Introduction

Coal mining plays a pivotal role in the global economy, but is not without significant occupational health risks, particularly due to exposure to coal dust1. China is rich in coal resources, and its coal output ranks first in the world2,3. The silica dust produced by coal miners in coal mining operations poses a major threat to their health4. Susceptible workers exposed to coal miners and silica dust may develop various lung diseases, the main example being classic pneumoconiosis, which is a nodular interstitial lung disease and may lead to progressive mass fibrosis in severe cases5,6,7. In addition, the specific environment in underground coal mines can disrupt the microecological balance of the throat, leading to a reduction in beneficial bacteria and an enrichment of harmful bacteria, which in turn can induce or aggravate respiratory and overall health damage to coal miners8. It can be seen that the long-term occupational exposure of coal miners has caused serious harm to their health.

The human microbiome may be influenced by environmental changes9, and the respiratory microbiome is increasingly recognized as an important driver of respiratory health10. As an important part of the front end of the respiratory system, the microbial composition of the pharynx is closely related to the progression of pulmonary diseases such as pulmonary fibrosis11. It has been reported that the throat is an important reservoir of respiratory flora, which may be used as a predictor of pulmonary microbial community status12. Imbalance of pulmonary microbiota, involving species such as Bacteroides and Prevotella, has been shown to promote pulmonary fibrosis in bleomycin-induced models13. In addition, changes in the composition and function of the microbiome are thought to be a key cause of metabolic disorders14. It has been reported that Marinifilaceae belongs to Bacteroideae, and its abundance is related to lipid metabolites, which has been proven to regulate lipid metabolism15. The microbiota can promote cholesterol synthesis and deposition by interfering with the metabolism of bile acids, a metabolite of microorganisms, and exacerbate lipid metabolism disorders16. In the relevant studies of coal miners, with the increase of working years exposed to coal dust, coal miners have a higher risk of dyslipidemia1. Studies have confirmed changes in serum lipid profile and abnormal lipid metabolism in patients with coal workers’ pneumoconiosis17. Therefore, it is of great significance for the prevention and control of occupational diseases to explore the potential markers or early warning factors before coal miners fall ill. However, the changes of microbial communities and lipid profiles of early coal miners have not been reported.

Therefore, based on 16sRNA gene sequencing technology and lipid metabolomics, this study explored the pharyngeal microbial community characteristics and plasma lipid profile changes of coal miners with different years of exposure to dust, and provided information for prevention and management strategies for high-risk occupational groups.

Materials and methods

Study population

Twenty-two coal mine workers who underwent occupational health examination in a hospital in Tangshan City, Hebei Province, China from May to June 2023 were selected as the study subjects. Among them, the company executives who do not come into contact with dust at work are taken as the control group, and the front-line miners engaged in occupational dust exposure work are taken as the dust exposure group. According to the years of exposure to dust, they were divided into four groups: control group (no exposure to dust, C), short-term exposure to dust group (exposure to dust less than 10 years, D-1), medium exposure to dust group (exposure to dust 10 to 20 years, D-2), and long-term exposure to dust group (exposure to dust greater than 20 years, D-3). Inclusion criteria for research subjects: Engaged in coal mine work, and aged 18 to 69. Exclusion criteria: Those who had pharyngitis within the 3 months prior to the investigation, those who had asthma and other respiratory diseases before joining the work, those who had diabetes, hypertension and heart disease, and those who had used antibiotics or drugs acting on the nervous system in the past month. This study was reviewed and approved by the Medical Ethics Committee of Kailuan Rehabilitation Hospital of Tangshan, and all subjects signed informed consent.

Sample collection and DNA extraction

Before sampling, the research subjects first rinsed their mouths with sterile 0.9% sodium chloride solution to clean their mouths. Then, professional samplers held sterile throat swabs and rotated them around the posterior pharyngeal wall for one cycle to collect samples. All samples were collected centrally within the same period of time. After collection, they were immediately stored in a -80℃ refrigerator until DNA extraction. The total genomic DNA extraction of the microbial community was carried out in accordance with the instructions of HiPure Soil DNA Kit (Magen, Guangzhou, China). The integrity of the extracted genomic DNA was detected by 2% agarose gel electrophoresis. In addition, plasma samples of the subjects were collected and stored at -80℃ for use.

PCR amplification and sequencing library construction

Using the extracted genomic DNA as a template, PCR amplification of the V3-V4 variable region of the 16 S rRNA gene was performed using primers 338 F (5 ‘-ACTCCTACGGGAGGCAGCAG-3’) and 806R (5 ‘-GGACTACHVGGGTWTCTAAT-3’) with Barcode sequences. The total volume of the PCR reaction system is 50 µL, including: 10 µL 5× Q5® Reaction Buffer, 10 µL 5× Q5® High GC Enhancer, 1.5 µL 2.5 mM dNTPs, 1.5 µL upstream primer (10 µM), 1.5 µL downstream primer (10 µM), 0.2 µL Q5® High-Fidelity DNA Polymerase and 50 ng template DNA (PCR-related reagents were all purchased from New England Biolabs, USA). The amplification procedure is: pre-denaturation at 95℃ for 5 minutes, 30 cycles (denaturation at 95℃ for 1 minute, annealing at 60℃ for 1 minute, and extension at 72℃ for 1 minute), Finally, extend at 72℃ for 7 minutes. For each PCR reaction, a negative control was set up with nucleic acid-free water instead of the template to monitor potential reagent contamination.

The PCR products were detected by 2% agarose gel electrophoresis (no band in the negative control), and then purified using AMPure XP Beads (Beckman Coulter, CA, USA). Subsequently, sequencing libraries were constructed using the Illumina DNA Prep Kit (Illumina, CA, USA) and sequenced in PE250 mode on the Illumina NovaSeq 6000 platform.

High-throughput sequencing and bioinformatics analysis of 16sRNA gene

The quality of the original sequencing sequences at both ends was controlled using FASTP software (version 0.18.0), including: (1) reads with an unknown nucleotide (N) removal ratio of ≥ 10%, (2) Remove reads containing headers, (3) Remove low-quality reads (where the proportion of bases with a quality value of ≤ 20 exceeds 50%). The paired reads were concatenated into Tags using FLASH software (version 1.2.11), with a minimum overlap length of 10 bp and a maximum misalignment rate of 2%. Subsequently, the Raw Tags were strictly filtered in accordance with the reference standard18 : they were truncated from the first point where the number of consecutive low-quality bases (mass value ≤ 3) reached 3, and the sequences whose truncated length was less than 75% of the original Tags were discarded. Eventually, high-quality Clean Tags were obtained.

The UPARSE algorithm in USEARCH software (version 11.0.667) was used to cluster the Clean Tags with OTUs with a similarity of 97%, and the sequences with the highest abundance were selected as the representative sequences of OTUs. The chimeric sequence was removed using the UCHIME algorithm. Taxonomic annotation of species was conducted using RDP Classifier (version 2.13) to compare with the SILVA 16 S rRNA gene database (version 138.2), and the confidence threshold was set at 0.8. To eliminate the influence of sequencing depth differences, the OTU abundance table was flattened to the minimum number of sequences for each sample before diversity analysis.

Alpha diversity index (ACE, Sob, Simpson and Shannon) formula mothur (https://mothur.org/wiki/calculators/) in R language (version 4.3.1) in the calculation. Beta diversity analysis was conducted using the R language vegan package (version 2.5.6) based on the Bray-Curtis distance for principal coordinate analysis (PCoA). Functional prediction uses the Tax4Fun package (version 1.0) to infer the abundance of the KEGG pathway based on the standardized OTU table.

Plasma lipid metabolomics

Lipid metabolites in plasma were extracted with extraction solution (methanol: acetonitrile = 1:1), and equal volume extracts from each group were mixed as quality control (QC) samples which were used to test the stability and reliability of the entire testing process, and QC and each group of samples were placed in the mass spectrometer analysis tube for computer detection.

Lipid analysis was performed by Waters I-Class Acquity UPLC in conjunction with Vion IMS QToF, and samples were separated using a BEH C18 1.7 μm column (2.1 × 100 mm). Mobile phase A consisted of 0.1% formic acid, while mobile phase B is a mixture of methanol and acetonitrile (volume ratio 1:1) and contains formic acid with a volume fraction of 0.1%. Fragment mass spectrometry was performed using the MSE mode. The lipid metabolites were separated by gradient elution under the following conditions. 0–1 min, 99 − 70% A; 2.5–6.5 min, 40 − 10% A; 7–10 min, 0% A. The column temperature was maintained at 45 ℃ and the flow rate was 0.4 mL/min. The following table describes the parameters. MS range, m/z 50-1000; scanning time 0.2s; collision energy (CE) 6 eV; desorption temperature 500 ℃; source temperature, 120 °C; desorption gas flow, 1000 L/h; cone gas flow, 50 L/h; capillary voltage, 2000 V; lock calibration (lock nebulizer reference: mass, 556.2766 m/z; sampling time: 0.5 min; CE, 6 eV; the flow rate was 10 µL/min)19.

Using Progenesis QI v3.0 (UK, Nonlinear Dynamics) software, the data are reviewed for alignment, peak detection, peak identification and deconvolution normalization to obtain the peak intensity of important retention time (RT) and mass-charge ratio (m/z) data pairs. Subsequently, the sample distribution was observed by principal component analysis (PCA), variable weight value (VIP > 1) was obtained by OPLS-DA analysis, and metabolites with significant differences among groups were obtained by ANOVA analysis of variance. P < 0.05 was considered statistically significant. Then, using Progenesis QI compound identification plug-in ChemSpider network chemical structure database, parameter Settings: Mass error tolerance 5ppm, fragment mass error tolerance 5ppm, isotopic similarity greater than 85%, Choose LIPID MAPS (https://www.lipidmaps.org/), the human metabolome database (HMBD) and the Kyoto encyclopedia gene and genome (KEGG) data sources and plug-in LipidBlast database for LIPID metabolites of recognition and identification. To further improve the reliability of compound identification, online databases HMDB and LipidCCS (www.metabolomics-shanghai.org/LipidCCS) were used to match the differential metabolite Collision Cross Section. A metabolite with an error tolerance of ± 5% for CCS values was eventually identified as a potentially significant differentially differentiated metabolite. Using MetaboAnalyst5.0 metabolic analysis (http://www.metaboanalyst.ca) online software to generate the hierarchical clustering analysis diagram analysis heat maps, enrichment and metabolic pathways.

Statistical analysis

In this study, SPSS21 software was used for statistical analysis. Categorical variables are expressed as constituent ratios. Pearson Chi-square test is used to evaluate differences in categorical variables. For cells with expected count < 5, Fisher exact test is used. In addition, one-way ANOVA was used to compare the mean values of more than two groups, then LSD test was performed, and independent sample t test was used to compare the mean values of the two groups. All statistical analyses of 16sRNA sequencing were mainly performed in the R language environment (version 4.3.1) and on the Omicsmart online platform (http://www.omicsmart.com). All 22 samples in this study were independently collected and no repeated measurement design was adopted. Therefore, one-way ANOVA was used for the comparison between groups of the Alpha diversity index and Tukey’s HSD post hoc tests were conducted. The screening of differential species was conducted using Indicator species analysis. The indicator values of each species in different groups were calculated using the R language labdsv package, and 10 × 10 cross-validation was used to assist in the evaluation. To avoid missing potential difference information in exploratory analysis, we chose nominal P < 0.05 as the screening threshold. The significance of the differences between Beta diversity groups was evaluated by ANOSIM analysis based on the Bray-Curtis distance matrix (999 permutations).

Results

Baseline characteristics of participants

Throat swabs and serum samples were included from 22 participants based on inclusion and expulsion criteria. As shown in Table 1, smoking, alcohol consumption, and fatty liver status were not significantly different between groups C, D-1, D-2, and D-3, and there were statistical differences in age distribution.

Table 1 Characteristics of the coal miner at baseline.
Full size table

α diversity and β diversity of pharyngeal microbiota

α diversity represents the richness and diversity of species within the microbiome. The ACE and Sob indices are used to reflect richness, and ACE and Sob index analysis showed (Fig. 1A-B) that compared with C, species richness in D-2 and D-3 groups was significantly reduced, while D-1 had no significant difference. Simpson and Shannon index analyzed species diversity (Fig. 1C-D). Compared with C, there was no significant difference between group D-1 and group D-2, and species diversity decreased in group D-3. The above shows that the richness and diversity of species in the microbiota of coal miners did not change during the 10 years after exposure to dust. The species diversity of coal miners’ pharynx changed during 10 to 20 years after exposure to dust. After 20 years of exposure to dust, the richness and diversity of species in the pharyngeal microbiota of coal miners changed.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

α diversity of pharyngeal microbiota of coal miners. (A) The microbial richness was measured by the Ace. (B) The microbial richness was measured by the Sobs. (C) The α-diversity was measured by the Simpson. (D) The α-diversity was measured by the Shannon. *P < 0.05, **P < 0.01, and ***P < 0.001.

Then, based on Bray, we performed PCoA analysis (Fig. 2A) for the complexity of microflora structure between samples to reflect the similarity of the colony structure among different samples, and Anosim test (Fig. 2C-I). The results showed that the difference between groups was greater than the difference within groups. In addition, Venn diagram (Fig. 2B) analysis found that a total of 3871 outs were screened. There were 176 common outs between groups C, D-1, D-2, and D-3. On the other hand, compared with group C, 753 unique outs were detected in the D-1 group, 920 unique outs in the D-2 group, and 302 unique outs in the D-3 group.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

β diversity of pharyngeal microbiota of coal miners. (A) PCoA analysis at OUT level based on Bray. (B) Venn diagram of pharyngeal microbiota analysis. (C–I) Anosim test of each group at the unweighted-unifrac OUT level.

Microbial composition of pharynx

The composition analysis of Phylum showed that Bacteroidota, Proteobacteria and Firmicutes were the common dominant bacteroidota in pharyngeal microflora of all pharyngeal swab samples (Fig. 3A). We then used the Tukey HSD test to analyze and compare differences in the abundance of the 4 groups of microbiomes (Fig. 3B). The results showed that compared with the C group, the relative abundance of Bacteroidota in the D-3 group was significantly increased (P < 0.05). Indicator analysis (Fig. 3C) found that Bacteroidota et al. were the difference markers between the four groups. In addition, Genus composition analysis showed (Fig. 3D) that the dominant bacteria in pharyngeal microbiota include 10 species, including Prevotella, Neisseria, Lactobacillus, Porphyromonas, etc. Tukey HSD analysis found (Fig. 3E) that compared with group C, The relative abundance of Prevotella in D-3 group was significantly increased (P < 0.05). Indicator analysis (Fig. 3F) found that Prevotella et al. were markers of difference between the four groups.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Microbial composition of pharynx of coal miners. (A) The relative abundance of microorganisms at the Phylum level. (B) Tukey HSD statistical test for Phylum level microorganisms. (C) Indicator analysis of microorganisms at Phylum level. (D) The relative abundance of microorganisms at the Genus level. (E) Tukey HSD statistical test for Genus level microorganisms. (F) Indicator analysis of microorganisms at Genus level.

Changes in pharyngeal microbial function

To explore changes in the function of pharyngeal microorganisms in different groups, we performed Tax4Fun analysis (Fig. 4). By analyzing the abundance composition of the pharyngeal microbiome Kyoto gene and genomic pathway (level2), The results showed that Membrane Transport, Cell Growth and Death, Lipid Metabolism and Endocrine System were changed in the dust-exposed workers. In addition, by analyzing the abundance composition of the pharyngeal microbiome Kyoto gene and genome pathway (level3), it was found that, compared with group C, Cell cycle-Caulobacter, Oxidative phosphorylation, Arginine and proline metabolism, ABC transporters decreased in workers exposed to dust. However, Amino sugar and nucleotide sugar metabolism increased in workers exposed to dust.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Microbial function of pharynx in coal miners. (A) function prediction heat map for Tax4Fun (level2). (B) function prediction heat map for Tax4Fun (level3).

Screening of plasma differential lipid metabolites

This study was based on ultra-high performance liquid chromatography ion mobility spectroscopy quadrupole-time of flight mass spectrometer (UPLC-IMS-QTOF) to analyze differential lipid metabolites in the plasma of coal miners. The results of two-dimensional ion strength maps (Fig. 5) show that the ion strength maps of different groups have similar ion distributions. The PCA results of positive ions (Fig. 6A) showed clear sample clusters in all 4 groups. Subsequently, OPLS-DA and S-plot analysis were used to further analyze the metabolic differences between the control group and the dust exposure group, and the plasma samples of group C and group D-1 were significantly separated (R2Y = 1, Q2 = 0.655). The plasma samples of group C and group D-2 were significantly separated (R2Y = 0.921, Q2 = 0.385), and the plasma samples of group C and group D-3 were significantly separated (R2Y = 0.862, Q2 = 0.423). The cross-arrangement test of 200 iterations showed that no overfitting occurred in the models, indicating that all models were reliable (Fig. 6B-J). In addition, compared with group C, four biomarkers of lipid metabolism were obtained in group D-1, and their expression was reduced (Fig. 7A). In the D-2 group, 12 biomarkers of lipid metabolism were screened, among which the expression of 5 metabolites increased and 7 metabolites decreased (Fig. 7B). In the D-3 group, 9 biomarkers of biological metabolism were obtained, of which 3 metabolites were increased and 6 metabolites were decreased (Fig. 7C). The results of Veen diagram (Fig. 7D) showed that there were two common differential metabolites in the differential metabolites of D-1, D-2 and D-3 groups, namely Cer(d18:0/14:0) and N-Hexadecanoylpyrrolidine, and their expression was down-regulated in all 3 groups.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Two-dimensional ionic strength map of plasma samples from coal miners.

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Multivariate statistical results of plasma UPLC-IMS-QTOF data. (A) Positive ion PCA score map. (B–D) OPLS-DA, S-plot and permutation score plot of C and D-1 positive ions. (E–G) OPLS-DA, S-plot and permutation score plot of C and D-2 positive ions. (H–J) OPLS-DA, S-plot and permutation score plot of C and D-3 positive ions.

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Differential lipid metabolites in the plasma of coal miners. (A) D-1 differential lipid metabolite hierarchical cluster analysis heat map. (B) D-2 differential lipid metabolite hierarchical cluster analysis heat map. (C) D-3 differential lipid metabolite hierarchical cluster analysis heat map. (D) Venn diagram of differential lipid metabolites in D-1, D-2, and D-3.

Enrichment analysis of plasma differential lipid metabolites

In order to explore the enrichment pathway of these differential metabolites, we conducted enrichment analysis of these differential lipid metabolites. The results showed that the differential lipid metabolites of D-1 group were mainly enriched in Ceramides and Fatty alcohols (Fig. 8A). Differential lipid metabolites in D-2 group were mainly concentrated in Glycerophosphocholines, Sesquiterpenoids and Ceramides (Fig. 8B). Differential lipid metabolites in D-3 group were mainly enriched in Glycerophosphocholines, Ceramides, and Glycerophosphates (Fig. 8C).

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

Enrichment analysis of plasma differential lipid metabolites in coal miners. (A) Bubble map of differential lipid metabolite enrichment analysis between group D-1 and group C. (B) Bubble map of differential lipid metabolite enrichment analysis between group D-2 and group C. (C) Bubble map of differential lipid metabolite enrichment analysis between group D-3 and group C.

Discussion

Coal mining environment will form a series of occupational hazards, including coal dust, high temperature, high humidity and noise, among which coal dust is the cause of many human diseases, such as asthma, pneumoconiosis and lung cancer20. For coal miners, the most common and serious disease is pneumoconiosis21. Pneumoconiosis is one of the most common occupational diseases in the world. At present, there is a lack of specific treatment for pneumoconiosis, so it has a great social and economic burden22. In addition, to date, the scientific community has not identified specific biomarkers for screening coal miners for sub-health conditions in coal mining environments20. Therefore, understanding the health effects of coal mining operations and the warning signs for miners before they are diagnosed with occupational diseases is critical to the health of coal miners.

Differences in the health and disease of bacterial populations in human airways have been identified as possible contributing factors to the pathogenesis of respiratory diseases8, while coal dust exposed to coal miners causes serious damage to workers by accumulating in the upper respiratory tract or entering the lungs23. Therefore, it is of great value for the health of coal miners to study the changes of upper respiratory tract microflora. In this study, throat swab samples of coal miners were collected, and the control group was administrative staff with no coal dust exposure history. In addition, front-line miners with coal dust exposure records were divided into 3 groups according to the time of exposure to coal dust. Baseline data showed no significant differences in smoking, alcohol consumption and fatty liver status among the four groups, and there were statistical differences in age distribution, which may be related to the further division of the dust exposure groups according to the time of dust exposure. In order to investigate the changes of microflora in the airway of coal miners, we investigated the microflora in throat swabs based on 16 S RNA gene high-throughput sequencing technology. The results showed that with the extension of dust exposure time, the OUT number of pharyngeal microflora of coal miners gradually decreased, and they were all lower than that of healthy controls, which was consistent with the previously reported conclusion that the intestinal microflora OUT number of silicosis patients was less than that of healthy controls24. In addition, there was no significant change in pharyngeal microflora when coal miners were exposed to short-term indirect dust, and the richness and diversity of microflora decreased gradually with the extension of dust exposure time. Based on the analysis of Phylum composition, it was found that the relative abundance of Bacteroidota increased after 20 years of exposure to dust. Previous studies have found that Bacteroidota is the dominant bacterial group in pneumoconiosis patients and dust-exposed people. Compared with dust-exposed people, Bacteroidota has a higher abundance25. The results of this study are consistent with the above research results, revealing that with the extension of dust exposure time, Bacteroidota plays an important role in the sub-health of coal miners, which may be the early warning marker before the diagnosis of occupational disease. Based on Genus composition analysis, the relative abundance of Prevotella increased significantly after 20 years of exposure to dust, and the indicator revealed that it may be a marker of difference among coal miners with different exposure times. Prevotella belongs to Gram-negative anaerobic bacteria, which is very abundant in various parts of the body, such as oral cavity, respiratory tract, alveolar lavage fluid, etc. Prevotella is a key player in the balance between health and disease25,26. It has been reported that silica nanoparticles lead to an increase in the abundance of Prevotella in mice27. Prevotella is also the most common bacterial genus in the pulmonary flora of dust exposed workers and pneumoconiosis patients, and its abundance in patients with pneumoconiosis is significantly higher than that of dust exposed workers25. The above study further revealed that Prevotella plays an important role in the sub-health of coal miners, and may be a warning marker before coal miners are diagnosed with occupational diseases.

In order to investigate the functional changes of pharyngeal microbiota in coal miners with different years of exposure to dust, the functional analysis was carried out by Tax4Fun. The function of Lipid Metabolism is changed in dust exposure group. The relative abundance of Prevotella increased significantly after 20 years of exposure to dust based on Genus composition analysis. Previous studies have confirmed that Prevotella activates the chronic inflammatory response of the host through the TLR4 and mTOR signaling pathways, significantly increasing the host’s fat deposition28. In addition, Prevotella can promote intestinal lipid absorption, upregulate the expression of lipid transport proteins (such as CD36 and FABP2), and interfere with lipid metabolism through inflammatory mediation29. The proposal of the “gut microbiota-lipid Metabolism Axis” also clearly indicates that the gut microbiota is a key metabolic regulatory factor, which is related to atherosclerotic lipid profiling through immune metabolic regulation and bioactive metabolite signal transduction30,31. It has been reported that patients with pneumoconiosis are accompanied by dyslipidemia32. In addition, studies have found that lipids are involved in the pathophysiology of human respiratory diseases33. The results of this study reveal that lipid metabolism plays an important role in workers exposed to different dust level. However, the mechanism of lipid metabolism in coal miners remains unclear.

In order to identify potential differential lipid metabolites in early sub-health of coal miners, we evaluated the changes in plasma lipid profiles of healthy controls and dust-exposed workers based on lipid metabolomics. The results showed that the differential lipid metabolites Cer(d18:0/14:0) and N-Hexadecanoylpyrrolidine were decreased in the dust exposure group, and their expressions were differentially expressed in each dust exposure period. Ceramide (Cer) is a class of sphingolipids, which are closely related to abnormal lipid deposition34. Sphingolipids and cholesterol are two lipid partners on cellular membranes where they form specific microdomains, named lipid rafts, which mediate specific cell functions, including lipid transport, etc35. The above results suggest that Cer(d18:0/14:0) may play an important role in the sub-health of coal miners. In addition, enrichment analysis found that Ceramides enrichment occurred in coal miners at different dust exposure times, and gradually enriched to glycerophosphocholines (GPC) with the extension of dust exposure time. GPC, also known as glycerophospholipids, are the most common and abundant lipid category36. As we all know, GPC is an important precursor for intracellular choline supply in phosphatidylcholine (PC) metabolism37. PC, as the source of lipid signal molecules, plays an important role in signal transduction38. It has been reported that PC can regulate lipid deposition by regulating peroxisome proliferator-activated receptors (PPARs)39. Previous studies have reported that PC is enriched in the lung tissue of silicosis rats and may play an important role in the progression of silicosis19. These results suggest that GPC play an important role in workers with different exposure years.

However, this study has some limitations. First, the sample size was small, which could affect the interpretation of the results. Large sample sizes should be considered to further eliminate differences between the control group and dust exposure groups of different dust exposure years. Second, we cannot rule out that confounding factors such as dietary differences may have an effect on the microbiota. In addition, we discovered the changes in the pharyngeal microbiota and plasma lipid metabolism profiles of coal miners with different dust exposure times. Regarding the functions of the Tax4Fun pre-strategy microbiota, including lipid metabolism pathways, we have not conducted experiments to confirm how the changes in microorganisms lead to lipid metabolism disorders. We will conduct further research in the future.

In summary, based on 16s RNA high-throughput sequencing technology, this study mined the characteristics of pharyngeal microbial flora of coal miners, and the results showed that long-term dust contact led to a decrease in the diversity and richness of pharyngeal microbial flora, an imbalance of pharyngeal flora, and may induce an imbalance of lipid metabolism in the body. Therefore, based on lipid metabolomics, the changes of plasma lipid profiles in coal miners were subsequently mined in this study, and it was found that Cer(d18:0/14:0) and so on were changed in dust exposed workers, and glycerophosphocholines were significantly enriched. It is further suggested that long-term dust exposure may lead to the imbalance of pharyngeal microflora in coal miners, and induce the changes of differential lipid metabolites. The above research provided information for prevention and management strategies for high-risk occupational groups.