all know, Bradyrhizobium, is a kind of Gram-negative bacteria, mainly parasitic on plant roots and often found in acidic soil to participate in nitrogen fixation(Shah and Subramaniam 2018; Cabral Michel et al. 2020). Interestingly, our results showed that Bradyrhizobium and its symbiotic bacteria Lotus_japonicus were significantly increased in the TD group. At present, there is no explanation for this phenomenon(Imai et al. 2020; Shrestha et al. 2020).
In order to distinguish the dominant taxa between the two groups, we used the combination of linear discriminant analysis (LEFSE) analysis and linear discriminant analysis (LDA) to identify specific bacteria related to Thiram-induced TD. The results suggested that the TD group was highly enriched in o__Corynebacteriales, f__Corynebacteriaceae and g__Corynebacterium_1. It is worth mentioning that corynebacterium (Coryneillusspp.) is a human opportunistic pathogen responsible for causing several infectious diseases including respiratory tract infections, urinary tract infections, endocarditis, and bone and joint infections (BJI)(Bernard 2012; Hacker et al. 2016; Badell et al. 2020; Jaén-Luchoro et al. 2020). In addition, it is also mentioned in the report of Cazanave et al that Corynebacterium is an occasional cause of bone and joint infections (BJI) as a typical orthopedic infection(Roux et al. 2004). (Kalt et al. 2018; Noussair et al. 2019)confirmed that the most frequently isolated strain of Corynebacterium (Corynebacum sp.) is a contaminant in the cohort of orthopedic patients. 55reported that Corynebacumsp has a significant tendency towards bones and joints. All in all, the existing reports have shown the influence of Coryneillus spp on the bones, and the significance of its significant increase in TD affected broilers is worthy of in-depth exploration. It may bring new insights into the pathogenesis of TD(Tu et al. 2021).
As a rising and increasingly popular disease diagnosis method in recent years, metabonomics has also been used to understand the changes of serum metabolites in Thiram-induced TD broilers. Our results showed that 10 metabolites in serum changed significantly. Nine metabolites such as N-Desmethylcitalopram, 33-Deoxy-33-hydroperoxyfurohyperforin, Imipramine, Corchorusoside B, Myricatomentoside I and Glucosylceramide were significantly up-regulated in TD group, accompanied by a decrease in the concentration of PE (DiMe (13L5) / MonoMe (11P3)). Pathway enrichment revealed that drug metabolism-cytochrome P450 and Sphingolipid metabolism, two metabolic pathways in which Imipramine and Glucosylceramide (GlcCer) engage, were important in the development of TD. It has been reported that thiram can significantly reduce the activity of cytochrome P450 and induce hepatotoxicity in mice(Dalvi et al. 2002). This can explain the occurrence of liver injury in TD broilers. Previous research has demonstrated that sphingolipid metabolism is important in the onset and progression of a range of liver disorders(Régnier et al. 2019; Li et al. 2020a; Simon et al. 2020). Sphingomyelin includes ceramide and its derivatives, such as ceramide-1-phosphate, glucose ceramide (GlcCer) and sphingosine-1-phosphate, which are important structural components of cell membrane(Kartal Yandım et al. 2013). It is well recognized that ceramide is the central molecule of sphingomyelin metabolism, which usually regulates anti-proliferation responses such as cell growth suppression, apoptosis induction, and/or aging control(Castro et al. 2014; Nganga et al. 2018). Glycosphingolipids such as Glucosylceramide (GlcCer) are a class of bioactive molecules with diverse structures, which have an important function in signal transduction and the pathogenesis of various diseases(Gan et al. 2021). GlcCer is the product of Glucosylceramidesynthetase (GCS) transferring glucose from UDP- glucose to ceramide. GCS regulates the physiological activity of cells by controlling the balance of Cer/GlcCer(Bleicher and Cabot 2002). The imbalance between Cer and GlcCer or other complex glycosphingolipids can induce malignant growth of cells. Highly active GCS accelerates the transformation of Cer to glycosylation, thus reducing apoptosis and rapid proliferation of cancer cells(Kartal Yandım et al. 2013; Fujiwara et al. 2020). In addition, studies have shown that many viruses, bacteria and bacterial toxins can bind to the carbohydrate components of glycosphingolipids on the surface of host cells. Glycosphingolipids have the potential to operate as functional receptors for microbes and bacterial toxins, and assist it in the process of invading the body(Radin 1994; Van Meer et al. 2003). In summary, we reasonably speculate that the significant up-regulation of GlcCer may increase the risk of viral infection and induce the rapid proliferation of cancer cells. The synthesis of GlcCer requires glucose, so its upregulation also explains the significant decrease in the concentration of Glu in serum biochemical tests(Liao et al. 2018). Glucose is not only the key energy source to sustain life, but also the most important form of energy supply in most organisms. This suggests that the disturbance of Sphingolipid metabolism in TD affected broilers can participate the development of TD by interfering with energy supply.
The single group analysis method can provide information about different life processes or biological processes that are different between the disease group and the normal group. However, these analyses often have limitations. The multi-group method integrates the information of several groups, provides more evidence for the biological mechanism, and excavates the candidate key factors from the deep level.
For this reason, we combined the microbiome and metabolic to carry out Spearman correlation analysis. The results showed that Glucosylceramide and PE (DiMe (13,5) / MonoMe (11,3)) have a deep relationship with the change of intestinal microbiota. The Spearman correlation between the 6 genera and 2 metabolites related to TD was further observed in the related network. We found that Glucosylceramide and PE (DiMe (13,5) / MonoMe (11,3)) are the hubs of the entire network. Corynebacterium_1, Lotus_japonicus, Bradyrhizobium and uncultured_bacterium_f_Beggiatoaceae were substantially favorably connected with Glucosylceramide rise, but inversely associated with PE (DiMe (13,5) / MonoMe (11,3)). CHKCI001 has a high negative correlation with Glucosylceramide, and a positive correlation with PE (DiMe (13,5) / MonoMe (11,3)). Collectively, these findings imply that metabolites can affect the structure and composition of intestinal microflora. In turn, the imbalance of intestinal microflora is the most critical factor in the regulation of serum metabolites.
In our experiment, we initially developed an animal model of thiram-induced TD and verified its use in biomarkers and histopathology. Afterward we presented the alterations in the composition and structure of the gut microflora of TD broilers induced by thiram,as well as the disturbance of Sphingolipid metabolism. We have also discovered the vast majority vital connected pairings of bacteria and lipid metabolites. The combined analysis of microbiome and metabolome revealed the pathogenesis of TD from a new perspective.
Ethical Approval
All animal tests were carried out in accordance with the ethical standards and procedures authorized by the Animal Care Committee of South China Agricultural University in China (permit number: 32002350).
CREDIT Author Statement
Yingwei Liu, Shouyan Wu and Kai Liu: Conceptualization, Formal analysis, Methodology, Writing - original draft. Yingwei Liu: Validation, Investigation. Ying Li, Jiaqiang Pan, Lianmei Hu, Jianzhao Liao, Mujahid Iqbal and Khalid Mehmood: Formal analysis, Writing - review & editing. Ying Li, Zhaoxin Tang and Hui Zhang: Project administration, Funding acquisition, Writing - review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Funding
This work was financially supported by the National Natural Science Foundation of China (No. 32002350), Guangdong Basic and Applied Basic Research Foundation (No. 2020A1515110149, 2021A1515010469), and the Youth Innovative Talents Project of Education Department of Guangdong Province (No. 2020KQNCX007).
Availability of data and materials
The data supporting the research results can be obtained by the corresponding authors upon reasonable request. The author believes that all data have been included in the supplementary materials.
- Albillos A, de Gottardi A, Rescigno M (2020) The gut-liver axis in liver disease: Pathophysiological basis for therapy. J. Hepatol. 72
- Badell E, Hennart M, Rodrigues C, et al (2020) Corynebacterium rouxii sp. nov., a novel member of the diphtheriae species complex. Res Microbiol 171:. https://doi.org/10.1016/j.resmic.2020.02.003
- Belaid C, Sbartai I (2021) Assessing the effects of Thiram to oxidative stress responses in a freshwater bioindicator cladoceran (Daphnia magna). Chemosphere. https://doi.org/10.1016/j.chemosphere.2020.128808
- Bernard K (2012) The genus Corynebacterium and other medically relevant coryneform-like bacteria. J. Clin. Microbiol. 50
- Bleicher RJ, Cabot MC (2002) Glucosylceramide synthase and apoptosis. Biochim. Biophys. Acta - Mol. Cell Biol. Lipids 1585
- Cabral Michel D, Azarias Guimarães A, Martins da Costa E, et al (2020) Bradyrhizobium uaiense sp. nov., a new highly efficient cowpea symbiont. Arch Microbiol 202:. https://doi.org/10.1007/s00203-020-01827-w
- Castro BM, Prieto M, Silva LC (2014) Ceramide: A simple sphingolipid with unique biophysical properties. Prog. Lipid Res. 54
- Caussy C, Reeder SB, Sirlin CB, Loomba R (2018) Noninvasive, Quantitative Assessment of Liver Fat by MRI-PDFF as an Endpoint in NASH Trials. Hepatology 68
- Chen X, Fang M, Chernick M, et al (2019) The case for thyroid disruption in early life stage exposures to thiram in zebrafish (Danio rerio). Gen Comp Endocrinol 271:. https://doi.org/10.1016/j.ygcen.2018.11.003
- Clavijo V, Flórez MJV (2018) The gastrointestinal microbiome and its association with the control of pathogens in broiler chicken production: A review. Poult. Sci. 97
- Dalvi PS, Wilder-Kofie T, Mares B, et al (2002) Effect of cytochrome P450 inducers on the metabolism and toxicity of thiram in rats. Vet Hum Toxicol 44:
- Fujiwara K, Yazama H, Donishi R, et al (2020) Inhibitory Effects of Glucosylceramide on Tumorigenesis Induced by a Carcinogen in Mice. Laryngoscope 130:. https://doi.org/10.1002/lary.28449
- Gan J, Zheng SJ, Mao XR, Li JF (2021) The role of glucosylceramide and glucosylceramide synthase in liver disease: from bench to bedside. Acta Biochim Pol 68:. https://doi.org/10.18388/abp.2020_5478
- H. H (2010) Fungicides and Their Effects on Animals. In: Fungicides
- Hacker E, Antunes CA, Mattos-Guaraldi AL, et al (2016) Corynebacterium ulcerans, an emerging human pathogen. Future Microbiol. 11
- Huang S cheng, Cao Q qin, Cao Y bing, et al (2021) Morinda officinalis polysaccharides improve meat quality by reducing oxidative damage in chickens suffering from tibial dyschondroplasia. Food Chem 344:. https://doi.org/10.1016/j.foodchem.2020.128688
- Imai A, Ohtani M, Nara A, et al (2020) The Lotus japonicus nucleoporin GLE1 is involved in symbiotic association with rhizobia. Physiol Plant 168:. https://doi.org/10.1111/ppl.12996
- Jaén-Luchoro D, Gonzales-Siles L, Karlsson R, et al (2020) Corynebacterium sanguinis sp. nov., a clinical and environmental associated corynebacterium. Syst Appl Microbiol 43:. https://doi.org/10.1016/j.syapm.2019.126039
- Kalt F, Schulthess B, Sidler F, et al (2018) Corynebacterium species rarely cause orthopedic infections. J Clin Microbiol 56:. https://doi.org/10.1128/JCM.01200-18
- Kartal Yandım M, Apohan E, Baran Y (2013) Therapeutic potential of targeting ceramide/glucosylceramide pathway in cancer. Cancer Chemother. Pharmacol. 71
- Kong A, zhang C, Cao Y, et al (2020) The fungicide thiram perturbs gut microbiota community and causes lipid metabolism disorder in chickens. Ecotoxicol Environ Saf 206:. https://doi.org/10.1016/j.ecoenv.2020.111400
- Li J, Bi D, Pan S, Zhang Y (2007) Effect of diet with thiram on liver antioxidant capacity and tibial dyschondroplasia in broilers. Br Poult Sci 48:. https://doi.org/10.1080/00071660701665858
- Li L, Wang H, Jones JW (2020a) Sphingolipid metabolism as a marker of hepatotoxicity in drug-induced liver injury. Prostaglandins Other Lipid Mediat 151:. https://doi.org/10.1016/j.prostaglandins.2020.106484
- Li Z, Li A, Zhang J, et al (2020b) Identification and expression analysis of microRNAs in tibial growth plate of chicken through thiram toxicity. Environ Sci Pollut Res 27:. https://doi.org/10.1007/s11356-019-06648-z
- Liao J, Guan Y, Chen W, et al (2018) ACBD3 is required for FAPP2 transferring glucosylceramide through maintaining the Golgi integrity. J Mol Cell Biol 11:. https://doi.org/10.1093/jmcb/mjy030
- Liu K, Li Y, Iqbal M, et al (2022a) Thiram exposure in environment: A critical review on cytotoxicity. Chemosphere 295
- Liu Y, Yi J, Li Y, et al (2022b) Residue of thiram in food, suppresses immune system stress signals and disturbs sphingolipid metabolism in chickens. Vet Immunol Immunopathol 247:. https://doi.org/10.1016/j.vetimm.2022.110415
- Lloyd-Price J, Arze C, Ananthakrishnan AN, et al (2019) Multi-omics of the gut microbial ecosystem in inflammatory bowel diseases. Nature 569:. https://doi.org/10.1038/s41586-019-1237-9
- Mazidi M, Rezaie P, Kengne AP, et al (2016) Gut microbiome and metabolic syndrome. Diabetes Metab. Syndr. Clin. Res. Rev. 10
- Nganga R, Oleinik N, Ogretmen B (2018) Mechanisms of Ceramide-Dependent Cancer Cell Death. Adv Cancer Res 140:. https://doi.org/10.1016/bs.acr.2018.04.007
- Noussair L, Salomon E, El Sayed F, et al (2019) Monomicrobial bone and joint infection due to Corynebacterium striatum: literature review and amoxicillin-rifampin combination as treatment perspective. Eur J Clin Microbiol Infect Dis. https://doi.org/10.1007/s10096-019-03542-x
- Radin NS (1994) Glucosylceramide in the nervous system - A mini-review. Neurochem. Res. 19
- Rath NC, Huff WE, Huff GR, Kannan L (2007) Induction of tibial dyschondroplasia by carbamate and thiocarbamate pesticides. Avian Dis 51:. https://doi.org/10.1637/0005-2086(2007)51[590:IOTDBC]2.0.CO;2
- Rath NC, Richards MP, Huff WE, et al (2005) Changes in the tibial growth plates of chickens with thiram-induced dyschondroplasia. J Comp Pathol 133:. https://doi.org/10.1016/j.jcpa.2005.01.005
- Régnier M, Polizzi A, Guillou H, Loiseau N (2019) Sphingolipid metabolism in non-alcoholic fatty liver diseases. Biochimie 159
- Roux V, Drancourt M, Stein A, et al (2004) Corynebacterium Species Isolated from Bone and Joint Infections Identified by 16S rRNA Gene Sequence Analysis. J Clin Microbiol 42:. https://doi.org/10.1128/JCM.42.5.2231-2233.2004
- Salam S, Arif A, Mahmood R (2020) Thiram-induced cytotoxicity and oxidative stress in human erythrocytes: an in vitro study. Pestic Biochem Physiol. https://doi.org/10.1016/j.pestbp.2019.12.003
- Shah V, Subramaniam S (2018) Bradyrhizobium japonicum USDA110: A representative model organism for studying the impact of pollutants on soil microbiota. Sci. Total Environ. 624
- Sharma A, Kumar V, Shahzad B, et al (2019) Worldwide pesticide usage and its impacts on ecosystem. SN Appl. Sci. 1
- Sharma VK, Aulakh JS, Malik AK (2003) Thiram: Degradation, applications and analytical methods. J. Environ. Monit.
- Shrestha A, Zhong S, Therrien J, et al (2020) Lotus japonicus Nuclear Factor YA1, a nodule emergence stage-specific regulator of auxin signalling. New Phytol. https://doi.org/10.1111/nph.16950
- Simon J, Ouro A, Ala-Ibanibo L, et al (2020) Sphingolipids in non-alcoholic fatty liver disease and hepatocellular carcinoma: Ceramide turnover. Int. J. Mol. Sci. 21
- Thingholm LB, Rühlemann MC, Koch M, et al (2019) Obese Individuals with and without Type 2 Diabetes Show Different Gut Microbial Functional Capacity and Composition. Cell Host Microbe 26:. https://doi.org/10.1016/j.chom.2019.07.004
- Tong X, Rehman MU, Huang S, et al (2018) Comparative analysis of gut microbial community in healthy and tibial dyschondroplasia affected chickens by high throughput sequencing. Microb Pathog 118:. https://doi.org/10.1016/j.micpath.2018.03.001
- Tripathi A, Debelius J, Brenner DA, et al (2018) The gut-liver axis and the intersection with the microbiome. Nat. Rev. Gastroenterol. Hepatol. 15
- Tu Y, Yang R, Xu X, Zhou X (2021) The microbiota-gut-bone axis and bone health. J. Leukoc. Biol. 110
- Van Meer G, Wolthoorn J, Degroote S (2003) The fate and function of glycosphingolipid glucosylceramide. In: Philosophical Transactions of the Royal Society B: Biological Sciences
- Waheed N, Elias-Smale S, Malas W, et al (2020) Sex differences in non-obstructive coronary artery disease. Cardiovasc. Res. 116
- Wang SZ, Yu YJ, Adeli K (2020) Role of gut microbiota in neuroendocrine regulation of carbohydrate and lipid metabolism via the microbiota-gut-brain-liver axis. Microorganisms 8
- Wiest R, Albillos A, Trauner M, et al (2017) Targeting the gut-liver axis in liver disease. J. Hepatol. 67
- Yang B, Liu Y, Li Y, et al (2021) Exposure to the herbicide butachlor activates hepatic stress signals and disturbs lipid metabolism in mice. Chemosphere 283:. https://doi.org/10.1016/j.chemosphere.2021.131226
- Zhang H, Mehmood K, Jiang X, et al (2018) Effect of tetramethyl thiuram disulfide (thiram) in relation to tibial dyschondroplasia in chickens. Environ Sci Pollut Res 25:. https://doi.org/10.1007/s11356-018-2824-2
- Zhang H, Wang Y, Mehmood K, et al (2020) Treatment of tibial dyschondroplasia with traditional Chinese medicines: “Lesson and future directions.” Poult. Sci. 99
- Zhang JJ, Yang H (2021) Metabolism and detoxification of pesticides in plants. Sci. Total Environ. 790
- Zhou L, Wang S Bin, Chen SG, et al (2018) Prognostic value of ALT, AST, and AAR in hepatocellular carcinoma with B-type hepatitis-associated cirrhosis after radical hepatectomy. Clin Lab 64:. https://doi.org/10.7754/Clin.Lab.2018.180532
- Zitvogel L, Ma Y, Raoult D, et al (2018) The microbiome in cancer immunotherapy: Diagnostic tools and therapeutic strategies. Science (80-. ). 359
No competing interests reported.